Citation
An approach for cadastral records reorganization and implementation of a topologically structured cadastral information system in Tanzania

Material Information

Title:
An approach for cadastral records reorganization and implementation of a topologically structured cadastral information system in Tanzania
Creator:
Derby, Francis W
Publication Date:
Language:
English
Physical Description:
xiii, 254 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Agricultural land ( jstor )
Information storage and retrieval systems ( jstor )
Land development ( jstor )
Land management ( jstor )
Land ownership ( jstor )
Land surveying ( jstor )
Land tenure ( jstor )
Land use ( jstor )
Recordings ( jstor )
Registry ( jstor )
Cadastres -- Tanzania ( lcsh )
Civil Engineering thesis, Ph.D ( lcsh )
Dissertations, Academic -- Civil Engineering -- UF ( lcsh )
Land titles -- Tanzania ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references (leaves 243-253).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Francis W. Derby.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030019306 ( ALEPH )
40921665 ( OCLC )

Downloads

This item has the following downloads:


Full Text










AN APPROACH FOR
CADASTRAL RECORDS REORGANIZATION
AND IMPLEMENTATION OF A TOPOLOGICALLY STRUCTURED
CADASTRAL INFORMATION SYSTEM IN TANZANIA















By

FRANCIS W. DERBY


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998











ACKNOWLEDGMENT

There are so many people, to whom I owe my gratitude for their assistance and

encouragement through this dissertation, it would be impossible to name them all here. I

would especially like to thank my graduate committee, who through their combined

knowledge and experience, inspired, motivated, and guided me to the end.

My utmost gratitude goes to Associate Professor David W. Gibson for giving me

the opportunity to study in this university and for believing in my ability to complete this

program successfully. The advice and guidance he gave to me during the initial stage of

my doctoral program helped me to develop the concepts for this research.

My sincere thanks goes Associate Professor Scot E. Smith, Ph.D., for the

financial support at the time when I needed it most. I am grateful for his contribution as a

major professor and the helpful ideas he offered during my most trying moments.

Sincere gratitude goes to my friend and fellow graduate student, Joe Aufinuth for

assisting me in analyzing my results and for preparing my figures. I thank him for the

agonizing moments he spent listening to my problems and complaints.

I would like to express my appreciation to Associate Professor Charles D. Ghilani,

Ph.D. and Assistant Professor Thomas A. Seybert, Ph.D., of The Pennsylvania State

University for their helpful ideas and companionship. I will always cherish the help and

advice that they offered .







To my wonderful and understanding wife, Angela, and our children, Yahan and

Pinkrah, I would like to express my appreciation for their love, devotion, and support. I

am really blessed to have a family like ours.

Finally, and most of all, I thank God for making it all possible.













TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ............................................... ii

LIST OF TABLES .................................................... vii

LIST OF FIGURES .................................................... ix

ABSTRACT ......................................................... xii

CHAPTERS

1 INTRODUCTION ............................................... 1
Research Objectives, Methodology and Scope .......................... 4
Definitions ..................................................... 7
Research Organization and Contribution ............................ 10

2 LAND MANAGEMENT AND CADASTRAL INFORMATION SYSTEM .. 13
Land Management .............................................. 14
Legal Framework ............................................... 16
Land Delivery ............................................ 18
Estate Management ....................................... 18
Regulation and Reform .................................... 19
Revenue Generation ....................................... 19
Operational Agencies ............................................ 20
Land Survey ............................................. 22
Land Titling ............................................. 23
Land Registration ......................................... 24
Valuation and Assessment ................................. 25
Information Support Systems ..................................... 28
Environmental Information System ........................... 31
Socioeconomic Information System .......................... 31
Infrastructure Information System ............................ 32
Cadastral Information Systems .............................. 32

3 CADASTRAL INFORMATION AND RELATED ISSUES .............. 36








Different Types of Cadastres .................
Existing Cadastral Information Models ..........
The North American Model (NRC model) .
Williamson's Model ..................
The Developing Country Model .........
The Wisconsin Land Information Model ...
Spatial Data ..............................
Spatial Data Capture .................
Parcel Identifiers ..........................
Data Management .........................
Flat Files ...........................
Hierarchical Files ....................
Networks ..........................
Relational Databases ..................

4 EXISTING CADASTRAL ARRANGEMENTS IN
Administrative Arrangements within Tanzania ....
Organizational Arrangements within MLHUD ....
The Urban Development Division ........
The Survey and Mapping Division .......
Land Development Division ............
The City Council ofDar Es Salaam ............
Existing Land Delivery Process ...............
Survey and Demarcation...............
Allocation and Registration.............
Cadastral Surveying Processes within Divisions ...
Surveying and Mapping Division ........
Property Valuation and Rent Assessment ..
Certificate of Occupancy ..............
Registration of Certificate of Occupancy ...


TANZANIA
..........
..........

.....,m....
.....a....
..........
..........
..........
..........

..........
..........
..........
.. .. ..


..........


5 CADASTRAL INFORMATION SYSTEMS APPROACH ..............
Discussion of Improvements to Current Cadastral Arrangements ..........
Urban Development Division ..............................
Surveys and Mapping Division ..............................
Land Development Division ................................
Document Processing Improvements ...............................
Reorganization of Existing Records ..........................
Processing of New Documents ..............................
Organization of Cadastral Information ..............................
Topologically Structured Cadastral Data Concept .....................
Topological Rules and Cadastral Index Mapping ..................
Application of the Rules to cadastral Surveying .......................


. 57
. 58
. 59
. 59
. 63
. 66
* 70
. 71
. 73
.74
. 76
. 76
. 83
. 85
. 88

.94
. 97
. 99
100
103
108
110
115
129
133
136
141







Boundary Definition ...................................... 142
Subdivisions ....................................... ..... 143
Topological Rules and Principles Illustrated ................... 146

6 CADASTRAL INFORMATION MODEL FOR TANZANIA...... ....... 153
Cadastral Index Map Compilation in Metric Space......................... 154
Cadastral Index Mapping in Topological Space.............. .......... 158
Cadastral Information System for Tanzania.. ......................... 160
Cadastral Index Map for Tanzania ........................... 160
Linkage Mechanism................................... 166


7 PILOT PROJECT ..............
Data Sources ..................
Hardware and Software ..........
Analogue image conversion .
Attribute Data Processing ........
Spatial Analysis ................
Summary and Analysis of Results ...


172
174
178
179
183
188
194

201
204


8 CONCLUSIONS AND RECOMMENDATIONS ..........
Recommendations ..................................


APPENDICES


A-1 TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND
O FFICE ............................................... 208

A-2 TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND
O FFICE ............................................... 211

B EVOLUTION OF LAND TENURE POLICIES IN TANZANIA .......... 214

C OBSERVATIONS AND CONCERNS WITH EXISTING SYSTEM ...... 233

REFERENCES ..................................................... 243

BIOGRAPHICAL SKETCH .......................................... 254


...........
...........











LIST OF TABLES


Table page

4-1 Administrative Regions in Tanzania.................................................................58

4-2 Land Registration zones in Tanzania................................................................68

5-1 Cadastral Survey Processing Tasks ................................................................. 117

5-2 Improved procedure for processing and Issuing Certificate of Occupancy ....... 123

5-3 Document processing Tasks at the Land Registry ...........................................126

5-4 Point Equivalence Table ................................................................................... 47

5-5 Line Equivalent Table .................................................................................. 148

5-6 Line-Node Topology ................................................................................... 148

5-7 Polygon-Line topology ................................................................................ 150

6-1 Coding structure of Parcel Identifiers................................................................169

6-2 Codes for Administrative regions in Tanzania .................................................170

7-1 Textual information associated with Individual Subdivision/Cadastral
Plans.....................................................................................................1... 75

7-2 Relevant Textual Information from Title Office and Land Registry....................183

7-3 Erroneous Records Identified during data entry at the Land Office..................186

7-4 Erroneous Records Isolated during data entry at the Land Registry................ 186

7-5 Results of Internal Consistency Check among the Land Office records .......... 187

7-6 Inconsistencies among Land Office and Land Registry records....................... 188







7-7 Query Types That Were Done on the Cadastral Information..........................189

7-8 Internal Inconsistencies Identified During manual Data Entry..........................196

7-9 Results of Internal Consistency Check on the Data..........................................197

7-10 Comparison of land Office and Land Registry Records....................................198


vii














LIST OF FIGURES


Figure page

2-1 Land Management and Land Information Systems.......................................... 15

2-2 Components of a cadastre ............................................................................... 27

2-3 Components of a Cadastral Information System...............................................34

3-1 NRC Model for a Multipurpose Cadastre.........................................................41

3-2 Williamson's Multipurpose Land Information model.......................................43

3-3 Cadastral Model for Developing Countries..................................................... 44

4-1 Organizational Chart of MLHUD.....................................................................60

4-2 Survey and Demarcation.................................................................................. 72

4-3 Allocation, Titling and Registration.................................................................73

4-4 Schematic Diagram of the Procedure for assessing Property.............................84

4-5 Land Office Procedure for Issuing a Certificate of Occupancy..........................86

4-6 Procedure for Registering a Certificate of Occupancy......................................89

5-1 Computerized Cadastral Data Management System........................................ 109

5-2 Procedure for Isolating Inconsistencies among Existing Land Records.......... 112

5-3 Improved Cadastral Survey Processing Procedure.............................................16

5-4 Revised Approach for Processing new Certificates of Occupancy................... 122

5-5 Improved Procedure for Processing Documents at the Land Registry............. 125


i







5-6 Topologically Structured Multipurpose Land Information Model ......................130

5-7 Node snapping precedence rule.........................................................................137

5-8 Two Definitions of a line ............................................................................... 137

5-9 A node-vertex precedence ...... .................................................................1... 38

5-10 Node-line precedence ............................................................................... 139

5-11 Node precedence after a line intersection........................................................140

5-12 V ertex snapping............................. ................................................................... 141

5-13 Two Topological representations of a Cadastral Boundary................................142

5-14 Representation of a Boundary with Vertices......................................................143

5-15 Topological errors in Cadastral Index Mapping .................................................144

5-16 Three separate cadastral surveys .......................................................................146

5-17 Two Topologically structured cadastral index maps of the same area..............151

6-1 Coordinate Transformation in Metric Space......................................................156

6-2 Topologically assembled cadastral index map..............................................159

6-3 Graphical data conversion................................................................................. 161

6-4 Creating a Cadastral Index Map for Tanzania...................................................163

7-1 Vector Drawing of Region Surrounding Pilot Area.................................... ....... 173

7-2 Topologically Generated Cadastral map of Pilot Area ......................................177

7-3 An Overlay of Cadastral map and a Geo-Referenced Aerial Photograph............ 182

7-4 Land Rent Analysis on Kijitonyama Block 44....................................................191

7-5 Land Rent Analysis on Sinza Block A..............................................................193

B-1 Effect of 1896 Land Tenure Amendment........................................................217







B-2 Tenure Structure after 1928 Amendment ..........................................................219

B-3 Land Tenure Structure after 1969 Amendment................................................ 223

B-4 The Land Policy in a Dilemma........................................................................ 227











Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfilment of the
Requirements for the Degree of Doctor of Philosophy

AN APPROACH FOR CADASTRAL RECORDS REORGANIZATION AND
IMPLEMENTATION OF A TOPOLOGICALLY STRUCTURED CADASTRAL
INFORMATION SYSTEM IN TANZANIA

By

Francis W. Derby

December 1998

Chairman: Dr. Scot E. Smith
Major Department: Civil Engineering

The government of Tanzania is in the process of changing the country's existing

land policies in favor of a land market economy. In preparation for the anticipated

increase in property conveyancing and land-related transactions, the government needed

among other things, to:

* review procedures for recording land information and to adopt methods for

strengthening administrative and cadastral capacity to support land registration

functions;

* review the mission, organizational structure, and staffing allocation of regional

offices, identify problems and institutional obstacles that prevent synergy of

information handling and record handling among related agencies and the national

registry.







* develop a well-designed administrative procedure for land registration as well as

procedures for developing a cadastral information system suitable for Tanzania.

This dissertation presents a study of the organizational arrangements and administrative

procedures for parcel allocation, parcel survey and demarcation, and the registration of all

particulars affecting the creation of a legal cadastre in Tanzania. Administrative problems

and bottlenecks that prevent the smooth flow of activities between the land management

agencies were identified.

Approaches were developed to eliminate the land records organizational problems,

improve administrative procedures for land allocation and title processing, and provide a

streamlined method for faster document processing and land record maintenance. A land

information system model was developed for Tanzania. The model uses topologically

structured graphical overlays, to provide information support for a cadastral, cadastral

information system. The graphical data are linked to the descriptive records through a

newly developed parcel identification system for Tanzania.

Procedures were developed for isolating inconsistencies in the existing records which are

in paper form and converting the error-free records into digital format. A procedure for

adding new records to the computerized system and for maintaining the records up-to-

date was developed.

A pilot project was initiated which successfully tested the topological approach for

producing the cadastral index map to support a cadastral formation system. The pilot

study highlighted the ability to perform analysis on the data and to obtain information to

support land management decisions, even in the absence of an accurately surveyed map.












CHAPTER 1
INTRODUCTION


An essential recipe for proper land management is up-to-date information

concerning location, extent, ownership and use of parcels of land. In many developing

countries, where a large percentage of the economy is tied to the land, land records are

vital for efficient land and natural resource management. When maintained properly, such

information can also help to facilitate transactions in land. In more developed countries

with dynamic land market activities, information derived from land records offers

substantial benefits to individual land owners as well as governments. For the individuals,

currency of the information ensures, among other things, faster, safer and less

cumbersome procedures for land-related transactions, protection of various rights to the

use and enjoyment of the property, and fair taxation on properties. For the government,

up-to-date and well maintained land-related information based on systematic recording of

rights in land are important in many sectors of government such as physical planning of the

land, revenue generation, infrastructure development, and environmental protection.

Advances in computer technology and data management procedures have provided

additional benefits such as the ability to perform complex statistical analysis on land

records, identify trends in land market activities, and assess impacts of developments on





2

the society. The same cannot be said about synergy of land-related information in

developing countries.

In 1974, a sub committee of the United Nations noted that,

"...systematic records of land and rights in land have great importance for public
administration, land planning, and land development, and private transactions in
land. This situation is particularly true in those developing countries where the
rapid growth of population has caused increasing pressure on rural land, while
simultaneously a massive migration of people to the cities and towns has led to the
uncontrolled growth of urban centers. Nevertheless, the need for accurate land
records is often ignored by policy-makers; and the cadastral systems of many
countries are, in consequence, highly defective..." (United Nations 1974, 25-26).


This situation resulted in improper planning and inefficient management of natural

resources, and was a catalyst to exacerbated social and economic problems. This

observation from the United Nations was probably the first to establish a link between

accurate land records and efficient resource management. Reports from the World Bank

(Feder and Davis 1991; Holstein 1990) indicate that although efforts have been made by

governments of developing countries to improve the quality of their land records, progress

has been slow and success stories are few.

Among developing African countries, the use of land as a source of livelihood

often takes precedence over its use as a marketable commodity. In some rural

communities of Africa, such as in Tanzania, Ethiopia, Zimbabwe, Uganda, Namibia, and

Kenya, parcels of land have traditionally belonged to tribes, ethnic groups, or clans.

Among rural communities, land titling and registration concepts, as are known and

practiced in the western world, are alien to established systems of communal ownership

and land stewardship. Common property rights in the traditional (indigenous) land tenure





3

system ensures that users hold land in trust for descendants of the community. In the

urban centers of these same countries, despite the relatively active land-related

transactions, many people view the processes of registration and titling as expensive, time

consuming, and offer no tangible benefits to them as individuals (Moreno in Dale and

McLaughlin, 1989, 27). The result is that only a small percentage of the land in most

developing countries is registered.

The system for recording individual rights to the use and enjoyment of parcels of

land in most developing countries function poorly. In the urban centers of most

developing countries, it is financially costly and extremely time consuming to establish

legal ownership to any land parcel, and to identify the type of limitations to the use of the

parcel (Dunkerley 1985; Dale and McLaughlin 1989). Part of the economic problem in

most developing countries has originated from improper allocation of land to its most

economic use, improper or ineffective documentation of rights to land, and consequently,

inability on the part of the government to monitor the use of the land and to invest in the

resources of the land. In addition, most developing countries have not had proper

procedures for recording transactions in land.

Experiences in developed countries indicate that private individuals, and the public

as a whole, are benefitting from the derivative information which is obtained from large

scale documentation of ownership and rights to land units. Such benefits are realized in

terms of effective land management, fast and efficient transactions in land, fair property

taxation system, and economic development.

Change, however, has begun to occur. Taylor (1991) noted that since 1986,





4

governments of many developing countries such as Mexico, Bolivia, St. Lucia, Zimbabwe,

Botswana, Thailand, and Honduras, with the help of international agencies, such as the

World Bank and the United Nations, have initiated programs to recompile and to maintain

comprehensive records of land ownership as well as transactions in land as initial stages to

effective land management. With the exception of Thailand, many of these programs are

in progress and successful results with regard to effective land and resource management

have not been well documented.

The research presented here is an effort to reorganize and modernize the land

records management system in Tanzania in a manner that will enable the government to

take control of land and resource management and to ensure economic development

through proper land and resource management.



Research Objectives, Methodology and Scope

In 1991, the government of Tanzania initiated a reform of its existing land policy in

favor of a land market economy. Adoption of a land market economy in Tanzania

required that the cadastral system should not only be aimed at documenting the legal

ownership of the parcel (as is the current situation), but it should contain information that

may be used for faster property conveyancing, fair property valuation, equitable tax

assessment and for monitoring land-related transactions and land use patterns. In

reforming the cadastral recording system to accommodate the changes in objectives, the

existing records and the recording process had to be purged of all the problems that

prevented efficient land allocation and title registration. The improved recording system





5

had to be designed to facilitate the registration process, provide efficient data recording,

storage, and retrieval methods, and where necessary, trace the sequence of land transfers

over a specified period of time.

This research has been conducted using the following basic assumptions :

* Land as a resource, is essential to all mankind and therefore needs to be managed

efficiently.

* The cadastre is a tool for land and resource management (Holstein 1990).

* The cost of compiling the cadastre should bear a direct relationship to the value of

the land and the objectives for government (Dale 1990).

* A complete and up-to-date cadastral information system can serve as a resource

for land management, and socioeconomic development and national development

(Holstein 1990).

* The cadastre is the building block for a multi-purpose land information system

(NRC 1983).

* The need to maintain land records is particularly important in developing countries

(United Nations 1974; Dale and McLaughlin 1989).

The objectives of this research are:

To investigate a suitable approach for reorganizing existing cadastral records by

studying the existing process and identifying problematic areas.

* Identify and recommend actions that will eliminate the problems and pave the way

for a more efficient cadastral data capturing and processing.





6

* Using the reorganized records, develop a cadastral information system which will

be the building block for a broader Multipurpose Land Information System.

* Design an improved land data processing procedures to eliminate bottlenecks and

to speed up the titling and registration processes.

* Develop an approach for cleaning and updating existing records.

* Develop procedures for incorporating new data into the system.



To achieve these objectives, a comprehensive study of the Tanzania land delivery

and cadastral record management system had to be done. The study involved visits to

administrative centers within Tanzania to study the land management procedures. The

study included, among other things, land allocation procedures, cadastral data capture and

processing, title registration and records management procedures, as well as the legislation

that identified the type, quality, standards, and format for the data. The record keeping

and maintenance procedures had to be studied in order to identify problematic areas so as

to develop improvements to the system.

The focus of this research is on organizational and administration issues pertaining

to the reorganization of cadastral records and development of a cadastral information

system for Tanzania. This research does not cover political, legal, policy or institutional

issues pertaining to cadastral information systems, land management, or land information

management in Tanzania, although it is recognized that these play a major role in land and

resource management.





7

Definitions

For this research, "Land" is defined to encompasses all things directly attached to

the surface of the earth, including those areas covered by water (Dale 1989). Land

management is defined as the process by which land and resources in land are put to good

effect. Land management includes resource management, which deals with facilitation of

economic development through inventory, extraction, conservation, and sale of natural

resources. Allocation and management of such resources are effected through instruments,

concepts, measures, and principles which are based on culture, land laws, land tenure and

property rights, registration of those rights, and transactions involving those rights

(Holstein 1990).

Land administration involves the development and use of the land in the manner

which has been prescribed by the instruments, land laws and property rights. The aim of

land administration is to define management procedures, regulations and legal framework

for agencies responsible for land delivery, estate management, revenue generation,

planning and control of land resources. Land administration, therefore, provides the

mechanism for land planning, parcel allocation, enforcing of rights and restrictions on the

use of land, impact assessment and policy reform. These activities are facilitated by the

ability to capture the relevant data to aid in monitoring and identifying areas where actions

and reforms are needed, planning appropriate courses of action, implementing the adopted

choice, and monitoring the results of the implementation for success and further

improvement.





8

In this dissertation, information is defined as the product of data analysis. Within

any organization, management decisions and actions arise from the flow of information

upward, downward, and laterally across the organization. Due to the complex nature of

decision making processes, vis-a-vis the external factors which influence the decision,

quantity and quality of the available information upon which the decision is based, and

possible impact on the organization and the community at large, there is often a need to

establish an information system to support decision making processes. An information

system is a group or pattern of associated activities which according to Anderson (1986),

will normally have the following elements:

* A common purpose.

* An identifiable objective.

* An established sequence of procedures and data flow with at least one but possibly

many elements of input, movement, action, storage and output.

* Feedback of information, giving control over the system.

* A boundary that defines the extent of the system.

* Dependence on specific data.



One such information system is a Land Information System (LIS). The purpose of

an LIS is to provide a decision support for land management. In achieving this objective,

the requisite data about the land, within the confines of the jurisdiction, are captured,

stored, and processed. As an information system, an LIS has an established sequence of

data input, data processing and output channels. For this dissertation, therefore, an LIS is





9

a system containing spatially referenced land data and requisite analytical tools for

querying the data to obtain information to support land and related management decisions.

The system may include human and technical resources which allow retrieval and

dissemination of the information. At the root of land information systems are parcel-level,

though not necessarily parcel-based, data.



A Cadastral Information System is a special type of land information system which

deals specifically with cadastral data. It is comprised of computer hardware, software, and

a database containing cadastral data such as the graphical layout of the parcel, ownership,

size, location, use, and encumbrances that affect the use and enjoyment of the parcel. The

system enables the performance of ad hoc queries on the data and may have graphical

capabilities for the display of the results. Cadastral records play an important role when it

comes to transactions in land and management of properties in the public and private

sectors.

Due to marketability and transferability of land parcels, ownership information

changes very often and hence consideration needs to be given as to how current this

information needs to be to meet user needs. Information has to be relevant, available and

timely, if they are to be of any use for land management. The purpose for which cadastral

information are to be used should control the accuracy and reliability standards for data

capture and management. Although in some circumstances one can be more important

than the other, such relative importance could change over time.







10

Research Organization and Contribution

This research begins with the notion that land policy as an institution, governs all

land management activities. From this standpoint, a land management taxonomy is

developed in Chapter 2. The taxonomy identifies the hierarchical structure of land

management and the role of information as a decision support tool in land management.

Chapter 3 discusses existing cadastral information models and current issues pertaining to

spatial, descriptive information and linkage mechanisms and reviews their focus and

relevance to the Tanzanian objectives. Applicability of existing cadastral models to these

objectives is discussed. This review helps to develop a cadastral information model which

uniquely addresses the Tanzania land management problem. Chapter 4 takes an in-depth

look at the existing land delivery system and cadastral arrangements in Tanzania and

identifies inherent problems. Cadastral data capture and computerization methods are also

reviewed with a view to remove bottlenecks. A discussion of the Tanzanian land delivery

problems is conducted in Chapter 5. Solutions and approaches for eliminating the

problems are discussed at this stage. A Multipurpose Land Information system which

utilizes topologically structured graphical overlays is presented as a model for managing

Tanzania land information. In Chapter 6, a pilot project aimed at implementing the

recommendations and to highlight the benefits of integrated cadastral information system

for Tanzania is discussed. Conclusions and recommendations are presented in Chapter 7.

This research makes an initial contribution to the field of cadastral record

organization, with specific reference to Tanzania, by conducting an extensive study of the

cadastral data capture, land allocation procedures, land registration system, and record





11

keeping and maintenance procedures in Tanzania. Organizational and technical, problems

associated with current practices and procedures are also identified. For the first time, the

legislation identifying the structure and responsibilities of land management agencies

within Tanzania are reconciled to identify areas of concern such as overlapping

responsibilities and inconsistent authorities between land management agencies.

In developing procedures to alleviate the identified problems, a cadastral

information model was conceptualized based on the information support system

components which were developed in Chapter 2, using topologically structured graphical

overlays for individual information support components. This approach which utilizes

base maps in topological space is a deviation from existing models which are based on

base maps in metric space. Some of the existing metric models have been reviewed in

Chapter 3. In addition, the data organization is designed with regard to the type of land

information support system (cadastral, infrastructure, socioeconomic, or environmental),

rather than the agencies that utilize the information, especially since within any

jurisdiction, the data requirements in terms of volume and accuracy, vary among agencies.

Another contribution is the development of a hierarchical parcel identification structure

based on the socio-political divisions in Tanzania and the public's perception of parcel

identification, as a linkage mechanism between the descriptive information support

systems and the graphical overlays.

In considering the records management aspect of this research, a systematic

approach for reorganizing existing cadastral records, and developing a cadastral

information system without drastically altering existing administrative procedures was the







12

objective. An approach for isolating inconsistent and obsolete data among existing

records was developed. Another procedure for recording new information and for

maintaining the integrity of the system was developed.

Finally, an approach for removing administrative bottlenecks within the system and

speeding up the land allocation and registration processes without compromising accuracy

or integrity of the data was developed. As would be observed, some of the contributions

of this research are specific to Tanzania while other contributions, such as the

topologically structured land information model and the parcel identification system may

be applicable elsewhere.










CHAPTER 2
LAND MANAGEMENT AND CADASTRAL INFORMATION SYSTEM


Disregarding minor additions through volcanic activities, earthquakes, and other

natural occurrences, land is finite in size. Land provides the resource base for most

human existence. Humans, plants, and animals have always depended on the land for

sustenance. As of yet, no substitute has been found to match the uniqueness of land both

as a resource base and as the platform upon which terrestrial activities are performed.

People have different concepts or attitudes about land. On one extreme land is

viewed as property which carries specific rights of ownership and use which is transferable

to other people. This concept is dominant in developed countries where land is viewed as

a marketable commodity and, as such, can be used as collateral for credit and economic

development (Marquis 1979).

Another extreme is the view of land as common property. This view implies that

rights of access and use of the land belong to members within a specific group or

community. Access and use by people outside of the group or community is restricted

(Bohannan 1973; Marquis 1979). This concept is mostly held in developing countries

where land is viewed as an interacting natural system whose integrity needs to be

protected, an whose primary qualities have to be preserved for future generations

(Andrews 1979).







14

Population increase, large scale mechanized farming, and urbanization have

imposed unanticipated pressures on the available land in both developed and developing

countries. For example in Tanzania, mechanized farming during "operation vijiji" (see

Appendix B) eliminated large portions of the breeding grounds of the Masai tribe and

introduced other socioeconomic problems besides alienation of their land. With

advancement of development and agriculture, industry and settlement compete for

available land. Usage of parcels of land undergoes changes, sizes of holdings change, and

land values change in response to social and economic factors. Land management

activities such as land use planning and control are therefore critical to the survival of any

nation or community that is undergoing development to ensure that any piece of land is

put to its most economic use. In this chapter, the role of information in decision-making is

reviewed in the context of land management. This chapter establishes the need to

organize land records in a manner which facilitates synergy of information in support of

land management decisions.



Land Management

Land management, in the context of this research, is viewed as an embodiment of

legal principles, administrative procedures, and operations which are associated with the

stewardship of land. The overriding principle for land management activities is the policy

(see Figure 2-1). Land policy is an institution which comprises of social, economic,

cultural, and legal prescriptions that define how the land and benefits from the land are to

be allocated, and the manner in which the land and resources are to be used (Dale and








LAND MANAGEMENT

LAND POLICY

I

I I
I Land Delivery Estate Management
I LEGAL
I FRAMEWORK
Revenue Generation Regulation and Reform


----------------Srvey

Planning and Control /j

Registration
_I- OPERATIONAL
I Adjudication COMPONENTS
I -- --I
I Tfling

Other operations Valuation and

I i I
n i ii n
^------------------------
Infrastructure Cadatral
Info. System Info. System

I -SUPPO RTSYiEMS
Socioeconomic Environmental
Info. System Info. System


Land Management and Land Information Systems


Figure 2-1:







16

McLaughlin 1989, 6). A land policy provides the guiding framework within which

interests in land may be held and the manner in which the land, as a resource, may be

managed.

Barnes (1994a), contends that in the ideal situation, the land policy should reflect

the practices of the people, even though this is not the situation in many developing

countries. Land policy prescriptions are defined through instruments, land laws, rules and

regulations which are executed by land administrators. Although the policy should

establish the legal framework within which land management operations are conducted, it

has been noted by Barnes (1994b) that in many Latin American countries the land laws are

drafted with little or no regard to the indigenous tenure practices of the people. Another

example is Tanzania, where existing land laws are the same as those that were enacted by

the colonial governments, with minor modifications such as replacing the word

"Governor" with "President." Existing land policies of Tanzania clearly do not reflect the

land tenure practices of the Tanzanian people (Shivji 1995, 4-8). The legal framework of

land management is the subject of the next section.



Legal Framework

Through the legal framework of land management (Figure 2-1), rights and rules

which guide the development and use of the land in accordance with the prescriptions of

the land policy are established. The objective of the legal framework is to define

guidelines for managing land resources, protecting individual rights to the use and

enjoyment of the land, consolidating land, protecting the environment, and controlling land





17

development and land degradation, which may result from uncontrolled urban migration,

excessive use of chemicals, pollutants, and other adverse uses of the land.

The legal framework defines the mechanism for land planning, parcel allocation,

enforcing of rights and restrictions on the use of land, impact assessment, and policy

reform. 'Rights,' in this context, should be distinguished from 'rules' governing the use of

the land or its resources. The Random House Dictionary of the English Language (1971)

describes a 'right' as a just claim, whether legal, prescriptive or moral, and defines 'rules'

as principle or regulation governing a conduct, action, or procedure. Rules, therefore

create regulations and thereby, authorizations. A property right is the authority to

undertake particular actions related to specific domain on the land. For every right an

individual holds, rules exist that authorize or require particular actions in exercising that

property right. In addition, all rights have complementary duties. To possess a right

implies that someone else has a commensurate duty to observe this right. Thus, rules

specify rights and duties.

Proper land management cannot be achieved by simply defining the legal

framework to guide the manner in which resources pertaining to land can be used, but also

by providing suitable conditions for individual social and economic development, such as

secure ownership, access to credit facilities, and participation in land markets transactions

(Holstein 1990; Palmer 1996). Legal prescriptions of land management are defined with

relevance to issues pertaining to land delivery, estate management, revenue generation,

and regulation and reform ( Dale and McLaughlin 1989; Nichols 1993). These are

described in the subsection below.









Land Delivery

This involves acquisition of land for public good or public use, government

projects, and new settlements particularly to accommodate the poor and landless, and the

assignment or delegation of interests in land parcels to individuals and organizations. An

ideal land delivery system should enable planned access to land in order to meet basic and

developmental needs of the people (Williamson 1990, 88). Land delivery procedures vary

from country to country. However, national interests and patriotic behavior requires that

those who have access to land, use it productively, and in a manner that will enhance

national development.

In Tanzania, where the land is regarded as belonging to the state, and land

ownership in perpetuity is not the practice, rights to the use and enjoyment must first be

obtained from the state. Allocation is normally for a fixed period with some reversionary

rights and some usufructuary conditions. There may be conditions attached to the use and

enjoyment of the land so as to ensure enhancement and proper use. One such condition

may be that the land must be used for a specific purpose, such as agriculture, residential, or

commercial. Another condition may be that the land should be developed within a

specified period of time to avoid forfeiture.



Estate Management

Estate management involves the management of large land holdings owned by

organizations and communities, such as tribes, clans, and families. In some jurisdictions,

the rules of land management within an estate may differ from those of other lands within







19

the state. For example, in Tanzania, it is required that individual owners of urban land

should have registered titles whereas an individual title is not required for property

ownership within community land. Another example is the situation whereby parcels in an

urban setting may be transferred to anyone, whereas community land may only be

transferred to members of the community. In villages and rural areas, where the land is

owned by the community, a 'Right of Occupancy' title is issued to the communities.



Regulation and Reform

Regulation and reform deals with issues affecting the manner in which land and its

resources are used. In developed countries, regulations in land use are imposed not only to

protect the land from excessive degradation and abuse, but also to protect the public and

the environment. As development progresses and priorities regarding the demand for

available land change, the need arises to review the uses for particular pieces of land and to

reform policies and conditions affecting certain uses of the land. Regulation and reform

activities are designed to monitor the land and to recognize changing objectives so as to

reform the land use plans, and to control excessiveness accordingly.



Revenue Generation

The revenue generation function is a way for the government to generate revenue

for land and infrastructure development. In developing countries, this is done in the form of

fees for processing transactions in land, such as sales, transfers, and registrations. In

situations where the land has been leased for a fixed term, a rental fee may also be imposed.







20

In some developed countries, such as the United States where the full bundle of rights is

conveyed during a sale or transfer, one of the methods for generating revenue from land is

to impose a tax on structural improvements which have been made on the land.

The legal framework of land management is implemented through operational

agencies which are responsible for adjudication, planning and control, land survey,

registration and titling, and valuation and assessment (see Figure 2-1). These are discussed

in the next section.



Operational Agencies

Operational agencies constitute the organizational arrangements which are made to

administer the prescriptions within the legal framework to support land management

activities. Operational agencies consist of divisions in management that are staffed by

personnel whose responsibilities include the provision of services in accordance with the

legal interpretation of the guiding principles of the land policy. The operational component

of land management is multi-disciplinary. Effective land management involves the

interaction and coordination of several government agencies and several operational units.

There is no definite pattern as to which agencies should be responsible for any

particular operation. However, for this research, land management operations have been

defined to include land use planning and control, adjudication, land survey, titling,

registration, and valuation and assessment (see Figure 2-1).

Those with decision-making responsibilities within these agencies are constantly

having to review their plans and actions, and to modify decisions in accordance with







21

changing conditions. Such decision makers need the information that may be derived from

analyzing the database, in order to make sound decisions. Decision-making within the

operational components are either to provide solutions to prevailing or impending

problems, or to capitalize on opportunities.

Problem detection and resolution arises when prevailing conditions indicate a

deviation from the expected results and therefore may impact adversely on expected

objectives. For example, if the data indicate a higher than normal growth in urban

population, the decision may be to allocate more land for residential purposes. This

decision may be an ad hoc one, however, a closer look at the data may reveal some

underlying cause, which may require a more permanent solution. Timely resolution of such

a problem can avoid some of the socioeconomic problems that are associated with

excessive urban migration.

Opportunity-seeking on the other hand, arises when the facts suggest that a

particular action may result in opportunities for the agency or the community. For

example, the strategic decision to construct a dam across a river may be aimed at providing

more agricultural land, jobs and development for the community. Also a decision to

computerize land records based on the knowledge that the information would be useful as

decision support resource for efficient land management operations. Both problem

detection and resolution, and opportunity-seeking strategies are based, not upon instincts

and wisdom alone, but on analysis of available data, together with external factors, such as

socioeconomic and environmental conditions within which the agency operates (McCloy

1995, 342-347).







22

Planning and control, in the context of land management deals with allocation and

monitoring of resources in land with a view to maximizing efficiency by putting land to its

best use while ensuring the welfare of the community and the sustainability of the available

resources.

Adjudication is the determination of rights in parcels of land. The procedure

involves identification of the types of rights in the land, the persons in whom those rights

are vested, and limitations to the enjoyment of those rights. The adjudication process is

used by the operational units to eliminate defects in land titles by judiciously applying the

legal principles that define land ownership (Dale and McLaughlin 1989).

Due to the broad scope of land management activities, the following section is

focused on the cadastral aspect of land management operations, which involves land

survey, land titling and registration, and in the case of fiscal cadastre, valuation and

assessment.





Land Survey

Land survey is a process for providing the geometric framework for mathematically

defining land parcels. As an essential land management tool, survey plans and maps are

used for planning and controlling development, for redefining disputed and uncertain

boundaries, and for defining property, political and geographic boundaries. Different types

of land survey activities are conducted for different purposes. For example, cadastral (land

inventory) surveys are used to establish property boundaries and to determine sizes and







23

shapes of the parcels over which individual rights exist. Cadastral survey of the mutually

accepted boundaries ensures that the boundaries can be replaced if they are destroyed.

Wherever possible, cadastral surveys should be tied to a network of pre-established

control points which were connected to a geodetic reference frame. The global positioning

system (GPS) and advances in modem surveying technologies have made it easier to

connect more survey works to the geodetic reference framework. With advancement of

computerized information management systems, survey plans and maps, in electronic

forms, have become an integral part of a land information system (LIS). These approaches

facilitate the integration of information because it is referenced to some spatial reference

framework.



Land Titling

Land titling is the process of issuing valid property titles by a recognized state

agency to existing occupiers of the land who do not have legitimate titles. The land titling

process confers official recognition of individual rights to the use of any particular parcel.

In a title registration system, such as the Torren's system in Australia, where some

guarantees are offered by the state against inadvertent loss due to an error in the

registration process, the title provides the unimpeachable proof of ownership and therefore

tenure security (Dale and McLaughlin 1989, 26) to the individual. Land title records

constitute a step towards land records compilation for the government. As documentary

evidence of ownership, the title may show types and limitations of any rights that are

exercised over the particular parcel of land. There is a general belief that secure property







24

rights, in the form of registered titles, act as an inducement for investment in real property

and, in the longer term, contribute to increased productivity by the individual (Holstein

1990; Lemel 1985). Feder, et al. (1988) also have presented evidence to support the

notion that increased security of tenure by having title documents increased agricultural

productivity in Thailand by between 11 and 27 percent over comparable non-secure

properties. Land titling should be accompanied by a registration system in order to legalize

the ownership of the property.



Land Registration

Land registration is the process of recording information about legal claims to

parcels of land. Registration is done to ensure clear and unambiguous titles and to avoid

fraud and disputes pertaining to conflicting claims concerning the right of use and

enjoyment of any piece of property. The types of registration include :

* private conveyancing whereby the records of the transaction are handled privately

between the individual parties, sometimes in the presence of witnesses.

* the deed system where the copies of the transaction records are kept in an official

registry of the government or state.

* title registration in which a state organization maintains the records, sometimes with

some guarantees in terms of security.

In the past, most forms of land registration were classified as either deed registration or

title registration. Over the years, adaptations of the two systems have been implemented by

governments according to their objectives and suitability of the system to their local







25

purposes. The main differences between deed and title registration systems have been

documented extensively (Dowson and Sheppard 1968; Simpson 1984; Dale 1990; Larsson

1991). Since the focus of this dissertation is more on the land records and the process for

acquiring such data than on the legal and institutional aspects of land registration systems,

this distinction will be ignored. However, it is important to note that as a land management

process, the system of land registration can influence not only the physical and legal, but

also the social and economic environments (Dale and McLaughlin 1989) of any jurisdiction.



Valuation and Assessment

In order to generate revenue for development, procedures are adopted to assess and

to tax land as a resource. Property rating and assessment are two methods that

governments use for generating revenue from land for development (Dale and McLaughlin

1989, 47). With the property rating system, revenue is generated as a result of a property

assessment. The tax is applied to improvements to buildings, and other structures that have

been erected on the land, and the uses to which the structures are being put. With land

valuation, tax is determined not on the structure that have been erected on the land nor on

the land use, but on the basis of the value of the land itself as determined from the

improved or unimproved state. The revenue generation objective necessitates classification

of the uses of land parcels, and where necessary, the yield. This ensures a fair assessment

of tax liability. Land records for such purposes have been an integral component of the

original cadastree'.







26

Cadastre. The word cadastre has its origins in antiquity when it referred to a

register containing descriptions of land parcels, value, use and proprietorship. The original

purpose for cadastral records was to assess the liability for tax and to determine

responsibility for payment. Over time, the records began to show evidence of land rights

(Simpson, 1984). Dowson and Sheppard (1968, 47), remarked that:

"... it is impossible to give a definition of cadastre which is both terse and
comprehensive, but its distinctive character is readily recognized and may be
expressed as the marriage of:
1. A technical record of the parcellation of the land through any given
territory, usually represented on plans of suitable scale, with
2. Authoritative documentary record, whether of a fiscal or proprietary nature
or of the two combined, usually embodied in appropriate associated
registers...."


Today, the term cadastre is used to imply a parcel-based up-to date record of rights, and

responsibilities in land (FIG 1996).

The cadastral record includes a graphical delineation of the land, to which other

descriptive records pertaining to ownership, types of rights, and sometimes the value of the

parcel and any record of improvements on the land are linked. Such pieces of data are

contained in the land survey, land registration, and the valuation and assessment records, as

shown in Figure 2-2. The use of cadastre has broadened to an extent that in situations

where standards for its creation and maintenance can be ascertained in the judicial system,

the cadastral record may have legal status which is recognized by the courts, not only as

property description, but also for confirming right of use to the registered owner (Larsson

1991). The cadastral map can also serve as an index to other legal records such as

mortgages and liens. Currently, there are three distinct types of cadastral records (Simpson











LEGAL CADASTRE 1


Figure 2-2: Components of a cadastre

1984; Dale and McLaughlin 1989);

1. Fiscal cadastre which refers to a register that has been compiled primarily for

property valuation and tax assessment as a source of revenue generation for the

government. As an information resource, it also serves as an instrument for

administering the policy on land taxation system. It provides the information base

for equitable and efficient tax assessment.

2. Juridical or legal cadastre which contains records of the legally recognized record

of land ownership as a means of avoiding conflicting claims to the use and

enjoyment of the parcel of land. It also provides a means for legally transferring

those rights or interests, either through sale, lease, or mortgage. As an information

resource, legal cadastres provide the information base on land ownership to assist

planners in their efforts to sustain development and curtail abuse of the land.


-- ---------------------------.-----

Land Survey Data

Land Registration Data

Land Titling Data


Valuation and
Assessment Data

FISCAL CADASTRE
mmmmm.mnmm.mm-mmmmmmmmmmmmmmmmmmmmmmmmm.mmmmmmmmmmm.I







28

3. Multipurpose cadastre is a combination of both the fiscal and the juridical as well

as other parcel-related information. It provides a variety of land tenure,

registration, and information services that are required by the community and other

land management agencies (McLaughlin 1975).



While the origins of cadastral record compilation may be tied to tax collection of the olden

days, principles underlying the cadastral register have been adopted around the world due

to the amount of information that can be derived from an analysis of the records.

The operational components of land management involve decision-making as well

as selection of choices among possible options. The availability of information in an

appropriate form can reduce the amount of uncertainty among the options. For efficient

performance, therefore, operational agencies require an information support system in

order to make sound land management decisions. Information support systems are

discussed in the next section.



Information Support Systems

Information is the basic ingredient for sound decisions. In a decision-making

situation, data alone can be overwhelming. Within any organization, better information

leads to a better understanding of a situation, and thereby, the possibility of a better

management decision. The need for a timely and informed decision calls for innovative

ways to not only access accurate and up-to-date information, but the tools to analyze the

available data and present the information in useful and easily comprehensible ways.







29

Activities within the operational units of land management in Figure 2-1 can be

viewed as a series of decision making processes. Information pertaining to the land are

needed to support decisions concerning the operational aspects of land management.

Through decision-making processes, rights, restrictions, and uses of the land are

continually revised so as to sustain development and maintain the maximum utility of the

land. For example, the rights and restrictions applied to peri-urban land in a developing

country may change once the area is integrated into the urban town. Administrative

control is transferred from the community to central or district government. Different rules

and regulations are applied to the transfer, sale, and use of what used to be village or

community land. Similarly, the use of a particular piece of land may change from residential

to commercial in order to keep pace with development and to satisfy the need to provide

amenities for the residents. Land managers need adequate information pertaining to the

land for planning, managing, and controlling its resources.

Many of the developing countries that maintain cadastres still use rudimentary filing

systems. With population increases and an associated upsurge in demand for land and

parcel-related transactions, there is an increased need for more refined information to

support decisions in land management. These situations call for a need to record details of

land parcels in an organized manner. Depending on the number of files, volume of

information, methods for cataloguing, and storage of files, information retrieval can be an

onerous task. In most developing countries, the inadequacies of information pertaining to

the land pose serious constraints on land administration and resource management.

Without knowledge about the land ownership and type of tenure, development programs







30

are difficult to initiate. Land-related information which is embodied in the cadastre, is being

increasingly recognized by governments as valuable resource for decision-making in land

management (Dale 1991).

Governments of many developing countries are reorganizing their cadastral records

so as to derive the benefits of effective land administration. The governments are doing this

by ensuring that rights in land are identified, recognized by the state, and are recorded in a

suitable form. Examples of such activities are found in Peru (Palmer 1996), Bolivia (World

Bank 1995), and Asia (Burns et al. 1996). Land records are also being converted into

digital format so as to harness the benefits of technological advances in data management

and information processing with the use of computers.

Advances in computer technology during the latter part of the 20"' century have

contributed immensely to the growth in the use of information handling technology. This

technology has offered decision makers the tools to adequately analyze voluminous

amounts of available data and the ability to model the effect of decisions, even before they

are implemented. This opportunity provides a means to perform ad hoc queries in order to

arrive at a viable option when it comes to decision making and allocation of resources.

In Figure 2-1, the information support system in land management has been

grouped into four broad categories: environmental, socio-economic, infrastructure, and

cadastral (Dale and McLaughlin 1989, 11). Each group can serve as an independent land

information system.







Environmental Information System

The focus of an environmental information system is to provide information about

factors which influence human health as well as ecological and economic impacts of land

use. The objective of the system is to protect and improve air quality as well as land and

water resources from degradation and abuse.

Human activities are not the only the only factors that can adversely affect the

quantity and quality of the resources. Naturally occurring phenomena such as volcanoes,

earthquakes, tornadoes, hurricanes and forest fires are all factors which impact the

environment and hence human survival. Therefore, an environmental information system is

used to address issues associated with human activities and programs that affect the

environment at local, regional, or global levels. Examples of environmental issues would be

wetland depletion due to rapid and unplanned development, impact of development on

endangered species and degradation of the ecological system, flood analysis, and pollution

control.



Socioeconomic Information System

Socioeconomic information systems include census, demographic, and statistical

data which are essential to government agencies for planning. Several different types of

data are gathered during a population census which are used by governments to monitor

progress of development and to plan for the future. The information derived from the data

helps to redefine priorities with regards to development objectives and to re-allocate

resources with respect to amenities such as schools, housing, infrastructure, and other







32

needs. By themselves, socioeconomic data may not qualify as land information. However,

when demographic data are referenced to a spatial location, the data become land-related

and therefore could be used as an input into a land information system.



Infrastructure Information System

Also referred to as Automated Mapping and Facilities Management (AM/FM), an

infrastructure information system is used for managing engineering and utility structures

such as pipe lines, telecommunications, transportation, and underground facilities. In

general, utility companies have the same requirement as parcel-based record keeping to

maintain records of their transmission and distribution networks, and to make decisions

regarding system capacities to meet public demand based on growth potential. With a

infrastructure information system, it is possible to forecast demand for utilities, plan

extension, locate plants for maintenance, and provide service connections. Land Managers,

in turn, benefit from such an information system by obtaining up-to-date information on the

development of the land so as to strategically plan for the location and magnitude of future

facilities.



Cadastral Information Systems

Information pertaining to land parcels and land ownership and use are always

needed by several agencies for land management purposes. Reliable land records on their

own, do not provide solutions to land management problems. However, they do provide

the resource through which solutions can be devised and implemented. Whether the







33

cadastral records are kept in a manual filing system or in a computerized format, a cadastral

information system provides the government and other interested parties with a complete

and up-to-date inventory of land holdings and land use patterns for a particular jurisdiction.

A cadastral information system, as illustrated in Figure 2-3, is a system comprised of

computer hardware, software, database, and human resource, which operates on cadastral

data.

The cadastral data may vary in accordance with the goals of the cadastre. Whereas

a legal cadastre will have information about the legal description of individual parcels, the

fiscal cadastre will contain such information as land value, land use, and tax liability. Other

information such as ownership, shape, area, location, and owner's address may be common

to both fiscal and legal cadastres. The system may be linked to other records such as data

at the land registry or utility records. The records are compiled from affidavits, signed or

notarized documents, maps, documented evidence, boundary identifiers, adjudication

records, files, and other legal documents that contain relevant information about the parcel.

In a dynamic land market, cadastral information is the most clearly documented type of

land information because ownership of properties changes every so often. Proper records

maintenance and updating procedures must be adopted in order to keep the data current.

A computerized cadastral information system allows analysis and synergy of information in

support of land management activities.

Grouped together, information support systems as shown in Figure 2-1, are referred

to as a Multipurpose Land Information System (MPLIS). Multipurpose land information

systems are recognized as a valuable resource for effective decision making in land














A





I
I

I
I

I


Figure 2-3 Components of a Cadastral Information System




management (Dale 1991). In most developing countries the inadequacies of information

pertaining to the land pose serious constraints on land administration and resource

management. Since 1985, emphasis in many countries has been placed on ensuring that

rights in land are identified, recognized by the state, and are recorded in a suitable form.

Examples of such activities are found in Peru (McLaughlin and De Soto 1994), Bolivia


Hardware
Software
Database
Human resource


Land Registration
Data

Ownership
Evidence
Historical data
Uses
Limitations
Liens
Other encumbrances


Cadastral Data

Legal Cadastre
Legal description

SShape of parcel
I Area
Monuments
Spatial Location
I Administrative location
Cadastral plans
Ownership
Owners's Address
Terms
Value
Use
Tax liability
I Fiscal cadastre







35

(Barnes 1994a; Barnes 1994b), Asia (Burns et al, 1996; Feder and Nishio 1996), and the

Republic of Belarus (Bloch 1996). Such activities involve graphical delineation of the

property boundaries followed by an association of the relevant descriptive information.

Several approaches have been adopted to produce a representation of the property

boundaries. The approaches range from rudimentary land survey methods, such as the use

of the surveyor's compass and linen tape, to photogrammetric methods with variations

dictated by circumstances such as cost, time and technology. The next chapter deals with

current issues relating to graphical data capture and attribute data compilation.










CHAPTER 3
CADASTRAL INFORMATION AND RELATED ISSUES


Advances in computer and information technology have revolutionized the way

management decisions are made. Information systems are being utilized as decision

support resources to minimize the uncertainty in the choices that are made by management.

With cadastral information, technology is again impacting the method for capturing the

graphical data, the structure of the descriptive data, and organization of the cadastral

records so as to meet the information requirements of the jurisdiction. Current research

has focused largely on three distinct areas of the cadastral information system. These are

the structure of the cadastral model, the methods used to capture the graphical data, and

the organization and association of the descriptive data. This chapter deals with a literature

review and identifies current thinking in these areas, even though the state of technology in

Tanzania does not lend itself very well to the application of the latest technology.

Historically, land records have been compiled for public use, by the State or by the

private sector. Simpson (1984, 124) distinguishes between land records which have been

compiled for the benefit of the State and those that are compiled by private entrepreneurs.

Whereas information required by the private sector are those which facilitate dealings in

land such as conveyancing and mortgage, those that are required by the state are related to

issues such as taxation, economic planning and land management.







37

A cadastral information system, as has been shown in Figure 2-1, is a type of land

information systems which is central to the land management process in any jurisdiction.

Development of cadastral information systems in the developed world have been influenced

by factors such as technological advances in the latter part of the 20' century, the need for

improved methods of managing land and related resources, and efforts to protect the

environment. In recent years, the type and detail of cadastral information that is needed to

support particular societal and administrative needs have been changing with respect to the

changing needs of the society. In the 1980s, researchers focused on cadastral models to

identify the complex interactions between the cadastral information and institutional,

political, and economic development of governments. Later, Williamson (1990, 81)

observed that cadastral models and related studies which evolved in the 1980s clarified

concepts, identified essential elements and broadened the use of cadastral information. A

description of the different types of cadastres is presented in the next section.



Different Types of Cadastres

Today, many types of cadastral systems are in operation with varying degrees of

resemblance to the classical fiscal and legal cadastres. Distinguishing characteristics of

cadastral information systems include factors such as the spatial resolution and scale of the

source map, the type and characteristics of the information that are recorded, and the

professional responsibility for managing the data. The F6deration Internationale des

G6ometres (FIG) (1996, 3) has identified the following means of categorizing cadastres:

* their primary function. (e.g. tax, juridical or multi-purpose land management).







38

* the type of rights that are recorded. (e.g. sub-surface rights, mineral leases, or

private ownership, timber concessions, etc.).

* the level of state responsibility in ensuring the accuracy and reliability of the data.

(For example, some cadastres may have complete state oversight and responsibility

and perhaps state guarantee of tenure security whiles other systems may be

implemented with varying levels of public and private sector responsibilities for data

and information management).

* location and jurisdiction. (Distinguishes between urban and rural cadastres as well

as centralized and decentralized cadastres).

* the way in which information about the parcel are collected. (Methods for

capturing cadastral information include digitizing from existing maps and plans,

aerial photogrammetry, ground surveys tied to a geodetic reference framework,

uncoordinated ground surveys and measurements, cadastral survey using Global

Positioning (GPS) methods, geo-referenced aerial photography, etc. Each method

contains a certain degree of positional accuracy which influences the spatial

resolution of the system).

Cadastral information systems are regarded as vehicles for economic growth and social

equity (see Palmer 1996; Holstein 1990; Holstein 1996). In consonance with the modem

concept of information synergism, funding agencies such as the World Bank, the United

Nations(UN), International Development Bank (IDB), and the United States Agency for

International Development (USAID) are working with governments of developing

countries to reform their cadastral systems and to modernize their land records. Cadastral







39

information models have been developed based on needs and priorities. Below are reviews

of some of the existing cadastral information models.



Existing Cadastral Information Models

Perhaps the earliest attempt to establish some connection between the information

provided by the central components of the cadastre and information required by the

government for socioeconomic development and land management was presented by Dale

(1976) in the United Kingdom and McLaughlin (1975) in Canada. Dale presented a

cadastral model that was comprised of central process elements which were influenced by

external factors. The external factors consisted of a broad spectrum of factors ranging

from government, education and professionalism, to legal issues and socioeconomic

factors. The central processes consisted of adjudication, demarcation, survey specifications,

survey methods and boundary description. The central processes were linked by output

elements and some feedback mechanisms. The output elements being the cadastral map,

title records, valuation and taxation, and planning and control elements. The feedback

elements for the model were the title legislation, land values, planning and control, and

boundary disputes.

An interesting point about Dale's model was the'fact that the boundary descriptions

served as the linkage mechanisms to the output elements. The key elements of the entire

model were the land survey and the boundary definition. It is not clear at this point

whether the model depicted a bias towards Dale's profession as a surveyor, but it is

obvious that at the time of its inception, the dominant factors regarding the cadastral







40

process were the boundary definition rather than land use or the resources within the

boundary. Other than simplified keywords, the use of long and verbose boundary

description as a linkage mechanism in a computerized environment would have presented

some coding problems, both in the amount of space required to store the code and the

processing time when it came to searching for an item. More importantly, since people use

different words and style in describing the same parcel, the descriptions of a boundary

could differs from one database to another. Dale's model did not envisage multipurpose

land information systems. McLaughlin (1975) presented the role of the cadastre within a

multipurpose land information management system and identified cadastral information as a

land management tool and a decision support resource. This became the cornerstone for

the cadastral models that are currently in existence.



The North American Model (NRC model)

McLaughlin's presentation was followed by a multipurpose cadastre model which

was developed by the National Research Council (NRC) in 1983 as a basis for a

multipurpose land information system for North America (see Figure 3-1). The model is

conceptualized as an integrated land information system for both administrative and public

uses. It was developed in response to growing concerns about how foreign land ownership

affects the balance of trade in the United States, land prices, access to farm land by young

farmers, intensity of land use, and community viability (NRC 1980, 12-13). The primary

objective was for land administrators to provide public administrators such as governors,

mayors, as well as Congress with information pertaining to land holdings, distribution and


































Figure 3-1: NRC Model for a Multipurpose Cadastre (Highlighted)
Source: NRC (1983)


use within the North American continent. The NRC model identifies the components of a

multipurpose cadastre. A distinction is made between natural land information system and

cultural land information system. The model combines the cultural with the natural land

information system to form the multipurpose land information system. Whereas the cultural

LIS is based on the cadastral parcel boundaries, the natural LIS is based on other natural

boundaries. The two systems are spatially connected by a unifying geodetic reference

framework and the base map with data exchange conventions between the two data types.

The NRC model advocates the use of base maps containing natural and cultural features


CULTURAL LIS

. .NATURAL LIS


Resource Environmental
Reoords Data

NATURAL AREA
I IDNTIFIER PARCEL NUMBERS

VARIOUS DATA-EX CHANGE
NATURAL T
BOUNDARY CONVEIONS
OVERLAYS







42

tied to each other in accordance with the level of accuracy with which the feature was

surveyed. The model proposes a reliance on the mapping standards that have been

established by the American Society for Photogrammetry and Remote Sensing and the

National Mapping Accuracy Standards which were established by the Office of

Management and Budget.



Williamson's Model

Williamson's model which was also developed with regard to the changing needs of

society, was designed in 1986 for Australia. The model is a centralized cadastral

information system which provides legal cadastral information support for local

government agencies. On the basis that a registration system is integral to the

implementation of the cadastre, Williamson attaches the same level of importance to land

registration as the cadastral overlay. The model attempts to show the importance of the

cadastral map and land registration in the design of a multipurpose land information system,

especially in Australia, where tenure security is guaranteed by the State. Here also the

topographic base map emphasizes the need for the base map on a geodetic reference

framework to ensure spatial consistency. The Williamson model obviously, does not

consider natural resource and environmental records as integral components of a

multipurpose land information system. Besides, the structure is identical to the cultural

LIS component of the NRC model except for the fact that Williamson's model proposes a

land information center as a link between the cadastral database and databases of

independent government authorities are established (Williamson, 1986).






































Figure 3-2: Williamson's Multipurpose Land information model
Source: Williamson (1986).




The Developing Country Model

This was developed by Williamson and Jeyanandan in 1990 with particular

reference to developing countries (see Figure 3-3). The model recognizes interaction

between people, land, social groups and cadastre. In this model, block parcels which seem

to be the underlying graphical layer are not tied to any geodetic reference framework.

There is no indication as to how the cadastral map will be registered with other maps


EXISTING
INDEPENDENT
GOVERNMENT
AUTHORITIES


LAND
INFORMATION
CENTER


CADASTRAL
DATABASE
(JURIDICAL
CADASTRE)




TOPOGRAPHIHC
MAPPING AND
GEODETIC
SURVEY
ORGANIZATION


































Figure 3-3: Cadastral Model for Developing Countries
Source: Jeyanandan and Williamson (1990)


produced by other agencies. Also this model focuses on cadastre and does not consider

natural features as a component of the land information that will be required by say, the

planning and development agency. The cadastral blocks are designed around the social

groups. Block boundaries, which may be individual communities, form the basic unit of the

cadastral map, and are relatively permanent and identifiable on the ground. Proprietary

land parcels within the blocks are also recognized. According to Jeyanandan and


z







CADASTRAL MAP

BLOCK PARCELS PROPRIETARY PARCELS







45

Williamson (1990, 91), the special features of the model are;

* Boundaries of blocks are relatively permanent, identifiable on the ground and

therefore recognized even without maps.

* Enables preferences for cadastral products through comparison between blocks and

selective education of people, who hold rights within a block.

* Enables evolution of cadastral system and facilitates selective intervention in land

issues.

* provides the basis for reorganizing land and other data on a geographical (block)

basis.

* Flexibility in, the size of the blocks, use of technology, and cadastral practices.

* Involves very little additional resources but provides for orderly improvement of

cadastral system in keeping with user demands in specific spatial areas.



The Wisconsin Land Information Model

The Wisconsin Land Information Program was initiated in 1985, in response to

demand for information about the land by both the private and public sectors of the

community. A multipurpose land information system for the state of Wisconsin was

established in recognition of the fact that:

a wide variety of land records exist in the form of record books, paper files,
maps, charts, and many other formats.
different land records might be collected at varying levels of detail and
accuracy or might be mapped at different scales (Merideth et al. 1990,
79)."
The system was developed to provide among other things, a standard foundation for







46

accurate geographic referencing of land information. The requirement was an accurate

large scale maps which show small areas in detail. To the committee, details were essential

for decision-making because of their ability to allow comparison of the areas and attribute

of various locations. To a large extent, the Wisconsin model followed the North American

MPLIS(1983) model except that instead of a base map, layers of homogeneous accurately

surveyed graphical layers such as parcels, zoning, flood plains, soils, and registered

together with a geodetic reference framework (NAD83). This approach provided the

option to form composite overlays by integrating layers as needed. For accurate graphical

overlays, Digital Line Graphs (DLGs) of 1:24000 scale and 7.5 minute topographic maps

were initially used with the understanding that more accurate, precise, or detailed

information would be incorporated into the system whenever they become available

(Wisconsin Land Records Committee 1987, 23).

The unique thing about the Wisconsin program is the separation of the layers and

the ability to form composite overlays. However, the need for accurate graphical layers

implies that, for a country like Tanzania, where data have been captured with different

levels of detail and accuracy would necessitate a re-survey of the jurisdiction in order to

obtain a map whose accuracy is uniform across the entire jurisdiction.



Other researchers (Fourie 1993; Davies and Fourie 1996) hold the view that the

model should be dictated by the social structure. Modem cadastral information systems

consist of three parts; a graphical database which contains information that depicts the

subject land, a descriptive part which may be one or more databases containing the list of







47

proprietors and other descriptive information that are relevant to different operational

agencies, and a linkage mechanism which in most cases is the parcel identifier that links the

graphical and descriptive databases together. Irrespective of the cadastral model,

institutional arrangements and management procedures, organization of these components

affect the usefulness and applicability of the cadastral information.



Spatial Data

"The base map is a graphical representation, at a specified scale, of selected

fundamental map information, used as a framework upon which additional data of a

specialized nature may be compiled (NRC 1983, 37)". Within the context of a cadastral

information system, the base map

"... provides a primary medium by which the locations of cadastral parcels can be
related to the geodetic framework; to major natural features such as bodies of
water, roads, buildings, and fences; and to municipal and political boundaries...
(NRC 1983, 39)."

Procedures for capturing and presenting graphical cadastral data as well as for producing

base maps have undergone technological changes. The biggest changes are occurring in

the application of aerial photography for base mapping purposes. Over the years, base

maps have been produced from rectified and unrectified aerial photographs, and digital

maps, in either vector or raster formats. Advances in computer technology are opening

new ways for processing aerial photographs for cadastral purposes, such as soficopy

photogrammetric methods. In the next section, issues related to current spatial data

capture methods are discussed.









Spatial Data Capture

In the past, cadastral boundary data have been captured by traditional ground

survey methods. Equipment for such types of surveys have ranged from the plane table,

tapes and compasses, and transit theodolites to electronic distance measuring (EDM)

devices. Whereas these instruments have not been eliminated completely, Total Stations

and Global Positioning Systems (GPS) technology have become equipment of choice.

Cadastral boundary data are depicted as vector representations of the spatial features.

Aerial photography has gained popularity due to cost and time savings when surveys of

large areas are involved, and also, the fact that photographs enhance the communication of

spatial information. People naturally relate better to information depicted on photographs

than on conventional line and symbol maps. GPS methods have also been used to extend

survey controls so that aerial photography or other ground survey methods can be used to

capture the data. Aerial and other photographic forms of data capture are often converted

into raster representations.

Although spatial accuracy is not the focus of this dissertation, it is worth

mentioning that the ground resolution of pixels in the raster image influences the spatial

accuracy of the features. The use of GPS for cadastral surveying is still in its infancy (see

Barnes and Eckl, 1996), although the results are promising.







49

Parcel Identifiers

In any computerized records system, special techniques are required to define and

uniquely identify the land objects or entities about which data are to be recorded in order to

associate the graphical database with the attribute information. In 1972 the committee on

Compatible Land Identifiers: Problems, Prospects and Payoffs (CLIPPP), noted that parcel

identifiers have to be unique, simple, permanent in utility, flexible, economical and

accessible (Fisher and Moyer, 1973). Recommendations from the CLIPPP committee

included the use of coordinate of the centroid of parcels, the maximum and minimum

values of the Easting and Northing coordinates of the parcel, and the block system,

whereby the blocks are given sequential numbers. The National Research Council of

Canada (1976) identified three methods for unique parcel identification; the hierarchical

system, grid-based identifiers, and hybrid identifiers.

Hierarchical identifiers represent parcel entity identification structure based on

stratified political or administrative units such as Federal, State, County, Town, Ward,

Block, and lots. Other forms of hierarchical structure are the Public Land Survey System

which is used in most of the United States, as well as the Census Tract and Block system.

Grid or graticule system involves the identification of parcel entity based on

geographical or Cartesian coordinate system in a spheroidal or ellipsoidal system. The

centroid or the maximum and minimum coordinates of the boundaries are used as the grid

identifier. The disadvantage is that, as the coordinates are adjusted occasionally due to

improved measurement technology and better mathematical model for the shape of the







50

Earth, the identifiers would have to be corrected with every modification of the coordinates.

The min-max approach also assumes that all parcels will be rectangular in shape.

For a hexagonally-shaped parcel however, the min-max coordinates will lie in someone

else's property. The same situation may occur for an L-shaped parcel if the coordinates of

the centroid of the parcel is used as the identifier.

The NRC (1983), considered name-related, alphanumeric and location-based

identifiers. Name-related identifiers associate individual names and the legal entities over

which their interests exist. The grantee-grantor index is an example of such association.

Ignoring the fact that duplicate names are common in many jurisdictions, name-related

identifiers require that identifiers be changed whenever an interest in the parcel gets

transferred.

Whereas there is no dominant choice for a unique parcel identifier, computer

technology and database management systems facilitate the use of multiple indexes for a

multi-purpose land information system. As noted by NRC (1983, 64), the ultimate choice

for a parcel identifier should be dictated by local needs and resources such as the need for

accessibility and effective management of the identifiers. In this regard, uniqueness,

simplicity and economy of maintenance are more important. Since the CLIPPP conference,

different jurisdictions such as Dade, Palm Beach, and Broward counties in Florida, have

developed suitable, yet independent, parcel identifiers internally.







51

Data Management

One of the responsibilities for Land Information Management personnel, is to

integrate the graphical record with descriptive information from other databases. The

technology for maintaining and manipulating database has been undergoing evolutionary

changes since the early 1960s. In this section, databases that are currently operational as

well as those that are in developmental stages are presented. Their applicability to land

information management and analysis are discussed in order to justify a suitable choice of

database technology for the Tanzania project.

The evolution of modem data organization methodologies began in the early 1960s

with the development of System Design Life cycles (SDLC) (Lee, 1997). Such systems

provided some control over the organization of the data but limited assistance with regard

to analytical operations (Rhine, 1995). Data were stored in flat files, hierarchical files,

network files, etc. Structured methodologies which evolved in the 1970s provided more

effective analytical tools and extended design methodologies. These were achieved by

structuring the data into elemental forms and focusing on the modeling of entities and data.

Codd's (1970; 1979) relational model, along with the structured query language (SQL)

gained support because it provided LIS users with query capabilities which were not

available before. The logical structuring of the data determined the degree of flexibility

within the system. However, the inherent deficiencies in the relational database model and

entity relationship concept influenced the object-oriented methodologies of the 1990s

(Yourdon 1994).







52

Object-oriented technology is regarded as the cutting-edge approach for data

modeling, analysis and software design (Lee 1997). The object-oriented principle is based

on the assumption that people generally think in terms of objects rather than entities or

functions. Objects are direct representations of real things that people perceive when

communicating or describing characteristics of entities or things (Yourdon 1994; Usery

1996). The object-oriented principle incorporate the inheritance and encapsulation

characteristics of data into an integrated whole (Yourdon 1994). Despite the power of

object-oriented technology, very few software packages have been developed to harness

the capabilities of the technology. Below is a description of the data file formats that are

currently in use for LIS operations and their advantages.



Flat Files

Flat files are traditional "spread-sheet type" systems without full database

management support. Flat files contain tabulated data in rows and columns. The rows

contain the records and the columns represent the fields or items. Ordering of the rows

within the table has importance for the ways in which data can be accessed. Data retrieval

is done by means of search keys which index the occurrence of values for a specific field.

Any item (column) within the table can be used as the search key. Searches within the

table may be done sequentially, by binary search, or by index search. The efficiency of

sequential search depends on the location of the record which is being searched. If the

search key is changed to another item within the table, the table has to be sorted again with

the new key.







53

A faster, yet resource intensive, method is the Indexed search which relies on a

separate index table to search for records. A separate table containing the key of every

record as well as an address pointing to the data location of each key is associated with the

data. The index is sequenced and the search is done on the index rather than the data itself.

Because the index table is smaller than the table of data itself, searching of records is faster.

Even though early LIS implementation used Flat files to associate the graphical

overlays, Flat files are severely limited in their utility to LIS applications due to their limited

flexibility. They are simple and efficient for specific repetitive tasks such as transaction-

based information (e.g. in the retail industry and banking activities).



Hierarchical Files

Hierarchical database system works like a "family tree" relationship. With

hierarchical files there is always more than one record in the file. One record is the

"parent" or master, and it can be associated with any number of "children" or detail records

through internally assigned pointers. The detail records can also have children assigned to

them. This establishes a one-to-many relationships among the files. The advantage of this

system is that it allows multiple sets of identical attributes to be associated with any given

record without storing those repetitive data in separate files. Linking of files is done with

pointers which provides some flexibility in relating data between records. One major

limitation with the hierarchical file system is the fact that data in the detail record can only

be accessed by first accessing the master record.









Networks

An improvement to hierarchical files are the Networked databases which allow

detail records to be accessed with more than one master record. This establishes a many-

to-many relationship between the files. The advantage is that if any record needs to be

updated, it can be done on only one file. The drawbacks to the networks are as follows:

* Logical linkages among files multiply as new databases are added to the system.

* In complex networks involving large databases, the amount of storage required for

the pointers can be larger than the database itself.

* Management of the pointers, as records are added, new fields created, and linkages

created as data values change can become cumbersome and in danger of being

inflexible.



Relational Databases

The relational database concept was developed by Codd in 1970 (Date 1991, xi;

Healey 1991, 257). The concept is based on the mathematical theory of relational algebra.

Relational databases allow related records from different tables to be associated without the

use of pointers. Relationships are established through common items within the structure

of the tables. Values in a column or columns in one table are matched to corresponding

values in the column or set of columns in another table. From the second table another set

of matching tables will be associated. The linking continues until all the databases have

been joined. In order to prevent data redundancy in the relational system due to the







55

commonality of items in the relations, relational designs follow Codd's theory of normal

forms (Healey 1991, 258-259) which specifies that:

1. All tables must contain rows and columns and atomicity (i.e. no repeating groups of

data) should be enforced among values within columns.

2. Every column which is not a part of the primary key must be fully dependent on the

primary key.

3. Every primary key must be non-transitively dependent on the primary key.



Normalization. The join mechanism matches column values between tables using

the common item. Normalization of the relational system is based on the principle that a

set, as mathematically defined, cannot have duplicate values. Since a table is a set, it

cannot have rows whose entire contents are duplicated. In addition, each row must be

different from any other. It follows that the values in a single column or a combination of

values in multiple columns can be used to define a primary key for the table. No column

that is part of a primary key can have null values since this could have the potential for

permitting duplicate values. Healey (1991) lists the advantages of the relational databases

as follows:

1. Rigorous design based on sound theoretical foundation.
2. All other forms of database structures can be reduced to a set of relational
tables, so they are most general form of data representation.
3. Almost unlimited flexibility in forming relationships among data items
without the limitation of linkage management.
4. Ease of use and implementation compared to other types of systems.
5. Modifiability which allows new tables and new rows of data within the
tables to be added without difficulty.







56

6. Flexibility in ad hoc data retrieval because of the relational join mechanism
and powerful SQL facilities (Healey 1991, 259)."

Due to the volume of data that may be associated with a typical Land Information System,

Database Management Systems (DBMS) have become integral to LIS. Data modeling

techniques, such as entity-relationship model, have become the key element in designing

spatial databases. The Relational model has dominated LIS database applications due to the

design of commercial software to harness the advantages of the relational construct over

older models such as inverted lists, hierarchical files and networked files. The relational

database model with the Structured Query Language (SQL) was chosen system for

Tanzania. It is recognized at this stage that as the object-oriented technology gains

popularity among land information managers, the land administrators might change to the

object-oriented system.

In the next chapter, attention will be focused on the administrative arrangements

which support land management activities in Tanzania. A study of these arrangements as

well as data processing procedures will be analyzed in order to develop approaches for

modernizing the cadastral records and implementing land information system for Tanzania.










CHAPTER 4
EXISTING CADASTRAL ARRANGEMENTS IN TANZANIA


Following the discussion on land management taxonomy in Chapter 2, where

typical land administration agencies were identified together with the necessary information

support systems, a review of current issues in the data capture and management activities

were discussed in Chapter 3. This chapter deals with the land management arrangements in

Tanzania, the responsibilities of the established agencies, and their data capture and data

management procedures. In this chapter, problems with existing land management

arrangements and procedures are identified. A study of the history of Tanzania reveals that

some of the problems associated with the Tanzania land delivery process can be traced to

colonial times. A brief review of the evolutionary process of the current land tenure system

in Tanzania has been given in Appendix B. Approaches are developed in the subsequent

chapters to eliminate or reduce the impact of the problems.

At the State and regional levels, land management administration in Tanzania is

handled by the Ministry of Land, Housing and Urban Development (MLHUD). At the

local level, the responsibility falls on the Ministry of local government. Reorganization of

land records in Tanzania required a study of the land management arrangements that have

been instituted by the government. This study was accomplished through field visits to

thirteen of the twenty regional capitals and interviews with land administration officials in







58

both the regional and local government offices. The study focused on the legal instrument

and operational guidelines that defined objectives and responsibilities, organizational

arrangements for meeting those objectives, and data capture, storage and processing

methods, record management practices, and information dissemination, both internally and

with other agencies. The responsibilities of agencies that operate on a national level were

compared with those that operate on a local level, in an effort to identify overlapping or

conflicting responsibilities between agencies. The results of this study provided the basis for

the land record reorganization approach and data processing methods. In order to

understand the land management arrangements in Tanzania, the administrative

arrangements within which land management activities are described.



Administrative Arrangements within Tanzania

Tanzania is divided into 20 administrative regions (see Table 4-1), each region

having a capital city. For administrative purposes, the city of Dar es Salaam, which is the

national capital is also a region by itself. So is Dodoma, the proposed State capital which

will replace Dar es Salaam once the infrastructure development has been completed.



Table 4-1: Administrative Regions in Tanzania

Dar es Salaam Arusha Mtwara Kigoma Iringa
Coast Tanga Lindi Mara Rukwa
Morogoro Dodoma Mwanza Shinyanga Ruvuma
Moshi Singida Kagera Mbeya Tabora







59

Each region has offices for regional planning, surveying and mapping, land and title

registry. However, all major land management decisions concerning all the regions are

made in Dar es Salaam, except Dodoma which by legislation, has autonomy in land

management decisions.

At the national level, the Ministry of Land, Housing, and Urban Development is

responsible for land planning, allocation, title processing and registration, property

valuation and assessment, and legal issues in land within Tanzania. These responsibilities

are handled by the Urban Development, Surveys and Mapping and Land Development,

Valuation and Legal divisions in the Ministry (see Figure 4-1). Sections with specific

responsibilities have been described below.



Organizational Arrangements within MLHUD



The Urban Development Division

The Urban Development Division currently operates under the Tanzanian Town

and Country Planning Act of 1956, which was revised in 1961. The Division is responsible

for defining and planning the use of all public land in Tanzania. Activities of this Division

include planning redevelopment areas, renewal of blighted urban areas, re-designation of

land use, and monitoring of development to ensure compliance with the development

program in accordance with master plans of cities. The division is headed by a Director of

Urban Development. The Director of Urban Development approves all town and village

layout plans prior to implementation. The Urban Development Division has five main















^.-.. ..-------. --._. -- .-.- -. --.- .- -- -
I Urban Design Urban land I
and Research CadastralDevelopment

Master Plans Topographic Rural land
and Geodetic Development I
I I
Sites and Village and Land Registry
SServices Hydrographic

Urban Dev. Map
Control e Production Legal


Village Land Vaation
Valuation
-use SECTIONS :



Figure 4-1: Organizational Chart of MLHUD

sections which are the Urban Design and Research, Urban Development Control, Master

Plan, Village Land Use Plan, and Sites and Services sections.



The Urban Design and Research section. This section is headed by a senior

principal town planner. The responsibilities of this section include planning redevelopment

areas (such as the central areas of Dar es Salaam), renewal of blighted urban areas, and

controlling the layout designs of the city. In addition, staff within this section conduct

research and designate high-, medium-, and low-density development areas. Classification

of development areas is based on socioeconomic studies of the communities within the







61

area. The result of the study becomes the basis for setting the size of individual lots within

the proposed area. High-density areas consist of small plots for low income people. The

parcels are usually 400 square meters in size. Medium-density plots are slightly larger with

sizes ranging between 400 and 800 square meters. Low-density classification is for

residential areas for high-income people. The plot sizes are between 800 and 1600 square

meters. The size of the land allocated for industrial use varies depending on the intended

use.



The Master Plans Section. This is the section where master plans are prepared. It

is the policy of the government of Tanzania, to have five-year development plans for all

rapidly growing urban areas in Tanzania. The development plans are graphical layouts

showing the allocation of land for various uses in accordance with the rate of expansion of

the city. The development plans are referred to as the master plans. The Urban Planning

Division schedules the areas to be planned in accordance with available funds however,

councils within fast growing municipalities can request priority consideration for physical

planning of their community. In such circumstances, the municipal councils provide some

of the cost of preparing the master plans

Currently, there are up-to-date master plans for all major cities except Dar es

Salaam which was last prepared in 1979. The reason for failing to update the master plans

for Dar es Salaam is purely it cost which would require a large proportion of the resources

of the Division. This implies that the master plans of other cities will not be updated for

some time.







62

The Sites and Services Section. The main activities of the sites and services

section are for planning of squatter settlements, ensure that basic infrastructure such as

roads and water are available to the residents of those settlements, and that development

plans are carried out according to the design. Activities performed by this section are multi-

disciplinary and not restricted to planning. This section operates on a project-by-project

basis. The activities of the section are concentrated in Dar es Salaam.



Urban Development Control Section. This section ensures that development

agencies adhere to the existing master plans and to ensure that the local councils operate

within the development program. The section is responsible for resolving all land conflicts,

including those that emanate as a result of planning or allocation. The Urban Development

Control section stipulates development conditions and declares areas as ready for urban

development. This section ensures that the local councils operate within the guidelines of

the development program. The section also supervises the preparation of designs for urban

towns. In the regions, the urban town plans are prepared by the municipal councils, but the

development control section ensures that the designs conform with the set standards. The

section may recommend a change in land use for any particular area. Such changes are

recommended to the director and have to be approved by the minister for lands.



Village Land Use Section. The Village Land Use Section assists regional planning

officers in preparing village land-use plans.









The Survey and Mapping Division

The Surveys and Mapping Division provides survey services to government

agencies, maintains a geodetic survey control network, and prepares and maintains

cadastral and topographic mapping statewide. The Director of Surveys is responsible for

coordinating all public sector mapping activities and for maintaining records of all maps,

plans and surveys completed by government agencies. As shown in Figure 4-1, the

Division has four sections that deal with various aspects of surveying. The sections are

cadastral, topographic and geodetic, village and hydrographic, and mapping. There is a

survey department in each of the twenty regions in Tanzania. Although the regional

surveyors are responsible for the cadastral surveys within their regions, all surveys must be

checked in Dar es Salaam before they are accepted. This is a major bottleneck in the data

processing procedures within the Division. Surveyors are required to submit their field

notes, computation sheets, and a plot of the survey to Dar es Salaam for checking. A six-

man field computation checking team in Dar es Salaam has the responsibility of checking

survey work from the regions. For a timely computation checking process, the regions in

Tanzania have been divided into six zones and a technician is responsible for each zone.



The Cadastral Section. The cadastral section deals with demarcation of plots in

accordance to town planning drawings. These are physical layout plans for a development

area. The cadastral section is responsible for the custody of all original field survey records

and checking of computations of surveys from the regions. The demarcation layout from

the Urban Development division is submitted to the cadastral section on a base map which







64

may or may not be current. In laying out the demarcation, the surveyor has the authority to

change the design if he or she encounters any obstacles, such as existing houses or roads

that conflict with the town planning design. The surveyor lays out the plots as closely as

possible to the town planning design, but resolves conflicting issues on site. He or she then

surveys the comer monuments to obtain final coordinates. Staff at the computation unit

check all control surveys and demarcation surveys for computational and drafting errors

and recommend approval or rejection by the Director of Surveys. Deed plans of individual

plots are prepared from approved survey plans after the lots have been allocated.



The Topographic and Geodetic Section is responsible for all survey projects that

are related with national mapping and the establishment and densification of national

geodetic control network. The section is further divided into four units:

1. The geodetic surveys unit is responsible for planning, monumentation, surveying,

documenting, and maintaining national geodetic control points. The national

geodetic controls are first order, second order, third order, and precise leveling for

national vertical bench marks. The section has not executed any control extensions

for many years. Planimetric control densification is currently conducted by

cadastral surveyors whenever they need to extend control to a project site. Besides

that, no precise leveling has been done in several years.

2. The topographical survey unit is also responsible for all topographic surveys,

national mapping, and for providing data for the production of cadastral base maps

and national maps at various scales.







65

3. The international boundary surveys unit maintains the national boundary

monuments and is also responsible for resolving national boundary disputes.

4. The stores and equipment unit is responsible for logistics. The unit purchases

survey equipment, stores and maintains surveying instruments and camping

equipment for the topographical and geodetic section. The unit is responsible for

keeping stock of stationery, such as field survey forms and other materials needed

by other sections for performing their normal tasks.



Village and Hydrographic Surveys Section. The village mapping section was

established in 1970 in response to the government's desire to institute village governments

that would be the nuclei of national planning and development. The Village and Ujamaa

Act of 1975 required that village boundaries be known. One of the responsibilities of the

village and hydrographic section is to demarcate and establish village boundaries so that

title may be issued to the village committees. The section is headed by a senior surveyor

who is responsible for the planning, coordinating, and monitoring of the implementation of

village mapping projects.

Village mapping activities involve a series of seminars with the residents of the

village that have has been earmarked for mapping. The rationale of the seminars is to

educate the villagers on the intent and purpose of the demarcation survey and for the

surveyors to learn about their customary land tenure system in order to ensure that

residents are not unduly dispossessed of their property. The village mapping team uses

1:50,000 base maps and aerial photographs to map out village boundaries. The team visits







66

the boundary marker (if one exists) or establishes a boundary identifier at the position that

is mutually accepted by the representatives of the adjoining villages. The photo interpreters

identify the boundary marker in the photograph or its location (if the point is not on the

photograph). A number is given to the point and a textual description of its location is

made. In the office, a map is produced by photogrammetric methods and submitted to the

director of surveys for checking and approval.



Map production section. This is the section that is responsible for the

cartographic production of thematic, topographic and special purpose maps and atlases.

Maps of all categories, with the exception of the 1:50,000 scale-map series, are produced

by the staff in this section. The 1:50,000 scale-map series are normally contracted to

overseas contractors such as the Ordinance Survey of The United Kingdom or Kenting

Surveys of Canada.



Land Development Division

The Land Development Division deals with allocation of parcels, preparation and

issuance of titles, valuation and assessment of properties, registration of titles and

encumbrances, and resolution of disputes involving ownership. Responsibilities and

activities within this Division are regulated by the Land Ordinance (Chapter 113) and the

Land Registration Ordinance (Chapter 334) of the laws of Tanzania. These laws declare

all lands to be under the control and subject to the disposition of the president.







67

Some of the powers of the President to administer the land have been passed to the

Minister for Lands and subsequently, to the Commissioner for Land. Government Notice

124 of March 22, 1963 extends the power of disposition to land officers. In the office of

the commissioner for lands, specific land officers have been assigned the responsibility to

allocate land.

The Land Development Division has five sections (see Figure 4-1). The division

deals with all public land administration matters in Tanzania. The division is headed by a

commissioner for lands, whose responsibility it is, to approve all land allocations and to

endorse all certificates of occupancy. The commissioner's office maintains records of all

land transactions that pass through the Land Development Division. There is, therefore, an

open registry where records of all transactions relating to lands in Tanzania are kept. There

are five main sections in the Land Development Division; urban land development, rural

land development, land registry, legal, and valuations sections.



Urban Land Development Section. This section handles all matters relating to

urban land including;

preparation of Certificate of Occupancy in urban areas;

processing of land allocations that are made from the commissioner's office;

resolution of ownership dispute; and

approves transfers, mortgages, and leases that last longer than five years.







68

Rural Land Development Section. The Rural Land Development section has the

same responsibilities as the Urban Land Development Section, but with regard to rural

areas. In addition, this section deals with matters concerning the allocation of land for large

farms, the processing of village titles, rights of occupancies in village areas, trading areas,

and subdistricts.



The Land Registry is headed by the registrar of titles, who is responsible for all the

titles that are issued in the country. For land registration purposes, Tanzania is divided into

six zones. Each zone has a registry that is headed by an assistant registrar. Zonal registries

are located in Dar es Salaam, Moshi, Mbeya, Mwanza, Mtwara, and Dodoma (see Table 4-

2). This grouping is inconsistent with the grouping at the Surveys and Mapping Division.

The zonal registrars have the mandate to register titles within their zones. The



Table 4-2: Land Registration zones in Tanzania
(Zonal headquarters in italics)

Zone 1 Zone 2 Zone 3 Zone 4 Zone 5 Zone 6
Dar es Salaam Moshi Dodoma Mtwara Mwanza Mbeya
Coast Arusha Singida Lindi Kagera Rukwa
Morogoro Tanga Kigoma Ruvuma
Mara
Shinyanga
Tabora


responsibilities of the registrars has been clearly defined in the Land Registration Ordinance

(Chapter 334) of the laws of Tanzania. Three classes of documents can be registered at







69

the land registry: Certificate of Occupancy, land-related transactions, and collateral.

Chapter 334 of the Tanzanian laws lists 33 items that may be registered. Of these, 21 have

so far been submitted for registration. Some of the items submitted for registration include

acquisition of Right of Occupancy, caveat, registration as legal personal representative,

change of ownership, change of name, deed of variation, mortgage, revocation of right of

occupancy, and land transfer. After the commissioner has endorsed the certificate of title,

the document is sent to the zonal registries for the title to be issued. There are two systems

for registering certificates of occupancy--the "old system" and the "new system." Both

systems are described later in this chapter, but it will be mentioned at this point that the

major differences between the two systems are the numbering system and the filing method.



The Legal Section. It is the responsibility of the legal section to review all the

land laws of Tanzania, identify loop holes, and conflicting areas of the law. The legal

section recommends modifications to the attorney general. The activities of the legal

section include the preparation of deeds of variation, deeds of surrender, and deeds of

rectification. In addition, the legal section prepares the background information and

strategy for defense in all land-related legal actions brought against the Ministry and to

coordinate with the attorney general in the defense of the Ministry. One of the legal

actions the section handles includes claims for annulment of revocations of certificates of

occupancy that were made between 1971 and 1992 that were signed by the minister for

lands. The legal standpoint is that the minister did not have the mandate to sign those

revocations. According to the law, only the president can revoke an allocation of land.







70

Other cases include resolution of legal action instituted as a result of double allocation and

failure to follow the correct procedure either for revocation or reallocation.



The Valuation Section. For first time allocations, a land assessment report is

required to establish the fee to be paid by the applicant upon acceptance of the offer from

the allocating committee. It is the responsibility of the valuation section to assesses the land

rent. Actual property valuation is done to establish the value of improvement or develop-

ments that have been carried out on the land when the values are needed for compensation,

mortgage, transfer, or the determination of rental charges on a house. The procedure for

assessing rent for first time allocations id described later in this chapter.



The City Council of Dar Es Salaam

The City Council is a completely independent authority that operates under the

Local Government Act (1982) to develop and manage the resources within the city. The

council operates under the Minister for Local Government. This enables the City Council

to operate independently of the Ministry for Lands, Housing and Urban Development. The

Dar es Salaam City Council has jurisdiction over three districts; Ilala, Temeke, and

Kinondoni. Each district has a district lands officer. The City Council has its own land

surveyors, town planners, valuers, architects, and land officers who operate independent of

the Ministry of Land Housing and Urban Development. However, the town planning

drawings that are prepared for demarcation have to be approved by the director of urban

development before allocation can begin. The City Council maintains and updates its own





71

set of standard sheets (1:2500 scale maps), which the council uses as base maps for the

preparation of town planning drawings.

The Local Government Act (1982) authorizes the City Council to establish the

requisite administrative divisions to enable the council to function efficiently and

effectively. The City Council is guided by the policies of the Ministry of Lands, Housing

and Urban Development. Although the functions of the two ministries are clearly defined,

there seems to be overlapping responsibilities. The effect is that the two ministries

sometimes have different views on the resolution of certain problems that are associated

with land administration.



Existing Land Delivery Process

Land Delivery in Tanzania is done in two stages. The first stage(see Figure 4-2),

which is referred to as survey and demarcation, deals with subdivision of the land, physical

demarcation and monumentation of the parcel boundaries, and survey of the parcels.

briefly, the process begins with a request from the Commissioner for Lands. Layout plans,

depicting how the designated area should be subdivided, are prepared by the Urban

Development Division. The Surveys and Mapping division is responsible for demarcation,

monumentation and survey of the parcels. In places where the land has never been sub-

divided before, adjudication precedes survey and demarcation, to establish ownership and

rights to the use of the property. Upon successful checking and approval process, copies of

the subdivision plan are passed to the relevant offices including an allocation committee





























Figure 4-2: Survey and Demarcation


which uses the map to allocate the individual parcels to applicants. The existing base maps

are updated with the new information.



The second stage involves allocation of the parcels to applicants, titling and

registration of the certificate of title (see Figure 4-3). It is the responsibility of the allocation

committee to assign the plots to successful applicants. The Land Development

Division is responsible for preparing and registering the certificate of occupancy. After

allocation has been done, certificates are prepared and sent to the Commissioner for Lands

who appends his seal to the title and endorses each certificate for authentication. The


jPrepare Cadastral
I n Plans

-I

1_ Update existing
Base Maps





































Figure 4-3: Allocation, Titling and Registration


certificate is finally registered by the Registrar of Titles. Copies of the registered documents

are then given to the respective owners.



Survey and Demarcation

The process begins with the town planner preparing the layout (which is referred to

as "town planning drawing") in accordance with the development phases of the master plan.

This is done using the base map of the area of interest. Normally, the town planners should







74

request the survey division for a current base map. In most cases, the base maps are so old

that it is a major task just to update them.

When the town planning drawing has been completed, the local Development

Committee has to approve and accept the layout. Upon acceptance, the plan is sent to the

director of urban planning for approval. A sepia copy of the approved layout is sent to the

region and the original is kept in Dar es Salaam.

A print of the town planning drawing is submitted to the Survey Division with a

request for demarcation. The task is assigned to a staff surveyor and execution is done in

accordance with survey instructions. The survey instructions are normally issued either by

the director of surveys or his representative. After demarcation, the layout is surveyed. Field

notes and computation results are submitted to the regional surveyor, who checks the work

and sends the completed results to the director of surveys for approval. The director of

surveys checks and approves the survey. Prints of the completed plan are made and

distributed to the relevant offices. The field notes become the property of the government. A

copy of the town planning drawing is sent to the regional lands officer who requested the

survey and demarcation. The print received by the regional lands officer is used by the

Allocating Committee to distribute plots to applicants.



Allocation and Registration

The land officer prepares a list of applicants who qualify to be considered for

allocation. There is a committee that has the responsibility for allocating lands in every

district and every town. The list is submitted to the committee that allocates the plots to the







75

applicants. In Dar es Salaam, there are three allocating agencies. These are the office of the

commissioner for lands, the urban planning committee, and land officers (under Government

Notice 124 of March 22, 1963). As there are always more applicants than available plots,

there is a genuine desire on the part of most members of the allocating committees to

distribute the land fairly. In most regions and districts, allocations are done on a first-in-first-

out basis. It is, however, difficult to overlook the request of a superior or a politician.

Once allocation has been done, the regional land officer requests the valuation office

for an assessment of land rent. There are criteria for assessing the rent for first-time

allocations. The valuation officer calculates the appropriate rent and prepares a report. The

regional valuation officer checks the figures and approves if everything is correct.

The land officer prepares a letter of offer that is sent to successful applicants. The letter

details the fees that need to be paid and the development conditions for that piece of land.

Some of the fees are to be paid at the land office while the other fees have to be paid

at the Inland Revenue office. The applicants are supposed to take the evidence of such

payments to the Land Officer, even though in most regions the Letter of Offer does not

mention it. The land officer prepares the Certificate of Occupancy and requests a cadastral

plan from the Surveys Division only when the advice of payments have been presented to

him. The cadastral plan is attached to the Certificate of Occupancy. The applicant signs the

Certificate of Occupancy before a land officer, a magistrate, or a commissioner of oaths.

The document is finally sent to the commissioner for lands (in Dar es Salaam) for checking

and endorsement.







76

At the office of the commissioner for lands, the certificate and all the relevant

documents are checked. Those that are found to contain errors are sent to the reception

desk pending further communication with the regional land development officer who

submitted the certificate. Certificates that are free from errors are stamped with the

commissioner's seal of approval and sent to the commissioner for signature. The documents

are finally returned to the zonal registry office of the region where the plot is situated, after

the commissioner for lands has approved the certificate and the statistics section has

extracted the necessary information from the documents.

The zonal registrar checks for any conflict of ownership, the correctness of the deed

plan, and anything that is required under Chapter 334 of the Laws of Tanzania. In the

absence of any adverse claims or reasons for objection, the title is registered.



Cadastral Surveying Processes within Divisions



Surveying and Mapping Division

The procedure for executing a cadastral survey in a rural area is slightly different

from the method that is used for urban surveys. Surveys in the rural areas are mainly for

physical demarcation of village boundaries, whereas urban cadastral survey are for plot

demarcation and survey. They both start with survey instructions from the office of the

director of surveys and mapping or any person appointed by him. In the regions, the survey

instruction may be written by the regional surveyor. The survey instructions contain the





77

coordinates of existing survey controls that may be used to connect and control the new

survey as well as a sketch of the area to be surveyed.



Survey of Rural Areas. The survey instructions contain details such as the location

and the shape of the land to be surveyed. Adjudication between neighboring communities is

done at the time of demarcation to ascertain the mutually accepted position of the village

boundaries. Sometimes it is found that the sketch that is attached to the director's

instructions is different from what the village communities perceive to be their boundaries.

In such circumstances, the boundaries as perceived by the communities are adopted.

Until recently, a photogrammetric method was used to demarcate village boundaries.

Using a 1:50,000 scale map and aerial photographs, the boundaries are identified and

marked on the photos. Photogrammetric methods are used to coordinate the boundary

markers and to produce maps. In situations where conventional theodolite traversing

method is used, accuracy requirements for village demarcations are 15 minutes of arc for

angular disclosure and 1:5000 in linear disclosure.



Surveys of urban lands are often for physical demarcation and survey of parcels.

The request for survey is preceded by an approved town planning drawing. The

commissioner for lands or his representative makes a request to the director of surveys and

mapping, who in turn, issues survey instructions. The survey instructions are either sent to

the Regional Survey Offices or to licensed survey firms. Requests for surveys may also be

made in the regions by the regional land development officer to the regional surveyor. The







78

survey instructions constitute an express authority for the land surveyor to enter upon the

land with his or her field assistants to carry out the survey.



Procedure for Carrying out Cadastral Survey. The field surveyor studies the layout

and identifies the location from an index map. Any existing cadastral control points in or

around the neighborhood are identified. Coordinates, reports, descriptions, and any relevant

information pertaining to the controls are extracted from files. All preliminary computations

needed to commence the survey task are done in advance prior to the actual field work.

The field procedure for setting out the parcels involves demarcation of the block

comers first. Measurements are made to ensure that the block comers have been located as

accurately as required in the survey instruction. A traverse is run to coordinate the block

comers. The traverse is computed to ensure that the positions of the points are within

acceptable misclosures. In urban areas, the allowable angular disclosure is 30n' seconds of

arc, where n represents the number of stations. Allowable linear disclosure is 1:6000.

Traverse adjustment is by the Bowditch method.

Setting out individual plots in a high- and medium-density area is done by extending

a steel tape horizontally straight along the line between two block comer points and driving

iron pins into the ground at the points where plot comers should be. Concrete mortar is

poured around the pins to make them more permanent.

In the case of low-density plots, the sides are measured rigorously and corrected for

slope, temperature, and where applicable, sag correction. Plot comers are marked with

concrete monuments and numbered sequentially. Such numbers are also shown on the







79

cadastral plans. In the office, the field notes are checked by re-computing all the data to

make sure that there are no errors. The surveyed plots are finally drafted. The original field

notes, computation sheets, and plan are sent to the director of surveys and mapping for

further checking and subsequent acceptance. A job that meets the required standards is

accepted and approved by the director. Field notes and all accompanying documents

become the property of the government.

In most cases where the survey instructions have been followed and proper survey

procedures applied, the resulting survey has been within acceptable tolerances and the job

has been accepted. Occasionally, the survey work has been rejected. In such situations, the

director of surveys and mapping requests that the work be done again.



Rejection of Cadastral Survey Job. Some of the reasons for rejecting a survey

include:

failure to follow the survey instructions

failure to conform with the layout as shown on the town planning drawing

without adequate reasons;

failure to comply with accuracy requirements, either as specified in the

survey instructions, the Surveyors Regulations, or technical circulars;

survey work extending over or overlapping an existing survey; or

non-maintenance of road parallelism.







80

Relative accuracies of the cadastral control points. Different methodologies have

been applied at different times in different parts of the country during survey control

densification. Four different approaches have been used to establish survey controls;

triangulation method, the use of theodolites and Electronic Distance Measuring (EDM)

instruments, theodolites and invar tapes, and lately, global positioning system (GPS)

methods. Each of the methods has inherent levels of accuracy. Surveyors in both the public

and private sectors expressed the need to readjust the survey control network and obtain

unified coordinate values for all the control points as well as established levels of accuracy

between the different measuring processes.

New settlements usually start from the edge of the road and extended inland, away

from the road. In demarcating plots, the procedure has been to use controls points from the

edge of the road and close on other points along the same road. The controls that were

established during one cadastral survey were used as starting and closing controls for

subsequent surveys. This has been the procedure for extending cadastral controls especially

in the urban areas. By doing that, the errors in one cadastral survey got carried over to

future surveys.

Over the years, different disclosure levels have increased between different town

subdivision blocks along different roads. In the current situation, it is not recommended to

start a survey project from controls derived from a particular road and close on controls that

have been established from a different road, due to the possibility of having unacceptable

disclosures. In Dar es Salaam, misclosures of about two meters have been observed in







81

certain areas. This may translate to a fraction of a millimeter on a 1:2500 base map, but

unacceptable for cadastral survey by the Director of Survey.



Land survey and demarcation within Dar es Salaam. The procedure for

demarcating and allocating land begins with a request for a planning scheme by the city

planner. A request is made for a base map (1:2500 map) of the area to be planned from the

surveys section of the council. The base map is updated by the town planners using steel

tape and compass. The proposed layout is done on the updated base map. This becomes the

town planning drawing. After the town planning drawing has been prepared, the Urban

Planning Committee reviews the design and recommends whatever changes that the

committee might consider necessary.

The design is adopted when both the Urban Planning Committee and the city

planner's office are satisfied. The drawing is then sent to the Director of Urban Development

for approval. At the director's office, the drawing is taken through the normal checking

procedures, including ensuring that the design conforms with the land use scheme as shown

in the master plan of the area. Once the design has been accepted as conforming with the

master plan, the drawing is approved by the director on behalf of the Minister for Lands and

returned to the City Council for implementation.

The land surveyors at the City Council demarcate the land according to the design.

Survey field sheets, computation sheets, and a list of the coordinates are sent to the Director

of surveys and mapping for checking. The office of the director of surveys checks the







82

quality of the reference controls that were used for the survey, the method of survey, the

computations, and the misclosures of the traverses.

If the results are within the required standards of accuracy, then individual plot num-

bers are assigned. Otherwise, the drawing is returned to the city land surveyor's office for a

resurvey. Having assigned unique numbers to all the individual lots, copies of the plan are

distributed to various departments, including the city planner, city land surveyor, and the

city lands officer. The city land surveyor's copy is used to update the 1:2500 standard

sheets that are maintained in the office. Another copy is archived for future reference. The

copy that goes to the city lands officer is used to allocate the plots to applicants.



Land allocation. In Dar es Salaam, allocation of plots is done by the Urban

Planning Committee. The copy of the subdivision plan that is sent to the Commissioner for

lands is meant to serve, among other things, a notice to the fact that the parcels have already

been allocated. Sometimes the notice gets delayed to the extent that the commissioner's

office inadvertently allocates that same plot to other applicants. Genuine mistakes of this

kind are easily noticed when the owners commence registration of the certificate of

occupancy.

Requests for land under Certificate of Occupancy may be submitted to the district

land officer or the city lands officer. Requests received by the district lands officer are

forwarded to the city lands officer. All requests are compiled by the city lands officer and

submitted to the Urban Planning Committee. The Planning Committee meets to allocate the

plots to the applicants. Allocation is supposed to be done in the order in which they are







83

received (i.e., on a first come, first served basis). After the committee has made the

allocations, the city lands officer requests an assessment of each of the newly allocated

parcels.

Upon receipt of the assessment information, the city lands officer prepares letters of

offer and sends copies to each of the applicants whose request was approved. The letter of

offer contains the terms of the offer, the fees that need to be paid, and the time within which

to pay the required amounts. Each letter of offer is prepared in quadruplicate. The original

goes to the applicant and copies sent to the commissioner for lands and the city lands

officer. The last copy is kept in a file at the District Land Office where the letter was

prepared.

The recipient normally has 30 days from the date of the letter to accept the offer and

pay the relevant fees. After accepting the offer, the recipient also has three years to develop

the plot in accordance with the intended use (e.g., by constructing a house on it).



Property Valuation and Rent Assessment

As mentioned earlier, it is the responsibility of the valuation section to assess the

land rent. Assessment is done after an offer has been accepted. The land officer submits a

request to the valuer for land rent assessment for each parcel of land before the letters of

offer are sent to the applicants. As shown in Figure 4-4, the request for valuation is

submitted in the form of a letter to the chief valuer who records the description of the

property, the plot number, location of the property, the purpose of the valuation, and the

date of the application. The chief valuer then passes the file to a valuation officer for action.





































Figure 4-4: Schematic Diagram of the Procedure for Assessing Property


The valuation officer asks the registry section to create a file for the job. A valuation number

is then assigned.

The file is returned to the valuation officer once a file number has been assigned. The

valuation officer then schedules a visit to the site of the property and collects the necessary

information. Subsequently, a report is written in which a value is assigned to the property.

The report is submitted to the chief valuer for checking and approval. Unless there are







85

obvious mistakes requiring a return to the valuation officer for amendments, the chief valuer

approves the valuation report. The fees for valuation are based on the assessed value of the

property and the direct costs of trips to the site. Upon payment of the required fees, the

report is given to the client and a copy is kept at the lands records office.



Certificate of Occupancy

The request for a Certificate of Occupancy may be presented by an individual, the

regional land officer (in the case of requests from the regions) or the city land officer. This

happens after the allocating committee has offered the plot of land to the applicant and the

applicant has accepted the offer.

There is no way of knowing if the offer has been declined by the applicant. As a

result of this flaw, there is no way of keeping track of the allocations that have been

accepted and those that have been declined. The only way to find out about an offer that

was not accepted is through the process of official search or if an interested party keeps

track of the allocation.

As shown in Figure 4-5, the approval process for tha certificate of occupancy begins

at the reception counter. The documents are first submitted to a receptionist at the

reception counter who stamps a date to indicate when the documents were first received.

The document is then sent to the open registry for a land division number. At the open

registry, the documents are put in a folder. The folder is then sent to another office within

the open registry (indexing room) where a unique land division number is issued to the

folder. The plot number, block number, location description, and name of the owner are




































Write to the owner or the
District Land Officer or the
City Land Officer for
clarification
The document is returned tc
reception desk for filing.


ZONAL REGISTRATION OFFICERS

ZONE I ZONE 2 ] ZONE 3 ZONE 4 ZONE 5 ZONE 6


Land Office Procedure for Issuing a Certificate of Occupancy


Figure 4-5:







87

recorded on an index card that is kept in the indexing room. The folder containing the

document is sent to the printing section where the land division number and description are

printed on the cover of the folder.

The folder is returned to the open registry again and given to the respective schedule

officer. The folder is placed among other folders from the same region that are waiting to be

processed by the schedule officer who is responsible for documents from that particular

region. The schedule officer checks for ownership, letter of offer, advise of payment,

evidence of payment of fees, correctness of the certificate, typography, and other relevant

letters and documents are included.

In the case of farms, a check is made whether the respective land allocation

committees from the village level to the regional level have all consented. If all the items

are correct, then the schedule officer sends the folder to the statistics section. Otherwise the

schedule officer writes a letter to the individual or the land officer who submitted the

document and requests more information or explanation as to why the documents were

submitted in that manner. Having written the letter, the schedule officer returns the folder to

the receiving desk to be filed until a reply to the letter is received.

At the statistics section, the folder is checked again for any errors that the schedule

officer may have missed. If a mistake or omission is found, the folder is returned to the

schedule officer. Assuming that everything is correct, the technician at the statistics office

takes the folder to the commissioner's office. The commissioner's seal is put on the

document and the folder is left in the commissioner's office where the commissioner checks

the document for any errors that the schedule officer and the technician from the statistics




Full Text

PAGE 1

AN APPROACH FOR CADASTRAL RECORDS REORGANIZATION AND IMPLEMENTATION OF A TOPOLOGICALLY STRUCTURED CADASTRAL INFORMATION SYSTEM IN TANZANIA By FRANCIS W. DERBY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998

PAGE 2

ACKNOWLEDGMENT There are so many people, to whom I owe my gratitude for their assistance and encouragement through this dissertation, it would be impossible to name them all here. I would especially like to thank my graduate committee, who through their combined knowledge and experience, inspired, motivated, and guided me to the end. My utmost gratitude goes to Associate Professor David W. Gibson for giving me the opportunity to study in this university and for believing in my ability to complete this program successfully. The advice and guidance he gave to me during the initial stage of my doctoral program helped me to develop the concepts for this research. My sincere thanks goes Associate Professor Scot E. Smith, Ph.D., for the financial support at the time when I needed it most. I am grateful for his contribution as a major professor and the helpful ideas he offered during my most trying moments. Sincere gratitude goes to my friend and fellow graduate student, Joe Aufinuth for assisting me in analyzing my results and for preparing my figures. I thank him for the agonizing moments he spent listening to my problems and complaints. I would like to express my appreciation to Associate Professor Charles D. Ghilani, Ph.D. and Assistant Professor Thomas A. Seybert, Ph.D., of The Pennsylvania State University for their helpful ideas and companionship. I will always cherish the help and advice that they offered. 11

PAGE 3

To my wonderful and understanding wife Angela and our childre' Yahan and Pinkr~ I would like to express my appreciation for their love devotio' and support. I am really blessed to have a family like ours. Finally and most of all, I thank God for making it all possible. iii

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS ............................................... ii LIST OF TABLES ......................................... .. ......... vii LIST OF FIGURES .................................................... ix ABSTRACT ......................................................... xii CHAPTERS 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . 1 Research Objectives, Methodology and Scope .......................... 4 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Research Organization and Contribution . . . . . . . . . . . . . . 10 2 LAND MANAGEMENT AND CADASTRAL INFORMATION SYSTEM .. 13 Land Management . . . . . . . . . . . . . . . . . . . . . . . 14 Legal Framework ............................................... 16 Land Delivery . . . . . . . . . . . . . . . . . . . . . . 18 Estate Management . . . . . . . . . . . . . . . . . . . 18 Regulation and Reform . . . . . . . . . . . . . . . . . . 19 Revenue Generation . . . . . . . . . . . . . . . . . . . 19 Operational Agencies . . . . . . . . . . . . . . . . . . . . . . 20 Land Survey . . . . . . . . . . . . . . . . . . . . . . 22 Land Titling . . . . . . . . . . . . . . . . . . . . . . 23 Land Registration ......................................... 24 Valuation and Assessment . . . . . . . . . . . . . . . . . 25 Information Support Systems ...................................... 28 Environmental Information System . . . . . . . . . . . . . . 31 Socioeconomic Information System . . . . . . . . . . . . . 31 Infrastructure Information System ............................ 32 Cadastral Information Systems ............................... 32 3 CADASTRAL INFORMATION AND RELATED ISSUES .............. 36 IV

PAGE 5

Different Types ofCadastres ...................................... 37 Existing Cadastrallnformation Models ............................... 39 The North American Model (NRC model) ...................... 40 Williamson's Model ....................................... 42 The Developing Country Model . . . . . . . . . . . . . . . 4 3 The Wisconsin Land Information Model ........................ 45 Spatial Data ................................................... 47 Spatial Data Capture ...................................... 48 Parcel Identifiers ............................................... 49 Data Management . . . . . . . . . . . . . . . . . . . . . . . 51 Flat Files . . . . . . . . . . . . . . . . . . . . . . . . 52 Hierarchical Files ......................................... 53 Networks ............................................... 54 Relational Databases . . . . . . . . . . . . . . . . . . . 54 4 EXISTING CADASTRAL ARRANGEMENTS IN TANZANIA .......... 57 Administrative Arrangements within Tanzania . . . . . . . . . . . . 5 8 Organizational Arrangements within MLHUD ......................... 59 The Urban Development Division ............................. 59 The Survey and Mapping Division . . . . . . . . . . . . . . 63 Land Development Division . . . . . . . . . . . . . . . . 66 The City Council of Dar Es Salaam ................................. 70 Existing Land Delivery Process . . . . . . . . . . . . . . . . . . 71 Survey and Demarcation .................................... 73 Allocation and Registration . . . . . . . . . . . . . . . . . 7 4 Cadastral Surveying Processes within Divisions . . . . . . . . . . . . 7 6 Surveying and Mapping Division ............................. 76 Property Valuation and Rent Assessment . . . . . . . . . . . 83 Certificate of Occupancy . . . . . . . . . . . . . . . . . 85 Registration of Certificate of Occupancy . . . . . . . . . . . . 88 5 CADASTRAL INFORMATION SYSTEMS APPROACH ............... 94 Discussion of Improvements to Current Cadastral Arrangements ........... 97 Urban Development Division ................................ 99 Surveys and Mapping Division . . . . . . . . . . . . . . . 100 Land Development Division . . . . . . . . . . . . . . . . 103 Document Processing Improvements . . . . . . . . . . . . . . . 108 Reorganization of Existing Records .......................... 110 Processing ofNew Documents .............................. 115 Organization of Cadastral Information . . . . . . . . . . . . . . . 129 Topologically Structured Cadastral Data Concept ..................... 133 Topological Rules and Cadastral Index Mapping ..................... 136 Application of the Rules to cadastral Surveying . . . . . . . . . . . 141 V

PAGE 6

Boundary Definition ...................................... 142 Subdivisions ............................................ 143 Topological Rules and Principles Illustrated .................... 146 6 CADASTRAL INFORMATION MODEL FOR TANZANIA ............. 153 Cadastral Index Map Compilation in Metric Space ....................... 154 Cadastral Index Mapping in Topological Space .................... ... 158 Cadastral Information System for Tanzania ... .. .................... 160 Cadastral Index Map for Tanzania . . . . . . . . . . . . . 160 Linkage Mechanism. . . . . . . . . . . . . . . ........ 166 7 PILOT PROJECT . . . . . . . . . . . . . . . . . . . . . . 172 Data Sources . . . . . . . . . . . . . . . . . . . . . . . . 174 Hardware and Software . . . . . . . . . . . . . . . . . . . . 178 Analogue image conversion . . . . . . . . . . . . . . . . 179 Attribute Data Processing . . . . . . . . . . . . . . . . . . . 183 Spatial Analysis . . . . . . . . . . . . . . . . . . . . . . . 188 Summary and Analysis of Results .................................. 194 8 CONCLUSIONS AND RECOMMENDATIONS ..................... 201 Recommendations ............................................. 204 APPENDICES A-1 TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND OFFICE ......................................... ..... 208 A-2 TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND OFFICE ............................................... 211 B EVOLUTION OF LAND TENURE POLICIES IN TANZANIA .......... 214 C OBSERVATIONS AND CONCERNS WITH EXISTING SYSTEM ...... 233 REFERENCES ..................................................... 243 BIOGRAPHICAL SKETCH ........................................... 254 V1

PAGE 7

LIST OF TABLES 4-1 Administrative Regions in Taozaoia .................................................................... 58 4-2 Land Registration zones in Taozania ................................................................... 68 5-1 Cadastral Survey Processing Tasks ................................................................. 117 5-2 Improved procedure for processing and Issuing Certificate of Occupancy ....... 123 5-3 Document processing Tasks at the Land Registry ............................................. 126 5-4 Point Equivalence Table ................................................................................... 147 5-5 Line Equivalent Table ..................................................................................... 148 5-6 Line-Node Topology ...................................................................................... 148 5-7 Polygon-Line topology ................................................................................... 150 6-1 Coding structure of Parcel Identifiers ................................................................ 169 6-2 Codes for Administrative regions in Tanzania ................................................... 170 7-1 Textual information associated with Individual Subdivision/Cadastral Plans ........................................................................................................ 175 7-2 Relevant Textual Information from Title Office and Land Registry .................... 183 7-3 Erroneous Records Identified during data entry at the Land Office .................... 186 7-4 Erroneous Records Isolated during data entry at the Land Registry ................... 186 7-5 Results ofinternal Consistency Check among the Land Office records ............. 187 7-6 Inconsistencies among Land Office and Land Registry records .......................... 188 Vil

PAGE 8

77 Query Types That Were Done on the Cadastral Information ............................. 189 7-8 Internal Inconsistencies Identified During manual Data Entry ............................ 196 7-9 Results ofinternal Consistency Check on the Data ............................................ 197 7-10 Comparison ofland Office and Land Registry Records ...................................... 198 Vl1l

PAGE 9

LIST OF FIGURES Figure 2-1 Land Management and Land Information Systems ............................................. 15 2-2 Components of a cadastre .................................................................................. 27 2-3 Components of a Cadastral Information System. ................................................ .34 3-1 NRC Model for a Multipurpose Cadastre ........................................................... .41 3-2 Williamson's Multipurpose Land Information model.. ........................................ .43 3-3 Cadastral Model for Developing Countries ........................................................ 44 4-1 Organizational Chart ofMLHUD ........................................................................ 60 4-2 Survey and Demarcation ..................................................................................... 72 4-3 Allocation, Titling and Registration ..................................................................... 73 4-4 Schematic Diagram of the Procedure for assessing Property ................................ 84 4-5 Land Office Procedure for Issuing a Certificate ofOccupancy ............................. 86 4-6 Procedure for Registering a Certificate of Occupancy ......................................... 89 5-1 Computerized Cadastral Data Management System ........................................... 109 5-2 Procedure for Isolating Inconsistencies among Existing Land Records ............. 112 5-3 Improved Cadastral Survey Processing Procedure ............................................. 116 5-4 Revised Approach for Processing new Certificates ofOccupancy ...................... 122 5-5 Improved Procedure for Processing Documents at the Land Registry ................ 125 IX

PAGE 10

5-6 Topologically Structured Multipurpose Land Information Model.. .................... 130 57 Node snapping precedence rule ......................................................................... 13 7 5-8 Two Definitions of a line .................................................................................. 137 5-9 A node-vertex precedence ................................................................................ 138 5-10 Node-line precedence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 9 5-11 Node precedence after a line intersection .......................................................... 140 5-12 Vertex snapping ................................................................................................ 141 5-13 Two Topological representations of a Cadastral Boundary ................................ 142 5-14 Representation of a Boundary with Vertices ...................................................... 143 5-15 Topological errors in Cadastral Index Mapping ................................................. 144 5-16 Three separate cadastral surveys ....................................................................... 146 5-17 Two Topologically structured cadastral index maps of the same area ................ 151 6-1 Coordinate Transformation in Metric Space ...................................................... 156 6-2 Topologically assembled cadastral index map .................................................... 159 6-3 Graphical data conversion ................................................................................. 161 6-4 Creating a Cadastral Index Map for Tanzania .................................................... 163 7-1 Vector Drawing ofRegion Surrounding Pilot Area ........................................... 173 7-2 Topologically Generated Cadastral map of Pilot Area ...................................... 177 7-3 An Overlay of Cadastral map and a Geo-Referenced Aerial Photograph ............ 182 7-4 Land Rent Analysis on Kijitonyama Block 44 .................................................... 191 7-5 Land Rent Analysis on Sinza Block A ............................................................... 193 B-1 Effect of 1896 Land Tenure Amendment .......................................................... 217 X

PAGE 11

B-2 Tenure Structure after 1928 Amendment.. ........ .. ................................... ......... .. 219 B-3 Land Tenure Structure after 1969 Amendment.. ................................................ 223 B-4 The Land Policy in a Dilemma ..................... .... .. ................. ........................... ... 227 XI

PAGE 12

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfilment of the Requirements for the Degree of Doctor of Philosophy AN APPROACH FOR CADASTRAL RECORDS REORGANIZATION AND IMPLEMENTATION OF A TOPOLOGICALLY STRUCTURED CADASTRAL INFORMATION SYSTEM INT ANZANIA By Francis W. Derby December 1998 Chairman: Dr. Scot E. Smith Major Department: Civil Engineering The government of Tanzania is in the process of changing the country's existing land policies in favor of a land market economy. In preparation for the anticipated increase in property conveyancing and land-related transactions, the government needed among other things, to: review procedures for recording land information and to adopt methods for strengthening administrative and cadastral capacity to support land registration functions; review the mission, organizational structure, and staffing allocation of regional offices, identify problems and institutional obstacles that prevent synergy of information handling and record handling among related agencies and the national registry. X1l

PAGE 13

develop a well-designed administrative procedure for land reg is tration as w ell as procedures for developing a cadastral information system suitable for Tanzania. This dissertation presents a study of the organizational arrangements and administrat iv e procedures for parcel allocation, parcel survey and demarcation, and the registration of all particulars affecting the creation of a legal cadastre in Tanzania. Administrative problems and bottlenecks that prevent the smooth flow of activities between the land management agencies were identified Approaches were developed to eliminate the land records organizational problems improve administrative procedures for land allocation and title processing and provide a streamlined method for faster document processing and land record maintenance. A land information system model was developed for Tanzania. The model uses topologically structured graphical overlays, to provide information support for a cadastral, cadastral information system. The graphical data are linked to the descriptive records through a newly developed parcel identification system for Tanzania. Procedures were developed for isolating inconsistencies in the existing records which are in paper form and converting the error-free records into digital format. A procedure for adding new records to the computerized system and for maintaining the records up-to date was developed. A pilot project was initiated which successfully tested the topological approach for producing the cadastral index map to support a cadastral i,nformation system. The pilot study highlighted the ability to perform analysis on the data and to obtain information to support land management decisions, even in the absence of an accurately surveyed map. Xlll

PAGE 14

CHAPTER I INTRODUCTION An essential recipe for proper land management is up-to-date information concerning location, extent, ownership and use of parcels of land. In many developing countries, where a large percentage of the economy is tied to the land, land records are vital for efficient land and natural resource management. When maintained properly, such information can also help to facilitate transactions in land. In more developed countries with dynamic land market activities, information derived from land records offers substantial benefits to individual land owners as well as governments. For the individuals, currency of the information ensures, among other things, faster, safer and less cumbersome procedures for land-related transactions, protection of various rights to the use and enjoyment of the property, and fair taxation on properties. For the government, up-to-date and well maintained land-related information based on systematic recording of rights in land are important in many sectors of government such as physical planning of the land, revenue generation, infrastructure development, and environmental protection. Advances in computer technology and data management procedures have provided additional benefits such as the ability to perform complex statistical analysis on land records, identify trends in land market activities, and assess impacts of developments on I

PAGE 15

2 the society. The same cannot be said about synergy of land-related information in developing countries. In 1974, a sub committee of the United Nations noted that ... systematic records ofland and rights in land have great importance for public administration, land planning, and land development, and private transactions in land. This situation is particularly true in those developing countries where the rapid growth of population has caused increasing pressure on rural land while simultaneously a massive migration of people to the cities and towns has led to the uncontrolled growth of urban centers. Nevertheless, the need for accurate land records is often ignored by policy-makers; and the cadastral systems of many countries are, in consequence, highly defective ... (United Nations 1974 25-26). This situation resulted in improper planning and inefficient management of natural resources, and was a catalyst to exacerbated social and economic problems. This observation from the United Nations was probably the first to establish a link between accurate land records and efficient resource management. Reports from the World Bank (Feder and Davis 1991; Holstein 1990) indicate that although efforts have been made by governments of developing countries to improve the quality of their land records, progress has been slow and success stories are few. Among developing African countries, the use of land as a source of livelihood often takes precedence over its use as a marketable commodity. In some rural communities of Africa, such as in Tanzania, Ethiopia, Zimbabwe, Uganda, Namibia, and Kenya, parcels of land have traditionally belonged to tribes, ethnic groups, or clans. Among rural communities, land titling and registration concepts, as are known and practiced in the western world, are alien to established systems of communal ownership and land stewardship. Common property rights in the traditional (indigenous) land tenure

PAGE 16

3 system ensures that users hold land in trust for descendants of the community. In the urban centers of these same countries despite the relatively active land-related transactions many people view the processes of registration and titling as expensive time consuming and offer no tangible benefits to them as individuals (Moreno in Dale and McLaughlin, 1989 27). The result is that only a small percentage of the land in most developing countries is registered. The system for recording individual rights to the use and enjoyment of parcels of land in most developing countries function poorly. In the urban centers of most developing countries, it is financially costly and extremely time consuming to establish legal ownership to any land parcel and to identify the type of limitations to the use of the parcel (Dunkerley 1985; Dale and McLaughlin 1989). Part of the economic problem in most developing countries has originated from improper allocation of land to its most economic use improper or ineffective documentation of rights to land and consequently inability on the part of the government to monitor the use of the land and to invest in the resources of the land. In addition, most developing countries have not had proper procedures for recording transactions in land. Experiences in developed countries indicate that private individuals, and the public as a whole, are benefitting from the derivative information which is obtained from large scale documentation of ownership and rights to land units. Such benefits are realized in terms of effective land management, fast and efficient transactions in land, fair property taxation system, and economic development. Change, however, has begun to occur. Taylor (1991) noted that since 1986,

PAGE 17

4 governments of many developing countries such as Mexico Bolivia St. Lucia, Zimbabwe Botswana, Thailand and Honduras, with the help of international agencies such as the World Bank and the United Nations have initiated programs to recompile and to maintain comprehensive records of land ownership as well as transactions in land as initial stages to effective land management. With the exception of Thailand many of these programs are in progress and successful results with regard to effective land and resource management have not been well documented. The research presented here is an effort to reorganize and modernize the land records management system in Tanzania in a manner that will enable the government to take control of land and resource management and to ensure economic development through proper land and resource management. Research Objectives, Methodology and Scope In 1991, the government of Tanzania initiated a reform ofits existing land policy in favor of a land market economy. Adoption of a land market economy in Tanzania required that the cadastral system should not only be aimed at documenting the legal ownership of the parcel (as is the current situation), but it should contain information that may be used for faster property conveyancing, fair property valuation, equitable tax assessment and for monitoring land-related transactions and land use patterns. In reforming the cadastral recording system to accommodate the changes in objectives the existing records and the recording process had to be purged of all the problems that prevented efficient land allocation and title registration. The improved recording system

PAGE 18

5 had to be designed to facilitate the registration process provide efficient data recording, storage and retrieval methods and where necessary trace the sequence ofland transfers over a specified period of time. This research has been conducted using the following basic assumptions: Land as a resource, is essential to all mankind and therefore needs to be managed efficiently. The cadastre is a tool for land and resource management (Holstein 1990) The cost of compiling the cadastre should bear a direct relationship to the value of the land and the objectives for government (Dale 1990). A complete and up-to-date cadastral information system can serve as a resource for land management, and socioeconomic development and national development (Holstein 1990). The cadastre is the building block for a multi-purpose land information system (NRC 1983) The need to maintain land records is particularly important in developing countries (United Nations 1974; Dale and McLaughlin 1989). The objectives of this research are: To investigate a suitable approach for reorganizing existing cadastral records by studying the existing process and identifying problematic areas. Identify and recommend actions that will eliminate the problems and pave the way for a more efficient cadastral data capturing and processing.

PAGE 19

6 Using the reorganized records, develop a cadastral information system which will be the building block for a broader Multipurpose Land Information System. Design an improved land data processing procedures to eliminate bottlenecks and to speed up the titling and registration precesses. Develop an approach for cleaning and updating existing records. Develop procedures for incorporating new data into the system. To achieve these objectives, a comprehensive study of the Tanzania land delivery and cadastral record management system had to be done. The study involved visits to administrative centers within Tanzania to study the land management procedures. The study included, among other things, land allocation procedures, cadastral data capture and processing, title registration and records management procedures, as well as the legislation that identified the type, quality, standards, and format for the data. The record keeping and maintenance procedures had to be studied in order to identify problematic areas so as to develop improvements to the system. The focus of this research is on organizational and administration issues pertaining to the reorganization of cadastral records and development of a cadastral information system for Tanzania. This research does not cover political, legal, policy or institutional issues pertaining to cadastral information systems, land management, or land information management in Tanzania, although it is recognized that these play a major role in land and resource management.

PAGE 20

7 Definitions For this research Land is defined to encompasses all things directly attached to the surface of the earth including those areas covered by water (Dale 1989 ). Land management is defined as the process by which land and resources in land are put to good effect. Land management includes resource management which deals with facilitation of economic development through inventory extraction, conservation, and sale of natural resources. Allocation and management of such resources are effected through instruments concepts measures and principles which are based on culture land laws land tenure and property rights registration of those rights and transactions involving those rights (Holstein 1990) Land administration involves the development and use of the land in the manner which has been prescribed by the instruments land laws and property rights. The aim of land administration is to define management procedures regulations and legal framework for agencies responsible for land delivery estate management revenue generation, planning and control ofland resources. Land administration, therefore provides the mechanism for land planning parcel allocation, enforcing of rights and restrictions on the use ofland impact assessment and policy reform. These activities are facilitated by the ability to capture the relevant data to aid in monitoring and identifying areas where actions and reforms are needed planning appropriate courses of action, implementing the adopted choice and monitoring the results of the implementation for success and further improvement.

PAGE 21

8 In this dissertation, information is defined as the product of data analysis Within any organization, management decisions and actions arise from the flow of information upward downward and laterally across the organization. Due to the comple x nature of decision making processes vis-a-vis the external factors which influence the decision, quantity and quality of the available information upon which the decision is based and possible impact on the organization and the community at large there is often a need to establish an information system to support decision making processes. An information system is a group or pattern of associated activities which according to Anderson (1986 ), will normally have the following elements: A common purpose. An identifiable objective. An established sequence of procedures and data flow with at least one but possibly many elements of input, movement, action, storage and output. Feedback of information, giving control over the system. A boundary that defines the extent of the system. Dependence on specific data. One such information system is a Land Information System (LIS) The purpose of an LIS is to provide a decision support for land management. In achieving this objective the requisite data about the land within the confines of the jurisdiction, are captured stored, and processed. As an information syste~ an LIS has an established sequence of data input, data processing and output channels. For this dissertation therefore an LIS is

PAGE 22

9 a system containing spatially referenced land data and requisite analytical tools for querying the data to obtain information to support land and related management decisions. The system may include human and technical resources which allow retrieval and dissemination of the information. At the root ofland information systems are parcel-level, though not necessarily parcel-based, data. A Cadastral Information System is a special type of land information system which deals specifically with cadastral data It is comprised of computer hardware, software, and a database containing cadastral data such as the graphical layout of the parcel, ownership, size, location, use, and encumbrances that affect the use and enjoyment of the parcel. The system enables the performance of ad hoc queries on the data and may have graphical capabilities for the display of the results. Cadastral records play an important role when it comes to transactions in land and management of properties in the public and private sectors. Due to marketability and transferability of land parcels, ownership information changes very often and hence consideration needs to be given as to how current this information needs to be to meet user needs. Information has to be relevant, available and timely, if they are to be of any use for land management. The purpose for which cadastral information are to be used should control the accuracy and reliability standards for data capture and management. Although in some circumstances one can be more important than the other, such relative importance could change over time.

PAGE 23

10 Research Organization and Contribution This research begins with the notion that land policy as an institution, governs all land management activities. From this standpoint, a land management taxonomy is developed in Chapter 2. The taxonomy identifies the hierarchical structure of land management and the role of information as a decision support tool in land management Chapter 3 discusses existing cadastral information models and current issues pertaining to spatial, descriptive information and linkage mechanisms and reviews their focus and relevance to the Tanzanian objectives. Applicability of existing cadastral models to these objectives is discussed. This review helps to develop a cadastral information model which uniquely addresses the Tanzania land management problem. Chapter 4 takes an in-depth look at the existing land delivery system and cadastral arrangements in Tanzania and identifies inherent problems. Cadastral data capture and computerization methods are also reviewed with a view to remove bottlenecks. A discussion of the Tanzanian land delivery problems is conducted in Chapter 5. Solutions and approaches for eliminating the problems are discussed at this stage. A Multipurpose Land Information system which utilizes topologically structured graphical overlays is presented as a model for managing Tanzania land information. In Chapter 6, a pilot project aimed at implementing the recommendations and to highlight the benefits of integrated cadastral information system for Tanzania is discussed. Conclusions and recommendations are presented in Chapter 7. This research makes an initial contribution to the field of cadastral record organization, with specific reference to Tanzania, by conducting an extensive study of the cadastral data capture, land allocation procedures, land registration system, and record

PAGE 24

11 keeping and maintenance procedures in Tanzania. Organizational and technical problems associated with current practices and procedures are also identified. For the first time the legislation identifying the structure and responsibilities of land management agencies within Tanzania are reconciled to identify areas of concern such as overlapping responsibilities and inconsistent authorities between land management agencies. In developing procedures to alleviate the identified problems a cadastral information model was conceptualized based on the information support system components which were developed in Chapter 2 using topologically structured graphical overlays for individual information support components. This approach which utilizes base maps in topological space is a deviation from existing models which are based on base maps in metric space Some of the existing metric models have been reviewed in Chapter 3 In addition, the data organization is designed with regard to the type of land information support system ( cadastral infrastructure socioeconomic, or environmental) rather than the agencies that utilize the information, especially since within any jurisdiction the data requirements in terms of volume and accuracy vary among agencies. Another contribution is the development of a hierarchical parcel identification structure based on the socio-political divisions in Tanzania and the public s perception of parcel identification, as a linkage mechanism between the descriptive information support systems and the graphical overlays. In considering the records management aspect of this research, a systematic approach for reorganizing existing cadastral records, and developing a cadastral information system without drastically altering existing administrative procedures was the

PAGE 25

12 objective. An approach for isolating inconsistent and obsolete data among existing records was developed. Another procedure for recording new information and for maintaining the integrity of the system was developed. Finally an approach for removing administrative bottlenecks within the system and speeding up the land allocation and registration processes without compromising accuracy or integrity of the data was developed. As would be observed, some of the contributions of this research are specific to Tanzania while other contributions such as the topologically structured land information model and the parcel identification system may be applicable elsewhere.

PAGE 26

CHAPTER2 LAND MANAGE:tvffiNT AND CADASTRAL INFORMATION SYSTEM Disregarding minor additions through volcanic activities earthquakes and other natural occurrences land is finite in size. Land provides the resource base for most human existence Humans, plants and animals have always depended on the land for sustenance. As of yet no substitute has been found to match the uniqueness of land both as a resource base and as the platform upon which terrestrial activities are performed. People have different concepts or attitudes about land On one extreme land is viewed as property which carries specific rights of ownership and use which is transferable to other people This concept is dominant in developed countries where land is viewed as a marketable commodity and as such, can be used as collateral for credit and economic development (Marquis 1979) Another extreme is the view of land as common property. This view implies that rights of access and use of the land belong to members within a specific group or community. Access and use by people outside of the group or community is restricted (Bohannan 1973; Marquis 1979). This concept is mostly held in developing countries where land is viewed as an interacting natural system whose integrity needs to be protected, an whose primary qualities have to be preserved for future generations (Andrews 1979). 13

PAGE 27

14 Population increase, large scale mechanized farming, and urbanization have imposed unanticipated pressures on the available land in both developed and developing countries. For example in Tanzania mechanized farming during operation vijiji" (see Appendix B) eliminated large portions of the breeding grounds of the Masai tribe and introduced other socioeconomic problems besides alienation of their land. With advancement of development and agriculture, industry and settlement compete for available land. Usage of parcels ofland undergoes changes sizes of holdings change, and land values change in response to social and economic factors. Land management activities such as land use planning and control are therefore critical to the survival of any nation or community that is undergoing development to ensure that any piece of land is put to its most economic use. In this chapter, the role of information in decision-making is reviewed in the context of land management. This chapter establishes the need to organize land records in a manner which facilitates synergy of information in support of land management decisions. Land Management Land management, in the context of this research, is viewed as an embodiment of legal principles, administrative procedures, and operations which are associated with the stewardship ofland. The overriding principle for land management activities is the policy (see Figure 2-1). Land policy is an institution which comprises of social, economic cultural, and legal prescriptions that define how the land and benefits from the land are to be allocated, and the manner in which the land and resources are to be used (Dale and

PAGE 28

15 LAND MANAGEMENT I LAND POLICY I I I r----------,------------, I I I I I I Land Delivery Revenue Generation I I I LEGAL I FRAMEWORK I I I I I .... ,.. .... ,,. Estate Management Regulation and Reform I I I I I I ____________ ..... ___________ __ I r----------~-----------I Planning and Control I Adjudication OPERATIONAL ~--1 COMPONENTS Survey Registration Titling Other operations L--------IM Valuation and .... ----------~-------------------------T----------------, I I I I I I I I Infrastructure Info System Socioeconomic Info. System ,J ,J I I I INFORMATION SUPPORT SYSTEMS I I ..... ..... Cadastral Info. System Environmental Info. System I I I I I I I _________________________________ _, Figure 2-1: Land Management and Land Information Systems

PAGE 29

16 McLaughlin 1989, 6). A land policy provides the guiding framework within which interests in land may be held and the manner in which the land, as a resource may be managed. Barnes (1994a), contends that in the ideal situation, the land policy should reflect the practices of the people, even though this is not the situation in many developing countries. Land policy prescriptions are defined through instruments, land laws rules and regulations which are executed by land administrators. Although the policy should establish the legal framework within which land management operations are conducted it has been noted by Barnes (1994b) that in many Latin American countries the land laws are drafted with little or no regard to the indigenous tenure practices of the people. Another example is Tanzania, where existing land laws are the same as those that were enacted by the colonial governments, with minor modifications such as replacing the word "Governor" with "President." Existing land policies of Tanzania clearly do not reflect the land tenure practices of the Tanzanian people (Shivji 1995, 4-8). The legal framework of land management is the subject of the next section. Legal Framework Through the legal framework of land management (Figure 2-1 ), rights and rules which guide the development and use of the land in accordance with the prescriptions of the land policy are established. The objective of the legal framework is to define guidelines for managing land resources, protecting individual rights to the use and enjoyment of the land, consolidating land, protecting the environment, and controlling land

PAGE 30

17 development and land degradatio~ which may result from uncontrolled urban migratio~ excessive use of chemicals, pollutants, and other adverse uses of the land. The legal framework defines the mechanism for land planning, parcel allocatio~ enforcing of rights and restrictions on the use ofland, impact assessment, and policy reform. 'Rights,' in this context, should be distinguished from 'rules' governing the use of the land or its resources. The Random House Dictionary of the English Language (1971) describes a 'right' as a just claim, whether legal, prescriptive or moral, and defines 'rules' as principle or regulation governing a conduct, actio~ or procedure. Rules, therefore create regulations and thereby, authorizations. A property right is the authority to undertake particular actions related to specific domain on the land. For every right an individual holds, rules exist that authorize or require particular actions in exercising that property right. In additio~ all rights have complementary duties. To possess a right implies that someone else has a commensurate duty to observe this right. Thus, rules specify rights and duties. Proper land management cannot be achieved by simply defining the legal framework to guide the manner in which resources pertaining to land can be used, but also by providing suitable conditions for individual social and economic development, such as secure ownership, access to credit facilities, and participation in land markets transactions (Holstein 1990; Palmer 1996). Legal prescriptions of land management are defined with relevance to issues pertaining to land delivery, estate management, revenue generatio~ and regulation and reform ( Dale and McLaughlin 1989; Nichols 1993). These are described in the subsection below.

PAGE 31

18 Land Delivery This involves acquisition of land for public good or public use government projects, and new settlements particularly to accommodate the poor and landless and the assignment or delegation of interests in land parcels to individuals and organizations. An ideal land delivery system should enable planned access to land in order to meet basic and developmental needs of the people (Williamson 1990, 88). Land delivery procedures vary from country to country. However, national interests and patriotic behavior requires that those who have access to land, use it productively, and in a manner that will enhance national development. In Tanzania, where the land is regarded as belonging to the state, and land ownership in perpetuity is not the practice, rights to the use and enjoyment must first be obtained from the state. Allocation is normally for a fixed period with some reversionary rights and some usufructuary conditions. There may be conditions attached to the use and enjoyment of the land so as to ensure enhancement and proper use. One such condition may be that the land must be used for a specific purpose, such as agriculture, residential, or commercial. Another condition may be that the land should be developed within a specified period of time to avoid forfeiture. Estate Management Estate management involves the management of large land holdings owned by organizations and communities, such as tribes, clans, and families. In some jurisdictions, the rules of land management within an estate may differ from those of other lands within

PAGE 32

19 the state. For example, in Tanzania, it is required that individual owners of urban land should have registered titles whereas an individual title is not required for property ownership within community land. Another example is the situation whereby parcels in an urban setting may be transferred to anyone, whereas community land may only be transferred to members of the community. In villages and rural areas, where the land is owned by the community, a 'Right of Occupancy' title is issued to the communities. Regulation and Reform Regulation and reform deals with issues affecting the manner in which land and its resources are used. In developed countries, regulations in land use are imposed not only to protect the land from excessive degradation and abuse, but also to protect the public and the environment. As development progresses and priorities regarding the demand for available land change, the need arises to review the uses for particular pieces of land and to reform policies and conditions affecting certain uses of the land. Regulation and reform activities are designed to monitor the land and to recognize changing objectives so as to reform the land use plans, and to control excessiveness accordingly. Revenue Generation The revenue generation function is a way for the government to generate revenue for land and infrastructure development. In developing countries, this is done in the form of fees for processing transactions in land, such as sales, transfers, and registrations. In situations where the land has been leased for a fixed term, a rental fee may also be imposed.

PAGE 33

20 In some developed countries, such as the United States where the full bundle of rights is conveyed during a sale or transfer, one of the methods for generating re enue from land is to impose a tax on structural improvements which have been made on the land. The legal framework of land management is implemented through operational agencies which are responsible for adjudication, planning and control land survey, registration and titling, and valuation and assessment (see Figure 2-1). These are discussed in the next section. Operational Agencies Operational agencies constitute the organizational arrangements which are made to administer the prescriptions within the legal framework to support land management activities. Operational agencies consist of divisions in management that are staffed by personnel whose responsibilities include the provision of services in accordance with the legal interpretation of the guiding principles of the land policy. The operational component of land management is multi-disciplinary. Effective land management involves the interaction and coordination of several government agencies and several operational units. There is no definite pattern as to which agencies should be responsible for any particular operation. However, for this research, land management operations have been defined to include land use planning and control, adjudication, land survey, titling, registration, and valuation and assessment (see Figure 2-1). Those with decision-making responsibilities within these agencies are constantly having to review their plans and actions, and to modify decisions in accordance with

PAGE 34

21 changing conditions. Such decision makers need the information that may be derived from analyzing the database, in order to make sound decisions. Decision-making within the operational components are either to provide solutions to prevailing or impending problems or to capitalize on opportunities. Problem detection and resolution arises when prevailing conditions indicate a deviation from the expected results and therefore may impact adversely on expected objectives. For example, if the data indicate a higher than normal growth in urban population, the decision may be to allocate more land for residential purposes. This decision may be an ad hoc one, however, a closer look at the data may reveal some underlying cause, which may require a more permanent solution. Timely resolution of such a problem can avoid some of the socioeconomic problems that are associated with excessive urban migration. Opportunity-seeking on the other hand, arises when the facts suggest that a particular action may result in opportunities for the agency or the community. For example, the strategic decision to construct a dam across a river may be aimed at providing more agricultural land, jobs and development for the community. Also a decision to computerize land records based on the knowledge that the information would be useful as decision support resource for efficient land management operations. Both problem detection and resolution, and opportunity-seeking strategies are based, not upon instincts and wisdom alone, but on analysis of available data, together with external factors, such as socioeconomic and environmental conditions within which the agency operates (McCloy 1995, 342-347).

PAGE 35

22 Planning and control, in the context of land management deals with allocation and monitoring of resources in land with a view to maximizing efficiency by putting land to its best use while ensuring the welfare of the community and the sustainability of the available resources. Adjudication is the determination of rights in parcels of land. The procedure involves identification of the types of rights in the land, the persons in whom those rights are vested, and limitations to the enjoyment of those rights. The adjudication process is used by the operational units to eliminate defects in land titles by judiciously applying the legal principles that define land ownership (Dale and McLaughlin 1989). Due to the broad scope of land management activities, the following section is focused on the cadastral aspect of land management operations, which involves land survey, land titling and registration, and in the case of fiscal cadastre, valuation and assessment. Land Survey Land survey is a process for providing the geometric framework for mathematically defining land parcels. As an essential land management tool, survey plans and maps are used for planning and controlling development, for redefining disputed and uncertain boundaries, and for defining property, political and geographic boundaries. Different types ofland survey activities are conducted for different purposes. For example, cadastral (land inventory) surveys are used to establish property boundaries and to determine sizes and

PAGE 36

23 shapes of the parcels over which individual rights exist. Cadastral survey of the mutually accepted boundaries ensures that the boundaries can be replaced if they are destroyed. Wherever possible cadastral surveys should be tied to a network of pre-established control points which were connected to a geodetic reference frame. The global positioning system (GPS) and advances in modem surveying technologies have made it easier to connect more survey works to the geodetic reference framework. With advancement of computerized information management systems, survey plans and maps, in electronic forms, have become an integral part of a land information system (LIS). These approaches facilitate the integration of information because it is referenced to some spatial reference framework. Land Titling Land titling is the process of issuing valid property titles by a recognized state agency to existing occupiers of the land who do not have legitimate titles. The land titling process confers official recognition of individual rights to the use of any particular parcel. In a title registration system, such as the Torren' s system in Australia, where some guarantees are offered by the state against inadvertent loss due to an error in the registration process, the title provides the unimpeachable proof of ownership and therefore tenure security (Dale and McLaughlin 1989, 26) to the individual. Land title records constitute a step towards land records compilation for the government. As documentary evidence of ownership, the title may show types and limitations of any rights that are exercised over the particular parcel of land. There is a general belief that secure property

PAGE 37

24 rights, in the form of registered titles, act as an inducement for investment in real property and, in the longer term, contribute to increased productivity by the individual (Holstein 1990; Lemel 1985). Feder, et al. (1988) also have presented evidence to support the notion that increased security of tenure by having title documents increased agricultural productivity in Thailand by between 11 and 27 percent over comparable non-secure properties. Land titling should be accompanied by a registration system in order to legalize the ownership of the property. Land Registration Land registration is the process of recording information about legal claims to parcels of land. Registration is done to ensure clear and unambiguous titles and to avoid fraud and disputes pertaining to conflicting claims concerning the right of use and enjoyment of any piece of property. The types of registration include : private conveyancing whereby the records of the transaction are handled privately between the individual parties, sometimes in the presence of witnesses. the deed system where the copies of the transaction records are kept in an official registry of the government or state. title registration in which a state organization maintains the records sometimes with some guarantees in terms of security. In the past most forms of land registration were classified as either deed registration or title registration. Over the years, adaptations of the two systems have been implemented by governments according to their objectives and suitability of the system to their local

PAGE 38

25 purposes. The main differences between deed and title registration systems have been documented extensively (Dowson and Sheppard 1968; Simpson 1984; Dale 1990; Larsson 1991). Since the focus of this dissertation is more on the land records and the precess for acquiring such data than on the legal and institutional aspects of land registration systems, this distinction will be ignored. However, it is important to note that as a land management process, the system of land registration can influence not only the physical and legaL but also the social and economic environments (Dale and McLaughlin 1989) of any jurisdiction. Valuation and Assessment In order to generate revenue for development, procedures are adopted to assess and to tax land as a resource. Property rating and assessment are two methods that governments use for generating revenue from land for development (Dale and McLaughlin 1989, 47). With the property rating system, revenue is generated as a result of a property assessment. The tax is applied to improvements to buildings, and other structures that have been erected on the land, and the uses to which the structures are being put. With land valuation, tax is determined not on the structure that have been erected on the land nor on the land use, but on the basis of the value of the land itself as determined from the improved or unimproved state. The revenue generation objective necessitates classification of the uses ofland parcels, and where necessary, the yield. This ensures a fair assessment of tax liability. Land reco.E_ds for such purposes have been an integral component of the original 'cadastre'.

PAGE 39

26 Cadastre. The word cadastre has its origins in antiquity when it referred to a register containing descriptions of land parcels, value, use and proprietorship. The original purpose for cadastral records was to assess the liability for tax and to determine responsibility for payment. Over time, the records began to show evidence of land rights (Simpson, 1984). Dowson and Sheppard (1968, 47), remarked that: ... it is impossible to give a definition of cadastre which is both terse and comprehensive, but its distinctive character is readily recognized and may be expressed as the marriage of: 1. A technical record of the parcellation of the land through any given territory, usually represented on plans of suitable scale, with 2. Authoritative documentary record, whether of a fiscal or proprietary nature or of the two combined, usually embodied in appropriate associated registers .... Today, the term cadastre is used to imply a parcel-based up-to date record of rights, and responsibilities in land (FIG 1996). The cadastral record includes a graphical delineation of the land, to which other descriptive records pertaining to ownership, types of rights, and sometimes the value of the parcel and any record of improvements on the land are linked. Such pieces of data are contained in the land survey, land registration, and the valuation and assessment records, as shown in Figure 2-2. The use of cadastre has broadened to an extent that in situations where standards for its creation and maintenance can be ascertained in the judicial system, the cadastral record may have legal status which is recognized by the courts, not only as property description, but also for confirming right of use to the registered owner (Larsson 1991). The cadastral map can also serve as an index to other legal records such as mortgages and liens. Currently, there are three distinct types of cadastral records (Simpson

PAGE 40

27 LEGAL CADASTRE -----------------------------------------------, ---Land Survey Data fl Land Registration Data .,. : ; : :: ,' ..... -,~~-'l" ,., ,. Land Titling Data Valuation and Assessment Data FISCAL CADASTRE -------------------------------------------------------' Figure 2-2: Components of a cadastre 1984 ; Dale and McLaughlin 1989) ; 1. Fiscal cadastre which refers to a register that has been compiled primarily for property valuation and tax assessment as a source of revenue generation for the government. As an information resource it also serves as an instrument for administering the policy on land taxation system It provides the information base for equitable and efficient tax assessment. 2. Juridical or legal cadastre which contains records of the legally recognized record of land ownership as a means of avoiding conflicting claims to the use and enjoyment of the parcel of land. It also provides a means for legally transferring those rights or interests, either through sale, lease, or mortgage. As an information resource, legal cadastres provide the information base on land ownership to assist planners in their efforts to sustain development and curtail abuse of the land.

PAGE 41

28 3. Multipurpose cadastre is a combination of both the fiscal and the juridical as well as other parcel-related information. It provides a variety ofland tenure, registration, and information services that are required by the community and other land management agencies (McLaughlin 1975). While the origins of cadastral record compilation may be tied to tax collection of the olden days, principles underlying the cadastral register have been adopted around the world due to the amount of information that can be derived from an analysis of the records. The operational components of land management involve decision-making as well as selection of choices among possible options. The availability of information in an appropriate form can reduce the amount of uncertainty among the options. For efficient performance, therefore, operational agencies require an information support system in order to make sound land management decisions. Information support systems are discussed in the next section. Information Support Systems Information is the basic ingredient for sound decisions. In a decision-making situation, data alone can be overwhelming. Within any organization, better information leads to a better understanding of a situation, and thereby, the possibility of a better management decision. The need for a timely and informed decision calls for innovative ways to not only access accurate and up-to-date information, but the tools to analyze the available data and present the information in useful and easily comprehensible ways.

PAGE 42

29 Activities within the operational units of land management in Figure 2-1 can be viewed as a series of decision making processes. Information pertaining to the land are needed to support decisions concerning the operational aspects of land management. Through decision-making processes rights restrictions and uses of the land are continually revised so as to sustain development and maintain the maximum utility of the land For example the rights and restrictions applied to peri-urban land in a developing country may change once the area is integrated into the urban town. Administrative control is transferred from the community to central or district government. Different rules and regulations are applied to the transfer sale and use of what used to be village or community land. Similarly the use of a particular piece ofland may change from residential to commercial in order to keep pace with development and to satisfy the need to provide amenities for the residents Land managers need adequate information pertaining to the land for planning managing and controlling its resources Many of the developing countries that maintain cadastres still use rudimentary filing systems. With population increases and an associated upsurge in demand for land and parcel-related transactions there is an increased need for more refined information to support decisions in land management. These situations call for a need to record details of land parcels in an organized manner. Depending on the number of files, volume of information, methods for cataloguing, and storage of files, information retrieval can be an onerous task. In most developing countries, the inadequacies of information pertaining to the land pose serious constraints on land administration and resource management. Without knowledge about the land ownership and type of tenure, development programs

PAGE 43

30 are difficult to initiate. Land-related information which is embodied in the cadastre is being increasingly recognized by governments as valuable resource for decision-making in land management (Dale 1991 ). Governments of many developing countries are reorganizing their cadastral records so as to derive the benefits of effective land administration. The governments are doing this by ensuring that rights in land are identified, recognized by the state, and are recorded in a suitable form. Examples of such activities are found in Peru (Pahner 1996), Bolivia (World Bank 1995), and Asia (Burns et al. 1996). Land records are also being converted into digital format so as to harness the benefits of technological advances in data management and information processing with the use of computers. Advances in computer technology during the latter part of the 20 th century have contributed immensely to the growth in the use of information handling technology. This technology has offered decision makers the tools to adequately analyze voluminous amounts of available data and the ability to model the effect of decisions, even before they are implemented. This opportunity provides a means to perform ad hoc queries in order to arrive at a viable option when it comes to decision making and allocation of resources. In Figure 2-1, the information support system in land management has been grouped into four broad categories: environmental, socio-economic, infrastructure, and cadastral (Dale and McLaughlin 1989, 11). Each group can serve as an independent land information system.

PAGE 44

31 Environmental Information System The focus of an environmental information system is to provide information about factors which influence human health as well as ecological and economic impacts ofland use. The objective of the system is to protect and improve air quality as well as land and water resources from degradation and abuse. Human activities are not the only the only factors that can adversely affect the quantity and quality of the resources. Naturally occurring phenomena such as volcanoes, earthquakes, tornadoes, hurricanes and forest fires are all factors which impact the environment and hence human survival. Therefore, an environmental information system is used to address issues associated with human activities and programs that affect the environment at local, regional, or global levels. Examples of environmental issues would be wetland depletion due to rapid and unplanned development, impact of development on endangered species and degradation of the ecological system, flood analysis, and pollution control. Socioeconomic Information System Socioeconomic information systems include census, demographic, and statistical data which are essential to government agencies for planning. Several different types of data are gathered during a population census which are used by governments to monitor progress of development and to plan for the future. The information derived from the data helps to redefine priorities with regards to development objectives and to re-allocate resources with respect to amenities such as schools, housing, infrastructure, and other

PAGE 45

32 needs. By themselves socioeconomic data may not qualify as land information However when demographic data are referenced to a spatial location, the data become land-related and therefore could be used as an input into a land information system. Infrastructure Information System Also referred to as Automated Mapping and Facilities Management (AM/FM) an infrastructure information system is used for managing engineering and utility structures such as pipe lines telecommunications transportation, and underground facilities. In general, utility companies have the same requirement as parcel-based record keeping to maintain records of their transmission and distribution networks and to make decisions regarding system capacities to meet public demand based on growth potential. With a infrastructure information system, it is possible to forecast demand for utilities plan extension, locate plants for maintenance, and provide service connections. Land Managers in turn benefit from such an information system by obtaining up-to-date information on the development of the land so as to strategically plan for the location and magnitude of future facilities. Cadastral Information Systems Information pertaining to land parcels and land ownership and use are always needed by several agencies for land management purposes Reliable land records on their own, do not provide solutions to land management problems. However, they do provide the resource through which solutions can be devised and implemented. Whether the

PAGE 46

33 cadastral records are kept in a manual filing system or in a computerized format a cadastral information system provides the government and other interested parties with a complete and up-to-date inventory of land holdings and land use patterns for a particular jurisdiction. A cadastral information system, as illustrated in Figure 2-3, is a system comprised of computer hardware software database and human resource which operates on cadastral data The cadastral data may vary in accordance with the goals of the cadastre. Whereas a legal cadastre will have information about the legal description of individual parcels the fiscal cadastre will contain such information as land value land use and tax liability. Other information such as ownership, shape area location, and owner s address may be common to both fiscal and legal cadastres. The system may be linked to other records such as data at the land registry or utility records. The records are compiled from affidavits signed or notarized documents maps documented evidence boundary identifiers adjudication records files and other legal documents that contain relevant information about the parcel. In a dynamic land market cadastral information is the most clearly documented type of land information because ownership of properties changes every so often. Proper records maintenance and updating procedures must be adopted in order to keep the data current. A computerized cadastral information system allows analysis and synergy of information in support of land management activities. Grouped together information support systems as shown in Figure 2-1, are referred to as a Multipurpose Land Information System (MPLIS). Multipurpose land information systems are recognized as a valuable resource for effective decision making in land

PAGE 47

34 Hardware ..... Software ,.,. Database Human resource A Land Registration Data Cadastral Data Legal Cadastre Legal description Ownership Evidence Historical data Uses Limitations Liens Shape of parcel Other encumbrances I I I I I I I I I I._' Area I Monuments I Spatial Location Administrative location I Cadastral plans I Ownership Owners' s Address I Terms I Value Use Tax liability Fiscal cadastre Figure 2-3 Components of a Cadastral Information System management (Dale 1991). In most developing countries the inadequacies of information pertaining to the land pose serious constraints on land administration and resource management. Since 1985, emphasis in many countries has been placed on ensuring that rights in land are identified, recognized by the state, and are recorded in a suitable form. Examples of such activities are found in Peru (McLaughlin and De Soto 1994), Bolivia

PAGE 48

35 (Barnes 1994a ; Barnes 1994b ), Asia (Burns et al, 1996; Feder and Nishio 1996) and the Republic of Belarus (Bloch 1996) Such activities involve graphical delineation of the property boundaries followed by an association of the relevant descriptive information. Several approaches have been adopted to produce a representation of the property boundaries. The approaches range from rudimentary land survey methods such as the use of the surveyor s compass and linen tape, to photogrammetric methods with variations dictated by circumstances such as cost time and technology The next chapter deals with current issues relating to graphical data capture and attribute data compilation.

PAGE 49

CHAPTER3 CADASTRAL INFORMATION AND RELATED ISSUES Advances in computer and information technology have revolutionized the way management decisions are made. Information systems are being utilized as decision support resources to minimize the uncertainty in the choices that are made by management. With cadastral information, technology is again impacting the method for capturing the graphical data, the structure of the descriptive data, and organization of the cadastral records so as to meet the information requirements of the jurisdiction. Current research has focused largely on three distinct areas of the cadastral information system These are the structure of the cadastral model, the methods used to capture the graphical data and the organization and association of the descriptive data. This chapter deals with a literature review and identifies current thinking in these areas even though the state of technology in Tanzania does not lend itself very well to the application of the latest technology. Historically, land records have been compiled for public use, by the State or by the private sector. Simpson (1984, 124) distinguishes between land records which have been compiled for the benefit of the State and those that are compiled by private entrepreneurs. Whereas information required by the private sector are those which facilitate dealings in land such as conveyancing and mortgage, those that are required by the state are related to issues such as taxation, economic planning and land management. 36

PAGE 50

37 A cadastral information syste~ as has been shown in Figure 2-1, is a type of land information systems which is central to the land management process in any jurisdiction. Development of cadastral information systems in the developed world have been influenced by factors such as technological advances in the latter part of the 20 th century, the need for improved methods of managing land and related resources, and efforts to protect the environment. In recent years, the type and detail of cadastral information that is needed to support particular societal and administrative needs have been changing with respect to the changing needs of the society. In the 1980s, researchers focused on cadastral models to identify the complex interactions between the cadastral information and institutional, political, and economic development of governments. Later, Williamson (1990, 81) observed that cadastral models and related studies which evolved in the 1980s clarified concepts identified essential elements and broadened the use of cadastral information. A description of the different types of cadastres is presented in the next section. Different Types of Cadastres Today, many types of cadastral systems are in operation with varying degrees of resemblance to the classical fiscal and legal cadastres. Distinguishing characteristics of cadastral information systems include factors such as the spatial resolution and scale of the source map, the type and characteristics of the information that are recorded, and the professional responsibility for managing the data. The Federation Internationale des Geometres (FIG) (1996, 3) has identified the following means of categorizing cadastres: their primary function. (e.g. tax, juridical or multi-purpose land management).

PAGE 51

38 the type of rights that are recorded. (e g. sub-surface rights mineral lease s, or private ownership timber concessions etc.). the level of state responsibility in ensuring the accuracy and reliability of the data. (For example some cadastres may have complete state oversight and responsibili ty and perhaps state guarantee of tenure security whiles other systems may be implemented with varying levels of public and private sector responsibilities for data and information management). location and jurisdiction. (Distinguishes between urban and rural cadastres as well as centralized and decentralized cadastres). the way in which information about the parcel are collected (Methods for capturing cadastral information include digitizing from existing maps and plans aerial photogrammetry, ground surveys tied to a geodetic reference framework uncoordinated ground surveys and measurements, cadastral survey using Global Positioning (GPS) methods geo-referenced aerial photography, etc. Each method contains a certain degree of positional accuracy which influences the spatial resolution of the system). Cadastral information systems are regarded as vehicles for economic growth and social equity (see Palmer 1996; Holstein 1990; Holstein 1996). In consonance with the modern concept of information synergism, funding agencies such as the World Bank the United Nations(UN), International Development Bank (IDB), and the United States Agency for International Development (USAID) are working with governments of developing countries to reform their cadastral systems and to modernize their land records. Cadastral

PAGE 52

39 information models have been developed based on needs and priorities. Below are reviews of some of the existing cadastral information models. Existing Cadastral Information Models Perhaps the earliest attempt to establish some connection between the information provided by the central components of the cadastre and information required by the government for socioeconomic development and land management was presented by Dale (1976) in the United Kingdom and McLaughlin (1975) in Canada. Dale presented a cadastral model that was comprised of central process elements which were influenced by external factors. The external factors consisted of a broad spectrum of factors ranging from government, education and professionalism, to legal issues and socioeconomic factors. The central processes consisted of adjudication, demarcation, survey specifications survey methods and boundary description. The central processes were linked by output elements and some feedback mechanisms. The output elements being the cadastral map, title records, valuation and taxation, and planning and control elements. The feedback elements for the model were the title legislation, land values, planning and control, and boundary disputes. An interesting point about Dale's model was the fact that the boundary descriptions served as the linkage mechanisms to the output elements. The key elements of the entire model were the land survey and the boundary definition. It is not clear at this point whether the model depicted a bias towards Dale's profession as a surveyor, but it is obvious that at the time of its inception, the dominant factors regarding the cadastral

PAGE 53

40 process were the boundary definition rather than land use or the resources within the boundary. Other than simplified keywords, the use of long and verbose boundary description as a linkage mechanism in a computerized environment would have presented some coding problems, both in the amount of space required to store the code and the processing time when it came to searching for an item. More importantly, since people use different words and style in describing the same parcel, the descriptions of a boundary could differs from one database to another. Dale's model did not envisage multipurpose land information systems. McLaughlin (1975) presented the role of the cadastre within a multipurpose land information management system and identified cadastral information as a land management tool and a decision support resource. This became the cornerstone for the cadastral models that are currently in existence. The North American Model (NRC model) McLaughlin's presentation was followed by a multipurpose cadastre model which was developed by the National Research Council (NRC) in 1983 as a basis for a multipurpose land information system for North America (see Figure 3-1). The model is conceptualized as an integrated land information system for both administrative and public uses. It was developed in response to growing concerns about how foreign land ownership affects the balance of trade in the United States, land prices, access to farm land by young farmers, intensity ofland use, and community viability (NRC 1980, 12-13). The primary objective was for land administrators to provide public administrators such as governors mayors, as well as Congress with information pertaining to land holdings distribution and

PAGE 54

41 I 7 CULTURAL LIS ,1 I NATURAL LIS I Natural Resource Records Other Environmental Data VARIOUS NATURAL : i .. I I I i I BOUNDARY CONVENTIONS OVERLAYS PARCEL NUMBERS Figure 3-1: NRC Model for a Multipurpose Cadastre (Highlighted) Source: NRC (1983) use within the North American continent. The NRC model identifies the components of a multipurpose cadastre. A distinction is made between natural land information system and cultural land information system. The model combines the cultural with the natural land information system to form the multipurpose land information system. Whereas the cultural LIS is based on the cadastral parcel boundaries, the natural LIS is based on other natural boundaries The two systems are spatially connected by a unifying geodetic reference framework and the base map with data exchange conventions between the two data types. The NRC model advocates the use of base maps containing natural and cultural features

PAGE 55

42 tied to each other in accordance with the level of accuracy with which the feature was surveyed. The model proposes a reliance on the mapping standards that have been established by the American Society for Photogrammetry and Remote Sensing and the National Mapping Accuracy Standards which were established by the Office of Management and Budget. Williamson's Model Williamson's model which was also developed with regard to the changing needs of society, was designed in 1986 for Australia. The model is a centralized cadastral information system which provides legal cadastral information support for local government agencies. On the basis that a registration system is integral to the implementation of the cadastre, Williamson attaches the same level of importance to land registration as the cadastral overlay. The model attempts to show the importance of the cadastral map and land registration in the design of a multipurpose land information syste~ especially in Australia, where tenure security is guaranteed by the State. Here also the topographic base map emphasizes the need for the base map on a geodetic reference framework to ensure spatial consistency. The Williamson model obviously, does not consider natural resource and environmental records as integral components of a multipurpose land information system. Besides, the structure is identical to the cultural LIS component of the NRC model except for the fact that Williamson's model proposes a land information center as a link between the cadastral database and databases of independent government authorities are established (Williamson, 1986).

PAGE 56

EXISTING INDEPENDENT GOVERNMENT AUTHORITIES FISCAL j ' 43 LOCAL GOVT. j ' UTILITIES OTHER j J ' LAND INFORMATION CENTER LINKAGE MECHANISM CADASTRAL DATABASE (JURIDICAL CADASTRE) TOPOGRAPI-ll-IC MAPPING AND GEODETIC SURVEY ORGANIZATION CADASTRAL OVERLAY j ' TOPOGRAPHIC BASEMAP GEODETIC REFERENCE FRAMEWORK j LAND REGISTRATION Figure 3-2: Williamson's Multipurpose Land information model Source: Williamson (1986). The Developing Country Model This was developed by Williamson and Jeyanandan in 1990 with particular reference to developing countries (see Figure 3-3). The model recognizes interaction between people, land, social groups and cadastre. In this model, block parcels which seem to be the underlying graphical layer are not tied to any geodetic reference framework. There is no indication as to how the cadastral map will be registered with other maps

PAGE 57

44 CADASTRAL MAP BLOCK PARCELS CADASTRAL SYSTEM PEOPLE PROPRIETARY PARCELS SOCIAL GROUPS LAND Figure 3-3: Cadastral Model for Developing Countries Source: Jeyanandan and Williamson ( 1990) produced by other agencies. Also this model focuses on cadastre and does not consider natural features as a component of the land information that will be required by say the planning and development agency. The cadastral blocks are designed around the social groups. Block boundaries which may be individual communities form the basic unit of the cadastral map and are relatively permanent and identifiable on the ground. Proprietary land parcels within the blocks are also recognized. According to Jeyanandan and

PAGE 58

45 Williamson (1990, 91 ), the special features of the model are; Boundaries of blocks are relatively permanent, identifiable on the ground and therefore recognized even without maps. Enables preferences for cadastral products through comparison between blocks and selective education of people, who hold rights within a block. Enables evolution of cadastral system and facilitates selective intervention in land issues. provides the basis for reorganizing land and other data on a geographical (block) basis. Flexibility in, the size of the blocks, use of technology, and cadastral practices. Involves very little additional resources but provides for orderly improvement of cadastral system in keeping with user demands in specific spatial areas. The Wisconsin Land Information Model The Wisconsin Land Information Program was initiated in 1985, in response to demand for information about the land by both the private and public sectors of the community. A multipurpose land information system for the state of Wisconsin was established in recognition of the fact that: a wide variety of land records exist in the form of record books, paper files, maps, charts, and many other formats. different land records might be collected at varying levels of detail and accuracy or might be mapped at different scales (Merideth et al. 1990, 79)." The system was developed to provide among other things, a standard foundation for

PAGE 59

46 accurate geographic referencing of land information. The requirement was an accurate large scale maps which show small areas in detail. To the committee, details were essential for decision-making because of their ability to allow comparison of the areas and attribute of various locations. To a large extent, the Wisconsin model followed the North American MPLIS(1983) model except that instead of a base map, layers of homogeneous accurately surveyed graphical layers such as parcels, zoning, flood plains, soils, and registered together with a geodetic reference framework (NAD83). This approach provided the option to form composite overlays by integrating layers as needed. For accurate graphical overlays, Digital Line Graphs (DLGs) of 1 :24000 scale and 7.5 minute topographic maps were initially used with the understanding that more accurate, precise, or detailed information would be incorporated into the system whenever they become available (Wisconsin Land Records Committee 1987, 23). The unique thing about the Wisconsin program is the separation of the layers and the ability to form composite overlays. However, the need for accurate graphical layers implies that, for a country like Tanzania, where data have been captured with different levels of detail and accuracy would necessitate a re-survey of the jurisdiction in order to obtain a map whose accuracy is uniform across the entire jurisdiction. Other researchers (Fourie 1993; Davies and Fourie 1996) hold the view that the model should be dictated by the social structure. Modem cadastral information systems consist of three parts; a graphical database which contains information that depicts the subject land, a descriptive part which may be one or more databases containing the list of

PAGE 60

47 proprietors and other descriptive information that are relevant to different operational agencies, and a linkage mechanism which in most cases is the parcel identifier that links the graphical and descriptive databases together. Irrespective of the cadastral model, institutional arrangements and management procedures, organization of these components affect the usefulness and applicability of the cadastral information. Spatial Data "The base map is a graphical representation, at a specified scale, of selected fundamental map information, used as a framework upon which additional data of a specialized nature may be compiled (NRC 1983, 37)". Within the context of a cadastral information system, the base map ... provides a primary medium by which the locations of cadastral parcels can be related to the geodetic framework; to major natural features such as bodies of water, roads, buildings, and fences; and to municipal and political boundaries ... (NRC 1983, 39)." Procedures for capturing and presenting graphical cadastral data as well as for producing base maps have undergone technological changes. The biggest changes are occurring in the application of aerial photography for base mapping purposes. Over the years, base maps have been produced from rectified and unrectified aerial photographs, and digital maps, in either vector or raster formats. Advances in computer technology are opening new ways for processing aerial photographs for cadastral purposes, such as softcopy photogrammetric methods. In the next section, issues related to current spatial data capture methods are discussed.

PAGE 61

48 Spatial Data Capture In the past, cadastral boundary data have been captured by traditional ground survey methods. Equipment for such types of surveys have ranged from the plane table tapes and compasses, and transit theodolites to electronic distance measuring (EDM) devices. Whereas these instruments have not been eliminated completely Total Stations and Global Positioning Systems (GPS) technology have become equipments of choice. Cadastral boundary data are depicted as vector representations of the spatial features. Aerial photography has gained popularity due to cost and time savings when surveys of large areas are involved, and also, the fact that photographs enhance the communication of spatial information. People naturally relate better to information depicted on photographs than on conventional line and symbol maps. GPS methods have also been used to extend survey controls so that aerial photography or other ground survey methods can be used to capture the data. Aerial and other photographic forms of data capture are often converted into raster representations. Although spatial accuracy is not the focus of this dissertation, it is worth mentioning that the ground resolution of pixels in the raster image influences the spatial accuracy of the features. The use of GPS for cadastral surveying is still in its infancy (see Barnes and Eckl, 1996), although the results are promising.

PAGE 62

49 Parcel Identifiers In any computerized records system, special techniques are required to define and uniquely identify the land objects or entities about which data are to be recorded in order to associate the graphical database with the attribute information. In 1972 the committee on Compatible Land Identifiers: Problems, Prospects and Payoffs (CLIPPP), noted that parcel identifiers have to be unique, simple, permanent in utility, flexible, economical and accessible (Fisher and Moyer, 1973). Recommendations from the CLIPPP committee included the use of coordinate of the centroid of parcels, the maximum and minimum values of the Easting and Northing coordinates of the parcel, and the block system, whereby the blocks are given sequential numbers. The National Research Council of Canada (1976) identified three methods for unique parcel identification; the hierarchical system, grid-based identifiers, and hybrid identifiers. Hierarchical identifiers represent parcel entity identification structure based on stratified political or administrative units such as Federal, State, County, Town, Ward, Block, and lots. Other forms of hierarchical structure are the Public Land Survey System which is used in most of the United States, as well as the Census Tract and Block system. Grid or graticule system involves the identification of parcel entity based on geographical or Cartesian coordinate system in a spheroidal or ellipsoidal system. The centroid or the maximum and minimum coordinates of the boundaries are used as the grid identifier. The disadvantage is that, as the coordinates are adjusted occasionally due to improved measurement technology and better mathematical model for the shape of the

PAGE 63

50 Earth, the identifiers would have to be corrected with every modification of the coordinates. The min-max approach also assumes that all parcels will be rectangular in shape. For a hexagonally-shaped parcel however, the min-max coordinates will lie in someone else's property. The same situation may occur for an L-shaped parcel if the coordinates of the centroid of the parcel is used as the identifier. The NRC (1983), considered name-related, alphanumeric and location-based identifiers. Name-related identifiers associate individual names and the legal entities over which their interests exist. The grantee-grantor index is an example of such association. Ignoring the fact that duplicate names are common in many jurisdictions, name-related identifiers require that identifiers be changed whenever an interest in the parcel gets transferred. Whereas there is no dominant choice for a unique parcel identifier, computer technology and database management systems facilitate the use of multiple indexes for a multi-purpose land information system. As noted by NRC (1983, 64), the ultimate choice for a parcel identifier should be dictated by local needs and resources such as the need for accessibility and effective management of the identifiers. In this regard, uniqueness, simplicity and economy of maintenance are more important. Since the CLIPPP conference, different jurisdictions such as Dade, Palm Beach, and Broward counties in Florida, have developed suitable, yet independent, parcel identifiers internally.

PAGE 64

51 Data Management One of the responsibilities for Land Information Management personnel, is to integrate the graphical record with descriptive information from other databases. The technology for maintaining and manipulating database has been undergoing evolutionary changes since the early 1960s. In this section, databases that are currently operational as well as those that are in developmental stages are presented. Their applicability to land information management and analysis are discussed in order to justify a suitable choice of database technology for the Tanzania project. The evolution of modem data organization methodologies began in the early 1960s with the development of System Design Life cycles (SDLC) (Lee, 1997). Such systems provided some control over the organization of the data but limited assistance with regard to analytical operations (Rhine, 1995). Data were stored in flat files, hierarchical files, network files, etc. Structured methodologies which evolved in the 1970s provided more effective analytical tools and extended design methodologies. These were achieved by structuring the data into elemental forms and focusing on the modeling of entities and data. Codd's (1970; 1979) relational model, along with the structured query language (SQL) gained support because it provided LIS users with query capabilities which were not available before. The logical structuring of the data determined the degree of flexibility within the system. However, the inherent deficiencies in the relational database model and entity relationship concept influenced the object-oriented methodologies of the 1990s (Yourdon 1994).

PAGE 65

52 Object-oriented technology is regarded as the cutting-edge approach for data modeling, analysis and software design (Lee 1997). The object-oriented principle is based on the assumption that people generally think in terms of objects rather than entities or functions. Objects are direct representations of real things that people perceive when communicating or describing characteristics of entities or things (Yourdon 1994; Usery 1996). The object-oriented principle incorporate the inheritance and encapsulation characteristics of data into an integrated whole (Yourdon 1994). Despite the power of object-oriented technology, very few software packages have been developed to harness the capabilities of the technology. Below is a description of the data file formats that are currently in use for LIS operations and their advantages. Flat Files Flat files are traditional "spread-sheet type" systems without full database management support. Flat files contain tabulated data in rows and columns. The rows contain the records and the columns represent the fields or items. Ordering of the rows within the table has importance for the ways in which data can be accessed. Data retrieval is done by means of search keys which index the occurrence of values for a specific field. Any item (column) within the table can be used as the search key. Searches within the table may be done sequentially, by binary search, or by index search. The efficiency of sequential search depends on the location of the record which is being searched. If the search key is changed to another item within the table, the table has to be sorted again with the new key.

PAGE 66

53 A faster yet resource intensive method is the Indexed search which relies on a separate index table to search for records. A separate table containing the key of every record as well as an address pointing to the data location of each key is associated with the data. The index is sequenced and the search is done on the index rather than the data itself Because the index table is smaller than the table of data itself, searching of records is faster. Even though early LIS implementation used Flat files to associate the graphical overlays Flat files are severely limited in their utility to LIS applications due to their limited flexibility. They are simple and efficient for specific repetitive tasks such as transaction based information (e g in the retail industry and banking activities). Hierarchical Files Hierarchical database system works like a family tree relationship. With hierarchical files there is always more than one record in the file. One record is the parent or master and it can be associated with any number of"children or detail records through internally assigned pointers. The detail records can also have children assigned to them. This establishes a one-to-many relationships among the files. The advantage of this system is that it allows multiple sets of identical attributes to be associated with any given record without storing those repetitive data in separate files Linking of files is done with pointers which provides some flexibility in relating data between records. One major limitation with the hierarchical file system is the fact that data in the detail record can only be accessed by first accessing the master record.

PAGE 67

54 Networks An improvement to hierarchical files are the Networked databases which allow detail records to be accessed with more than one master record. This establishes a many to-many relationship between the files. The advantage is that if any record needs to be updated, it can be done on only one file. The drawbacks to the networks are as follows: Logical linkages among files multiply as new databases are added to the system. In complex networks involving large databases, the amount of storage required for the pointers can be larger than the database itself Management of the pointers, as records are added, new fields created, and linkages created as data values change can become cumbersome and in danger of being inflexible. Relational Databases The relational database concept was developed by Codd in 1970 (Date 1991, xi; Healey 1991, 257). The concept is based on the mathematical theory of relational algebra. Relational databases allow related records from different tables to be associated without the use of pointers. Relationships are established through common items within the structure of the tables. Values in a column or columns in one table are matched to corresponding values in the column or set of columns in another table. From the second table another set of matching tables will be associated. The linking continues until all the databases have been joined. In order to prevent data redundancy in the relational system due to the

PAGE 68

55 commonality of items in the relations, relational designs follow Codd's theory of normal forms (Healey 1991, 258-259) which specifies that: 1. All tables must contain rows and columns and atomicity (i.e. no repeating groups of data) should be enforced among values within columns. 2. Every column which is not a part of the primary key must be fully dependent on the primary key. 3. Every primary key must be non-transitively dependent on the primary key. Normalization. The join mechanism matches column values between tables using the common item. Normalization of the relational system is based on the principle that a set, as mathematically defined, cannot have duplicate values. Since a table is a set, it cannot have rows whose entire contents are duplicated. In addition, each row must be different from any other. It follows that the values in a single column or a combination of values in multiple columns can be used to define a primary key for the table. No column that is part of a primary key can have null values since this could have the potential for permitting duplicate values. Healey (1991) lists the advantages of the relational databases as follows: 1. Rigorous design based on sound theoretical foundation. 2. All other forms of database structures can be reduced to a set of relational tables, so they are most general form of data representation. 3. Almost unlimited flexibility in forming relationships among data items without the limitation of linkage management. 4. Ease of use and implementation compared to other types of systems. 5. Modifiability which allows new tables and new rows of data within the tables to be added without difficulty.

PAGE 69

56 6. Flexibility in ad hoc data retrieval because of the relational join mechanism and powerful SQL facilities (Healey 1991, 259)." Due to the volume of data that may be associated with a typical Land Information System, Database Management Systems (DBMS) have become integral to LIS. Data modeling techniques, such as entity-relationship model, have become the key element in designing spatial databases. The Relational model has dominated LIS database applications due to the design of commercial software to harness the advantages of the relational construct over older models such as inverted lists, hierarchical files and networked files. The relational database model with the Structured Query Language (SQL) was chosen system for Tanzania. It is recognized at this stage that as the object-oriented technology gains popularity among land information managers, the land administrators might change to the object-oriented system. In the next chapter, attention will be focused on the administrative arrangements which support land management activities in Tanzania. A study of these arrangements as well as data processing procedures will be analyzed in order to develop approaches for modernizing the cadastral records and implementing land information system for Tanzania.

PAGE 70

CHAPTER4 EXISTING CADASTRAL ARRANGEMENTS IN TANZANIA Following the discussion on land management taxonomy in Chapter 2 where typical land administration agencies were identified together with the necessary information support systems a review of current issues in the data capture and management activities were discussed in Chapter 3 This chapter deals with the land management arrangements in Tanzania the responsibilities of the established agencies and their data capture and data management procedures. In this chapter problems with existing land management arrangements and procedures are identified A study of the history of Tanzania reveals that some of the problems associated with the Tanzania land delivery process can be traced to colonial times. A brief review of the evolutionary process of the current land tenure system in Tanzania has been given in Appendix B Approaches are developed in the subsequent chapters to eliminate or reduce the impact of the problems. At the State and regional levels land management administration in Tanzania is handled by the Ministry of Land Housing and Urban Development (MLHUD). At the local level, the responsibility falls on the Ministry of local government. Reorganization of land records in Tanzania required a study of the land management arrangements that have been instituted by the government This study was accomplished through field visits to thirteen of the twenty regional capitals and interviews with land administration officials in 57

PAGE 71

58 both the regional and local government offices. The study focused on the legal instrument and operational guidelines that defined objectives and responsibilities organizational arrangements for meeting those objectives and data capture storage and processing methods record management practices and information dissemination, both internall y and with other agencies. The responsibilities of agencies that operate on a national level were compared with those that operate on a local level, in an effort to identify overlapping or conflicting responsibilities between agencies. The results of this study provided the basis for the land record reorganization approach and data processing methods. In order to understand the land management arrangements in Tanzania, the administrative arrangements within which land management activities are described. Administrative Arrangements within Tanzania Tanzania is divided into 20 administrative regions (see Table 4-1) each region having a capital city. For administrative purposes the city of Dar es Sal~ which is the national capital is also a region by itself So is Dodoma the proposed State capital which will replace Dar es Salaam once the infrastructure development has been completed. Table 4-1: Administrative Regions in Tanzania Dar es Salaam Arusha Mtwara Kigoma Iringa Coast Tanga Lindi Mara Rukwa Morogoro Dodoma Mwanza Shinyanga Ruvuma Moshi Singida Kagera Mbeya Tabora

PAGE 72

59 Each region has offices for regional planning, surveying and mapping, land and title registry. However, all major land management decisions concerning all the regions are made in Dar es Salaam, except Dodoma which by legislation, has autonomy in land management decisions. At the national level, the Ministry of Land, Housing, and Urban Development is responsible for land planning, allocation, title processing and registration, property valuation and assessment, and legal issues in land within Tanzania. These responsibilities are handled by the Urban Development, Surveys and Mapping and Land Development, Valuation and Legal divisions in the Ministry (see Figure 4-1). Sections with specific responsibilities have been described below. Organizational Arrangements within MLHUD The Urban Development Division The Urban Development Division currently operates under the Tanzanian Town and Country Planning Act of 1956, which was revised in 1961. The Division is responsible for defining and planning the use of all public land in Tanzania. Activities ofthis Division include planning redevelopment areas, renewal of blighted urban areas, re-designation of land use, and monitoring of development to ensure compliance with the development program in accordance with master plans of cities. The division is headed by a Director of Urban Development. The Director of Urban Development approves all town and village layout plans prior to implementation. The Urban Development Division has five main

PAGE 73

rban Development Division r Urban Design and Research Master Plans Sites and Services Urban Dev Control Village Land use 60 MINISTRY OF LAND, HOUSING & URBAN DEVELOPMENT Surveying and Ma in Division Cadastral Topographic and Geodetic Village and Hydro graphic Map Production SECTIONS and Development Division Other Dividion s 1 Urban land I Develo ment Rural land Development Land Registry Legal Valuation Figure 4-1: Organizational Chart of MLHUD sections which are the Urban Design and Research, Urban Development Control Master Plan, Village Land Use Plan, and Sites and Services sections. The Urban Design and Research section. This section is headed by a senior principal town planner. The responsibilities ofthis section include planning redevelopment areas (such as the central areas of Dar es Salaam) renewal of blighted urban areas and controlling the layout designs of the city. In addition, staff within this section conduct research and designate high-, medium-, and low-density development areas. Classification of development areas is based on socioeconomic studies of the communities within the

PAGE 74

61 area. The result of the study becomes the basis for setting the size of individual lots within the proposed area. High-density areas consist of small plots for low income people. The parcels are usually 400 square meters in size. Medium-density plots are slightly larger with sizes ranging between 400 and 800 square meters. Low-density classification is for residential areas for high-income people. The plot sizes are between 800 and 1600 square meters. The size of the land allocated for industrial use varies depending on the intended use. The Master Plans Section. This is the section where master plans are prepared. It is the policy of the government of Tanzania, to have five-year development plans for all rapidly growing urban areas in Tanzania. The development plans are graphical layouts showing the allocation of land for various uses in accordance with the rate of expansion of the city. The development plans are referred to as the master plans. The Urban Planning Division schedules the areas to be planned in accordance with available funds however, councils within fast growing municipalities can request priority consideration for physical planning of their community. In such circumstances, the municipal councils provide some of the cost of preparing the master plans Currently, there are up-to-date master plans for all major cities except Dar es Salaam which was last prepared in 1979. The reason for failing to update the master plans for Dar es Salaam is purely it cost which would require a large proportion of the resources of the Division. This implies that the master plans of other cities will not be updated for some time.

PAGE 75

62 The Sites and Services Section. The main activities of the sites and services section are for planning of squatter settlements, ensure that basic infrastructure such as roads and water are available to the residents of those settlements, and that development plans are carried out according to the design. Activities performed by this section are multi disciplinary and not restricted to planning. This section operates on a project-by-project basis. The activities of the section are concentrated in Dar es Salaam. Urban Development Control Section. This section ensures that development agencies adhere to the existing master plans and to ensure that the local councils operate within the development program. The section is responsible for resolving all land conflicts, including those that emanate as a result of planning or allocation. The Urban Development Control section stipulates development conditions and declares areas as ready for urban development. This section ensures that the local councils operate within the guidelines of the development program. The section also supervises the preparation of designs for urban towns. In the regions, the urban town plans are prepared by the municipal councils, but the development control section ensures that the designs conform with the set standards. The section may recommend a change in land use for any particular area. Such changes are recommended to the director and have to be approved by the minister for lands. Village Land Use Section. The Village Land Use Section assists regional planning officers in preparing village land-use plans.

PAGE 76

63 The Survey and Mapping Division The Surveys and Mapping Division provides survey services to government agencies, maintains a geodetic survey control network, and prepares and maintains cadastral and topographic mapping statewide. The Director of Surveys is responsible for coordinating all public sector mapping activities and for maintaining records of all maps, plans and surveys completed by government agencies. As shown in Figure 4-1, the Division has four sections that deal with various aspects of surveying. The sections are cadastral, topographic and geodetic, village and hydrographic, and mapping. There is a survey department in each of the twenty regions in Tanzania. Although the regional surveyors are responsible for the cadastral surveys within their regions, all surveys must be checked in Dar es Salaam before they are accepted. This is a major bottleneck in the data processing procedures within the Division. Surveyors are required to submit their field notes, computation sheets, and a plot of the survey to Dar es Salaam for checking. A six man field computation checking team in Dar es Salaam has the responsibility of checking survey work from the regions. For a timely computation checking process, the regions in Tanzania have been divided into six zones and a technician is responsible for each zone. The Cadastral Section. The cadastral section deals with demarcation of plots in accordance to town planning drawings. These are physical layout plans for a development area. The cadastral section is responsible for the custody of all original field survey records and checking of computations of surveys from the regions. The demarcation layout from the Urban Development division is submitted to the cadastral section on a base map which

PAGE 77

64 may or may not be current. In laying out the demarcation the surveyor has the authority to change the design ifhe or she encounters any obstacles, such as existing houses or roads that conflict with the town planning design. The surveyor lays out the plots as closely as possible to the town planning design, but resolves conflicting issues on site. He or she then surveys the corner monuments to obtain final coordinates. Staff at the computation unit check all control surveys and demarcation surveys for computational and drafting errors and recommend approval or rejection by the Director of Surveys. Deed plans of individual plots are prepared from approved survey plans after the lots have been allocated. The Topographic and Geodetic Section is responsible for all survey projects that are related with national mapping and the establishment and densification of national geodetic control network. The section is further divided into four units: 1. The geodetic surveys unit is responsible for planning, monumentation, surveying documenting, and maintaining national geodetic control points. The national geodetic controls are first order, second order, third order, and precise leveling for national vertical bench marks. The section has not executed any control extensions for many years. Planimetric control densification is currently conducted by cadastral surveyors whenever they need to extend control to a project site. Besides that, no precise leveling has been done in several years. 2. The topographical survey unit is also responsible for all topographic surveys national mapping, and for providing data for the production of cadastral base maps and national maps at various scales.

PAGE 78

65 3. The international boundary surveys unit maintains the national boundary monuments and is also responsible for resolving national boundary disputes. 4. The stores and equipment unit is responsible for logistics. The unit purchases survey equipment, stores and maintains surveying instruments and camping equipment for the topographical and geodetic section. The unit is responsible for keeping stock of stationery, such as field survey forms and other materials needed by other sections for performing their normal tasks. Village and Hydrographic Surveys Section. The village mapping section was established in 1970 in response to the government's desire to institute village governments that would be the nuclei of national planning and development. The Village and Ujamaa Act of 1975 required that village boundaries be known. One of the responsibilities of the village and hydrographic section is to demarcate and establish village boundaries so that title may be issued to the village committees. The section is headed by a senior surveyor who is responsible for the planning, coordinating, and monitoring of the implementation of village mapping projects. Village mapping activities involve a series of seminars with the residents of the village that have has been earmarked for mapping. The rationale of the seminars is to educate the villagers on the intent and purpose of the demarcation survey and for the surveyors to learn about their customary land tenure system in order to ensure that residents are not unduly dispossessed of their property. The village mapping team uses 1 :50,000 base maps and aerial photographs to map out village boundaries. The team visits

PAGE 79

66 the boundary marker (if one exists) or establishes a boundary identifier at the position that is mutually accepted by the representatives of the adjoining villages. The photo interpreters identify the boundary marker in the photograph or its location (if the point is not on the photograph). A number is given to the point and a textual description of its location is made. In the office, a map is produced by photogrammetric methods and submitted to the director of surveys for checking and approval. Map production section. This is the section that is responsible for the cartographic production of thematic, topographic and special purpose maps and atlases. Maps of all categories, with the exception of the 1 :50,000 scale-map series, are produced by the staff in this section. The 1 :50,000 scale-map series are normally contracted to overseas contractors such as the Ordinance Survey of The United Kingdom or Kenting Surveys of Canada. Land Development Division The Land Development Division deals with allocation of parcels, preparation and issuance of titles, valuation and assessment of properties, registration of titles and encumbrances, and resolution of disputes involving ownership. Responsibilities and activities within this Division are regulated by the Land Ordinance (Chapter 113) and the Land Registration Ordinance (Chapter 334) of the laws of Tanzania. These laws declare all lands to be under the control and subject to the disposition of the president.

PAGE 80

67 Some of the powers of the President to administer the land have been passed to the Minister for Lands and subsequently, to the Commissioner for Land. Government Notice 124 of March 22, 1963 extends the power of disposition to land officers. In the office of the commissioner for lands, specific land officers have been assigned the responsibility to allocate land. The Land Development Division has five sections ( see Figure 4-1 ) The division deals with all public land administration matters in Tanzania. The division is headed by a commissioner for lands, whose responsibility it is, to approve all land allocations and to endorse all certificates of occupancy. The commissioner's office maintains records of all land transactions that pass through the Land Development Division. There is therefore, an open registry where records of all transactions relating to lands in Tanzania are kept. There are five main sections in the Land Development Division; urban land development, rural land development, land registry, legal, and valuations sections. Urban Land Development Section. This section handles all matters relating to urban land including; preparation of Certificate of Occupancy in urban areas; processing of land allocations that are made from the commissioner's office; resolution of ownership dispute; and approves transfers, mortgages, and leases that last longer than five years.

PAGE 81

68 Rural Land Development Section The Rural Land De v elopment section has the same responsibilities as the Urban Land Development Section, but with regard to rural areas In addition this section deals with matters concerning the allocation of land for large farms the processing of village titles rights of occupancies in village areas trading areas and subdistricts. The Land Registry is headed by the registrar of titles who is responsible for all the titles that are issued in the country. For land registration purposes Tanzania is divided into six zones. Each zone has a registry that is headed by an assistant registrar Zonal registries are located in Dar es Salaam, Moshi, Mbeya Mwanza, Mtwara and Dodoma (see Table 42) This grouping is inconsistent with the grouping at the Surveys and Mapping Division. The zonal registrars have the mandate to register titles within their zones. The Zone 1 Dar es Salaam Coast Morogoro Table 4-2: Land Registration zones in Tanzania (Zonal headquarters in italics) Zone2 Zone 3 Zone4 Zone 5 Moshi Dodoma Mtwara Mwan z a Arusha Singida Lindi Kagera Tanga Kigoma Mara Shinyanga Tabora Zone 6 Mb ey a Rukwa Ruvuma responsibilities of the registrars has been clearly defined in the Land Registration Ordinance (Chapter 334) of the laws of Tanzania. Three classes of documents can be registered at

PAGE 82

69 the land registry: Certificate of Occupancy, land-related transactions, and collateral. Chapter 334 of the Tanzanian laws lists 33 items that may be registered. Of these, 21 have so far been submitted for registration. Some of the items submitted for registration include acquisition of Right of Occupancy, caveat, registration as legal personal representative, change of ownership, change of name, deed of variation, mortgage, revocation of right of occupancy, and land transfer. After the commissioner has endorsed the certificate of title, the document is sent to the zonal registries for the title to be issued. There are two systems for registering certificates of occupancy--the "old system" and the "new system." Both systems are described later in this chapter, but it will be mentioned at this point that the major differences between the two systems are the numbering system and the filing method. The Legal Section. It is the responsibility of the legal section to review all the land laws of Tanzania, identify loop holes, and conflicting areas of the law. The legal section recommends modifications to the attorney general. The activities of the legal section include the preparation of deeds of variation, deeds of surrender, and deeds of rectification. In addition, the legal section prepares the background information and strategy for defense in all land-related legal actions brought against the Ministry and to coordinate with the attorney general in the defense of the Ministry. One of the legal actions the section handles includes claims for annulment of revocations of certificates of occupancy that were made between 1971 and 1992 that were signed by the minister for lands. The legal standpoint is that the minister did not have the mandate to sign those revocations. According to the law, only the president can revoke an allocation of land.

PAGE 83

70 Other cases include resolution of legal action instituted as a result of double allocation and failure to follow the correct procedure either for revocation or reallocation. The Valuation Section. For first time allocations a land assessment report is required to establish the fee to be paid by the applicant upon acceptance of the offer from the allocating committee. It is the responsibility of the valuation section to assesses the land rent. Actual property valuation is done to establish the value of improvement or develop ments that have been carried out on the land when the values are needed for compensation, mortgage, transfer, or the determination of rental charges on a house. The procedure for assessing rent for first time allocations id described later in this chapter. The City Council of Dar Es Salaam The City Council is a completely independent authority that operates under the Local Government Act (1982) to develop and manage the resources within the city. The council operates under the Minister for Local Government. This enables the City Council to operate independently of the Ministry for Lands, Housing and Urban Development. The Dar es Salaam City Council has jurisdiction over three districts; Ilala, Temeke, and Kinondoni. Each district has a district lands officer. The City Council has its own land surveyors, town planners, valuers, architects, and land officers who operate independent of the Ministry of Land Housing and Urban Development. However, the town planning drawings that are prepared for demarcation have to be approved by the director of urban development before allocation can begin. The City Council maintains and updates its own

PAGE 84

71 set of standard sheets (1 :2500 scale maps), which the council uses as base maps for the preparation of town planning drawings. The Local Government Act ( 1982) authorizes the City Council to establish the requisite administrative divisions to enable the council to function efficiently and effectively. The City Council is guided by the policies of the Ministry ofLands, Housing and Urban Development. Although the functions of the two ministries are clearly defined, there seems to be overlapping responsibilities. The effect is that the two ministries sometimes have different views on the resolution of certain problems that are associated with land administration. Existing Land Delivery Process Land Delivery in Tanzania is done in two stages. The first stage(see Figure 4-2), which is referred to as survey and demarcation, deals with subdivision of the land, physical demarcation and monumentation of the parcel boundaries, and survey of the parcels. briefly, the process begins with a request from the Commissioner for Lands. Layout plans, depicting how the designated area should be subdivided, are prepared by the Urban Development Division. The Surveys and Mapping division is responsible for demarcation, monumentation and survey of the parcels. In places where the land has never been sub divided before, adjudication precedes survey and demarcation, to establish ownership and rights to the use of the property. Upon successful checking and approval process, copies of the subdivision plan are passed to the relevant offices including an allocation committee

PAGE 85

Land Development Division Urban Planning Division Surveys and Mapping Division Distribution I I 7 I I I 72 Prepare Cadastral Plans Update existing Base Mas Figure 4-2: Survey and Demarcation which uses the map to allocate the individual parcels to applicants. The existing base maps are updated with the new information. The second stage involves allocation of the parcels to applicants, titling and registration of the certificate of title (see Figure 4-3). It is the responsibility of the allocation committee to assign the plots to successful applicants. The Land Development Division is responsible for preparing and registering the certificate of occupancy. After allocation has been done, certificates are prepared and sent to the Commissioner for Lands who appends his seal to the title and endorses each certificate for authentication. The

PAGE 86

73 Allocation Committee r--------------1 Land Development Division I I I I I I I I I I Preparaton of Certificate .J Cadastral Plan Attached ,r fide Sealed by Commissioner ,, Registration of Certificate L_ _J ---; ,-------Owner Figure 4-3: Allocatio' Titling and Registration certificate is finally registered by the Registrar of Titles. Copies of the registered documents are then given to the respective owners. Survey and Demarcation The process begins with the town planner preparing the layout (which is referred to as "town planning drawing") in accordance with the development phases of the master plan. This is done using the base map of the area of interest. Normally, the town planners should

PAGE 87

74 request the survey division for a current base map. In most cases the ba s e map s ar e s o old that it is a major task just to update them. When the town planning drawing has been completed the local Development Committee has to approve and accept the layout. Upon acceptance the plan is sent to the director of urban planning for approval. A sepia copy of the approved layout is sent to the region and the original is kept in Dar es Salaam. A print of the town planning drawing is submitted to the Survey Division with a request for demarcation. The task is assigned to a staff surveyor and execution is done in accordance with survey instructions. The survey instructions are normally issued either by the director of surveys or his representative. After demarcation, the layout is surveyed. Field notes and computation results are submitted to the regional surveyor who checks the work and sends the completed results to the director of surveys for approval. The director of surveys checks and approves the survey. Prints of the completed plan are made and distributed to the relevant offices. The field notes become the property of the government. A copy of the town planning drawing is sent to the regional lands officer who requested the survey and demarcation. The print received by the regional lands officer is used by the Allocating Committee to distribute plots to applicants. Allocation and Registration The land officer prepares a list of applicants who qualify to be considered for allocation. There is a committee that has the responsibility for allocating lands in every district and every town. The list is submitted to the committee that allocates the plots to the

PAGE 88

75 applicants In Dar es Sal~ there are three allocating agencies These are the office of the commissioner for lands the urban planning committee and land officers (under Government Notice 124 of March 22 1963). As there are always more applicants than available plots there is a genuine desire on the part of most members of the allocating committees to distribute the land fairly. In most regions and districts allocations are done on a first-in-first out basis It is however difficult to overlook the request of a superior or a politician. Once allocation has been done the regional land officer requests the valuation office for an assessment ofland rent. There are criteria for assessing the rent for first-time allocations. The valuation officer calculates the appropriate rent and prepares a report. The regional valuation officer checks the figures and approves if everything is correct. The land officer prepares a letter of offer that is sent to successful applicants. The letter details the fees that need to be paid and the development conditions for that piece of land. Some of the fees are to be paid at the land office while the other fees have to be paid at the Inland Revenue office. The applicants are supposed to take the evidence of such payments to the Land Officer even though in most regions the Letter of Offer does not mention it. The land officer prepares the Certificate of Occupancy and requests a cadastral plan from the Surveys Division only when the advice of payments have been presented to him The cadastral plan is attached to the Certificate of Occupancy. The applicant signs the Certificate of Occupancy before a land officer, a magistrate or a commissioner of oaths. The document is finally sent to the commissioner for lands (in Dar es Salaam) for checking and endorsement.

PAGE 89

76 At the office of the commissioner for lands the certificate and all the relevant documents are checked. Those that are found to contain errors are sent to the reception desk pending further communication with the regional land development officer who submitted the certificate. Certificates that are free from errors are stamped with the commissioner's seal of approval and sent to the commissioner for signature. The documents are finally returned to the zonal registry office of the region where the plot is situated, after the commissioner for lands has approved the certificate and the statistics section has extracted the necessary information from the documents. The zonal registrar checks for any conflict of ownership, the correctness of the deed plan, and anything that is required under Chapter 334 of the Laws of Tanzania. In the absence of any adverse claims or reasons for objection, the title is registered. Cadastral Surveying Processes within Divisions Surveying and Mapping Division The procedure for executing a cadastral survey in a rural area is slightly different from the method that is used for urban surveys. Surveys in the rural areas are mainly for physical demarcation of village boundaries, whereas urban cadastral survey are for plot demarcation and survey. They both start with survey instructions from the office of the director of surveys and mapping or any person appointed by him. In the regions, the survey instruction may be written by the regional surveyor. The survey instructions contain the

PAGE 90

77 coordinates of existing survey controls that may be used to connect and control the new survey as well as a sketch of the area to be surveyed. Survey of Rural Areas. The survey instructions contain details such as the location and the shape of the land to be surveyed. Adjudication between neighboring communities is done at the time of demarcation to ascertain the mutually accepted position of the village boundaries. Sometimes it is found that the sketch that is attached to the director's instructions is different from what the village communities perceive to be their boundaries. In such circumstances, the boundaries as perceived by the communities are adopted. Until recently, a photogrammetric method was used to demarcate village boundaries. Using a 1 :50,000 scale map and aerial photographs, the boundaries are identified and marked on the photos. Photogrammetric methods are used to coordinate the boundary markers and to produce maps. In situations where conventional theodolite traversing method is used, accuracy requirements for village demarcations are 15 minutes of arc for angular misclosure and 1 :5000 in linear misclosure. Surveys of urban lands are often for physical demarcation and survey of parcels. The request for survey is preceded by an approved town planning drawing. The commissioner for lands or his representative makes a request to the director of surveys and mapping, who in turn, issues survey instructions. The survey instructions are either sent to the Regional Survey Offices or to licensed survey firms. Requests for surveys may also be made in the regions by the regional land development officer to the regional surveyor. The

PAGE 91

78 survey instructions constitute an express authority for the land surveyor to enter upon the land with his or her field assistants to carry out the survey. Procedure for Carrying out Cadastral Survey. The field surveyor studies the layout and identifies the location from an index map. Any existing cadastral control points in or around the neighborhood are identified. Coordinates, reports, descriptions, and any relevant information pertaining to the controls are extracted from files. All preliminary computations needed to commence the survey task are done in advance prior to the actual field work. The field procedure for setting out the parcels involves demarcation of the block comers first. Measurements are made to ensure that the block comers have been located as accurately as required in the survey instruction. A traverse is run to coordinate the block comers. The traverse is computed to ensure that the positions of the points are within acceptable misclosures. In urban areas, the allowable angular misclosure is 30n seconds of arc, where n represents the number of stations. Allowable linear misclosure is 1 :6000. Traverse adjustment is by the Bowditch method. Setting out individual plots in a highand medium-density area is done by extending a steel tape horizontally straight along the line between two block comer points and driving iron pins into the ground at the points where plot comers should be. Concrete mortar is poured around the pins to make them more permanent. In the case of low-density plots, the sides are measured rigorously and corrected for slope, temperature, and where applicable, sag correction. Plot comers are marked with concrete monuments and numbered sequentially. Such numbers are also shown on the

PAGE 92

79 cadastral plans. In the office, the field notes are checked by re-computing all the data to make sure that there are no errors. The surveyed plots are finally drafted. The original field notes, computation sheets, and plan are sent to the director of surveys and mapping for further checking and subsequent acceptance. A job that meets the required standards is accepted and approved by the director. Field notes and all accompanying documents become the property of the government. In most cases where the survey instructions have been followed and proper survey procedures applied, the resulting survey has been within acceptable tolerances and the job has been accepted. Occasionally, the survey work has been rejected. In such situations, the director of surveys and mapping requests that the work be done again. Rejection of Cadastral Survey Job. Some of the reasons for rejecting a survey include: failure to follow the survey instructions failure to conform with the layout as shown on the town planning drawing without adequate reasons; failure to comply with accuracy requirements, either as specified in the survey instructions, the Surveyors Regulations, or technical circulars; survey work extending over or overlapping an existing survey; or non-maintenance of road parallelism.

PAGE 93

80 Relative accuracies of the cadastral control points. Different methodologies have been applied at different times in different parts of the country during survey control densification. Four different approaches have been used to establish survey controls; triangulation method, the use of theodolites and Electronic Distance Measuring (EDM) instruments, theodolites and invar tapes, and lately, global positioning system (GPS) methods. Each of the methods has inherent levels of accuracy. Surveyors in both the public and private sectors expressed the need to readjust the survey control network and obtain unified coordinate values for all the control points as well as established levels of accuracy between the different measuring processes. New settlements usually start from the edge of the road and extended inland, away from the road. In demarcating plots, the procedure has been to use controls points from the edge of the road and close on other points along the same road. The controls that were established during one cadastral survey were used as starting and closing controls for subsequent surveys. This has been the procedure for extending cadastral controls especially in the urban areas. By doing that, the errors in one cadastral survey got carried over to future surveys. Over the years, different misclosure levels have increased between different town subdivision blocks along different roads. In the current situation, it is not recommended to start a survey project from controls derived from a particular road and close on controls that have been established from a different road, due to the possibility of having unacceptable misclosures. In Dar es Salaam, misclosures of about two meters have been observed in

PAGE 94

81 certain areas. This may translate to a fraction of a millimeter on a 1 :2500 base map, but unacceptable for cadastral survey by the Director of Survey. Land survey and demarcation within Dar es Salaam. The procedure for demarcating and allocating land begins with a request for a planning scheme by the city planner. A request is made for a base map (1 :2500 map) of the area to be planned from the surveys section of the council. The base map is updated by the town planners using steel tape and compass. The proposed layout is done on the updated base map. This becomes the town planning drawing. After the town planning drawing has been prepared, the Urban Planning Committee reviews the design and recommends whatever changes that the committee might consider necessary. The design is adopted when both the Urban Planning Committee and the city planner's office are satisfied. The drawing is then sent to the Director of Urban Development for approval. At the director's office, the drawing is taken through the normal checking procedures, including ensuring that the design conforms with the land use scheme as shown in the master plan of the area. Once the design has been accepted as conforming with the master plan, the drawing is approved by the director on behalf of the Minister for Lands and returned to the City Council for implementation. The land surveyors at the City Council demarcate the land according to the design. Survey field sheets, computation sheets, and a list of the coordinates are sent to the Director of surveys and mapping for checking. The office of the director of surveys checks the

PAGE 95

82 quality of the reference controls that were used for the survey the method of survey the computations, and the misclosures of the traverses. If the results are within the required standards of accuracy, then individual plot num bers are assigned. Otherwise, the drawing is returned to the city land surveyor's office for a resurvey. Having assigned unique numbers to all the individual lots, copies of the plan are distributed to various departments, including the city planner, city land surveyor, and the city lands officer. The city land surveyor's copy is used to update the 1 :2500 standard sheets that are maintained in the office. Another copy is archived for future reference. The copy that goes to the city lands officer is used to allocate the plots to applicants. Land allocation. In Dar es Salaam, allocation of plots is done by the Urban Planning Committee. The copy of the subdivision plan that is sent to the Commissioner for lands is meant to serve, among other things, a notice to the fact that the parcels have already been allocated. Sometimes the notice gets delayed to the extent that the commissioner's office inadvertently allocates that same plot to other applicants. Genuine mistakes of this kind are easily noticed when the owners commence registration of the certificate of occupancy. Requests for land under Certificate of Occupancy may be submitted to the district land officer or the city lands officer. Requests received by the district lands officer are forwarded to the city lands officer. All requests are compiled by the city lands officer and submitted to the Urban Planning Committee. The Planning Committee meets to allocate the plots to the applicants. Allocation is supposed to be done in the order in which they are

PAGE 96

83 received (i.e., on a first come, first served basis). After the committee has made the allocations, the city lands officer requests an assessment of each of the newly allocated parcels. Upon receipt of the assessment information, the city lands officer prepares letters of offer and sends copies to each of the applicants whose request was approved. The letter of offer contains the terms of the offer, the fees that need to be paid, and the time within which to pay the required amounts. Each letter of offer is prepared in quadruplicate. The original goes to the applicant and copies sent to the commissioner for lands and the city lands officer. The last copy is kept in a file at the District Land Office where the letter was prepared. The recipient normally has 30 days from the date of the letter to accept the offer and pay the relevant fees. After accepting the offer, the recipient also has three years to develop the plot in accordance with the intended use (e.g., by constructing a house on it). Property Valuation and Rent Assessment As mentioned earlier, it is the responsibility of the valuation section to assess the land rent. Assessment is done after an offer has been accepted. The land officer submits a request to the valuer for land rent assessment for each parcel of land before the letters of offer are sent to the applicants. As shown in Figure 4-4, the request for valuation is submitted in the form of a letter to the chief valuer who records the description of the property, the plot number, location of the property, the purpose of the valuation, and the date of the application. The chief valuer then passes the file to a valuation officer for action.

PAGE 97

START CHIEF VALUER Receives Request 84 ~~VALUATION OFFICER Creates a job file RECORDS OFFICE Assigns a file number ALUATION OFFICE.l"41-------..... Assesses the property CHIEF V ALDER Checks the Report Yes Copy is kept on file Client receives the Original Report END Figure 4-4: Schematic Diagram of the Procedure for Assessing Property The valuation officer asks the registry section to create a file for the job. A valuation number is then assigned. The file is returned to the valuation officer once a file number has been assigned. The valuation officer then schedules a visit to the site of the property and collects the necessary information. Subsequently, a report is written in which a value is assigned to the property. The report is submitted to the chief valuer for checking and approval. Unless there are

PAGE 98

85 obvious mistakes requiring a return to the valuation officer for amendments, the chief valuer approves the valuation report. The fees for valuation are based on the assessed value of the property and the direct costs of trips to the site. Upon payment of the required fees, the report is given to the client and a copy is kept at the lands records office. Certificate of Occupancy The request for a Certificate of Occupancy may be presented by an individual, the regional land officer (in the case of requests from the regions) or the city land officer. This happens after the allocating committee has offered the plot of land to the applicant and the applicant has accepted the offer. There is no way of knowing if the offer has been declined by the applicant. As a result of this flaw, there is no way of keeping track of the allocations that have been accepted and those that have been declined. The only way to find out about an offer that was not accepted is through the process of official search or if an interested party keeps track of the allocation. As shown in Figure 4-5, the approval process for tha certificate of occupancy begins at the reception counter. The documents are first submitted to a receptionist at the reception counter who stamps a date to indicate when the documents were first received. The document is then sent to the open registry for a land division number. At the open registry, the documents are put in a folder. The folder is then sent to another office within the open registry (indexing room) where a unique land division number is issued to the folder. The plot number, block number, location description, and name of the owner are

PAGE 99

No RECEPTION COUNTER OPEN REGISTRY PRINTING ROOM OPEN REGISTRY SCHEDULE OFFICER Write to the owner or the District Land Officer or the City Land Officer for clarification The document is returned to reception desk for ftling. COMMISSIONER APPROVES STATISTICS 86 Yes STATISTICS Apply the Commissioner's seal COMMISSIONER No No ZONAL REGISTRATION OFFICERS ZONE 1 I .__I _zo_NE_2___,II .__ _zo_NE_3___,I l._____zo_NE_4___,I I ZONE 5 I L ZONE 6 I Figure 4-5: Land Office Procedure for Issuing a Certificate of Occupancy

PAGE 100

87 recorded on an index card that is kept in the indexing room. The folder containing the document is sent to the printing section where the land division number and description are printed on the cover of the folder. The folder is returned to the open registry again and given to the respective schedule officer. The folder is placed among other folders from the same region that are waiting to be processed by the schedule officer who is responsible for documents from that particular region. The schedule officer checks for ownership, letter of offer, advise of payment, evidence of payment of fees, correctness of the certificate, typography, and other relevant letters and documents are included. In the case of fanns, a check is made whether the respective land allocation committees from the village level to the regional level have all consented. If all the items are correct, then the schedule officer sends the folder to the statistics section. Otherwise the schedule officer writes a letter to the individual or the land officer who submitted the document and requests more information or explanation as to why the documents were submitted in that manner. Having written the letter, the schedule officer returns the folder to the receiving desk to be filed until a reply to the letter is received. At the statistics section, the folder is checked again for any errors that the schedule officer may have missed. If a mistake or omission is found, the folder is returned to the schedule officer. Assuming that everything is correct, the technician at the statistics office takes the folder to the commissioner's office. The commissioner's seal is put on the document and the folder is left in the commissioner's office where the commissioner checks the document for any errors that the schedule officer and the technician from the statistics

PAGE 101

88 section may have missed. If any errors are found the folder is returned the statistics office for forwarding to the respective land officer for corrections to be made. If the document passes the commissioner s scrutiny then the commissioner appends his signature. The folder is returned to the statistics section. At this stage a copy of the document is retained at the land records office Documents approved by the commissioner and returned to the statistics section are sorted according to zones. The twenty regions are grouped in six zones. The document is finally placed among those that belong to the same zone, either to be collected by the zonal registrar or be dispatched through official channels. Registration of Certificate of Occupancy Documents for registration may be submitted at the registration counter of the land registry, where a document begins its journey through the processes shown in Figure 4-6 The documents are stamped as having been received at the counter. An entry is made in the counter book as well to indicate that the document was received. The counter is closed daily at 13.00 hrs. to give the counter clerks time to record all the documents received that day. The documents are deposited at the office of the registrar of titles before the end of each working day. The registrar reviews each document the previous day and determines what needs to be done on each document. The documents are then distributed to the respective officers who need to work on the documents. Those to be registered are given to the staff at the

PAGE 102

RECOMMEND REFUSAL REGISTRAR REFUSES WITH REASONS CERTIFICATE OF OCCUPANCY UNIT COMMISSIONER FOR NECESSARY ACTION START REGISTRATION COUNTER REGISTRAR PERUSES CERTIFICATE OF OCCUPANCY UNIT NO YES ISSUE TITLE 89 DRAFTING SECTION INDICATE AS SUCH NO STAMP DOCUMENT AND INDICATEDATE& TIME REGISTRAR ENDORSES ERTIFICATE OF OCCUPANCY YES CERTIFICATE OF OCCUPANCY UNIT NO INDICATE AS SUCH RECOMMEND APPROVAL REGISTRAR APPROVES CERTIFICATE OF OCCUPANCY UNIT ORIGINAL IS INDEXED -------l 1----111 UNIT APPLICANT RECEIVES COPY BIND ACCORDING TO THE NUMBERING SYSTEM AND THE REGION WHERE T ...,,__ YE S ---1~ PARCEL OF LAND IS SITUATED -Ii----------, END FILE THE DOCUMENT IN THE GREEN ..._ ___ N o ___ _. 1 (PROPERTY) FILE ALL TRANSACTIONS 1----------' AND DEEDS ARE KEPT IN nns FILE Figure 4-6 : Procedure for Registering a Certificate of Occupancy

PAGE 103

90 certificate of occupancy unit. The staff at the certificate of occupancy unit checks that the following procedures are complete: The stamp duty has been paid and that the correct amount was paid. The Registration fee has been paid. A unique land office number has been issued. The document has been signed by the commissioner. A land records property index card does not already exists for the property. The name of the owner has been written out in full (no initials accepted). From the Certificate of Occupancy unit, the document is passed to the drafting section for the cadastral plan to be checked for errors and omissions. The observations of the drafter (whether the cadastral plan contained errors or not) are indicated on the document and returned to the certificate of occupancy unit. Based on the results of the checks by the staff at the drafting section and those at the certificate of occupancy unit, a recommendation is made to the registrar indicating whether the document should be approved for registration or not. If documents are then returned to the registrar with recommendations for refusal, that recommendation is put in writing together with reasons. The document is the returned to the certificate of occupancy unit. The certificate of occupancy unit passes the document to the commissioner's office along with the registrar's comment. At this stage, the commissioner decides whether to correct the errors or to return the document to the source for corrections to be made. After the corrections have been made, the document then has to got through the process all over

PAGE 104

91 agam. In order for documents that have been approved by the certificate of occupancy unit to be registered, the recommendation must be passed to the registrar and he gives the approval for the process of registration to begin. The document is then returned to the certificate of occupancy unit where the actual registration process begins. At this stage, if the document is for the old system of registration, then a unique title number is issued at the certificate of occupancy unit. All documents, whether under the new system or the old system, are stamped The date and time of registration are indicated on the document. The document is then sent to the registrar's office for endorsement. After endorsement, the document is returned to the certificate of occupancy unit. At the Certificate of Occupancy unit, a copy of the document is given to the owner and the original is indexed. What happens to the original document next depends on whether the document is to be registered under the "old system" or the "new system." The old system. Under the old system, the details of the Certificate of Occupancy are entered into a Land Registration book. The book has columns for 'counter' date and time, date of registration, and title number, district, parcel description, and other columns. Documents filed under this system are numbered sequentially in accordance with the numbering system of the region in which the document parcel ofland belongs, e.g. 43803, 43804, 43805, etc. The title number is sequential within the zonal area of the Registry. Other Registries may have their own numbering sequence. It is possible therefore, to have title number 43 9 l 7MTW in Mtwara and 43917:MBY in Mbeya.

PAGE 105

92 The title number and the date and time when the documents were first presented are stamped on the back cover of the folder. The documents (in duplicate) are sent to the Registrar (for the third time) and the Registrar finally signs them. The documents are then returned to the Certificate of Occupancy unit where an index card is prepared for the certificate. The original copy together with all the supporting documents are filed away and the duplicate copy is given to the owner. The certificate of Occupancy is bound together with others from the sane region in numerical order according to the title. The new system. The difference between the 'old' and the 'new' systems is that Chapter 334, section 30 of the Tanzanian laws stipulates the preparation of a register folios Identical records are kept in either system, however, the documents are sent to the Registrar together with the folio. The registrar has to sign the folio too. The signed documents are sent to the Certificate of Occupancy unit where index cards are prepared. The duplicate is given to the owner and the originals are filed in the following manner. The Certificate of Occupancy together with the cadastral plan goes into a parcel file known as a "green" file. The certificates are filed in numerical order according to the title number (e.g. 186001/15, 186001/16, 186001/17, etc). The register folio is put into a binder along with folios in the sane block. The binders are filed in sequential order. All correspondence concerning the particular title are also filed by the title number in a correspondence file. This process makes it easy to locate material pertaining to those parcels stored under the "new" system. An analysis on this study is given in Chapter 5. Organizational problems and concerns that were noticed in the divisions within MLHUD have been listed in Appendix C.

PAGE 106

93 Following the discussion of the problems and concerns, approaches to reorganize the cadastral records within the Ministry are presented.

PAGE 107

CHAPTERS CADASTRAL INFORMATION SYSTEMS APPROACH Following the review of the land delivery arrangements in Tanzania an analysis o f the problems concerns and observations that were raised during the study is made in this chapter. Problems and concerns with the existing arrangements have been listed in Appendix C. However recommendations for remedial actions and solutions to the problems are addressed in this chapter so that solutions could be included in the pilot study. Although some of the problems and concerns can only be addressed through institutional reform, most of the administrative bottlenecks can be removed by streamlining the document handling processes. Those problems that can only be addressed through legislation are mentioned here for completeness, since modification of legal instruments is the responsibility of the cabinet of the government of Tanzania. Problems that can be resolved through a modification of the document handling and data processing methods are addressed here and tested during the pilot project which is described in Chapter 7. The legal cadastre is still the primary system for maintaining land registration in Tanzania. The type of data that are recorded the completeness of the register access and availability of such information to other government agencies businesses in the private sector, as well as the public in general are all factors that can influence the benefits that can be derived from the cadastral information. The object of the research has been to prepare 94

PAGE 108

95 the infrastructure to support the perceived increase in land-related transactions following the adoption of the land market policy In order to develop a comprehensive cadastral information system, which is the ultimate goal of the government of Tanzania, the cadastral records have to be complete accurate and up-to-date. As such, approaches are developed, in this chapter for removing obsolete and inconsistent records from existing data and reorganizing the records in a manner that they can be exported into a computerized database. Procedures are then developed for adding new records to the database and finally an approach is developed for implementing a cadastral information system. Problems and concerns need to be addressed so as to ensure an efficient and smooth running system. Important issues that were raised during the study of the Tanzania land delivery process have been summarized as follows: The base maps which are referred to as standard sheets are not updated regularly. There are not enough cadastral survey control points within the urban centers. Also any survey controls that were established as control extension should be submitted to the director of surveys for approval before they are used for subsequent survey work. National survey controls have been established at different times with different equipment and different methodologies, the result being that while internal con sistencies of certain controls are within acceptable tolerances, the absolute position is not very consistent.

PAGE 109

96 Manual computation of survey data and manual survey drafting processes are time consuming and prone to human errors. Because infrastructure is not provided before allocation, bulldozer operators pass their vehicles through any convenient route to get to the parcel of interest destroying boundary markers in the process. Multiple allocation of the same parcel has occurred due to poor communication between allocating agencies and improper record keeping and checking mechanisms. There are areas within the Local Government Act (Chapter 113 of the Tanzania laws) and the Town and Country Planning Act (Chapter 378 of the Tanzania laws) that either conflict with each other or assign the same responsibilities to the Urban Planning Division and the City Council of Dar es Salaam. It takes too long for surveyed plots to get allocated. Current method for determining who gets allocated a plot leads to corruption. It takes too long to prepare the individual Cadastral Plans. Current procedure for allocating plots for low-density areas, commercial, large agricultural and large industrial projects interferes with the normal process for land administration by the local authorities. People neglect to pursue the process to obtain title. No mechanism for checking and enforcing the 30-day period within which people can accept or decline the offer of a plot.

PAGE 110

97 The three-years development condition in which land, once allocated, must be developed is not enforced. Revocation of certificates takes too long to process. It takes too long to locate documents in the Open Registry. Discussion of Improvements to Current Cadastral Arrangements In order to effectively reorganize cadastral information to serve the objectives of the government and other interested agencies, a review of the administrative arrangements to support cadastral records compilation was conducted. The review included the responsibilities of the agencies within the operational components (see Figure 2-1), the extent and effectiveness of the mandate which has been given to the agencies by the legislation under which they operate. The agencies in this case are the surveys and mapping division, the land office, the land registry, the valuation and assessment office and the revenue office. In Tanzania, the revenue office operates within the ministry of inland revenue, which is different from the Ministry of Lands, Housing and Urban Development. The study of institutional arrangements included efforts by the government to encourage registration by reducing the distance that people have to travel to registration offices, as well as the costs involved such as transfer fees, stamp duty, and transfer tax that people have to pay when transferring or registering their property were also noted. Technical considerations included appropriateness of innovative graphical data capture technologies such as GPS, Remote Sensing and Soft-copy photogrammetric methods as opposed to survey control densification followed by traditional land survey methods.

PAGE 111

98 It was important to recognize the concerns of the user community too. Concerns of the user community included the time it takes to execute a survey, the convoluted registration process, and the length of time it takes to obtain a title. In proposing an approach for reorganizing the cadastral records this research focused on approaches that will improve: cadastral management flow line by reviewing procedures for survey data processing, land delivery procedures, land registration, valuation and assessment land registration flow line data processing linkages between land allocation, titling, and registration so that a smooth flow is ensured data entry and checking procedures the entire land survey, land allocation, titling, registration, and compilation of cadastral information procedures to identify areas where delays are created so as to remove those bottlenecks in the new system. Other considerations included the introduction of computerized data archiving within the system, and standardization of the data format within the system so as to ensure data integration between interested parties. Operational needs of the staff were assessed, which was followed by recommendations that were aimed at removing administrative bottlenecks and eliminating or minimizing technical problems. This assessment was done according to divisions within the Ministry of Land, Housing and Urban Development and are described below.

PAGE 112

99 Urban Development Division The responsibilities of the Urban Development Division have been described in Chapter 4. The responsibilities of this division complement those of the local councils by providing the guidelines for local councils to manage their resources and ensure proper land administration. While studying the activities of the Urban Development divisio' it became apparent that the Urban Development division has assumed a greater role than the legislation provides for, since it is the responsibility of the City Council to ensure that basic infrastructure such as roads and water are available to residents of the squatter settlements. It appears the responsibilities of the two ministries overlap with respect to control of development in Dar es Salaam in particular. This causes confusion in the execution of responsibilities by both agencies and sometimes results in misunderstandings over choice actions in land administration. In preparing the layout for areas within the master pl~ the existing procedure is to send the proposed layout to the director of urban planning for approval. The layout is checked to ensure that the land is being used for the purpose that has been indicated in the master plan. For areas not included in the master p~ the director of urban development demands that layouts be submitted for approval even though the Town and Country Planning Act does not stipulate so. While this requirement may be a mechanism for controlling the planning of areas without master plans, there really is nothing to check if the planning committee for that particular jurisdiction has accepted the layout. The delay caused by the checking procedure in Dar es Salaam would be removed if the regional town

PAGE 113

100 planning officers are given the responsibility for approving the layout of areas within the unplanned areas. Surveys and Mapping Division It was mentioned in Chapter 4 that land surveying activities are regulated by National Council of Professional Surveyors which is a statutory body The study elucidated the fact that membership requirements of the council do not include any knowledge of the advances in the surveying technology beyond the use of a transit theodolite and a steel tape, although survey education in Tanzania includes subjects such as photogrammetry GIS, remote sensing, computerized mapping methods and GPS technology. It seems that the objectives of the National Council of Professional Surveyors have not kept pace with the trends in the changing phases of the surveying profession. As such, the council has not modified the survey practice to include these evolving areas of specialization. The National Council of Professional Surveyors should encourage students to specialize in any of the evolving areas and become members of the professional body by modifying the membership requirements to include knowledge and experience in modem surveying techniques. The Survey Regulations. The staff at the Surveys and Mapping Division, as well as private surveyors in Tanzania, are still operating with survey regulations that were written in 1957. The regulation stipulates positional accuracies for surveys conducted in the urban

PAGE 114

101 areas and for rural areas. The stipulation places emphasis on survey methods rather than providing solutions to problems. With the use of digital graphical data in Geographic Information Systems and Land Information Systems, it is important to ensure data integrity by including metadata standards for digital survey data. Quality of the survey controls. Good quality survey control points are essential for any form of cadastral information system. Over the years, cadastral surveys have been done for small areas at a particular time. In dealing with a cadastral information system, the volume of data for the system can be rather large. It is important to ensure that parcels that are contiguous on the ground remain contiguous both physically and mathematically. To achieve this, the surveys have to be tied to some unifying spatial framework. However, the need for survey control densification is obviated by the fact that GPS technology has reduced the need to have a network of accurately surveyed controls. As GPS technology is not commonly used in cadastral surveying activities in Tanzania, it is recommended that survey control points be established by GPS methods as an aid to continued use of existing technology as well as a transition to large scale application of GPS technology. The idea of having survey offices at the regions is to reduce the work load at the Dar es Salaam office. However, the stipulation for the director of surveys to approve and endorse all survey control points before tying new surveys to them defeats the entire objective for establishing the regional offices since the approval process takes time. It would be better for the control survey work to be checked by the regional surveyor. It is recommended that instead of having the director of surveys check, approve, and certify all

PAGE 115

102 survey controls, a survey control densification unit be established at the regions who will be responsible for establishing survey controls within the regions. The team should be equipped with transportation, global positioning system units, total stations, a laptop computer, and all the necessary accessories. The teams will consist of qualified and experienced surveyors who will be able to establish the controls, perform all the necessary field checks and computations, and certify the controls upon the request of the regional surveyors. The results may be sent to the director of surveys for his records. For the working environment of the Surveys and Mapping Division, it is recommended that: the Survey and Mapping Division invests in survey equipment for all the regions, preferably modem ones; adapt the survey practice to modem technology and define accuracy requirements with regard to economic viability of the survey rather than absolute numbers; store current maps, plans, and survey field data in atmospherically controlled environment; microfilm all old documents to make space for new ones; adopt a parcel referencing and indexing system that can be easily computerized, easy to implement, and unique for each parcel; and initiate a process to convert the standard sheets and base maps which are in paper format into digital maps.

PAGE 116

103 In a fully computerized records management system, electronic scanning of the old documents would be a better option than microfilming. However due mitigating factors such as to the amount of computer memory required to store all the scanned documents the speed of computer processors in 1995 the low level of computer knowledge in Tanzania, and the lack of locally available computer hardware maintenance skills it was more appropriate at that time to recommend microfilming of the documents. Land Development Division Most of the administrative problems associated with the land delivery process are in the Land Development Division (see appendix C). Duplicate allocation of parcels and poor record keeping were the biggest problems in this division of the Ministry It was obvious during the entire study that one of the biggest problem that the government faced was insufficient surveyed land to meet the demand for allocation. As long as the demand for surveyed parcel exceeds supply p eople who have the political power money or some influence would attempt to circumvent the established allocation procedure For this reason, stringent methods for controlling the allocation and registration process should be adopted A first step would be a revocation of the Government Notice 124 of March 22 1963. In addition to this revocation, the authority to allocate parcels should be given to a single committee such as the Urban Planning Committee or the District Land Allocating Committee. This will minimize, if not eliminate, incidents of duplicate allocation of the same parcel.

PAGE 117

104 The procedure for preparing and approving a Certificate of Occupancy has been described in Chapter 4 with Figure 4-5. In preparing the Letter of Offer, it is the responsibility of the Land Officer to specify the fees and the amounts that the applicant should pay. Upon payment, the advice of payment is taken to the Land Officer before the Certificate of Occupancy is prepared. The responsibility for ensuring the payment of the correct fees should belong to the land officer. When the document goes to the commissioner's office, the document is checked by a schedule officer, who is also a land officer. After that, the document goes to the statistics office to be checked. After all the checking processes through which the document passes, there is no reason to find mistakes such as underpayment of fees. Yet, the registrar's office has returned several documents due to reasons which could have been found at any of the four check points. It is the view that the statistics office is not performing any real function other than counting the documents that go to the commissioner for signature. The functions of the statistics office can be performed by the staff at the reception counter. The following recommendations are presented so as to remove bottlenecks in the document processing and record management procedures: The land officers at both the local councils and the commissioner's office should be responsible for ensuring that the correct amount has been paid to the revenue office. There should be responsible officers (Desk Officers), to whom the schedule officers report, who will be responsible for checking the work of the

PAGE 118

105 schedule officers before they are submitted to the commissioner for signature The seal could be stamped by the Desk Officer to whom schedule officers report. The statistics office should be removed from the certification process. After the commissioner has signed the certificate a responsible officer (for example, the Desk Officer) should take the documents to the registrar's office where he/she will go through the documents with a responsible officer at the registrar's office. Those documents that do not contain obvious errors will be accepted. The others will be taken back to the schedule officer who worked on the document. A mailing and/or holding room should be created where documents that have been approved will be sent for despatching to the regions. This may be a temporary process until either the recommendation to establish a Regional Land Administration Officer is adopted or the responsibilities of the regional land development officer have been extended to include the approval of certificates. If this responsibility is extended then documents will not have to be sent to the headquarters to be approved. Old documents should be microfilmed to make room for new ones. Important documents such as property files should be stored in an air conditioned room. Documents should be kept in file cabinets.

PAGE 119

106 Checking and documentation of Certificates of Offer should be computerized. Records should be organized in a manner that they could be found when they are needed. A senior staff person should be responsible for the Open Registry and control and movement of documents and files form the Open Registry The officer should maintain a register for tracking the movement of files and documents. The Registry of Titles. The operations at the land registry where titles are registered, are extremely well organized. The land registry is know by other names as title office and title registry. For this research, therefore the use of such names refer to the same office. This is the only section where documents are easily found. A proper checking and recording system is maintained. Any problems in the registry section have evolved simply because the number of documents have increased to extent where they cannot be main tained easily by manual methods. With some modernization and computerization, the records will be easier to maintain. Recommendations for facilitating the registration process and records management are presented here as follows : The reception counter should be headed by a responsible member of staff who will be able to do a preliminary check on the documents to ensure that all the obvious mistakes such as signatures correct amount of fees and

PAGE 120

107 other mistakes are corrected before they are accepted. This will avoid delays . m processmg Only documents that pass the preliminary check should be stamped with date and time. In the future this activity can also be computerized to eliminate human intervention The checking process should be simplified by using a standard letter containing boxes which will be checked for missing items. The reception counter should be equipped with computers so that official searches will be done quicker. A computer would avoid the registration or unintended duplication of ownership to a parcel. Such errors would be noticed at the entry point. The reception counter should be partitioned to create a waiting room for people who are waiting for assistance. A queuing system should also be adopted to ensure an orderly process which avoids several people clamoring for the attention of the officers. For the same reasons why electronic scanning of documents was not the most appropriate option at this moment, old land related documents should be microfilmed to make room for new ones. A separate computer should be assigned for indexing all documents that are registered On a regular basis the information from that computer should be used to check what is actually in the records office. This will eliminate errors.

PAGE 121

108 Cadastral maps should be converted into electronic format either by digitizing or processing the data in some coordinate geometry software for ease of reference. Document Processing Improvements An approach to limit human intervention, avoid duplication, and to ensure data integrity by introducing checks at critical stages during document processing is discussed here. Reduction in duplication of efforts and bottlenecks in document processing are also considered in this sub section. For example, entries for location, block number and the parcel numbers, which are first made at the Surveys and Mapping division do not have to be repeated at the title and land registry offices if the records are communicated to those offices that need to record them. This is more applicable to new documents that are presented for processing. An approach for reorganizing existing records as well as new documents with special attention to surveying, allocation, titling, and registration are discussed. A streamlined procedure is described which eliminates the convoluted approach which existed prior to this research. Cognizant of these requirements, a computerized cadastral data management system is presented in Figure 5-1. This figure represents the central focus of the computerized land records managements system which would provide cadastral information support for the land delivery arrangements in Tanzania. The figure depends for its efficiency and effectiveness on the ability to incorporate the recommendations and improvements that have been presented in this chapter.

PAGE 122

109 r-. -. . r:-------. Data acquisition Cadastral survey Textual data t---.i Check/Update cadastral data I Hardware and I Software I support chival systems I '-------....1 I n------__ ........ ___ 1 Record update Off line ..,___.M and storage access I I I I I I ...... l-----iJ --___ J Real time access Graphical data i ~---------------------------------------,nformation dissemination User access points Information Processing and Dissemination '-----------------------Figure 5-1: Computerized Cadastral Data Management System In this figure, the need for update or addition of a record will be instigated by the presentation of a certificate of occupancy to be processed. If the transfer involves a change in the spatial extent of the parcel, such as a subdivision or amalgamation, then a surveyed cadastral plan is required. The graphical data component deals with the survey and checking

PAGE 123

110 of the cadastral plan. The cadastral database is updated after the information on the cadastral plan has been checked. Change of ownership or and other descriptive information is made after the cadastral plan has been updated. Both the graphical and textual information are recorded into the hardware support component which includes hardware for data storage and archival, and software for data processing. The hardware support component is also used for storing historic data. For integration into other land-related information systems, the cadastral information system needs to be designed to provide both online and oflline data access capabilities. Public access is also possible through terminals linked to real time access nodes. Data capture and processing procedures in Tanzania have been discussed in Chapter 4. Following this discussion on problems and concerns, practical procedures for improved document processing are presented in the next section. Reorganization of Existing Records An important criterion for implementing a functional Information System to support land administration activities is for the records to be complete. Whereas new records can be incorporated into the system during data processing, the importance of existing records for comprehensive querying and analysis should not be ignored. A procedure incorporating existing records into the system should be adopted. Some of the possible problems with existing land records in Tanzania are multiple ownership of the same plot (duplicate allocation), allocated but unregistered plots, transfer of ownership without proper documentation and plots with outdated ownership information such as owner's address. It is necessary to identify problems associated with current practices and to define procedures to

PAGE 124

111 eliminate or minimize them. With regard to the land records it is necessary to simplify the data capture maintenance storage and retrieval processes without compromising accuracy or integrity of the data and to organize the records in a manner that will ease the flow of information between various divisions within related agencies. A procedure for re-organizing and purging existing records of erroneous data is described below with reference to a diagram in Figure 5-2. Cadastral maps and plans should be digitized and assigned identifiers (subdivision numbers block numbers and parcel identifiers) producing a graphical data layer. The procedure for producing this graphical cadastral data follows the topological structure which is discussed later in this chapter. It is important to ensure that parcel identifiers are not duplicated. 1. Data from the land registry (also known as the title office) should be entered into the database. Internal consistency should be maintained to ensure that errors are not introduced. To avoid multiple registration of the same parcel the database should be queried using descriptive location (from the textual database) block number and parcel identifier to identify parcels that have more that one legal owner except homogeneous groups of people such as families communities or organizations. Any plots identified this way would have to be isolated so that an adjudicating team may investigate and establish the legal owner. In Tanzania, where the state guarantees title, the aggrieved owner may have to be compensated for the loss. 11. Data from the land office should be entered into the database, making sure that errors are not introduced into the records. Detailed description of the data types in any

PAGE 125

112 particular record are given later (in Table 6-2). Again, the database should be queried using location as identified in the textual database district block number and parcel omputerize Land egistry records Relate databases and query for discrepancies No Ye igitize graphical adastral plans Computerize Land Office records L--------M Relate databases and query for Isolate for Adjudication Relate databases and query for discrepancies Partially purged records Systematic verification and updating of records discrepancies Isolate for Adjudication No Figure 5-2: Procedure for Isolating Inconsistencies Among Existing Land Records

PAGE 126

113 identifier to ensure that the same parcel has not been legally allocated to more than one perso~ family or organization. Any plots that violate this condition should be isolated so that the legal title can be established. m. Relating the computerized Land Office records with the computerized graphical data, query the databases to ensure that each record in the Land Office database has been referenced to a parcel in the graphical cadastral layer. Two possible errors may arise from this query; unallocated parcels and certificates without valid reference to the cadastral map. Since all parcels must be assigned by the allocation committee in accordance with the cadastral pl~ it may be that those parcels were either not allocated or that the allocations were declined by applicants who failed to communicate their decision to the committee. Certificates that do not have valid ground reference may indicate the possibility of fraud. Any discrepancies should be isolated from the database and verified. At this stage, each record in the Land Office database should be referenced to a parcel in the cadastral map. Remembering that the Land Office prepares the Certificate of Occupancy after the parcel has ben allocated and the Land Registry ( or Title Office) registers the Certificate of Occupancy ( the title) after the commissioner has signed and sealed the certificate, the record purging exercise is repeated with the records in the Land Registry. 1v. Textual records from the Land Registry (Title office) database should be related with the records in the graphical cadastral map. Using the locatio~ district, block number, and the parcel identifier, discrepancies between the two records should be identified,

PAGE 127

114 removed from the database, and resolved. Because Certificates of Occupancy that need to be registered are always referenced to the cadastral map before approval is given for registration, deliberate errors may not occur within the two databases. This condition was confirmed during the pilot study. The results form this query may comprise of parcels that have not been allocated and those that have been allocated but not registered. The next stage isolates the parcels that have not been allocated from those that have not been registered. v. By relating the results from the Land Office query (step (iv)) and the Land Registry (Title office) query (step (v)), the records from the Land Registry database that are not referenced to the records at the Land Office would be the parcels that have not been allocated. The unallocated parcels should be isolated and verified before a decision can be made regarding the allocation. The final result indicates purged and partially accurate information with respect to legality of the allocation, uniqueness of the location, identifier, and the identity of the owner. However, the records are incomplete unless they are up-to-date in all respects. A schedule should be drawn for systematic verification of the currency of the ownership information, preferably by blocks. This process requires extensive education and public relation activities to educate the parcel owners on the importance of the exercise. An announcement should be made requiring owners of parcels within designated blocks to submit their documents for verification in accordance with the schedule. Upon presentation, the information in the documents should be checked against the entries in the computer. A current mailing address should be entered in the computer.

PAGE 128

115 v1. Upon completion of this exercise the existing records for the designated area can be regarded as complete and current. Processing of New Documents In the past the practice at the Survey and Mapping division title office and the land registry has been to accept all documents that are presented for processing. Whenever a document was found to be incomplete processing was stopped until the owner submitted the missing details ] The improved system seeks to address some of the problems that were hindering the progress of the previous system. It was important to streamline current processes by removing unnecessary procedures and ensuring efficiency without compromising either data quality or data integrity Considering that current land allocation and title registration processes have been in existence for several years it was important that modifications be implemented gradually and systematically in order to confusion among the staff and perhaps total breakdown of the entire process. An important objective was to avoid introducing too many modifications into the existing processes to the point where the land allocation and title registration become a completely new processes for the staff to master. Surveys Processing. The process begins at the Surveys and Mapping division with the newly demarcated and surveyed cadastral plan. The cadastral plan is submitted for approval, together with field notes and all survey computations. A general scheme of the approval process is indicated in Figure 5-3. In addition, the activities to be performed at

PAGE 129

116 various stages in the new and improved system have been listed in Table 5-1. After the cadastral plan and the supporting documents have been checked and found to be complete, entries are made in the computer regarding date of submission and other relevant information. ZONE I COMPUTER COPIES DISTRIBUTED No ZONE 2 COMPUTER DOCUMENT RECEPTION COMPUTER QUALITY CONTROL Yes SURVEY COMPILATION ZONE 3 ZONE 4 COMPUTER COMPUTER L. I. S COMPUTER DIRECTOR APPROVE URVEY RECORDS COMPUTER ZONE 5 COMPUTER CAD TRAL ECTIO MICROFILMING AND DOCUMENTATION COMPUTER FOR DEED PL TO B PREPAR D Figure 5-3: Improved Cadastral Survey Processing Procedure 0

PAGE 130

117 The documents are submitted for quality checking. At this stage, the cadastral plan is reviewed for overall quality and neatness. The field notes are also reviewed for clarity and proper referencing. The next stage is a review of the compilation method where the documents are checked with regard to the choice of survey controls that were used during the survey and proper documentation of any intermediate survey controls that were established Table 5-1: Cadastral Survey Processing Tasks ACTIVITY TASKS Document 1. Receive all in-coming surveys Reception 2. Check to ensure that all supporting documents are included. Incomplete surveys or those without supporting documents must not be accepted 3. Register the survey as confirmation of receipt into a computer. Incomplete documents that are returned should not be logged into the computer. 4. Pass the documents to the Quality Assessor. Notes: 1 The officer at this reception desk must be an experienced Records officer. 2. He/she should be familiar with the documents that are needed to support all new surveys that are presented for checking and approval and the supporting documents that are required when seeking approval for the acceptance of a newly established survey control. 3. He/She should be able to log the details of the survey into the computer.

PAGE 131

118 Survey Quality 1. Check the quality of the job to ensure that it is worth processing. Control 2. That the job was done in accordance with the survey instructions. 3. The correct survey controls were used. 4. Acceptable angular closing errors were obtained. 5. Acceptable linear misclosure errors were obtained. 6. If in general, the work is of a professional quality then 7. Pass the documents to the compilation officer. Surveys that do not meet quality specifications should be returned with written reasons for rejection. Notes: Preferably the quality control assessor should be a senior surveyor whose judgement on the quality of any survey will be respected. Survey 1. Receive the documents from the quality control officer. compilation 2. Retrieves all adjacent surveys that have been done before. 3. Retrieves all previous computation files for adjacent surveys. 4. Retrieves all documents that have been referenced in the survey. 5. Puts all the documents together and passes the compiled documents to the examiner responsible for the zone in which the survey was carried out. Zonal 1. Re-do the computations by going through the same procedures computing that the original surveyor used. This duplication is necessary at exammer this stage until the computation process is computerized. 2. At each stage, compare the results that are obtained with the original results. Documents that contain erroneous calculations should be returned with written comments. 3. Compare the final coordinates with those that were obtained by the original surveyor. 4. Return the documents with a written report giving reasons for rejection. If the calculations are OK, then 5. Produce a plot with the new coordinate values and compare with the original plot which was submitted. 6. If everything is correct, then recommend for the Director's approval.

PAGE 132

119 L. I. s 1. Receive a copy of the approved plan from the Director of Surveys. 2. If digital data for the plan are available, then convert to an LIS readable format. Otherwise digitize the approved drawing and covert into an LIS readable format. 3. Attach plot numbers to the individual plots. 4. Merge the digital data with neighboring data. 5. Use the digital information to update the standard sheet. 6. Submit a copy of the digital data to the office of the Land Development Division so that the land Registry personnel would use to update their title records 7. Submit a copy of the digital data to the office of the Registrar of Titles. 8. Submit a copy of the digital data to the Microfilming and Documentation unit. Notes: The person responsible for this activity should be a surveyor with a working knowledge in Land Information Systems. That person does not have to be a GIS analyst yet have a very good understanding of spatial accuracy in a digital environment. Survey records 1. Receive the final computed coordinates of the survey from the Director of Surveys. 2. Update the relevant databases that have been designed by the Land Records Specialist; vis a vis jobs file cadastral control file geodetic control file etc. Notes: In addition to being a surveyor the person who is assigned this responsibility must have a working knowledge about databases. He/She must have a working knowledge of database management systems. Cadastral Plan 1. Receive the approved master plan from the Director of Surveys. section 2. Prepare deed plans for the individual plots. 3. Submit to the Director of Surveys for safe keeping.

PAGE 133

Microfilming and Documentation 120 1. Receive the computations file from the Survey record section. 2. Microfilm the observation sheets and the final coordinate list. 3. Make two copies of each document. 4. Catalogue the documents by assigning unique numbers. 5. Enter the catalogued information into the computer. 6. Store the duplicate copy outside of the Ministry building. 7. Retain the original copy for reference at the Microfilming and Documentation unit. Notes: The microfilming and documentation unit is a new unit to be created. Persons selected for this job must be knowledgeable on Land Records Management procedures; microfilming, cataloguing and archiving. Finally, to ensure that the entire survey has been conducted in accordance with the survey regulations, the documents are then passed to the respective zonal officer whose responsibility it is to check the survey calculations and the accuracy of the final plot. Only documents which pass these checks are sent to the Director of surveys for approval. The Director's signature at this stage is just to complete the formality. The Director assumes that all the regulations and standards have been complied with. Upon approval, the cadastral map is converted into an electronic format and the requisite information are associated. Other information such as coordinates and description of any newly constructed survey controls are added to the appropriate databases. Copies of the cadastral plan and associated information are distributed to relevant offices. A copy of the cadastral plan is also submitted to the allocation committee.

PAGE 134

121 Title Processing. It is the responsibility of the allocation committee to assign the parcels to successful applicants. After allocation, the list of successful applicants is sent to the title office where designated officers prepare the title for the signature of the Commissioner for Lands. After allocations have been done, processing and registering of titles are done at the Land Development office. As supporting documents are needed for each certificate that needs to be issued or registered, templates which contain checklist items that should be submitted with each document have been designed (see Appendix A). The improved procedure for processing documents at the Land Office is described below with the aid of the diagram in Figure 5-4. At the reception counters each document that is presented for processing is verified to ensure that the required items have been included with the document before the document is accepted. Incomplete documents are immediately returned to the owner together with a copy of the template showing all the missing items. This action eliminates the delays which had characterized document processing activities at the Land Office. Documents submitted with all the supporting details are accepted and recorded into the computer at the reception counter. The entries would be used to keep track of the movement of the files. Entries at the reception counter may be used to analyze the number of documents that are received in a day the number that have been processed or dispatched over a specified period the number of titles that have been issued per region per year, as well as other statistical operations that would be needed on occasion. With computers at the appropriate locations within the system, the statistics section, which is the source of major

PAGE 135

122 DOCUMENT RECEPTION COMPUTER I ZONE I 11 ZONE2 11 ZONE3 11 ZONE4 11 ZONE5 DARES SALAAM ZONAL SCHEDULE OFFICERS DATAENTRY COMPUTER L. I. S. DATAENTRY COMPUTER L. I. S. DATAENTRY COMPUTER COMPUTER COMPUTER I ZONAL REGISTRIES '------~ Connnissioner' s .,..1--__, Seal Connnissioner' s Signature I DISPATCH Figure 5-4: Revised Approach for Processing new Certificates of Occupancy bottleneck, is no longer needed. The entries also provide the current status of every document that has been submitted for processing. Processed documents were again catalogued into the computer before they were dispatched to respective owners. Internal procedures for processing titles were also revised to ensure rapid processing, computerized documentation and easy tracking of documents. Responsibilities and activities at each of the stages in the Figure 5-4, are described in Table 5-2. As usual, the Commissioner signs the title

PAGE 136

123 and it becomes official. Documents that belong to properties within the regions are sent to the assistant registrars in the zones. For properties within Dar es Salaam region, the prepared title and supporting documents are submitted to the Registrar's office for processing. Table 5-2: Improved procedure for processing and Issuing Certificate of Occupancy ACTIVITY TASKS Reception Counter 1. Receive files 2. Check that all supporting documents are included and are complete. 3. IfO.K., log into computer and log book. 4. If incomplete, indicate missing documents on template and return to the person who submitted the application. 5. Place accepted applications in folders. 6. Pass folder to Schedule Officer responsible for the region. 7. If checking for documents, enter Land Office number and date, if available. It is also possible to search the database by name. Schedule Officers 1. Receive folder containing application for title. 2. Check for payment of appropriate fees. 3. Check validity of the certificate. 4. Check that the parcel has not been assigned to someone already. 5. Enter data into computer if it is a new allocation. 6. Assign a Land Office (L-O) number 7. Print Land Office number on the folder. 8. Pass to LIS section. LIS 1. Check the physical location of the parcel and ensure that it has not been assigned to someone already. 2. Associate data to the appropriate parcel on the map supplied by the Surveys and Mapping after ensuring that it is a new allocation. 3. Pass folder to the Commissioner's assistant. Commissioner's 1. Apply the commissioner's seal to the title document Assistant 2. Keep folder until the commissioner is ready to sign the title. 3. The Commissioner's assistant should be responsible for assigning L-O numbers to the regional land offices.

PAGE 137

124 The Commissioner 1. Receive documents for signing. for Lands 2. Review and sign the Title. 3. Pass the folder to Documentation section. Document 1. Microfilm documents. Archiving 2. Catalogue microfilm and enter information into the computer. 3. Scan deed plan. 4. Submit digitalized deed plan to the LIS section. 5. File documents. 6. Store microfilm. Dispatch 1. Log the date of approval of the title. 2. Log the method of dispatch into the computer. 3. Log the dispatch date into the computer. Title Registration. With the title registration process, accuracy of data and the validity of the records are important. Checking the accuracy and validity of the data are the responsibilities of senior members of staff. In order to remove the delays that are caused due to insufficient information, it is necessary that the documents are checked at the entry point to ensure that any obvious omissions are corrected before the documents are accepted. Because several different types of documents may be registered at the land registry, each requiring different supporting documentation, there should be a senior member of staff at the reception, who is familiar with the registration requirements for all, or most of the documents, at the reception counter to check the documents against the requisite checklist before acceptance. Several other documents are processed there, including caveats, mortgages, power of attorney, etc. Each transaction requires different supporting documents. As such, different templates of checklists were prepared for registration purposes. It is imperative that documents that are presented for registration are thoroughly checked for legality, validity and

PAGE 138

125 WINDOW 1 COMPUTER WINDOW 2 COMPUTER ... ---~ PUBLIC ACCESS ASSISTANT REGISTRAR Title related Transactions Other types of transactions OTHER TRANSACTIONS Mortgages Liens Power of Attorney, etc. ,, DOCUMENT CORRECT? Yes DOCUMENT PROCESSING DATA ENTRY COMPUTER (A) No CERTIFICATE OF OCCUPANCY (C/O) UNIT I COMPUTER (B) I OWNERSHIP CHECK l MANUAL CHECK l I COMPUTER(A)I DEED PLAN _ DEAT AIL ACCURACY CHECK Yes REGISTRAR DOCUMENTS ENDORSES CORRECT? I No Lt MICROFILMING MAILING AND .-Dispatch to Owner DOCUMENTATION Figure 5-5: Improved Procedure for Processing Documents at the Land Registry

PAGE 139

126 accuracy. Again, the objectives are to remove bottlenecks by streamlining the convoluted process such as a document being presented to the Certificate of Occupancy (C/O) unit four times during a single registration process. The schematic diagram of the title registration process is shown in Figure 5-5. Activities and responsibilities at various stages of the process are shown in Table 5-3. Table 5-3: Document processing Tasks at the Land Registry ACTIVITY TASKS Reception I. Receive approved certificates. 2. Check that all supporting documents are included 3. Incomplete documents should be given back to the person who submitted them with an attached template showing all the missing supporting documents. 4. Documents that are complete should be entered into the computer. (This computer also contains status report of all documents, including dispatches). 5. Pass the document to the Asst. Registrar. Asst. Registrar I. Receive the documents. 2. Check what action need to be taken on the application. 3. Check that all supporting documents are submitted. 4. Incomplete documents should be returned to the mailing room with an attached template showing the missing items. 5. Complete documents should be passed to the appropriate section for processing. If document is for other transactions, pass to the appropriate personnel. If application is for title registration, pass to Certificate of Occupancy (C/O) unit. C/O Unit I. Receive the documents from the Asst. Registrar. 2. Check that the parcel has not been allocated already. If there is an evidence of duplicate allocation, retrieve that duplicate document and pass both documents to the Asst. Registrar for action. 3. Whenever fraud is suspected, the documents should be referred to the Registrar for necessary action. 4. If O.K. process document and pass to Deed Plan.

PAGE 140

127 Deed Plan 1. Receive documents from C/O unit. 2. Check that the Deed Plan has been correctly and accurately drawn. 3. Check the quality of the cartography. 4 Check that the Block# Plot# North arrow Scale and description of the property agree with what has been recorded Ill the certificate. 5. If O. K. then pass to Computer A else indicate what is wrong on the template and pass to mailing room. If in doubt consult the Assistant. Registrar. Computer A 1. Receive document from Deed Plan 2. Check from inside the computer that the spatial location of the Deed Plan is the same as has been presented on the Deed Plan. 3. Ensure that the same plot has not already been identified with different boundary markers. 4. Instances where fraud is suspected should be referred to the Registrar for appropriate action. 5. If everything O. K. then pass to the Registrar Registrar 1. Receives the document from the C/O unit 2 Endorses the title and approves the registration of the title. 3 Pass the document to Registration unit. Note: At the C/O unit (Computer B) and the Spatial Accuracy (Computer A) Suspected incidents of fraud should be handled with all seriousness. Whenever there is sufficient reason to believe that fraud has been attempted the Registrar should exercise his discretion to prosecute. In such situations the documents should be passed to the legal section. Registration 1. Receive the documents from the registrar. Unit 2. Register the title under whatever system that is appropriate. 3. Send original to mailing room( dispatch) 4. Send copy to Documentation and Archiving Legal Section 1. Receive documents and evidence of suspected fraud from the Registrar. 2. Investigate and recommend necessary action to the Registrar. The Registrar should make the final decision to prosecute. The Legal section should support the decision of the Registrar.

PAGE 141

Mailing (Dispatch) 128 1. Receive registered documents from the registration section. 2. Log details of the title into the computer and Log book. 3. Send to the owner and indicate method of dispatch into the computer at the reception counter. Entries are made into the computer regarding the ownership of the document and the date of submission. The document is then passed for processing. Besides checking for evidence of ownership, validity, and accuracy of information, checks are also made to ensure that the correct fees have been paid. Templates have been designed with a view to reduce the processing time (see Appendix A). Documents with incorrect details are returned to their owners with copies of the template showing missing items and any incorrect data that need to be corrected. Properly prepared documents with evidence of payment of all requisite fees are entered into the appropriate databases. Again, after the documents have been checked against the appropriate template for necessary supporting documents, the next stage depends on whether the document is for a title registration or other transaction (see Figure 5-5). The Assistant Registrar reviews the document to verify the type of registration that is required and passes it to the appropriate officer. For title registration the officer checks to ensure that names have been entered correctly, and that the information provided corresponds with the information that was obtained from the Land Office ( where the original certificate was prepared). A manual check is done on the deed plan to ensure that it meets the appropriate standard for registration. The next stage is to ensure that the parcel that is referenced does not belong to any other person. The parcel information is checked against the computerized cadastral plan to ensure its spatial location.

PAGE 142

129 If the title passes all the checking processes, it is then sent to the Registrar for endorsement. Date and time of registration are then recorded. The records are then associated with the cadastral plan which has already been converted into a digital format by the Surveys and Mapping division. The Registrar of Titles then signs the title and the necessary information are recorded into the register. Signed titles are catalogued and sent to the zonal land registries. Copies of the document are retained for archival purposes. The owner receives a copy of the title. Organization of Cadastral Information Having addressed administrative, technical, and public concerns about the land delivery system in Tanzania, attention is now directed towards approaches for removing the bottlenecks in the land management process, and ultimately, for developing a cadastral information system for Tanzania. A cadastral information model, which is a component of a topologically structured multipurpose land information model (Figure 5-6), has been developed for Tanzania. The topologically structured multipurpose land information model which is presented here, is based on graphical overlays that are constructed topologically. For example, the cadastral information support system component will have a topologically structured cadastral index map as an overlay. Other graphical overlays will be created topologically to support respective information system components. For example, topologically structured graphical overlays of road network and utility lines will be created to support the infrastructure information system component of the model.

PAGE 143

130 r-------, : LAND POLICY : L-----.-------J ,-------------------, :LEGAL FRAMEWORK : L------------------~ r------------*------------, I I 1 OPERATIONAL COMPONENT: L------------.------------. .. .. .. .. .. .. .. .. .. .. .. INFORMATION SUPPORT SYSTEMS ENVIRONMENTAL INFRASOCIOCADAS'l'RAL INFORMATION STRUCUTRE ECONOMIC INFORMATION INFORMATION INFORMATION VALUATION JURIDICAL ENVIRONRESOURCE AND RECORDS MENTAL lITILmEs MANAGEMENT TAXATION RECORDS AND PLANNING CADASTR.AL SURVEY DATA LINKAGE :MECHANISMS OTHER TOPOLOGICAIL Y PAltCEL IDENTIFIER. STRUCTURED GRAPHICAL OVERLAYS TOPOLOGICAILYSTllUCTUIUID GEODETIC REFERENCE FRAMEWORK CAT\ .a .. T: 4.1 -:"'"-;--.-=-.MAP Figure 5-6: Topologically Structured Multipurpose Land Information Model

PAGE 144

131 The topologically structured multipurpose land information model is developed from the perspective of the information support systems component ofFigure 2-1 and on the fact that accurate base maps are not always necessary in decision-making situations pertaining to land management. While acknowledging the need for accurate base maps in certain land management activities, the expense for capturing the accurate data has always been weighed against the value of the land. In many developing countries with scant financial resources and competing economic and political projects, accurate base mapping is not always a priority (United Nations 1974; Dunkerly 1985; Holstein 1990). A Cadastral Index Map ( CIM) is a map showing a graphical index of parcels within a jurisdiction or a geographic region. Cadastral Index maps are used for administrative purposes and contain references to information such as ownership, value, and uses of the parcel. A topologically structured cadastral index mapping approach presents a different way oflooking at boundary delineation and base mapping as a whole, without the expense that normally accompanies such ventures. The model for the Tanzanian cadastral information system is distinct from other models that have hitherto, been developed with respect to the graphical data structure, the linkage mechanism, and constituent data within the system. Distinctive differences between the topologically structured model and the existing models which have been described in Chapter 3 are that: 1. Unlike the model for developing countries (Jeyenandan and Williamson, 1990), the topologically structured model is independent of the social grouping even though the distinction can be made from demographic data.

PAGE 145

132 2. The four models described earlier advocate base maps in metric space with relevant spatial accuracies. For example the NRC model prescribes spatial accuracies as prescribed by the American Society for Photogrammetry and Remote Sensing and the United States National Mapping Accuracy Standards. This model on the other hand relies on spatial relationships in topological space. However in order to establish location in geographic space, the individual cadastral maps, which constitute the components of the cadastral index map, are tied together within a reference framework, which may or may not be geodetic in nature. 3. Unlike the existing models which have been designed to support existing government agencies the topologically structured model is designed to support information types. 4. By defining the spatial entities as nodes, lines and polygons, attribute information can be associated with each of the entities. For example, by regarding the corner monuments as nodes, descriptive information such as the type of monument and individual identification numbers may be associated with them. 5. Accurately measured data in metric space can be associated with the topological entities. For example, accurately surveyed parcel boundary data can be associated with the parcel polygon, through its parcel identification number. This makes it possible for interested parties to access the accurate data and produce a true graphical representations of the parcels, if the need arises. 6. The topologically structured graphical overlay can provide a great deal of information without the expense involved in accurate survey measurements. For example the cadastral index map can provide information pertaining to parcel ownership and

PAGE 146

133 thereby, responsibility for tax liability, without the need know the exact boundaries of the parcel. 7. In places where accurate measurements of the individual parcels are not readily available, a topologically derived cadastral index map is an invaluable resource. The accurately surveyed information can still be associated when the data become available. Some land management activities can still be conducted without the need for accurate surveys. In the following section, topological concepts are explained and the rules regarding the application of topological entities for cadastral index mapping, to support cadastral information system, are illustrated. The topological entities are defined in terms of the types of data to which the entities can be associated. The cadastral information system component is a subset of topologically structured multipurpose land information model. Topologically Structured Cadastral Data Concept In the computerized environment, procedures for representing graphical data are either the metric space or the topological space. In metric space, the information are presented in three dimensional space using lengths (Croom 1989, 55), which may be derived from Cartesian or Euclidean space (coordinates, angles, azimuths and distances). Most land surveying measurements and calculations are performed in metric space. With different measuring equipment and methodologies, land survey computations always contain some imprecision due to measuring device, methodology,

PAGE 147

134 or ambient conditions such as changes in temperature or topography. Additional procedures and sometimes statistical analysis are performed to reduce or eliminate the effect of such factors in measured quantities. The problem of imprecision in measures quantities due to measuring devices and procedures is eliminated in a topological representation. Topology deals with geometric properties which are dependent only upon relative positions of the component entities and not upon such properties as length, size, or magnitude. Topology deals with properties which are not altered by continuous transformations like bending, shrinking and twisting (Croom 1989, 1). As a mathematical procedure for representing spatial relationships, topological principles are applied to spatial entities by defining them in their basic forms as points, lines and areas and establishing relationships such as adjacency (contiguity), connectivity, and areas among them. The concept of topologically structured graphical overlays is adopted in this dissertation as the underlying structure of the graphical component of the multipurpose land information model that was developed for Tanzania. The topologically structured land information as applied to Tanzania is described in the next section however, the principles and application are given here. By applying topological principle to vectorized cadastral plan or map, the spatial entities can be defined as nodes, lines, or polygons (areas). It is therefore possible to establish relationships, such as adjacency, connectivity, and polygon definition among the entities. Node features. For this research, a node is defined as the intersection or meeting point of two or more lines. It is also the starting or terminal point of a line. In cadastral mapping

PAGE 148

135 environment, a node will be the intersection of two or more boundary lines. Such intersections are normally identified by corner monuments. The corner monument of the parcel may therefore be regarded as a node entity. Attribute information can be associated with that node. Such information may be the type of monument, its location within a geodetic reference framework, identification number, list boundary lines that are radiating from the node, etc. Linear features. Lines are spatial features which are too narrow to be represented as areas. In topology, linear features are considered to have beginning and end points. The terminal points are referred to as nodes. In terms of a cadastral plan, the linear feature is the boundary line, and the terminal nodes are the corner monuments. If lines, in topology, have starting points and end points, then a line can have a direction, a left side and a right side. By analogy, a boundary line, in topological terms, has a starting point, and end point, and a direction. As a topological entity, attribute information may be associated with the property boundary line such as, type of boundary marker (for example, a hedge, brick wall, or wooden fence), identity of adjacent parcels, type and identity of the monuments at the terminal points, etc. Polygon features. Polygons, in this situation are the actual parcels. A polygon or an area is obtained if a line, such as a boundary line, completely surrounds a homogeneous feature such as a parcel. When such line surrounds the feature and terminates at the same

PAGE 149

136 starting point an area or polygon is formed. Examples of polygons include land parcels forest stands and subdivision blocks In the next section topological principles are reviewed. Procedures and rules for applying topology to cadastral mapping are presented. These rules are then applied to a pilot study for producing a cadastral index map for Tanzania. Topological Rules and Cadastral Index Mapping In dealing with geometric features certain rules have to be applied. These rules define how topology is generated from the data that are presented. Due to the importance of cadastral boundaries these rules have to be applied carefully to ensure proper definition of the geometry of cadastral boundaries that have been generated topologically In this section topological rules and hierarchies of entity definition as are applied to this research are presented. Node snapping rule. The position of the first node to be entered takes precedence over subsequent nodes. This defines a node-to-node precedence. This means that if a node is created within a defined proximity from a pre-existent node the second node should snap on to the first node (see Figure 5-7). For this research, boundary situations are considered positive For example a node that falls exactly on the boundary of a snapping region is considered to be within the region.

PAGE 150

.... ... N1 : .. .. ... .. A: The first node and its snapping region 137 .... .. N1 ... d .. .... B: A second node is placed within its snapping region .. .. / N1 .\ . : \ j .. .. ............... C: The second node snaps onto the first node Figure 5-7: Node snapping precedence rule Nodes JI t-------4 .-vertices \ A B Figure 5-8: Two Definitions of a line Line (Arc) definition rule. A line (arc) is made up of two nodes which define the terminal points and, in addition, may contain zero to any number of vertices (see Figure 5-8). Minimum line definition rule. A line should have a length greater than the node snapping distance in order to be retained topologically. Node-Vertex precedence rule. The position of a node takes precedence over a vertex. Therefore, if a vertex is within the snapping region of a node, the vertex should snap

PAGE 151

. . ..... .: "'-:, ; : :. .: ..... ............................ A: A pre-existent node 138 .. . B : A vertex within the snapping region of anode C : The vertex snaps onto the node Figure 5-9: A node-vertex precedence to the node (see Figure 5-9). This is illustrated in figure 5-9 where a pre-existent node is shown in A. A line a drawn with one of its vertices falling within the snapping region of the node, as shown in B. The vertex snaps onto the pre-existent node as shown in C during the generation of topology. Node-Line {Arc) precedence rule. If a line falls within a defined snapping region of an existing node, the line should snap to the node (see Figure 5-10). This is explained with reference to Figure 5-10. The situations A or B occurs when a line falls within the snapping region of a node. In both cases the line snaps onto the node. A typical situation with respect to cadastral mapping is depicted in C, where a right of way exists between two neighboring parcels. As the nodes are within the snapping regions of

PAGE 152

139 r .. 11 . ... \ :; . ~ "' .. .... . ~ ,, ., ( ~) -:. ... .., u I A B 0 Vertex Node ( ... ~ .. .. I C : Figure 5-10: Node-line precedence neighboring boundary lines, the lines snap onto the nodes. Two adjoining pentagonal parcels are formed instead of two neighboring rectangular parcels. To avoid such situations during topological mapping of cadastral parcels, snapping distances have to be set with caution. Line intersection rule. If two lines cross each other, a node should be placed at point of intersection (see Figure 5-11). In Figure 5-1 lA, the lines 1-2-3 and 4-5 cross each other. By rule, a node should be placed at their intersection point. The node precedence rule takes effect after the new node has been placed in position. If the new node is outside the snapping region of any of the pre existent nodes. The final result is shown in B. Supposing that the line 1-2-3 was the first to

PAGE 153

140 4 ---3 A C Figure 5-11: Node precedence after a line intersection be entered, and the new node is within the snapping region of node 2, as shown in C, then node 6 snaps into node 2, as shown in D. Vertex snapping rule. When two vertices are within a certain proximity from each other the second vertex should snap onto the first vertex. This rule also define line snapping (see Figure 5-12). In Figure 5-12, the two lines are so close that some vertices fall within the snapping regions of other vertices. The affected vertices snap together, nodes are placed at their intersection points, and part of the second line is removed.

PAGE 154

141 ,. ,. I I I 9 I I I 0 I I b ', .. A .... B C Figure 5-12: Vertex snapping Application of the Rules to Cadastral Surveying In cadastral surveying boundaries are generally defined by the lines joining the comer monuments. This implies that the comer monuments have the highest integrity. As the comer monuments are represented by nodes this means that by definition, nodes should have the highest level of importance By the rules presented above vertices have the lowest integrity which can cause problems. Unless otherwise defined the line joining the boundaries should be rectilinear. A typical situation where curvilinear lines would occur in cadastral surveying is when the bank or center line of a stream is defined as a boundary of a parcel. Topologically curvilinear boundaries would be defined with vertices and nodes at the terminal points. In cases where monuments have been established along the curvilinear boundary, those monuments are established as nodes during topological representation so as to retain their integrity.

PAGE 155

142 Boundary Definition Cadastral surveys begin with a survey of the external boundary and followed by the internal partitions. Topologically, the external boundary may be represented by a single line which begins on a node and closes on the same node, with vertices at intermediate corners (see Figure 5-13). The situation represented as A, shows vertices as intermediate stations. This provides a poor representation of the external boundary, since the position of the vertices can be influenced by neighboring nodes and lines. In B, the boundary monuments are represented with nodes. This provides the highest integrity of the monuments. If internal partitions need to be drawn, those lines will begin as nodes and end as nodes. In A, the nodes of the internal lines will tend to move the vertices around. Movement of vertices will not occur in B since the nodes on the external boundary will have precedence over the nodes that define the internal partitions. Therefore, in order to have an accurate representation of the external boundary, the boundary monuments have to be represented with nodes. A B Node 0 Vertex Figure 5-13: Two Topological representations of a Cadastral Boundary

PAGE 156

143 0 Verte x Node Figure 5-14: Representation of a Boundary with Vertices A situation where a property boundary may not topologically be represented by nodes is when the boundary lies in the middle of a river (see Figure 5-14). In the figure one of the boundaries of property A lies in the middle of the river. Since no precise measurements were made directly to the middle of the river vertices are interpolated from mid-points between banks. Subdivisions Property boundaries in urban areas are generally rectangular in shape. For rural areas where parcels are irregularly shaped, partitions (or subdivision lines) need to be positioned precisely As such they have to be digitized as nodes instead of vertices. Figure 5-15 shows a typical urban subdivision. Here, the comer monuments of the external boundary have been represented by nodes. In this Figure, it is assumed that no digitizing errors occurred at the terminal points of the line L 1 as shown in Figure 5-15B (highly unlikely).

PAGE 157

144 Line L 2 extended over the boundary line but the boundary line was within the snapping region of the node. As shown in the Figure 5-15B, this condition caused the line to snap onto the node. A: Before Applying Topology B : After Applying Topology Figure 5-15: Topological errors in Cadastral Index Mapping Line L 3 also extended past the boundary but the line was outside the snapping region of the node. According to the line intersection rule, a node will be placed at the intersection point of the two lines (see Figure 5-15B). Lines L 4 fell short of the external boundary line. Again, the boundary line was within the snapping region of the node. Hence, the line was moved inward (see Figure 5-15B). These are possible scenarios that can occur during topological representation of cadastral parcels.

PAGE 158

145 From these conditions, the rules for adequately representing contiguous cadastral parcels are summarized as follows: 1. Every point which is physically surveyed or accurately measured should be denoted by a node. 2. Snapping tolerances should be set as minimum as possible in order to avoid unwanted displacement of vertices. 3. It is better to have overshoots that are longer than the minimum allowable length of an arc. This will prevent unwanted snapping of nodes. 4. Removal of unwanted dangle lines should follow generation of topology. These rules have been applied in the pilot study which is presented in Chapter 7. Since topology involves non-metric relationships, measurement accuracies are not essential in topologically structured maps. This principle is applied in the compilation of a cadastral index map. In situations where different parts of the jurisdiction have been surveyed using different equipment and different methodologies, and thereby, obtaining varying degrees of spatial accuracy in different parts of the map, topology provides an approach for tying the maps together. In its basic form, the cadastral plan consists of comer monument descriptions, boundary lines, the polygon which defines the parcel, and the label which identifies the parcel spatially.

PAGE 159

146 Topological Rules and Principles Illustrated An application of topology to the creation of a cadastral index map is illustrated by a situation shown in Figure 5-16 where three adjacent subdivision plans have been produced at different scales in different coordinate systems and different spatial accuracies. In the figure the numbers denote the comer monuments and the alphanumeric characters identify individual parcels. The boundary line is defined by the line joining the comer monuments. By establishing topological relationships between the spatial entities the three plans can be 'joined" together. This is achieved by establishing point equivalence, line equivalence line-to node topology and polygon definition. Point equivalence is basically, used to identify nodes that are identical in the two plans. Similarly, line equivalence identifies lines that are identical in the two systems. Before the matching can be accomplished, one basic rule that needs to be established is that the location of a pre-existent point has priority over subsequent points. For 1 9 10 A B C 12 5 ~1 D E F 4 8 7 34 35 30 __________ 31 L M N 33 37 36 32 2 6 3 21 22 ----G 25 1----------1 26 27 40 24 H J K z 28 41 23 Figure 5-16: Three separate cadastral surveys

PAGE 160

147 e x ample if a polygon needs to be added to the cadastral index map and the new polygon has two identical comers in the master cadastral index map the new polygon will be enlarged or shrunk until the two comers match the pre-existent ones. Point and line equivalence tables for Figure 5-16 have been shown in Tables 5-4 and 5-5 respectively. Table 5-4: Point Equivalence Table Point number Equivalent point numbers 2 21 6 25 3 27 31 32 40 4 30 8 34 7 35 A line-node topology establishes connectivity between nodes and lines. This establishes direction and therefore the left and right sides of the line. The line-node topology for Figure 5-16 is shown in Table 5-6. A polygon-line topology identifies the lines which surround individual parcels to isolate them from one another (see Figure 5-16).

PAGE 161

148 Table 5-5: Line Equivalent Table Line identifier Equivalent Line identifiers 2-6 21 25 6-3 25 27 31 32 27 -40 4-8 30 34 8-7 34 35 7-3 35 31 Table 5-6: Line-Node Topology From Number To Number Left Polygon Right Polygon 1 9 z A 9 10 z B 10 2 z C 2 22 z G 5 11 A B 11 12 B E 12 6 C F 6 26 G H 4 8 D L 8 7 E M 7 3 F N 3 28 H J 33 37 L z 37 36 M z

PAGE 162

149 36 40 N z 40 41 J K 32 23 K z 1 5 A z 5 4 D z 4 33 L z 9 11 B A 11 8 E D 8 37 M L 10 12 C B 12 7 F E 7 36 N M 2 6 G C 6 3 H F 3 40 J N 40 32 K z 22 26 z G 26 28 z H 28 41 z J 41 23 z K

PAGE 163

150 Table 5-7: Polygon-Line topology Polygon Connecting Lines A 1-9-11-5 B 9-10-12-11 C 10-2-6-12 D 5-11-8-4 E 11-12-7-8 F 12-6-3-7 G 2-22-26-6 H 6-26-28-3 J 3-28-41-40 K 40-41-23-24 L 4-8-37-33 M 8-7-36-37 N 7-3-40-36 After applying the topological principles and rules, the final product depends on factors such as the RMS error that was obtained after the transformation, digitizing errors node snapping tolerance that was set, and methods for applying the rules to the nodes and vertices that were digitized. Figure 5-17 shows two hypothetical results of Figure 5-16. Results of the pilot study have been shown in Chapter 7. Distortions have occurred for two main reasons. Firstly through the affine transformation, and secondly, by constraining the nodes and lines to maintain their defined

PAGE 164

151 1 9 10 2 22 1 9 10 2 22 A B C G A B C I G 12 6 111 5 26 5 12 6 11 26 D E F H D E F H 'i 3 L M N J 4 28 4 8 3 7 L M N J 28 ~o 37 36 33 41 3 \ Aln 3 37 36 41 K K z 23 z j 32 32 23 Figure 5-17: Two Topologically structured cadastral index maps of the same area relationships. However these distortions are secondary provided the spatial relationships between the individual parcels are maintained. The question arises as to the accuracy of the map in metric space. Again if the individual maps have good spatial accuracy then the resulting cadastral index map would have good spatial accuracy As an index map the actual field survey measurements and the individual survey plan can be associated with the index map Association can be achieved through linkage mechanisms such as a parcel identifier. The file containing the actual survey records can be accessed with the parcel identification number. Another advantage ofthis approach is the fact that attribute data can be associated with the comer monuments the boundary line, and the parcel independently. The procedure for developing a cadastral information system for Tanzania including the creation of a topologically structured cadastral index map is discussed in Chapter 6. The

PAGE 165

152 distinction between the cadastral index map in metric space and the same map in topological space is also clarified in Chapter 6.

PAGE 166

CHAPTER6 CADASTRAL INFORMATION MODEL FOR TANZANIA Information system components for land management have been described in Figure 2-1 (in Chapter 2) as cadastral, socioeconomic infrastructure and environmental. Each of these components is comprised of graphical overlays and associated descriptive records. For cadastral information, the graphical overlay is the cadastral index map which is comprised of the individual cadastral plans. Descriptive data which are associated with the cadastral index map include title records valuation and assessment records and the land registry data. Land related records in the land registry of Tanzania include land registration records, mortgages liens and several other records which can be associated with a cadastral index map. Since the topologically structured cadastral index map is not exact in metric space the original cadastral survey field data and the cadastral plans are integral to the system. The survey field data and the accurate (in metric space) cadastral plans should be linked to the appropriate parcels in the cadastral index map so that an accurate cadastral map can be constructed from this data. Other information support components will also have the requisite graphical and textual data types which are relevant to the objectives of the model. The infrastructure information component for example consists of the graphical data pertaining to road center lines pavement, water and electricity lines together with descriptive records such as road surface type, road maintenance schedule, etc. 153

PAGE 167

154 Socioeconomic information would include human and natural resource maps and the associated demographic records. The environmental component would contain natural resource data and hazard maps together with records of environmental impact studies. It must be mentioned at this point that whereas some of these data types and records may exist in paper format, no attempts were made to develop the other information support systems. As such, no effort was made, either to extract or convert data to support those information systems in this research. A spatially referenced land information model includes a spatial referencing framework, upon which the graphical layers within the individual information systems depend for positional accuracy and coincidence. Existing models of multipurpose land information systems employ a geodetic reference framework. A brief description of the procedure for compiling a cadastral index map within a geodetic reference framework is given below in order to highlight the major differences from a topologically structured model. Cadastral Index Map Compilation in Metric Space Compilation of cadastral index map in a geodetic reference framework requires a transformation of the individual cadastral maps and plans from their original referencing system into the geodetic reference framework. The Affine transformation function is generally adopted for performing such transformations, unless there is justification to use a different transformation. The Affine transformation function is of the form:

PAGE 168

155 X' =Ax+By+C Y'=Dx+Ey+F To transform from one system into a Geodetic Reference Framework (GRF) the transformation function will be of the form: XaRF = Ax+ By+ C Y GRF = Dx + Ey + F The unknown quantities in this function(~ B C D E and F) can be solved uniquely with three common non-collinear points in the two systems. Additional points provide a check and allow application of a least squares approach to the solution. The results provides equivalent coordinates in the GRF system for each point that was used, and some residual. From the residuals a Root Mean Square error (RMS) of the transformation can be calculated. The RMS error is calculated from the function : RMS e = n + e 2 n where n is the number of points that were used in the transformation and e denotes the individual residuals e is obtained by the equation e = J ( v} + v /), where v represents the individual residuals in x and y. The RMS error gives an indication of how good the transformation is. The lower the RMS error the better the transformation. In situations where there are several cadastral plans, each plan has to be transformed independently into the GRF system. The effect of these residuals, together

PAGE 169

156 with other errors such as methodology and plotting accuracies can, and often do result in gaps between parcels or overlapping parcels instead of contiguous parcels (see Figure 61). In Figure 6-1, two separate plans of contiguous parcels have been drawn in different referencing systems. Both plans have controls in a geodetic reference framework. To reproduce the plans in the required geodetic framework, they both have to be 'v A9 A8 Al 'v A2 A4 A7 AS A6 SYSTEM A 'v Transformation from System A into a GRF using the control points \ \ \ \ \ \ \ \ \ I I ', I I ---, I I I I I I I I I I 'v ',J 'v 'v 'v I I I Bl 'v B8 SYSTEM B 'v ,' Transformation from ,' System B into a GRF ,' using the control points GEODETIC REFERENCE FRAMEWORK 'v Control point Figure 6-1: Coordinate Transformation in Metric Space

PAGE 170

157 transformed. A hypotheticaL yet common occurrence is shown below, whereby the two parcels are not matching exactly even though they are supposed to be contiguous. Different methods have been adopted to make the plans match exactly. Such methods include averaging out the discrepancies or "rubber-sheeting" one drawing to fit the other. With reference to Tanzania, it has been mentioned earlier that whereas relative accuracies between neighboring survey projects are within acceptable tolerances (with regard to the stipulations in the Survey Regulations of Tanzania), excessive misclosures are observed with respect to absolute positions. This implies that a coordinate transformation would result in large RMS errors. In a measurement based (metric space) modeL such discrepancies would require re-adjustment of the survey control network, followed by a re-compilation of the cadastral index map. Both of these activities are expensive propositions which cannot be met with the meager financial and human resources that are available in Tanzania. In addition to these mitigating factors, the production of an accurate cadastral maps for Tanzania would be a time consuming project. Evidence from this research indicates that the standard sheets in Tanzania are outdated due to lack of maintenance. There is no reason to believe that even when a mechanism for updating the standard sheets is implemented, compliance can be enforced. The adoption of a topological approach is a way to curtail expenditure, provide a method for creating an index map which is consistent in detail to what exists on the ground, and robust to allow the performance of spatial analysis on the associated data.

PAGE 171

158 Cadastral Index Mapping in Topological Space In topology, the important criterion is the consistency of the data in topological space. This implies ensuring that spatial relationships between individual parcels are consistent with what exists on the ground. For a cadastral index map a geodetic reference framework is not critical as adjacency relationships are more important than absolute positioning. In order to merge the individual cadastral plans common points are required for transformation. However, as shown in Figure 6-2, the individual plans are transformed not with control points on the plans, but with common points on adjacent external boundaries of the plans. In the Figure 6-2 the cadastral maps A, B, and Care contiguous cadastral plans that need to be assembled together topologically. Cadastral map A is chosen as the base layer, onto which the other plans are going to be linked. The identical points Al to A6 on map A, and Bl to B6 on map B, are used as the control points for the transformation. With six control points to be used in an Affine transformation, a least squares solution will yield some residuals. Point equivalency is established between the two maps by connecting points B 1 through B6 to the corresponding points Al through A6. This forces the corresponding points to snap together. The position of the remaining points is a result of application of the transformation coefficients. The resultant map is shown in D. In a similar fashion, the cadastral plan, C, is connected to the composite map, D, with the common points A8, A9, Al, and B8 connecting to Cl, C2, C3, and C4. The final cadastral index map is shown in E.

PAGE 172

A9 A8 A7 AS A A6 Al A2 159 Transform and establish Topological relations -+ Result A9 L Result Transform and establish Topological relations C6 I : i A9 C6 CI COMPOSITE CADASTRAL INDEX MAP B8 B C Figure 6-2: Topologically assembled cadastral index map cs As expected, the resultant map has all the points connected. There are no gaps or overlaps. As such, there are no more adjustments to be made on the final figure, as was the case with the cadastral index map in metric space. This approach for producing the

PAGE 173

160 cadastral index map has been adopted in this research. In the next sectio~ an application of this procedure for developing a cadastral information system for Tanzania is presented. Cadastral Information System for Tanzania In Figure 5-6, a model for developing a topologically structured multipurpose land information system was presented. The highlighted portion of Figure 5-6 shows the cadastral information systems component of the model. The distinct feature about this model is the fact that the cadastral information systems component is independent of the geodetic reference framework. Beside hardware, software, and human resources, the components of cadastral information system include the cadastral index map, a parcel identification structure through which parcels can be uniquely and easily be identified, and the descriptive records. In Chapter 5, a procedure for isolating erroneous and inconsistent records in Tanzania was developed. In this sectio~ a procedure for developing the topologically structured cadastral index map and the development of parcel identification system are presented. These approaches are then applied in a pilot project to assess their effectiveness. Cadastral Index Map for Tanzania By applying the topologically structured cadastral index mapping concept for the Tanzanian land information model, the available graphical data need to be converted into a digital format in order to apply the topological principle. The different types of graphical

PAGE 174

161 data in Tanzania include subdivision plans in local coordinate systems, subdivision plans in a national geodetic coordinate system, and survey field notes. Figure 6-3 shows a schematic diagram for converting the different types of graphical data that were available in the Surveys and Mapping division in Tanzania. The cadastral index map for the pilot study in Tanzania was created from the following graphical data sources: Subdivision Plans in a local coordinate system Coordinate transfonnation or digitizing Subdivision Plans in the National Geodetic Framework : Coordinate : j transfonnation : Survey Field notes Coordinate geometry and keyboard entry Individual cadastral plans and other maps Appropriate conversmn method ESTABLISH TOPOLOGICAL RELATIONSHIPS I TOPOLOGICALLY STRUCTUREDADASTRAL INDEX MAP Figure 6-3: Graphical data conversion options 1. An existing standard sheet at a scale of 1 :2500 which was created in 1980 by photogrammetric methods. The inap was in paper format and contained coordinates in the Universal Transverse Mercator projection.

PAGE 175

162 2. Existing subdivision maps at a scale of 1 :2500. As a regulation in Tanz.ani~ all subdivision surveys have to be tied to the national geodetic coordinates syste~ unless prior pennission has been given by the Director of Surveys. 3. Computed coordinates of field survey data. Although it was not by design, the data for the pilot study did not include graphical data in a local coordinates system. Even though certain parts ofTanz.ania have been surveyed on local coordinate systems. Among the graphical data sets that were available, the standard sheet contained the most property boundary points. This meant that contiguity had already been established among those parcels that were on the base map. All the other pieces of graphical data were inserted into the standard sheet. Figure 6-4 shows a schematic diagram of the process for converting the different data types into a digital format and maintaining spatial relationships. The base layer. Since the standard sheet contained the most data in terms of property boundaries, the process began with the existing but outdated 1 :2500 scale standard sheet. It was assessed for quality in terms of line work, creases and smears, and determined that the base map was in good enough condition to be digitized. The grid intersections on the 1 :2500 scale standard sheet were used as control points to register the map with the digitizer through an Affine transformation. The parcel corners were digitized according to the rules which were established in Chapter 5.

PAGE 176

Existing Standard sheet Digitize Generate Topology Store in Master file Subdivision plans Retrieve Original survey data Compute Coordinates Keyboard Entry of coordinates 163 No Raw field data Cannot be included et Extract coordinates of common points from the Master file. Use coordinates to compute the values for the raw data Individual cadastral plans Use common points as controls and digitize Generate Topology Figure 6-4: Creating a Cadastral Index Map for Tanzania Due to the imprecision of the digitizer and the rounding-off errors of the computer (when working in single precision), a threshold of 1 meter was set so that nodes that fell within 1 meter of each other were considered identical, and therefore snapped together. By the node snapping rule, the new node snapped to the old one. The digital cadastral map was then checked to ensure that the boundary lines closed to form polygons. The software that was used (Arclnfo) contained a commands for isolating and correcting non-closing polygons as well as overshoots at the nodes. Topological relationships between the nodes, boundary

PAGE 177

164 lines and the individual polygons were established with the appropriate commands in the software. The finished digital cadastral map became the master index map. Although the standard sheet contained some geodetic control points the traverse boundaries of the original subdivision plans from which the standard sheets were prepared were not drafted. In order to merge the subsequent subdivision and cadastral plans to the digitized standard sheet the traverse boundaries of subdivision plans which produced the standard sheet were drawn and digitized. The traverse points provided the needed common points between adjacent subdivision plans. This process was described in Chapter 5. Subdivision plans. The subdivision plans which had not been used to update the hard copy standard sheet had all been drawn on mylar at a scale of I : 2500. Suitable and well maintained subdivision plans were digitized into the master index map Although the subdivision plans were drafted in the Tanzania geodetic referencing framework the coordinates were irrelevant at this stage. Registration of the subdivision plans to the master index map was accomplished by selecting at least three common points between the master index map and the subdivision plan, and performing the transformation. The reason for using common points between the two maps method was to avoid having to deal with too many overlapping parcels and boundary lines that crossed one another. In order to achieve this it was necessary to find at least three points that were common between the two maps. The use of topology ensured that the residuals that remained as a result of the least squares transformation did not affect the final outcome of the map. This was accomplished through the node equivalence and line

PAGE 178

165 equivalence relationships that were discussed in Chapter 5. A snapping distance of 1 meter was set for the subdivision plans too. Because there was contiguity among the parcels within any subdivision, minimal effort was needed to ensure that the spatial relationships were being maintained. However attention had to be focused on the individual parcels within the subdivisions It was necessary to guard against the influence of neighboring nodes and lines on vertices. Overshoots were introduced wherever necessary to ensure that the external lines did not snap to node unduly. Although the situation did not occur during the pilot study the intention was follow the procedure shown in the Figure 5-10 to compute coordinates for subdivisions that were too creased to be digitized and then transform the computed coordinates into the master index map. In situations where the common points had different coordinates due to errors in the survey controls the procedure would have been to transform the subdivision coordinates into the master index coordinates before doing the data entry with by the keyboard. Raw field data. Raw survey field data were treated in the same manner as subdivision plans that were unsuitable for digitizing. Raw field data that was based on local coordinate system were processed (through coordinate geometry) with controls that had been extracted from the master index map. In situations where there were no common points between the two maps, it was impossible to incorporate the data into the master

PAGE 179

166 index map. It had to wait until the data become available for doing the required computations. Individual cadastral plans. The easiest approach to incorporate individual cadastral plans that were plotted in a local coordinate system into the master index map was to identify common points in the master index map and to use those to digitize and to generate topology. Topologically structured overlay using a combination of existing cadastral maps was the most appropriate option since that did not involve large-scale production of a cadastral maps (which would be accurate in metric space), either by aerial photogrammetric methods or traditional surveying methods. A topologically structured cadastral index map has a great deal of potential in this regard. The effects of measurement errors and inconsistencies due to instrument type and methodology are secondary as a result of establishing spatial relationships in the topological structure. Linkage Mechanism Actual cadastral plan and field survey data may be associated with the parcel through a linkage mechanism. In a parcel based system, the linkage mechanism may be the same as the parcel identifier The developmental approach for a choice of linkage mechanism for cadastral data in Tanzania is discussed later in this chapter.

PAGE 180

167 Through linkage mechanisms the individual information support systems are related to each other. Descriptive information which may be also be associated with the cadastral index map are the legal property description of the parce4 valuation and assessment data, mortgage information, and other pieces of data that are registered in the land registry Similarly graphical data that may be associated with infrastructure information systems are road center lines pavements, and utility lines. Although each division and each section within the Ministry of Housing and Urban Development has a numbering procedure for its paper files there is no mechanism for linking files from one section to another. For example in the Land Development Division, all files pertaining to land allocations are given land office numbers which are issued by the statistics office A range of numbers is issued to each zone as they are needed. There is no way of isolating files from any particular zone or region without first knowing the range of numbers that have been assigned to that region over the years. After the range of numbers has been allocated to the region, the numbers are issued to transactions in the order that the transactions are recorded The zonal officers do not incorporate intelligence into the numbering procedure. Similarly file numbers are issued at the land registry. The numbers are sequential and are issued to the transactions in the order that they are received. Because the files that are created at the land office are separate from those at the land registry, some types of information are duplicated.

PAGE 181

168 Parcel identifiers. In defining a system for identifying parcels the cri t eria were to develop a numbering system which was simple easy to understand unique and manageable within the computerized environment. A general method for people in the urban areas of Tanzania to locate or describe their residence or property is the use oflocation name block number, and parcel number. For example Sinza, block 46 plot 124 is a valid and easil y understandable location description in Dar es Salaam. A recommended approach for assigning parcel identifiers is to use a hierarchical approach involving the administrative region, district location, block number and plot number. There are 20 administrative regions and 104 districts in Tanzania. The coding structure for a parcel identifier was for a region to have a 3 digit code beginning at 100 whiles districts would have a 3 digit code starting at 001. There are no districts with more than 99 townships or cities, therefore locations will be assigned 2 digit codes. Block numbers will have 3 digit codes, implying that there could be a maximum of 999 blocks within a location. Within each block plot numbers are assigned a 3 digit codes starting from 001. This means there could be a maximum of 999 parcels within a block. The approach for assigning hierarchical numeric codes to parcel identifiers is shown in Tables 6-1 and 6-2. There are 3 districts in Dar es Salaam; Ilala, Kinondoni, and Temeke The codes for these districts are 001, 002, and 003, respectively. To identify Kinondoni in Dar es Salaam, the code will be 102002. By assigning the area called Kijitonyama a code of 06 a parcel number 35, inside Block 44, in the Kijitonyama area, in the district ofKinondoni will have an identifier of 1020020604403500. The last two digits are reserved for any further

PAGE 182

169 subdivision, especially in the case of agricultural property. In the case of amalgamations, an entirely new identifier is assigned. A remark is made in the register to indicate that the new number replaces the two previous numbers as a result of an amalgamation. Table 6-1: Coding structure of Parcel Identifiers Item Numeric Code (digits) Region 3 District 3 Location 2 Block 3 Plot 3 Further subdivision 2 The parcel identifiers will be assigned by the survey division after the survey has been approved by the Director of surveys. Since the approved survey plan is the one that is used for land allocation, this means that all parcels will have identifiers before they are allocated. This does not prevent other sections or divisions within the Ministry from assigning their own numbering system.

PAGE 183

170 Table 6-2: Codes for Administrative Regions in Tanzania REGION COD E Arusha 100 Coast 101 Dar es Salaam 102 Dodoma 103 Iringa 104 Kagera 105 Kigoma 106 Lindi 107 Mara 108 Mbeya 109 Morogoro 110 Moshi 111 Mtwara 112 Mwaoza 113 Rukwa 114 Ruvuma 115 Shinyanga 116 Singida 117 Tabora 118 Tanga 119 In this chapter the problems and concerns that were identified during the stud y of the Tanzanian land delivery system have been addressed. A cadastral information model

PAGE 184

171 has also been developed for Tanzania. A topologically structured approach for developing a cadastral index map has been presented. A unique parcel identification system for Tanzania has been presented as part of the solution to the information management problems These approaches and solutions are tested during the implementation of a pilot project in chapter 7.

PAGE 185

CHAPTER 7 PILOT PROJECT This chapter begins by incorporating the solutions recommendations and approaches developed to address the problems and concerns that were identified during the study of the land delivery arrangements in Tanzania. The solutions developed include a topologically structured approach for developing a cadastral index map from varied sources of cadastral plans and subdivision maps that are available at the Surveying and Mapping division. Approaches were developed for removing inconsistencies and erroneous data from the descriptive records at the land registry as well as the land office. The "error free" records were linked to the cadastral index map and some spatial analysis was made on the data. The objectives of the pilot project were to assess the viability of the topologically structured cadastral index mapping approach and the subsequent ability to perform spatial analysis, and the records reorganization and moderniz.ation approaches that have been put forward. For this purpose, a 15 square kilometer area in Dar es Salaam was chosen for the pilot project. The area contained 11 subdivision blocks and approximately 5000 parcels (see Figure 7-1). Figure 7-1 shows an area containing the site for the pilot study. The site was chosen because that part of Dar es Salaam was developed in 1986 so the area was 172

PAGE 186

ow 0 ~: 0 1 0 0 I 2 Kilometers Figure 7-1: Vector Drawing showing Pilot Area N w E s Legend I>/. Property Boundary N PiiotArea

PAGE 187

174 believed to have relatively complete and up-to-date records. The implies that the results obtained from the pilot project would reflect the "best-case" scenario. Data Sources Data for the pilot project consisted of the relevant graphical data and textual information for the pilot area. The original data were compiled from multiple sources comprising of: 1. direct keyboard entry of mathematically computed survey coordinates, a digitization of hard copy cadastral plans. Some of the coordinates were based on the National Geodetic Coordinate system of Tanzania while other map coordinates were on local systems; 2. direct plotting from original survey field books; 3. digitization of hard copy cadastral plans, some of which were without coordinates, others were on a local coordinates system, while the remainder were on the national geodetic coordinate system. The pieces of graphical data were "stitched" together using topological relationships between identical parcel comer monuments and common boundaries between adjacent parcels. The composite vector map was incomplete as a result of missing records and maps that were too creased to be digitized. A geo-referenced aerial photograph was used as a backdrop to update the vector data. The aerial photographs were taken in 1992, using a metric aerial camera with a focal length of 152 mm. The flying height of the aircraft above mean terrain was 1900 m., yielding an average photo scale of 1: 12500.

PAGE 188

175 The descriptive records for the parcels were all obtained from the files of the Surveys and Mapping division. The records contained minimal information regarding the locatio' block name, and plot numbers for each parcel. A combination of the locatio' block name, and the plot number identified each parcel uniquely. Table 7-1 shows the attribute data that were obtained from the Surveys and Mapping division. Table 7-1: Textual information associated with Individual Subdivision/Cadastral Plans Location District City Block name Parcel number Calculated Area After the cadastral index map had been converted into the appropriate format, topology generated, and with all the errors removed, attribute information was added from the records at the Land Development division. Figure 7-2 shows a plot of the selected pilot project area after the spatial data had been converted into a software readable format. The type of attribute information which was associated with the cadastral map is shown in Table 7-1. A look up table was created to associate the data with the new parcel identifiers. The numbers were assigned on the basis of the information contained in the register. Relevant textual information regarding ownership, address, and rent assessment value, were obtained from the land office and the land registry in the Land Development division. The reason for using two sources of data was to reduce the chances of errors by

PAGE 189

Figure 7-2: Topologically Generated Cadastral Index Map of Pilot Area

PAGE 190

PORTION OF KIJITONYAMA/SINZA RESIDENTIAL AREAS s /\/ Blocks / v Parcel Boundaries Parcels F Derby

PAGE 191

178 isolating as many errors and inconsistencies as possible. Since each parcel has to be allocated before it can be registered it was logical to presume that the land office records will have more data pertaining to land allocation whiles the land registry will have only the records pertaining registered titles. The records showed that more parcels had been allocated than those that had actually been registered. The data from the land office were extracted from the paper files which were maintained at the office while the land registry data were extracted from the land register. All the textual data from both offices were entered manually into the computer. Hardware and Software As part of this research, computer hardware requirements were determined during the needs assessment phase. The choice of hardware configuration was based on the objectives of the project, compatibility with the software, ease of maintenance and cost. Local knowledge in computer technology eliminated the use of UNIX based workstations as well as any technology that required networking capabilities. Furthermore, existing infrastructure in Dar es Salaam, with regard to smooth (spike-free) electrical current, reliable telephone lines, and fire protected buildings imposed restrictions on the hardware choices that were made at the time of the pilot study. In 1995, the best configuration was a desktop Personal Computer which was equipped with an Intel DX4 chip that ran at 100 MHz. The available operating system was Windows 3.1 by Microsoft Corporation. Other hardware options included 321\,1B of Random Access Memory (RAM), and 1 Gigabyte of hard drive storage space.

PAGE 192

179 The software of choice included PC versionArclnfo and Arc View from Environmental Systems Research Institute (ESRI), Autocad from Autodesk Incorporated, and Dbase V for Windows from Borland Corporation. Analogue Image Conversion. Single aerial photographs normally have planimetric distortions such as those due to the camera attitude at the time of exposure, distortions due to manufacture, and image displacements due to ground relief As the mapping process was based on spatial relationships, spatial accuracy in the graphical overlay did not pose a problem since the focus was on relative positions of the parcels. In order to obtain a digital image with a ground resolution of 0.3 m., a micro-densitometer was used. For a camera focal length of 0.152 m., average flying height of 1900 m., the ground coverage on a .23 m. (9 inch paper) is ]900 X 0.23 = 2900 m. 0.152 On a .23 m. (9 inch) format, the ground resolution of .3 m. would require the micro densitometer to be set at 0.3 X 0.23 = 24 m. 2900 For a single photograph, the resultant digital image is approximately 83.7 MB. m stze. The image was saved in a Tagged Image File Format (TIFF).

PAGE 193

180 To geo-reference the image, control points which could be identified on both the image and the cadastral index map were selected. Corresponding pixels were assigned the coordinates that were obtained from the cadastral index map. An Affine transformation was performed after the location of the points had been identified in the image, to geo reference the image to the coordinates system of the cadastral index map. Although a Projective transformation would have been more accurate, the software that was used in this research did not have Projective transformation capabilities. Pixel values were obtained through a neighbor resampling method. As a result of the simplicity of the nearest neighbor resampling method, the possibility of pixels being displaced by up to one-half pixel was considered to be insignificant for the purpose of this research. A subset of the image was selected for the pilot project. An overlay of the property boundary map and the geo-referenced image was made. This exercise demonstrated the possibility of utilizing existing aerial photographs to update obsolete vector maps (see Figure 7-3). As can be evidenced at the south-west comer of Figure 73, there are structures on individual properties but the property boundaries have not been shown on the vector layer. The image may therefore be used to update the cadastral index map if evidence of a boundary such as a fence line or a hedge can identified on the image. On the other hand surveyors could visit the location and update that portion of the map without having to do an expensive and accurate survey.

PAGE 194

Figure 7-3: Topologically structured cadastral index map on a geo-referenced image

PAGE 195

182 Cadastral Map on Aerial Photograph 300 0 300 Meters N w E / Property Boundary s

PAGE 196

183 Attribute Data Processing The textual data which were extracted for the 5000 parcels were entered into the computer using Dbase V for Windows. Erroneous data such as a parcel allocated to more than one person or records that had erroneously been entered more than once. Because the land office and the land registry had independent numbers for each record (see Table 7-2 ), it was necessary to ensure that those internal numbers had not been duplicated. However, difficulty arose if the same parcel had been assigned two different internal numbers. This could happen if the same parcel was allocated to two or more people at different times. In a manual processing syste~ such mistakes can occur. During the manual data entry process, problems ranging from duplicate allocations to missing files were identified. Table 7-2: Relevant Textual Information from Title Office and Land Registry Land Office Data Land Registry Data Location Location District District City City Block name Block name Parcel number Parcel number Calculated Area Calculated Area Land Office number Land Office Number Rent Title Number Lease Term Lease Term Start Date Start Date Ending Date Ending Date Owner's Name Owner's Name Owner's Address Owner's Address Property Description Date of Registration Time of Registration Property Description

PAGE 197

184 An internal consistency check was conducted within the database to ensure that city name, block name, and parcel number were unique for all parcels. Parcels which were found to violate such conditions were isolated for verification. The procedures for checking and isolating errors followed simple rules which were described with the diagram in Figure 5-3. Although the procedure isolated several of the parcels with inconsistent records, there was no record of the number of parcels that are already pending adjudication before the courts. Some of the missing files could be in the custody of the courts, pending resolution of legal ownership. Parcel identification. In order to establish a unique identifier for parcels within the pilot area, utilizing the relational database concept, it was necessary to establish primary keys which uniquely identified each parcel. Table 7-2 indicates several items have been recorded twice between the two databases. Although the land office provided a unique identifier, it was also duplicated in the records at the land registry. The primary search key for the land office records, therefore, was a combination of location, district, city, block name, parcel number, and the land office number. This combination established a tuple which could be used to access the records within the land office records. To access records from the land registry database, a secondary key, which was the title number was used. The property description was long and verbose which made it unsuitable for parcel identifier. The newly created identifier had not been assigned to the descriptive data at this stage.

PAGE 198

185 The procedure for isolating inconsistencies were done by using the following simple rules: I. Perform internal checks on the database to ensure that no two records had the same location name, subdivision name, block name or parcel identifier. Sample pseudo code for checking record consistency is as follows: Begin name = city + block name + parcel number ; i = I; end. while (not end of list){ J = 1; } while (not end of list) { increment j; if (name(i) = name G)) then print "record number (i) is the same as record number G)''; } increment I; Inconsistent records were isolated for checking. If such records had the same owner's name and address, then it was just a recording error and one of the records was discarded. Those with different owners were forwarded for verification. The checking was done for the graphical data as well as the two textual databases. Table 7-3 shows a break down of the problematic files which were identified for two of the test areas during the manual data entry process.

PAGE 199

186 Table 7-3: Erroneous Records Identified during data entry at the Land Office Kijitonyama Block 44 Sinz.a Block A No. of Parcels 694 820 Missing Files 51 66 Files with inconsistent data 42 59 Total 93 125 Percentage 13.4 15.2 2. The land registry database was associated with the cadastral map and checked for consistency. Parcels which could not be associated with the appropriate records had not been registered although they may have been allocated to applicants and titles had been issued. Table 7-4 shows the number of records that were isolated during this checking process. Table 7-4: Erroneous Records Isolated during data entry at the Land Registry Kijitonyama Block 44 Sinza Block A No. of Parcels 601 695 Missing or Unregistered records 148 177 Files with inconsistent data Total 148 177 Percentage 24.6 25.5 3. After associating the land office records with the cadastral maps, parcels that could not be associated with records had either not been allocated to anyone or that the

PAGE 200

187 file for that particular parcel was missing (or misplaced). In this case, the only way to rectify this problem would be a site visit to obtain the original documents from the owner. Table 7-5 shows the results of the internal checking among the records at the land office. Table 7-5: Results of Internal Consistency Check among the Land Office records Kijitonyama Block 44 Sinza block A No. of Files 601 695 Parcels with Different owners 23 17 Parcels with identical numbers 15 6 Total 38 23 Percentage 6 3.3 4. The two databases from the land office and the land registry were related to each other with parcel identifiers. A check on the resultant information was performed to ensure that for each parce4 the name of the owner in the title office and that in the land registry are the same. Any discrepancies were again isolated for verification. Reasons were that a land transfer occurred for which the land office information was changed but the title has not yet been registered. Another possible reason could be the unlikely situation that a fraudulent transaction had taken place. Table 7-6 shows the results of this checking process.

PAGE 201

188 Table 7-6: Inconsistencies among Land Office and Land Registry records Kijitonyama Block 44 Sinza Block A No. of Parcels 563 672 Inconsistent L. 0. and L. R. names 17 29 No. of Error-free Records 546 643 Percentage 3 4 Overall Percentage of good records 78.7 78.41 5. Any of the remaining parcels that remained after all these checks could be considered to have been appropriately, allocated, titled and registered. Although the final result may be considered wholesome, the records still need to be verified because of possible unofficial transfers which have not been recorded anywhere. Furthermore, owners may have changed mailing addresses. In Tanzania, in the 1980s, several people used the office of the affiliated political party as their mailing address. Those political parties are now defunct but the records have not been rectified. It is therefore recommended that systematic record verification and address correction be done on the final list. Spatial Analysis Having combined the spatial data with the attribute information, ad hoc queries were performed on the resultant descriptive data to derive spatially referenced responses

PAGE 202

189 to the queries Various types of spatial queries were performed on the data. Table 7-7 shows the types of queries that could be performed on the spatial data and the associated descriptive information with explanations on the objective of the query. Some of the queries that were made on the data include: 1. Logically querying the records to find the number of unregistered parcels in selected locations within the pilot project area. This is important for the land managers to identify missing data within the land registry database. For effective land management, the cadastral information must be complete. Table 7-7: Query Types That Were Done on the Cadastral Information SPATIAL ANALYSIS GROUP QUERY TYPE Proximity analysis When information are needed about areas 2. surrounding existing features. Boundary operations When the information that are needed can be selected by their geographic locations. Logical operations When the information that are required can be I selected by their descriptive attributes only. Missing information also implies that the government has no way of collecting property taxes from the owners of those properties. Although this observation was not part of the original objective, the secondary benefit of this formation system is particularly important for economy of the country. Finding the variation in the land rent that owners of parcels within any particular block in the pilot project area are supposed to pay (see Figures 7-4 and 7-5).

PAGE 203

Figure 7-4: Land Rent analysis on Kijitonyama Block 44

PAGE 204

191 KIJITONYAMA BLOCK 44 SHOWING RENT DISTRIBUTION BY PARCELS IN SHILLINGS 250 LEGEND PARCELS 0 250 CJ (148) UNREGISTERED PLOTS CJ (23) 0 100 Shillings ( 418) 100 500 Shillings CJ ( 16) 500 1000 Shillings c=] (89) 1000 15000 Shilllings 500 750 Meters N F Derby

PAGE 205

Figure 7-5: Land Rent Analysis on Sinza Block A

PAGE 206

193 SINZA BLOCK A RENT DISTRIBUTION BY PLOTS IN SHILLINGS 2 50 LEGEND PARCELS 0 ( 177 ) U NREGISTERED PARCEL 0 ( 4 2 3 ) 0 100 SHILLINGS 0 ( 190 ) 100 500 SHILLINGS D ( 17 ) 500 1000 SHILLINGS 0 ( 13 ) 1000 42000 SHILLINGS Total : 820 Plots 250 500 Meter s N E s

PAGE 207

194 The result of this query identifies the weaknesses in the land assessment procedures especially when it was observed that wide discrepancies exist between identical parcels within the same block. The information from the database indicate the amount of revenue that the government should realize if an efficient revenue collection system is implemented. It also indicates that the amount of tax per parcel is too little for the government to be able to provide adequate infrastructure without supplementary with finance from other sources. The results of a query of this nature based on logical operation on the database, provides a justification for a land tax review. Summary and Analysis of Results In this section the results of the pilot project were reviewed with respect to the objectives and the focus of the research. The results were then analyzed with respect to the applicability to the Tanzania problems, the ability of the long term impact on Tanzania. Recommendations were then presented to ensure continuity and sustainability of the project in Tanzania. This analysis deals with the problems that were identified during the study of the land delivery processes in Tanzania and the solutions that were presented in the preceding chapters. The next chapter dealt with the conclusions and recommendations for further research with regard to the topologically structured cadastral model and other contnbutions to field of cadastral information management. The pilot project was developed to test the solutions and recommendations that were presented in Chapter 5 to address the problems and concerns regarding the Tanzania land delivery process. This pilot project demonstrated the following:

PAGE 208

195 an approach for creating a topologically structured cadastral index map to support land related information systems. The concept incorporates cadastral plans and maps from diverse sources with different spatial accuracy of the features. In establishing topological relationships rules and procedures have been developed for achieving the best results. a systematic procedure for isolating the errors in the parcel allocation process and the preparation of Certificates of Occupancy at the land office. a procedure for isolating erroneous and inconsistent records from the land registry records in Tanzania. an approach for maintaining consistency between the records of approved Certificates of Occupancy and the records to registered titles among the Tanzanian land records. The use of geo-referenced aerial photograph as a backdrop for digitizing and updating the cadastral index map. a flow-line for isolating discrepancies between the cadastral index map land office records, and the land registry records. a unique parcel identification coding system for Tanzania land records. a procedure for streamlining data processing within the land management agencies in Tanzania. templates to be used at the document reception counters to facilitate document processing within the Ministry of Lands, Housing and Urban Development. computers at strategic locations within MLHUD which will be used to monitor

PAGE 209

196 status of documents and to provide the necessary data for the land information system. Although partial results were presented earlier in this chapter for two of the eleven blocks the final results for the initial 5013 parcels are presented here for discussion. During the manual data entry process both at the land office and the land registry some erroneous records were identified and rejected. Table 7-8 shows the number that were identified at both offices. This is an indication that just the simple act of entering the records in the computer offered an opportunity to isolate 14.2% of the records at the land office. At first glance it may seem that the 31 % of the records at the land registry were either unaccounted for but the number indicates titles that have not been registered. Some of these may be titles waiting to be registered. Table 7-8: Internal Inconsistencies Identified During Manual Data Entry Land Office Land Registry Number of parcels 5013 5013 Missing records 455 1554 Files with inconsistent data 258 0 Total rejected 713 1554 Percentage 14.2 31

PAGE 210

197 After the data had been entered into the computer internal checks which were made to isolate duplicate records showed records which contained either duplicate land office numbers or parcels with duplicate owners. A total of 222 records were rejected. Table 7-9 shows results of the internal consistency check on the data for the pilot study. Table 7-9: Results of Electronic Inconsistency Consistency Check on the Data Land Office Land Registry Number of Records 4300 3459 Inconsistent Owners 137 0 Duplicate Numbers 85 0 Total 222 0 Percentage 5 2 0 These results indicate the diligence of the staff at the land registry. But the fact that no errors were isolated at the land registry does not indicate that the names of the owners are consistent with the records at the land office. The next check would identify inconsistencies in the records between two offices. The final check was to compare the records from the land office with the records at the land registry. Table 7-10 shows the results of the comparison.

PAGE 211

198 Table 7-10: Comparison ofland Office and Land Registry Records Land Office Land Registry Number of Records 4078 3459 Missing records 619 Inconsistent Owners 58 58 Total 677 58 Percentage 16.6 1.7 The comparison shows the possibility that there are 619 Certificates of Occupancy that seemingly have accurate information and are possibly waiting to be registered at the land registry. Altogether 3401 records out of the initial 5013 records passed the accuracy test. This constitutes approximately 68% of the data. This implies that even for the best case scenario only about 68% of the records may be considered error-free. The next action will be to check the validity of the owner's mailing address and other pertinent information within the databases. This could be done through systematic verification of the records. As far as purging the records of possible inconsistencies, the results indicate that the records at the land registry are reasonably reliable except for the current particulars of the owners. The results of the rent analysis show the effectiveness of the parcel identification system and the topological structure which was adopted to create the cadastral overlay. The rent analysis is one of many analyses to which the records can be subjected. This simple test shows large disparity between land rents for identical parcels in the same

PAGE 212

199 location. The results of the test suggest that a land rent review should be conducted so that fair and equitable land rent will be paid by all. Although some of the problems and concerns mentioned in this research are specific to Tanzania, the topologically structured cadastral index mapping approach is a solution that can be beneficial to many countries that are contemplating the production of a cadastral index map. I,n this pilot study the topologically structured cadastral index mapping approach has been successfully tested. Until now the approach for producing base maps for cadastral and land information management purposes has been in favor of accurate base maps which necessitated comprehensive aerial photography followed by stereo compilation and map production. Very often the map production was done at two or more scales (see Sonnenberg 1994 ; NRC 1983) one scale for urban areas and another for rural and agricultural areas. Not only is the topological approach economical (no extensive field surveys involved) it is also a time saver. The results of the pilot project shows that the solutions and recommendations that were presented are feasible. The method for producing a cadastral index map has great potential application in developing countries. It was produced with no field survey work. It therefore has some economic advantage over full scale base map production. The procedure for reorganizing existing records isolated a substantial amount of the inconsistent records. Although a large portion of those records were identified during the data entry process, the systematic approach presented here ensures consistent records

PAGE 213

200 and promotes the integrity of the records at the Surveying and Mapping divisio~ the Land Office and the Land Registry.

PAGE 214

CHAPTER& CONCLUSIONS AND RECOMMENDATIONS In Tanzania availability of land is being gradually reduced by natural disasters such as coastal erosion, and increased pressure from population increase. The need to closely manage the land and its resources is becoming greater with time. In this research, it was shown that for some land management activities accurate cadastral maps are not essential. It has also been shown that whereas computerization of land records is essential for informed land management decisions mere computerization does not resolve inherent problems of inconsistencies incomplete and non-current data caused by maintaining records on paper. Procedures to remove such inconsistencies in the records were a part of the focus of this research. The contributions of this research to the cadastral information field are the establishment of rules and procedures for compiling a topologically structured cadastral index map that is consistent in detail and robust enough to facilitate spatial analysis including the development of a topologically structured cadastral index map. Whereas this research does not reject the importance of spatial accuracy in cadastral mapping, the topologically structured approach for creating the cadastral index map is an initial step for countries with limited financial resource to have an inventory of land distribution and to promote better land and resource management. With this modei spatial accuracy of the cadastral index map is not paramount. This approach is not 201

PAGE 215

202 restricted to geographic region or location, and therefore may be adopted in situations where available data are not accurate enough to apply the other models that have been available until now. The pilot study proved that by associating descriptive information to the "less accurate" cadastral index map, it is possible to perform analysis on the data and obtain information that is spatially accurate and pertinent to land and resource management. This was shown, for example, by assessing the land rent distribution within the pilot study area. In conducting this research in Tanzania, certain observations and recommendations were made together with procedures to improve land management activities in Tanzania. These recommendations and procedures can contribute a greatly deal to improved land delivery processes in Tanzania. If adopted, the recommendations and procedures that were developed in this research will greatly improve land management activities in Tanzania. With regard to Tanzania, this research has looked at the responsibilities of public agencies that deal with land management in Tanzania and showed: that record keeping and maintenance, which is the cornerstone of any information system, has been given a low priority in Tanzania. that overlapping responsibilities between the city councils and the division of urban development exist. multiple agencies have authority to allocate land.

PAGE 216

203 A great contribution of this research to Tanzania is in the design of a process for cadastral index mapping. This research developed procedures for reorganizing land records without the need to initiate a large scale survey, titling, registration compilation from the start. The approach for developing a topologically structured cadastral index map has great potential for developing countries where resources for such large scale initiatives are scarce. Within the Ministry of Lands Housing and Urban Development organizational arrangements aimed at faster data processing and approaches for efficient record keeping and maintenance were developed. The objective was achieved by: Using existing data. Developing a procedure for computerizing existing land records. Developing a procedure for isolation of inconsistent data from the land office and the land registry. Developing an approach for improving the integrity of existing records. Developing a cadastral information model for Tanzania. Developing an approach for implementing a cadastral information system as a foundation for an integrated geographic information system for Tanzania. Developing procedures for streamlining data processing and facilitating data capture within land management agencies. Developing an integrated approach for incorporating new records into the cadastral information system.

PAGE 217

204 The development of a land parcel identification system for Tanzania, based on the Tanzanian way of describing residential addresses. A move to a adopt land market policy is a major milestone in the land policy reformation of the Tanzania government. While catering to the expected increase in the registration of land related transactions as a result of the move, computerization of land records became the most appropriate option. It was important to begin with the cadastral records since those records form the core data for a land information management system. With continued reduction in the cost of computer hardware and software, increases in computer hardware and software capabilities, and the associated benefits of complex analysis to support land administration, many developing countries are considering converting their existing land records from paper format into a computerized Land Information System. Success of implementing a cadastral information system will depend, not only on the design and application, but also on the economic and political climate in which the system operates, the manner and skills with which the potential benefits are presented to the public, and the intangible benefits that the government and the land managers derive from knowing that a decision has been made with the support of the most up-to-date and complete information resources that are available. Recommendations Access to an accurate cadastral map is still the objective of many land managers. Whereas the "less accurate" topologically structured cadastral index map provides a

PAGE 218

205 suitable starting point for land managers, an accurate map would have greater potential for spatial analysis in support of land related decision making. Having successfully developed the cadastral index map by the topologically structured approach, future research could improve the spatial accuracy of the topologically structured cadastral index map. One possible approach is to strengthen localized regions of the map as and when more accurate data become available. Another approach is to distort the more accurate data to fit the map in its current state, and to periodically transform the entire map with a combination of all the available accurate data. These approaches require further investigations, which may involve the development of mathematical models for improving the spatial accuracy of a topologically structured cadastral index map. This research has focused on the urban lands in Tanzania. It is important that similar research be conducted on the rural lands. As rural land in Tanzania is shared by communities, individual property boundaries in rural communities of Tanzania are not as well defined as properties in the urban areas. Records management in rural areas may require further restructuring of the database. Despite the falling cost of mass storage media in the computer industry, large volumes of data will be processed with heads-up digitizing and more so, with softcopy photogrammetry. This requires large memory for storage, retrievai and processing of the data. The need to improve processor speeds, especially for graphic processing is becoming more apparent as application of raster data to mapping gains popularity. Although agencies such as National Aeronautics and Space Administration (NASA) is

PAGE 219

206 conducting research into image compression algorithms, it may be appropriate to develop image processing algorithms and methodologies that are relevant to developing countries and priced within the financial capabilities of such countries. This means developing image processing algorithms that do not require high end computer hardware to operate. Analysis of land records is only the beginning of the numerous management options that are available for land managers. This research is incomplete unless it is linked with economically feasible and faster ways of determining property boundaries, adjudicating actual ownership, and resolving boundary disputes. During the research, inconsistent records were isolated but no attempt was made to resolve them. Further research could develop methods and guidelines for resolving boundary disputes and adjudicating property ownership. These are problems that exist in many developing countries. The computerized cadastral information system opens up wide areas of research which can be conducted to generate revenue and to ensure an equitable land rent for the community. This research evolved through an objective to provide the infrastructure for the government of Tanzania to facilitate land related transactions as a result of the move towards a land market policy. In the process, a topologically structured approach to cadastral index has been developed which could benefit developing countries with meager financial resources to manage their land and to keep consistent records.

PAGE 220

APPENDIX A-1: TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND OFFICE

PAGE 221

208 MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT LAND REGISTRY P.O. BOX 1191 DAR ES SALAAM. REGISTRATION OF CERTIFICATE OF OCCUPANCY RECEPTION COUNTER Reference your letter No . .... .......... .. .. . . ... .. ... .... dated .. . . .. .............. ......... with the enclosures the documents cannot be accepted for failing to submit the following : Copy of certificate of incorporation. Consent of the Administrator General. Evidence of payments of stamp duty/registration fees. An approved plan of the parcel ofland farm or plot. Other: ................ ... ... .. . ..... .. .... .... .. . . . . .. .. . .. . .. .... . ......... ... .. .. .. ...... ... ....... ..... .. .. . You are therefore advised to resubmit the documents together with the items indicated on th i s list. for: REGISTRAR OF TITLES DOCUMENT PROCESSING Reference your letter No ... .. ....... ........................... dated ............. . .. .... . ...... ... with the enclosures the documents cannot be registered for the following reasons: Documents were not sealed/ signed by Commissioner for Lands. Document(s) not dated. Ownership not witnessed by Authorized Person. Area of plot given in deed plan differs from that in schedule as well as Units of Area Used. Land description not uniform on cover page schedule and in deed plan. Distance from main road to the farm to be given. Type of ownership not specified i.e. Tenants in common/Joint Tenants. Other: .......... ... . . .. .. ... ... .. .. .. ...... ... ....... ............ ..... . .................................... .......... Please resubmit the documents after corrections have been made. for REGISTRAR OF TITLES

PAGE 222

209 MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT LAND REGISTRY P.O. BOX 1191 DAR ES SALAAM. REGISTRATION OF GENERAL DISPOSITION RECEPTION COUNTER Reference your letter No ......................... dated .............................. with enclosures, your documents cannot be accepted because the following were not submitted: Duplicate copy of the Documents. Title Deed. Capital gains tax clearance certificate. Land rent and service charge clearance certificate. Certificate of incorporation. Certificate of registration of a charge. Other: ...................................................................................................... Please resubmit the documents together with the item(s) marked on this list. for REGISTRAR OF TITLES DOCUMENT PROCESSING Reference your letter No ......................... dated .............................. with enclosures, your documents cannot be processed for the following reasons: Signature/Name of the occupier differs from that in the land register. Documents not in prescribed form. Description of land differs from that in the land register. Underpayment registration fees/stamp duty. Stamp duty paid after the due date. Attestation and execution clauses are improper i.e. document not dated or signed. Ownership not specified i.e. Tenants in common/Joint Tenants. Other: ...................................................................................................... Please resubmit the documents after the corrections have been made. for REGISTRAR OF TITLES

PAGE 223

APPENDIX A-2: TEMPLATES FOR PROCESSING DOCUMENTS AT THE LAND OFFICE

PAGE 224

211 MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT OFFICE OF THE COMMISSIONER FOR LANDS CERTIFICATE OF OCCUPANCY RECEPTION COUNTER Reference your application dated ......................... with the enclosures, the documents cannot be accepted for registration because the following were not submitted: D Approved survey plan D Official letter from the cell/ward leader. D Consent by Administrator General. D Village/District/Regional committee minutes. D Certificate of incorporation to be provided D Advice of payment/receipts to be produced D Letter of offer to be produced Other ........................................................... You are therefore advised to resubmit the application together with the item(s) indicated on this list. for COMMISSIONER FOR LANDS DOCUMENT PROCESSING Reference your application dated ......................... with the enclosures, the Certificate cannot be approved due to the following reasons: D Letter of offer is invalid D Deed plans not valid D Certificate not dated D Certificate to be signed/sealed by the owner D Land description to be uniform on Certificate of Occupancy cover, Schedule and on the Deed Plan D Occupier names to be same in all three places in the Certificate of Occupancy D Ownership to be witnessed by an authorized person D Type of Ownership/Share to be specified e.g. Tenants in common/joint tenancy D Improper drafting/typing of Certificate of Occupancy D Underpayment of fees D Fee for Certificate of Occupancy D Registration Fee D Survey Fee D Deed Plan Preparation Fee D Stamp Duty D Other .................. .. .................................................................................................... Please resubmit the application after corrections have been made. for COMMISSIONER FOR LANDS

PAGE 225

212 MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT OFFICE OF THE COMMISSIONER FOR LANDS TRANSFER OF OWNERSHIP RECEPTION COUNTER Reference your application dated ... .. ........... . .. .... with the enclosures the documents cannot be accepted for registration because the following were not submitted: D Certificate of Occupancy D Transfer Deeds D Land Rent Clearance Certificate D Capital Gains Tax Clearance Certificate. D Evidence of payment of fees D Registration Fee D Letter of Offer D Certificate of Incorporation D Application for Consent D Valuation Report D Consent fee D Stamp duty D Other ... .. ... . . ....... .... . . . .. ...... ..... .. . You are therefore advised to resubmit the application together with the item(s) indicated on this list. for COMMISSIONER FOR LANDS MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT OFFICE OF THE COMMISSIONER FOR LANDS MORTGAGES RECEPTION COUNTER Reference your application dated ....... ... ............. with the enclosures the documents cannot be accepted for registration because the following were not submitted: D Certificate of Occupancy D Cover letter from Bank/Grantor D Valuation Report D Evidence of payment of fees D Consent Fees D Stamp Duty Other .... .... ....... ... ............................. . D Application Letter from property owner D Mortgage deeds D Certificate of incorporation D Certificate of Registration of a Charge D Registration Fee You are therefore advised to resubmit the application together with the item(s) indicated on this list. for COMMISSIONER FOR LANDS

PAGE 226

213 MINISTRY OF LANDS HOUSING AND URBAN DEVELOPMENT OFFICE OF THE COMMISSIONER FOR LANDS LEASES RECEPTION COUNTER Reference your application dated .. .. .. . .. . . . . . . with the enclosures, the documents cannot be accepted for registration because the following were not submitted: D Certificate of Occupancy D Valuation Report D Contract / Agreement / Lease document D Consent Fee D Registration Fee D Land rent Clearance Certificate D Letter of Consent D Evidence of Payment of Fees Dstamp Duty 0 Other .............................................. You are therefore advised to resubmit the application together with the item(s) indicated on this list. for COMMISSIONER FOR LANDS

PAGE 227

APPENDIXB EVOLUTION OF LAND TENURE POLICIES IN TANZANIA Over the last hundred years land policies have been implemented without regard to the rights of the indigenous Tanzanians. The social structures of some communities have been destroyed by moving them away from their economic bases. The customary rights the people have been voided and replaced with western system of tenure. The result has been slow economic development, illegal land markets conflicts over land loss of pastoral rights and squatting. In reforming land policies, recognition cannot be given to the customary rights without first knowing what those rights were and the events that led to current situation This paper looks at how the rights of the indigenous people were taken away by the first European settlers and how post-independent governments have not been able to rectify the mistakes of the past. Pre-independent T anz.ania. Prior to colonization by the Germans in 1885 the land tenure structure in Tanz.ania (then Tanganyika) was based on traditional law and the culture of the respective tribes within each community. The components of the communities were clans families and individuals. The individual, as a member of the group acquired rights in the land that he and his family could clear, cultivate and manage. Whenever the land showed signs of 214

PAGE 228

215 deterioration, a new piece of land was carved elsewhere in a fallow area while the soil on the old cultivation was allowed to re-nourish itself Each family could cultivate approximately five acres for subsistence farming. The system worked very well since land was abundant and population was low. Each tribe consisted chiefs, elders and headmen who controlled and allocated land to members of the tribe on a fiduciary basis. Members were prohibited from disposing of land to members of other tribes or families, except with permission from the grantor's family, heirs and sometimes from the chief of the community. People from other tribes could gain membership to another tribe only by marriage. Both paternal and maternal inheritance existed in various parts of Tanganyika. In 1884, a German explorer and adventurer Dr. Karl Peters was granted large tracts ofland by the local chiefs in consideration for trinkets. Soon after, the German ruler, Kaiser, extended Germany's protection to all its territories. The period between 1884 and 1891, was when Europeans were scrambling for land in Africa. The European objective was primarily to exploit the African resources in land and minerals, promote European settlement and promote plantation agriculture in sisal, rubber, cotton, coffee etc. In an effort to obtain protection from the German government, Karl Peters transferred the lands into his company "German East African Company". The Europeans were partitioning the continent and dividing the inhabitants among their spheres of influence. East Africa fell to Britain and Germany. Germany was granted the region that is now Rwanda, Burundi and Tanzania. The people resisted the subjugation by engaging in

PAGE 229

216 bloody wars. In 1905, the Maji-Maji rebellion was crushed by the superior firepower of the Germans. Land alienation and land grabbing thus became the central core of their objectives. The Imperial Decree ofNovember 26, 1895 In its capacity as the conqueror, and with ultimate interest in land resources the Germans initiated the tenurial device for declaring all lands "Crown lands" and vesting them in the state. In 1896, the Decree was revised to distinguish between ownership and 'rights of occupation'. Ownership could be proved by documentary evidence whiles Right of Occupation may be exercised by the fact of cultivation and continuous possession Under the Decree, indigenous lands were unowned and title was vested in the political sovereign ( the state) except lands alienated to settlers and evidenced with documentary title. Settler lands became owned lands while indigenous lands were unowned and therefore became state owned by virtue of the decree. By this Decree the Germans succeeded in merging sovereignty with property. The state therefore not only ruled, but owned land. This gave the Germans absolute power over the indigenous people of German East Africa. The implications of the decree were that only settlers could prove their title and therefore enjoy the attendant security of tenure. Indigenous people were left with mere rights of occupation. Under the Decree land could be alienated either by outright sale or by lease. The policy always changed between plantation agriculture which was run by settlers and peasant cultivation. Land was alienated in favor of the settlers. The contract of sale contained certain conditions

PAGE 230

217 which permitted the state to take back the land upon payment of the appropriate part of the purchase price together with compensation for improvements. This was referred to as the 'resumptions clause'. Leases were for definite or indefinite periods. Definite period leases were for a maximum of99 years. Both methods contained the option to purchase upon fulfilhnent of development covenants. By the time the Germans left in 1923, about 1.3 million acres had been alienated to immigrants. In densely populated areas, the alienation had already caused local land shortages and resulted in impoverishment. LAND TENURE OWNERSHIP RIGHTS OF OCUPATION Customary Tenure Figure B-1: Effect of 1896 Land Tenure Amendment The British rule 1923 to 1961 With the defeat of Germany in World War I. Tanganyika passed into the British while Rwanda and Burundi were governed by Belgium. The British government was made the administrator of Tanganyika as a Trust Territory of the League ofNations. After World War II, the international status of Tanganyika continued under United

PAGE 231

218 Nations agreement. Mandates were set for governing Tanganyika. Articles 6 7 and 8 of the United Nations Mandate says among other things that: ... in reforming the laws relating to the holding and transfer ofland and natural resources, the administering authority shall take into consideration native laws and customs and shall respect the rights and safeguard the interests both present and future of the native population. No land or natural resources may be transferred except between natives, save with previous consent of the competent authority. No real rights over native land or natural resources in favor of non-natives may be created except with some consent of a competent authority .... This provision came to play an important role in shaping the land tenure laws during the British period. Land policy under the British rule was influenced by two conflicting interests; first of all, the status of Tanzania under international law as a Mandate and a Trust Territory, and secondly, the colonial policy to develop the country as a plantation/peasant economy (as opposed to settler economy) producing essentially cheap agricultural raw materials. It was the need to resolve the conflict between peasant and plantation sectors which underpinned much of the land policy and administrative actions of colonial British. About 1.2 million acres of land alienated by the Germans were sold as enemy property after the war. Of this, 75% was bought by the British, 20% was bought by Indians, 4.5% was bought by other Europeans and 0.2% was bought by Africans. It was believed that most of the Indians acted as front men for the Germans to repurchase their properties because the Germans were prohibited from buying these properties. By this process, land alienated to the Germans continued to remain in the hands of non-natives. German freeholds were recognized by the the British as English freehold. Leaseholds continued to carry the resumptions clause which was initiated by the Germans. The

PAGE 232

219 LAND TENURE OWNERSHlP RIGHTS OF OCUPATION Customary Tenure DEEMED RlGIITS GRANTED RlGIITS Figure B-2: Tenure Structure after 1928 Amendment British instituted the Land Tenure Legislation of 1923 which was referred to as the Land Ordnance. The mandates 6, 7 and 8 of the United Nations were recognized only in the preamble. As the preamble was not considered a part of the law, tenure had to be judged on the strength of the Ordinance. The Ordinance was very much a repetition of the German Decree except that the words "Crown lands" and "Empire" had been replaced with "Public lands" and "Governor" respectively In 1928, the Ordinance was amended to give statutory recognition to customary law titles. The customary law was extended to include the title of a native community lawfully using or occupying land in accordance with native law and custom. Title was vested in a corporate body such as the clan, tribe or community. A customary rights holder has indefinite interest during "good behavior". Since then the customary law title has become "Deemed Rights of Occupancy". Rights of occupancy was granted for a definite period to individuals or families. It was defined as a title to the use and occupation of the land. "Use" was further defined by attaching development conditions

PAGE 233

220 to the grant. Security of tenure depended upon use. The amendment did not specify whether the Deemed Rights of Occupancy enjoyed the same security of tenure as the Granted Rights. This issue was left to the interpretation of the administrators By virtue ofthis amendment, customary rights were governed more by administrative policy than by legal stipulations. In two court cases the courts observed that the occupation of land by Africans was merely "permissive" and did not establish any rights against the government. In short, the provisions of the law provided virtually no legal security to customary rights holders. Practically the state could and did alienate indigenous lands to settlers as and when it so desired, depending on its policy and often contrary to the present and future interests of the indigenous population. The Ordinance of 1928 gave the government full powers to deal with land according to whatever administrative powers were adopted at the time. So long as the state desired peasant production, the occupants would be considered to have deemed or permissive rights. When land was needed for the purpose of alienation the governor would withdraw 'consent' and alienate it as a granted right to a settler, presumably against the interests of the inhabitants. Indigenous people were regulated by administrative circulars rather than the subject of guarantee or security in law. Alienation of Indigenous People Very few grants were made during the British period. Under the Land Ordinance the governor was empowered to lease the land either to a native or non-native through grants of occupancy for 99 years. In practice, this power was used almost exclusively to

PAGE 234

221 alienate land to non-natives since the indigenous people occupied the land by virtue of customary law (deemed rights of occupancy). By 1954, only 8 Africans held 2,482 acres ofland under long term rights of occupancy. The following year the number fell to 2 Africans holding 136 acres between them. In 1926, the alienation process was extended to the southern part of Tanganyika. The Southern Highlands were opened to white settlers. Within 2 years 222,000 acres of land were alienated from the indigenous inhabitants. In a token effort to recognize the united Nations mandate, and deflect international criticism, an amendment was attached to the Land Ordinance in 1950. It stipulated the consultation of the native authority of the area before any land was disposed. In 1953, an administrative circular made it clear that the obligation to consult the Native authority did not imply an obligation to obtain consent. Following recommendations by the East African Royal commission in a 1955 report, a government paper entitled ''Review ofLand Tenure Policy'' was issued. The colonial government proposed the introduction of "a form of tenure which is individual, exclusive, secure, unlimited in time, and negotiable" (freehold). The paper briefly reviewed the defects of customary tenure. The most important reason was that the development of customary tenure had been too slow to keep pace with economic advancement. The then President of the Tanganyika African National Union (TABU), Julius Nyerere, argued against the proposal in an article. Excerpts of the paper are given below: "If people are given land to use as their property, then they have a right to sell it. It will not be difficult to predict who, in fifty years time, will be the landlords and who the tenants. In a country such as this, where generally speaking Africans are

PAGE 235

222 poor and the foreigners are rich, it is quite possible that within eighty or a hundred years, if the poor African were allowed to sell his land all the land in Tanganyika would belong to the wealthy immigrants and the local people would be tenants. But even if there were no rich foreigners in this country there would emerge rich and clever Tanganyikans. Ifwe allow land to be sold like a robe within a short period there would be only a few Africans possessing land in Tanganyika and all others would be tenants. If two groups emerge a small group of landlords an a large group of tenants we would be faced with a problem which has created antagonism among our peoples and led to bloodshed in many countries of the world. Our forefathers saved themselves from danger by refusing to distribute land on freehold basis. The government proposals were dropped. Tanganyika went on to become independent in 1961 and changed its name to Tanzania. Land Tenure After Independence The new Tanzania government also inherited the conceptual and the major framework on land tenure from the colonial period. The most significant amendment to the Land Ordinance was to replace the term 'Governor' wherever it appeared in the Land Ordinance with the term 'President'. With the President as the head of the executive the allocation and administration of the land fell under the executive of the government Some of the major policy changes that have had an explicit impact on land tenure in post independence Tanzania are to do with freehold the semi-feudal system (Nyarubanja) and the Landlord Tenant relationship The freeholds were converted into government leaseholds in 1963. About one million acres of land including some prime properties in urban areas were converted. The leaseholds were for a maximum of99 years starting form the day of conversion. In 1969 the Conversion of Rights of Occupancy Act was used to convert all leaseholds into Granted Rights of Occupancy. Thus freehold finally

PAGE 236

223 came to rest in one form of statutory tenure the Rights of Occupancy System under the Land Ordinance. The semi-feudal system of tenancy existed mainly in the Lake District. LAND TENURE DEEMED RIGHTS GRANTED RIGHTS Figure B-3: Land Tenure Structure after 1969 Amendment The system existed under various names including Nyarubanja. The system operated just like the feudal system of ancient times. A new Act (The customary Leasehold Act of 1968) was enacted. It had wider implications than simply abolishing the feudal type of tenure. The Act sought to enfranchise all types of customary tenancies. The Arusha Declaration The Arusha Declaration of 1967 is a significant divide between two major periods. Land tenure policies of the respective periods stand in sharp contrast with each other. Two major attempts at land reform stand out in the pre-Arusha period. The village settlement scheme. The Range development schemes of the 1960's.

PAGE 237

224 Two other attempts in policy reform which have impacted the land tenure system in the post-Arusha period are: The villagization program of 1973-74 (which started as 'operation vijiji' and later came to be known as 'operation Tanzania'). The granting of village rights of occupancy (village titling) in the late 1980's. The Pre-Arusha Period Immediately after independence two reports were crucial in guiding the approaches of the newly independent state to rural development. The Village settlement report was prepared by the International Bank for Reconstruction and Development (IBRD). The report was the basis for a strategy which was adopted in the setting up of the rural Settlement Commission in 1963. It was followed by the Land Tenure (Village settlement) Act of 1965. The second report was for the development of the Pastoral Sector This report was prepared by the United States Agency for International Development (USAID) in 1964. The Village settlement Program. The village settlement program offered what was referred to as Derivative Rights. Occupancy was granted under the Land Ordinance Act to the rural settlement Commission for the purpose of village settlement. The settlement rights were among other things, license for use and occupation, leases, easements way leaves and profits. The village communities had no powers to transfer, exchange, mortgage or change the land held under

PAGE 238

225 settlement right or to grant any lease therein. Derivative rights had little security of tenure. Tenure could be forfeited for any number of reasons including breaches of rules made by the commission on cultivation of land, harvesting or marketing of crops. The settlement schemes were not a great success and were abandoned or closed down. Pastoral Sector Development {Range Development) The government envisaged a Statutory Commission to oversee and develop resources in areas which would be selected for Range development by the Minister. The expectations were that Ranch associations would be formed which would be granted rights of occupancy. Subsequently, in range development land where associations held rights of occupancy, the existing customary rights would be extinguished. Again, there was a maze of regulations and bye-laws which controlled the members and the use of the land. Penalties ranged form seizure of stock to ultimate expulsion from the range land. The range development also failed for the same reasons s the settlement scheme. The Post-Arusha Period The policy statement of President Nyerere in his pamphlet "Socialism and Rural Development' (1968), formed the basis of the immediate post-Arusha villagiz.ation. Living and working together with some form of communal ownership of land was considered the cornerstone of 'Ujamaa' villages. Such villages were to be formed voluntarily by the villages themselves. The role of the government was to encourage and facilitate the process ofUjamaa rather than coerce people into such villages. The government selected

PAGE 239

226 and demarcated the sites. The government also helped to move people into the villages. Between 1969 and 1972, the process was slow and unsatisfactory from the point of view of the ruling party. On November 6, 1973, the President declared that living in the villages was no longer voluntary and that by the end of 1976, the whole rural population should have moved to villages. Thus began the large scale movement into villages. Operation Tanzania was born. The Ujamaa Villages Act was repealed in 1982. Currently it is not known what legislation, if any, govern the villages. It is assumed to be customary law. However, where pre-villagization customary land rights collide with post villagization allocations, the inclination of the courts have been to uphold the customary rights. Another amendment to the Local Government Laws (1992), following the multi party system has made the position of the village chairman an elected office. The Village Assembly consists of all members of age 18 years and over. The Village Council is composed of not less than 15 members and no more than 25 elected members, of whom at least one quarter must be women. Current Land Tenure Policy in Tanzania Current land law of Tanzania is essentially what was introduced by the Germans and later incorporated into the Land Ordinance by the British in 1923. A few changes have been made since independence such as the abolition of freehold and landlord/tenant relationships. Under the statute, All land is publicly owned and under the control of the state.

PAGE 240

227 Land rights and titles are based on use Commoditization and speculation in land are illegal Rights of occupancy the only recognized tenure are held in two ways: a. Under granted rights of occupancy which is give' subject to development conditions for up to 99 years. b. Deemed rights of occupancy or customary tenure which subject to use are held in perpetuity. According to the statute cultivated land is either held under granted rights of occupancy characteristically by a commercial fanner a corporation or a parastatal using capital intensive technology or is held under deemed or customary rights by a peasant farmer living in one of Tanzania's village communities I LAND POLICY! LAND TENURE I WESTERN TENUREH CUSOMARY TENURE I (formal) (informal) Figure B-4: The Land Policy in a Dilemma

PAGE 241

228 In an effort to administer the land, governments have implemented land policies to guide the :framework for land tenure and land administration. As has been demonstrated above, the policies that are implemented by the Tanzanian government have little or no significance to the customary tenure. In most cases the policies are adaptations of the western tenure model. Because the western model is controlled by statutes and regulations, it always gains recognition and hence protection by the law. The majority of the citizens are familiar with the customary tenure and consider the western system an alien principle. The implication is that people with customary tenure fail to register their property. Therefore do not have the security that is offered by the western system. A vast majority of Tanzanians cannot use the land as collateral. The most important problem is that the government has no way of knowing how many acres of land are in private hands or whether the maximum economic benefits are being derived form the land. The solution to an effective land tenure policy and hence efficient land administration in Tanzania lies in the ability to reconcile the two systems of tenure. The Tanzania government recognizes that current land policies are causing problems in the following areas: Land law, administrative process and access to natural resources Administrative practices of regulating access to land in Tanzania, lacks coherence. Land allocation does not take into adequate account such issues as existing patterns ofland use, customary rights, or land value and hence do not ensure justice or foster long-term investment. Land use planning.

PAGE 242

229 Land use plans were prepared from aerial photographs with brief visits to the villages concerned. The method took little or no account of location-specific patterns of resource use individuals' customary rights and common property regimes. In some cases there have been arbitrary denial of access to resource as a result of land use planning. Pastoral land use and common property regimes Since independence large land grants have been made to parastatals for the production of wheat. This has resulted in encroachment in by cultivators on dry season pastures, displacement of herds and people and heightened inter-group conflict and litigation. Forest law and tree management The statutes governing trees do not take into account the complex relationships between people and trees. All trees are classified among protected forests. All protected species are placed under state control. The use of protected trees are regulated through a system of permits and royalties. This process creates poor incentives for local tree and woodland management. Land markets Although illegal, land markets are developing and affecting land under customary as well as under registered tenure. The growth of land markets in the absence of supporting laws and legal institutions is contributing to economic social and environmental problems that will have long-term consequences on Tanzania. Land registration While the introduction of registration of leasehold title may yield anticipated

PAGE 243

230 benefits such as security and mortgage credit for the small holder in the city it can have negative effects on the village communities where customary land tenure has prevailed in terms of equity, land use and production. This may be attributable to the government's failure to recognize that customary land rights assure households of access to strategic mix of natural resources and enables the people to pool risks and cope with climatic and economic uncertainties. Conclusions Since the early 1980's the need to have a comprehensive land policy that would guide land ownership, leasehold and allocation has been recognized in both private and official circles. The village settlement program (Operation Vijiji) disrupted the customary land tenure system in many rural areas to the extent that it is difficult to determine with certainty the kind of tenure system that is currently operating in some rural communities. Increase in human population have increased the land use requirement of the people. This has resulted in the demand and hence the competition for land, especially around the urban areas. Rapid increase in livestock population has increased the demand for grazing land, especially in Dodoma, Shinyanga and Mwanza regions. Government policies favoring agriculture have resulted in the extension of the agricultural areas into lands which used to be grazing grounds. The effect has been a reduction in the land available to pastoralists especially in Mwanza, Shinyanga Arusha and Shingida regions. Increased movement of livestock and their keepers from traditionally livestock keeping areas like Mwanza, Shinyanga, Tabora and Dodoma regions to other such as Mbeya, Iringa, Morogoro,

PAGE 244

231 Rukwa Ruvuma regions have started to cause land use conflicts in the affected areas. Increased urbanization and high urban growth have increased the requirement for more land for settlements industries and commerce. The need to preserve valuable agricultural land around the urban centers to meet the food supply has intensified the competition for land in and around urban centers. Development of land markets in an around urban centers is occurring although it is prohibited. There needs to be a legislature to recognize and regulate land markets. Increased awareness among the people of the value of land and property has led to conflicts in urban areas as a result of more people competing for the limited number of lots in the urban centers. The present evolution of customary tenure towards more individualized tenure and development of markets in and around major urban centers In the older and high agricultural potential areas of Kilimanjaro, Bukoba and Rungwe districts where land is scarce cash crops such as coffee and tea are grown on individual holdings. This calls for a pragmatic policy and legal framework that supports such process. These problems are rooted in the attempt to import western culture over and above indigenous customs The solution at the moment lies not only in re introducing the old system but by modifying the western system to recognize the customary law which existed before the Europeans arrived. The present land policies pose problems for the intensification of agriculture, equitable access to land and sound natural resource management. Changes have begun to occur in the Tanzanian land policy. The objective for a land policy reform is to ensure the optimal use of the land and its resources without upsetting or endangering the ecological balance of the environment. Other derivative benefits include:

PAGE 245

232 Ensuring an equitable distribution of land. Recognizing customary rights in land and making the rights secure in law With current trend towards market economy avoiding the possibility of land being concentrated in the hands of a few individuals or organizations. Ensuring rapid socioeconomic development by allocating land to its most productive use Streamlinioe existing land management system and improving the efficiency of land management. Implementing sound land information management. Protecting land resources from degradation by encouraging development of sustainable resources.

PAGE 246

APPENDIXC OBSERVATIONS AND CONCERNS WITH EXISTING SYSTEM Urban Development division Issue Problem 1 Base Mapping Due to the lack of current and up-to-date base maps town planners often have to prepare the town planners drawings using old base maps. Sometimes the town planners have to update the base maps themselves This causes delays in the preparation of town planners drawings Surveying and Mapping Division Issue Problem 1 Base mapping Standard sheets are in paper format. They are too old and have not been revised for a long time. 2 Density of survey controls There are not enough survey controls available in the urban towns. Surveys are done using local origins At the moment control densification is done by basing surveys on previous cadastral controls. Newly established controls have to be certified by the director of surveys before they can be used for cadastral surveys. Sometimes approval takes too long to be made. 3 Quality of survey controls The national controls have been densified at different times using different types of equipment. 233

PAGE 247

234 4 Manual computation and Survey processing is done using hand-held drafting of field data calculators and plotting is done manually In so me regions e.g. Mbeya the surveyors use facit machines for survey processing a time-consuming process. 5 Town planners drawings This is because the base map was not updated do not always depict the before the town planners drawing was prepared exact situation on the site. The result is that when the surveyor encounters any discrepancies while demarcating the layout, he has to adjust the sizes of the plots to accommodate the discrepancy. Sometimes it may be existing structures. For this reason, experienced surveyors who can move boundaries in accordance with planning regulations have to be given the task of demarcating the plots and, thereby limit the number of surveyors who can be assigned the task. 6 The use of old and While theodolite and steel tapes are used in some damaged survey equipment regions for surveys works chain and compass surveys are used in other regions. In Songea for example, all the tapes are broken and there are only one T2 theodolite and two very old Kern instruments. There is no level instrument in Songea.

PAGE 248

235 C-1 Administrative Problems associated with Land Allocation Urban Development Division Problem Cause and Effect 1 Misidentification of plots The time between allocation and when the applicant develops the land can be so long that the owner forgets the actual plot that was allocated to him or her. The owner may not feel the need to contact the land officer or the surveyor. So the owner depends on his memory and chooses the plot he thinks is his. The land officer is supposed to show the owners their plots but because too much time has elapsed since the land officer was first shown the plots by the surveyor the land officer himself may have forgotten. There are instances where people have built houses either in the road reservation or on another person's plot. 2 Different allocating This problem is more prevalent in Dar es Salaam agencies with inadequate where plots are allocated from the office of the information exchange commissioner for lands the Urban Planning mechanism Committee and by some land officers. The office of the commissioner allocates land for large industrial uses low-density plots commercial buildings and farms larger than 5000 acres. There are no proper means of informing each other when an allocation is made. Attempts are made by the commissioner s office to inform the City Council and vice versa but due to poor communication system, the notice is received too late. This creates a problem with multiple allocation of the same plot.

PAGE 249

236 3 Lack of infrastructure The provision of utilities such as roads water and before allocation electricity are essential of any land development process. Although a proper road network may b designed by the town planner unl ss the road are made passable, people are not going to use them. Plot owners who are ready to develop their properties may not be familiar with the road reservation. They may use bulldozers to clear a path to their plot and thereby cause damage to other peoples' plots or even remove the boundary markers. 4 Over lapping responsibilities The Local Government Act defines the role of the between the planning City Council as the administrators of the city's redivision and the municipal sources. The Town Planning Act defines the Urban council in Dar es Salaam Planning Division in the Ministry of Lands Housing and Urban Development as policy makers. The City Council operates within the policies that are established by the Ministry. The two roles should not be ambiguous in their definitions. 5 Approval of layouts for For areas outside the planned areas in the regions, areas outside the master the Town Planning Act does not indicate that the plan layouts have to be approved by the director for urban planning. As there are no master plans of such areas with which to check for conformity there is no reason why the layout should be sent to Dar es Salaam for approval. The important part of the whole process is that the layout should be acceptable to the Urban Development committee. An experienced regional town planner can work with the committee to produce an acceptable layout. This will reduce the process by a considerable amount of time.

PAGE 250

237 Surveying and Mapping Division Problem Cause and Effect 1 Destruction of Boundary This occurs because the access roads within the markers. plots are not made usable. People usually drive bulldozers along the shortest path to their plots and destroy some boundary markers in the process. This means that the surveyors have to spend time replacing those markers. Not enough surveyed lands Several reasons may be given for this problem. The to meet the public demand most important one being lack of funds to pay for plots compensation. Unavailability of town planner drawings, survey equipment, transportation, and even stationery may be secondary reasons. Land Development Division Problem Cause and Effect 1 Multiple allocation of the Very few duplicate allocations occur deliberately. same parcel A major cause ofthis problem is poor record keeping. In Dar es Salaam, multiple allocations of the same parcel sometimes occurs because the same parcel may have been allocated by the Urban Planning Committee, but the information was not communicated to the commissioner's office on time. Meanwhile the commissioner's office allocated the same parcel to someone else. Poor communication between the two offices is a problem. One agency is not aware of the allocations that have been made by someone else. Sometimes land officers allocate parcels without proper investigation into its ownership.

PAGE 251

238 2 It takes too long for Because the members of the allocating committee surveyed plots to be are all holding responsible positions it is difficult allocated. to schedule a meetings. The result is that it takes a long time for allocations to be made once the plots have been surveyed. Some financial incentive is already being given for the members to attend meetings. The indications are that very often the meetings are not scheduled because there is no money to pay the allowances. People waiting for allocations may be tempted to squat on the plots because the cannot wait any longer. 3 Current methods for There is generally a desire to allocate on firstdetermining who gets an come-first-served basis. Because there are always allocation leads to few surveyed plots to be allocated, people who are corruption desperate for plots are tempted to seek favors through superiors of land officers, the influence of politicians, or even try bribing their way to get an allocation. If there are enough surveyed plots to distribute, then there will be no need to use political influence or any other means to obtain an allocation. 4 It takes too long to receive The current procedure is that the survey plan is deed plans. kept at the Surveys and Mapping office. Every time a request is made for a deed plan someone has to go and collect the master plan from the survey office and produce the deed plan. Meanwhile if another request is received, the request has to wait until the master plan is returned to the surveys office. If there were a hundred plots, this would mean that the plan will travel one hundred times between the survey office and the drafting office, meanwhile people are waiting to process their title. It has been said that people have had to wait 2 years just to obtain a deed plan.

PAGE 252

239 5 Neglecting to pursue the In most regions, the Letter of Offer does not process to obtain title. request the applicant to submit the advice of Failure to submit the advice payments to the land officer. It may be assumed of payments to the land that payment of the fees constitutes ownership and officer. title. Even if the letter of offer did request the presentation of the receipts to the land officer, it is the responsibility of the Ministry to ensure that certification and title follow so as to ensure record keeping. The Ministry should not leave that responsibility to the applicant. 6 No communication Communication is generally poor between offices. between allocating In the case where different people are allocating agencies from the same group of parcels, instant communication is essential to avoid giving the same parcel to two or more people. 7 It takes too long to get a People have to personally chase after their certificertificate approved cates to get them approved. Those who rely on the system to work never receive their certificates. The movement of documents from the schedule officers desk to the statistics and to the commissioner's office is too slow, although there aren't too many things that need to be done other than checking the entries. 8 The procedure to send all All allocations in Tanzania have to be signed by the documents to the commissioner for lands. The commissioner for commissioner for approval lands has more important responsibilities than just is too cumbersome signing certificates. There is the tendency to postpone signing any certificates whenever more important issues need to be dealt with. This could be a bottleneck in the whole allocation process. 9 No means for sending the The regional offices have no means of transportacertificates in the regions to tion and no money to send the certificates through the commissioner's office the post. Applicants are often asked to carry the for processing prepared certificates to Dar es Salaam by themselves. 10 Certificates that have been This is the same as the issue about getting approved remain in the documents to the headquarters. Transportation is headquarters for too long the problem.

PAGE 253

240 11 In Dar es Salaam, Documents can remain in the office for as long as documents that have been six months to be sent to the respective land officer returned by the registrar of for corrections to be made. Because of poor filing titles due to errors or methods, documents get misplaced easily. In most underpayment of fees cases, the land officer may not even know that the remain at the document has been returned for corrections to be commissioners office for made. too long 12 No way of knowing if an There is no feedback if the offer has been rejected. offer has been rejected The letter of offer assumes that everyone will accept the offer. For this reason and because there is no proper method for tracking allocations, an offer that has been rejected may never be reallocated. 13 No effective way to verify This issue is the same as not knowing if an offer if the 30-day period for has been rejected. paying the fees has elapsed. 14 No effective way of It is estimated that less than 20 percent of allocaknowing if the tions actually get developed. Of this, only about development conditions are half of them are done within the required period. being complied with, e.g., Technically, the remaining 80 percent can have the 3-year period within their certificates revoked and the plots reallocated. which the plot must be Almost everyone knows that the development developed conditions are not enforced. 15 Revocation of certificates Revocations can only be recommended. The takes too long to process recommendation must first come from the land officer to the commissioner and to the minister for lands. Every one has to investigate the circumstances and recommend action. According to the law, only the president of Tanganyika can revoke Certificates of Occupancy. Since the president is so busy with national issues he does not have time to deal with revocations. In the meantime, few comply with the development conditions.

PAGE 254

241 16 It takes too long to locate a The files in the open registry have grown beyond document in the open what the registry was designed for. The documents registry were supposed to be stored in pigeon holes in a wooden shelf That was when there were a couple of hundred files to work with. There are currently several thousands of files in the registry. Space has become so critical that files are kept on the floor. Documents get misplaced even in the open registry. There is no control on the movement of files. Land officers can walk into the registry and take any files that they need. Fearing that the file may not be found the next time it is needed, the land officer chooses to keep the file in his or her drawer. Because ofthis it takes a long time to locate files in the open registry. 17 Lack of stationery for File covers and index cards are always in short processing certificates supply in all the land offices. Without file covers, it is impossible to process the Certificate of Title. Sometimes applicants are advised to buy their own file covers before their certificates can be processed.

PAGE 255

242 Land Registry Problem Cause and Effect 1 Incorrect or incomplete According to section 22 of chapter 334 of the documents still have to be laws, the registrar should give reasons for sent to the registrar for objecting the registration of any title. This endorsement condition, however, does not apply to incomplete documents or underpayment of the required fees. The fact that a document is returned to be corrected does not constitute total objection. If documents that contain errors are not sent to the registrar, the time that the document stays at registrar's section can be saved. It is when the documents return to the commissioner's office that the delay really occurs. 2 Lack of stationery for This problem is the same as that of the open processing documents for registry. registration

PAGE 256

REFERENCES Aaronso~ S. A. (1996), Information Systems technology Considerations in the Establishment and Growth of a Land Cadastre System for a Rapidly Developing Economy. International Conference on Land Tenure and Land Administration. Orlando, Florida. American Society of Photogrammetry (1980), Manual of Photogrammetry. 4 th ed., Falls Church, Virginia. Anderso~ R. ( 1986), Management, Information Systems and Computers: An Introduction. Macmmao, London. Andrews, R. N. L. (1979), Land in America: A Brief History. In Land in America., Ed. R. Andrews. Lexington Books, Massachusetts. Angus-Lep~ P. (1989), The Thailand Land Titling ProjectFirst Steps in Parcel-based LIS. International Journal of Geographic Information System. Vol. 3, No. 1, 5968. Azad, B. (1996), Lebanon A Case Study in Registry and Cadastre Modernization: Lessons for Process-Based Land Information Automation. International Conference on Land Tenure and Land Administration. Orlando, Florida. Barnes, G. (1993), The Role of GIS in The Development of cadastral Information Systems. GIS Today, July 1993, 17-22. Barnes, G. (1994a), LIS Challenges in Latin America (Part One). International Journal for Surveying, Mapping and Applied GIS. Vol. 8, No. 4, 32-35. Barnes, G. (1994b), LIS Challenges in Latin America (Part Two). International Journal for Surveying, Mapping and Applied GIS. Vol. 8, No. 5, 29-31. Barnes, G., Eckl, M. (1996), Pioneering a GPS Methodology for Cadastral Surveying: Experience in Albania and Belize. International Conference on Land Tenure and Land Administration. Orlando, Florida. 243

PAGE 257

244 Barney, C., Salamanca, P. (1996), Las Bases de Datos Especiales de Catastro: Recur so Fundamental para la Titulacion Masiva y Modernizacion de/ Registro en Colombia International Conference on Land Tenure and Land Administration Orlando Florida. Bloch, P. C. (1996), Potential for Land market Development in Albania. International Conference on Land Tenure and Land Administration. Orlando Florida. Bohannon, P. (1973), Land', 'Tenure ', and Land Tenure Land Tenure Center University of Wisconsin, Madison, Wisconsin. Branscomb Anne W. (1994), Who owns Information? Harper Collins Publishers New York. Brown, C. M. Robillard, W. G., Wilson, D. A. (1995), Brown's Boundary Control and Legal Principles. ( 4 th Ed). John Wiley and Sons, Inc. New York. Bruce J. (1994), Review of the Report of the Presidential Commission of Inquiry into Land Matters in the Republic of Tanzania-199 2. Document Prepared for The World Bank. University of WisconsinMadison. Bruce, J. W. And Migot-Adholla, S. (1994), Searching for Land Tenure Security in Africa. Kendall/Hunt Publishing, Dubuque, Iowa. Burns, A., Eddington, R. Grant, C., and Lloyd, I. (1996) Land Titling Experience in Asia. International Conference on Land Tenure and Land Administration. Orlando Florida. Butler, S. B. (1996), Registration of Immovable Property in the Republic of Belarus : Issues of Law and Administration in the former Soviet Union. International Conference on Land Tenure and Land Administration. Orlando Florida. Buyong T. B., Kuhn, W., Frank, A. U. (1994), A Conceptual Model of Measurement Based Multipurpose Cadastral Systems. Urban and Regional Information Systems Association, 35-49. Cassettari, S. (1993), Introduction to Integrated Geo-Information Management. Chapman & Hall, London. Clapp, J. L. (1976) Concepts of a Multi-Purpose Cadastre Problems of design and Implementation. Proceedings of the Symposium on Land Records. University of Maine at Orono Orono, Maine.

PAGE 258

245 Codd, E. F. (1970), A Relational Model of Data for large Shared Data Banks. Communications of the Association for Computing Machinery., Vol 13, No. 6, 377-387. Codd, E. F. (1979), Extending the Database Relational Model to Capture more Meaning. Association for computing Machinery Transactions on Database Systems., Vol. 4., No. 4., 397-434. Cooke, D. (1980), A review of geoprocessing systems and a look at their future. In Computers in Local Government Urban and Regional Planning, eds. Kramer, K., King, J., Auerbach Publishers Inc., Pennsauken Publishers Inc., pp. (2.4.1) 1-16 Cotlear, D. (1992), Empowering Villages to Manage their Natural Resources: Rural Land Policy in Tanzania. The World Bank White Cover Paper. Southern Africa Department. Cox, J., Behnke, R. ( 1996), Privatizing African Rangelands : A Case of Enclosure in Northern Namibia. International Conference on Land Tenure and Land Administration. Orlando, Florida. Cracknell, A. P., Hayes, L. W. B. (1991), Introduction to Remote Sensing. Taylor and Francis Ltd., London, England. Croom, F. H. (1989), Principles of Topology. Saunders College Publishing, Philadelphia, P.A. Dale, P. F. (1976), Cadastral Surveys within the Commonwealth. Her Majesty's Stationery Office (HMSO), London. Dale, P. F. (1990), Guidelines for the improvement of Land-Registration and land Information Systems in Developing Countries. United Nations Center for Human Settlement (HABITAT), Nairobi, Kenya. Dale, P. F. (1991), Land Information Systems. In Geographical Information Systems. Volume 2., Eds. Maguire, D. J., Goodchild, M. F., Rhind, D. W., Longman Scientific & Technical, New York, 85-99. Dale, P. F., McLaughlin J. D. (1989), Land Information Management: An introduction with special reference to cadastral problems in Third World countries. Clarendon Press, Oxford.

PAGE 259

246 Davies, C. J. and Fourie, C. (1996), Incorporating Social Change Model into Land Records Design. International Conference on Land Tenure and Land Administration. Orlando, Florida. Date, C. J. (1991), An Introduction to database Systems. Volume 1. 5 th Edition, Addison-Wesley Publishing Company, New York. Derby, F. W. (1995a), Cadastral Operational, Equipment and Needs Assessment/or Tanzania. Report submitted by Tropical Research and Development for the United Republic of Tanzania. Derby, F. W. (1995b), Organization of Land Records in Tanzania. Report submitted by Tropical Research and Development for the United Republic of Tanzania. Derby F. W. (1996), Reorganization of Land records in Tanzania. International Conference on Land Tenure and Land Administration. Orlando, Florida. DeSoto, H. (1990), The Other Path The invisible Revolution in the Third World. Harper & Row, New York. Dowson, E., and Sheppard, V. L. 0. (1968), Land Registration. 3 rd Edition. Her Majesty's Stationery Office (HMSO), London. Dunkerly, H. B. (1985), Land Information Systems for Developing Countries. Commonwealth Association of Surveyors and Land Economy (CASLE). Kuala Lumpur, Malaysia. du Plessis, W. (1996), Historical Overview : Evolution of Land Tenure Systems in South Africa. International Conference on Land Tenure and Land Administration. Orlando, Florida. Egenhofer, M. and Frank, A. (1993), Object-Oriented Modeling/or GIS. Journal of the Urban and Regional Information Systems Association, Vol. 4, No. 2 3-19. Epstein, E. F. (1991), Legal Aspects of GIS. In Geographical Information Systems: Principles and Applications. Longman and Scientific Technical, New York. Feder,G., Tongroj, 0., Yongyuth, C.,Chira, H. (1988) Land policies and Farm Productivity in Thailand. Baltimore, MD. Johns Hopkins University Press. Feder, G., and Davis, G. (1991), Land Tenure and Property Right: theory and Implications/or Development policy. The World Economic Review. Vol 5 No I, 135-153.

PAGE 260

247 Feder, G. and Nishio, A. (1996) The Benefits of Land Registration and Land Titling: Economic and Social Perspectives. International Conference on Land Tenure and Land Administration. Orlando, Florida. Federation Internationale Des Geometres (FIG) (1996), FIG Commission 7 Statement on the Cadastre., http://sunspot.sli.unimelb.edu.au/fig7 /cadastre/statement on cadastre.html Fisher, K. P. and Moyer D. D. (1973), Land Parcel Identifiers for information Systems. Proceedings of the Committee on Compatible Land Identifiers: Problems, Prospects and Payoffs. American Bar Foundation, Chicago. Fotheringham, S. Rogerson, P. (1994), Spatial Analysis and GIS. Taylor and Francis Ltd., London, England. Fourie, C., (1993) A new Approach to the Zulu Land Tenure System: An Historical Anthropological Explanation of the Development of an Informal Settlement. Unpublished Ph.D Thesis. Rhodes University. Grahamstown, South Africa. Fourie C. (1996), The Role of Local Land Administrators and land Managers in Decentralization and land Delivery Registration and Information management in Developing Countries. International Conference in Land Tenure and Land Administration. Orlando Florida. Government of Tanzania (1956), Land Registration Ordinance: Chapter 334 of the Laws. The Government Printer. Dar es Salaam, Tanzania. Government ofTanzania (1967), Land Ordinance: Chapter 113 of the Laws. The Government Printer. Dar es Salaam, Tanzania. Government of Tanzania (1967), The Town and Country planning Ordinance: Chapter 378 of the Laws. The Government Printer. Dar es Salaam, Tanzania. Government of Tanzania (1957), The Land Survey Ordinance: Chapter 390 of the Laws. The Government Printer. Dar es Salaam, Tanzania. Government of Tanzania (1959), Land Survey and Surveyors Regulations. The Government Printer. Dar es Salaam, Tanzania. Graham, L. N., Ellison, K., Riddell, S. (1997), The Architecture of a Softcopy Photogrgammetry System. Photogrammetric Engineering and Remote Sensing, Developments inland Information management. Vol. 63, No. 8, 1013-1020.

PAGE 261

248 Haldrup, N. 0. (1996), Land Registration in Context of Informal Settlements. International Conference on Land Tenure and Land Administration. Orlando Florida. Hanson, H. (1996), "Showing the Land" Survey, and Registration: Mechanisms of Land Transfer in Buganda Fifty Years after Individualization of Tenure. International Conference on Land Tenure and Land Administration. Orlando. Florida. Hawkey, W. (1985), Land Information New Zealand: The experience of a Developing Nation. Conference on The Management of and development of Natural Resources. Kuala Lumpur, Malaysia. Healey, R. G. (1991), Database Management Systems, in Geographical Information Systems. Volume 1. Eds Maguire, D. J., Goodchild, M. F., Rhind D. W. Longman Scientific & Technical, New York, 251-267. Hendrix, S. E. (1996), Land Records Management in the Context of Hemispheric Trade Integration in the Americas. International Conference on Land Tenure and Land Administration. Orlando, Florida. Hoben, A., Bruce, J., Johansson, L. (1992), Rural Land Policy in Tanzania; Issues Paper Document Prepared for The World Bank. Land Tenure Center. University of Wisconsin, Madison. Holstein, L. (1985), Land Registration for Less Developed Countries -A Checklist. Commonwealth Association of Surveyors and Land Economy (CASLE). Kuala Lumpur, Malaysia. Holstein, L. (1990), The Cadastre as a Tool of Resource management in Developing Countries. Paper presented at the 6 th General Assembly and Technical Seminar Royal Institute of Chartered Surveyors (RICS). London. Holstein, L. (1996), What are the Roles of the Public and Private Sectors in Land Titling and Registration Systems? International Conference on Land Tenure and Land Administration. Orlando, Florida. Hunter, G. J. (1988), Data Archiving in a Digital Cadastral Data Base. Australasian Conference on Urban and Regional Information Systems. Sydney, Australia. Hunter, G. J. (1989), Non-Cu"ent Data and Geographical Information Systems : A Case for Data Retention., International Journal of Geographic Information Systems. Vol. 2, No. 3.

PAGE 262

249 Huxhold, W. E. (1991), An introduction to Urban Geographic Information Systems. Oxford University Press, New York. Hyden, G. (1980), Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry. Heinemann Press, London. Jefferess, G. A. (1996), Value of the Legal cadastre in a free Market Economy International Conference on Land Tenure and Land Administration. Orlando, Florida. Jeyanandan, D., and Williamson, I. P. (1990), A Cadastral Mode/for Developing Countries. Proceedings of the National cadastral Reform Conference, University ofMelboume, Melbourne, Australia. Kennedy, M. (1996), The Global Positioning System and GIS: An Introduction. Ann Arbor Press Inc.,Chelsea, Michigan. Kuang, S. (1996), Geodetic Network Analysis and Optimal Design: Concepts and Applications. Ann Arbor Press Inc.,Chelsea, Michigan. Larsson, G. ( 1977), The evolution of existing cadastres towards the multipurpose cadastre. FIG Congress, Stockholm. Larsson, G. (1991), Land Registration and Cadastral Systems. John Wtley & Sons, New York. Leberl, F. W. (1996), Practical concerns in Softcopy Photogrammetry Processing Systems. In Digital Photogrammetry: An addendum to the Manual of Photogrammetry. Ed. Chris Greve. American Society for Photogrammetry and Remote Sensing. Bethesda, Maryland. Lee, J. (1997), An Object-Oriented Approach to the Analysis of Land Information System. Proceedings of the 1997 GIS/LIS conference., Cincinnati, Ohio. Lemei H. W., (1985), Land Titling : Conceptual, Empirical, and Policy Issues. Land Tenure Center. University of Wisconsin. Madison, Wisconsin. le Roux, P. (1996), Information management Technology in Support of the Establishment, Development, and Maintenance of Land Management and Administration Infrastructure. International Conference on Land Tenure and Land Administration. Orlando, Florida.

PAGE 263

250 Liesz, S. J. (1996), Aframeworkfor Understanding and Integrating Community-Based and National Land Tenure systems in Africa. International Conference on Land Tenure and Land Administration. Orlando, Florida. Lilesand, T. M., Kiefer, R. W. (1994), Remote Sensing and Image Interpretation (3 rd Ed.). John Wiley and Sons, Inc., New York. MacLeod, F. (1996), Crofting: An Example of Interacting Land Tenure and Rural Development Models. International Conference on Land Tenure and Land Administration. Orlando, Florida. Marquardt, M.A. (1996), Land Law Reform in Uganda: The Changed Role of the State. International Conference on Land Tenure and Land Administration. Orlando, Florida. Marquis, S. D. (1979), Some thoughts on Concepts on Land, in Land in America. Ed. Richard Andrews. Lexington Books. McCloy, K. R. (1995), Resource Management Information Systems. Taylor & Francis, PA, USA. McLaughlin, J. D. (1975), The Nature, Function. and Design Concepts ofMultipurpose Cadastres. Ph.D Dissertation. University of Wisconsin. Madison, Wisconsin. McLaughlin, J. D., de Soto, H. (1994), Property Formalization: The Proform Solution. Geomatica. Vol 48, No. 4, 307-314. McLaughlin, J. D., Nichols, S. E. (1989), Resource management: The Land Administration and Cadastral Systems Component. Surveying and Mapping. Vol.49, No. 2, 77-85. McLaughlin, J. D., Williamson, I. P. (1985), Trends in Land Registration. The Canadian Surveyor, Vol 39, No. 2, 95-108. Methley, B. D. F. (1986), Computational Models in Surveying and Photogrammetry. Blackie & Sons Limited, Glasgow. Merideth, Jr., R. W., Clapp, J. L., Niemann, B. J. Jr., Weisman, B. E. (1990), Modernizing Land Records: An Example of Issues and Processes in Wisconsin. In Geographic and Land Information Systems for Practicing Surveyors: A Compendium. Eds. Onsrud, H. J, and Cooke, D. W., American Congress on Surveying and Mapping, Bethesda, Maryland.

PAGE 264

251 Michener W. K. Brunt, J. W., Stafford, S. G. (1994), Environmental Information management and Analysis; Ecosystem to Global Scales. Taylor and Francis, New York. Ministry ofHousing (1985), Land Information Systems in Denmark Experiences and Strategies. Copenhagen, Denmark. Moyer, D. D. and Fisher, K. P. (1973), Land Parcel Identifiers for Information Systems. American Bar Foundation, Chicago. Muller, J., Lagrange, J. Weibe4 R., (1995), GIS and Generalization: Methodology and Practice. Taylor and Francis. London, England. National Research Council of Canada (1976), Land Unit Identification; An Analysis. NRC 15736 Ottawa, Canada. National Research Council (1980), Need for a Multipurpose Cadastre. National Academy Press. Washington, DC. National Research Council (1983), Procedures and Standards for a Multipurpose Cadastre. National Academy Press. Washington, DC.,. Nichols, Susan (1993), Land Registration: Managing Information for land Administration. PhD. Dissertation. University ofNew Brunswick. Fredericton, Canada. Opadeyi, J. (1996), The Application of Entity-Relationship Model in the Design of an Integrated Data-Driven Land Administration System. International Conference on Land Tenure and Land Administration. Orlando, Florida. Palmer, D. W. (1996), Incentive-Based Maintenance ofLand Registration Systems. PhD Dissertation. University of Florida, Gainesville, Florida. The Random House Dictionary of English Language: the Unabridged Edition. Eds. Stein, J. and Urdang, L. (1971). Random House, New York. Ratcliffe, J. (1974), An introduction to Town and Country Planning. Hutchinson, London. Rhine, D. (1995), Reading in Object-Oriented Systems and Applications. IEEE Computer Society Press, Los Alamitos, California. Robinson, V. B. (1988), Modeling Consistency Preserving land Transactions in a Logic Programming Environment. GIS/LIS: Accessing the World, San Antonio, Texas.

PAGE 265

252 Scheepers, T. (1996), Land Administration: Facing Realities of Implementing Land Reform Programmes in South Africa. International Conference on Land Tenure and Land Administration. Orlando, Florida. Shivji, I. G. (1995), Problems of Land Tenure in Tanzania: A Review of the Report of the Presidential Commission of Inquiry into Land Matters 1992. Paper Presented at the Land Policy Workshop. Arusha, Tanzania. Shutt, A. K. (1996), "Everyone has a Right to the Farm": Generational Conflict in the African Freehold Areas of Zimbabwe. International Conference on Land Tenure and land Administration. Orlando, Florida. Sickle J. V. (1996), GPS for Land Surveyors. Ann Arbor Press Inc., Chelsea, Michigan. Simpson, S. R. (1984), Land Law and Registration. Surveyors Publications, London. Sonnenberg, W. Ir. (19944), Digital Base Map Opens Up Future for Cura~ao. International Journal for Surveying, Mapping and Applied GIS., vol. 8, No. 4, 72-73. Stanfield, D. (1996), Creation of Land Markets in Transition Countries : Implications for the Institutions of Land Administration. International Conference on Land Tenure and Land Administration. Orlando, Florida. Stanley, D. L., Mohamed, M. (1996), Creating and Using GIS Data Sets: An Honduran Example. International Conference on Land Tenure and Land Administration. Orlando, Florida. Taylor, D.R. F. (1991), GIS and developing nations. In Geographical Information Systems. Eds. Maguire, D, Goodchild, M. F. and Rhind, D. W., Vol. 2, National Academy Press. Longman Scientific and Technical, U. K. United Nations (1974), Report on the meeting of the Ad Hoc Group ofExperts on Cadastral Surveying and Mapping. New York. United Nations (1987), Guidelines for Development Planning: Procedures, Methods and Techniques. Department of Technical Co-operation for Development, New York. United Republic of Tanzania (1993), Report of the Presidential Commission oflnguiry into Land Matters. (2 Volumes), Dar es sal~ Tanzania. Usery, L. (1996), A Feature-Based Geographic Information System Model. Photogrammetric Engineering and Remote Sensing., Vol. 62, No. 7, 833-838.

PAGE 266

253 Walijat~ D. Grant C. (1996) Land Registration Reform in Indonesia. International Conference on Land Tenure and Land Administration Orlando Florida. Williamson, I. P. (1986) Trends in land Information System Administration in Australia in Developments in land Information management Eds. Dahberg, R. E., McLaughlin, J. D. Niemann, B. J. Institute for Land Information, Washington D.C. Williamson, I. P ., Hunter G. J. (1990) Conceptual Modeling and its Role in the Design of Land Geographic Information Systems., Proceedings of the FIG XIX International Congress Helsinki, Finland. 500-508. Wisconsin Land Records Committee (1987), Modernizing Wisconsin's Land Records: Final Report of the Wisconsin Land Records Committee. Madison, Wisconsin. Wolf: Paul. R. (1983), Elements of Photogrammetry. McGraw Hill New York. Wolf: P.R. Brinker R. C. (1994), Elementary Surveying. (9 th Ed.). Harper Collins. College Publishers New York. Worboys M. F. (1995) GIS: A Computing Perspective. Taylor and Francis Ltd London, England World Bank (1995) National Land Administration Project for Bolivia. A Staff Appraisal Report. World Bank Report No 13560-BO World Bank (1989), Debate over Land Registration Persists. The Urban Edge: Issues and Innovations, A World Bank Publication. Yourdon,E. (1994) Object-Oriented Systems Design, An Integrated Approach. Yourdon Press Upper saddle, New Jersey.

PAGE 267

BIOGRAPHICAL SKETCH Francis Derby obtained a Higher Diploma in Geodetic Engineering from the University of Science and Technology, Ghana, in 1976. In 1982, he graduated from the Polytechnic ofEast London, England, having earned a Bachelor of Science degree in Surveying and Mapping Sciences. Between 1982 and 1992, he practiced surveying in Ghana, England, and the Caribbean, during which he held positions of increasing responsibility. In 1992, he enrolled in the Master of Science program in Surveying and Mapping, at the University of Florida. He transferred to the Doctoral program after two semesters. 254

PAGE 268

I certify that I have read this study and in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philo so ~ ~,) Scot'. Smith, Chair / Associate Professor 6f Civil Engineering I certify that I have read this study and in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. ,, Grenville Barnes Cocha Associate Professor of Civil Engineering I certify that I have read this study and in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. /In\. Vt' g;j., '::t Bon Dewitt Associate Professor of Civil Engineering David G. Bloomquist Associate Professor of Civil Engineering I certify that I have read this study and in my opinion it conforms to acceptable standards of scholarly presentation and is fully a qua ope and quality as a dissertation for the degree of Doctor of Philo phy. an and Regional Planning

PAGE 269

This dissertation was submitted to the Graduate Faculty of the College of Engineering and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. ) December 1998 r / \.. .------..., ______ Winfred M. Phillips Y Dean, College ofEngineering M. J. Ohanian Dean Graduate School

PAGE 270

L 178 1 9 s --. I UNIVERSITY OF FLORIDA 1111111111111111111111111111111111111111111111111111111111111111 3 1262 08554 4244 I


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E35RWQN86_4ZA3P6 INGEST_TIME 2013-03-21T13:12:12Z PACKAGE AA00013549_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES