INFORMATION: THE MOST UN-MANAGED RESOURCE
IN WATER MANAGEMENT
Donald W. McEwen
QUAD Data Corporation
P.O. Box 2097
Tallahassee, Florida 32316
Water management is an information intensive service. It is
meant to benefit future as well as present generations. As the
population continues to grow, problems will become much more
critical. But will future generations of water managers chide us
for not effectively managing our information such that it will be
readily accessible to support future decisions?
From a corporate view, our management of information is some-
what haphazard compared to the way we manage other vital
resources within the organization. Synonymous data elements are
frequently not defined or structured consistently between project
leaders, much less between larger organizational units. While
adequate for first time use of data, accessibility for later use is
generally too cumbersome and labor intensive. The lack of
commitment to a cohesive organizational data plan and structure
along with too little leadership (and at too low a level) contribute
heavily to this situation.
This paper addresses the need to implement structures and
practices that facilitate the strategic planning, management, and
future accessibility of key data. The roles, attitudes, and
responsibilities of key information workers are discussed as well
as their placement in the organization's structure. The focal
point is the evolution of the Chief Information Officer (CIO).
Water Management has come a long way during the 1980's with
respect to information technology. Most of the agencies in water
management have embarked on increasingly aggressive strategies
to utilize information technology. The basic strategy is to use
appropriate information technology to contain staff growth, to
increase productivity, to improve the quality of work products, to
improve responsiveness to informational needs, and to improve
decision support and policy formation.
r.. .','.n'. on orn tli. sheet s proprnetac y Ito UAD DAIA Corp Disclosure restrc',on are stated in this document
In the early 1980's, it was considered innovative to introduce
personal computers and networks into the water management
districts. Today, personal computers for most professional staff
members was no longer an issue. We have accumulated many new
computer systems and are comfortable in the use of these tools.
We have evolved significantly in our computer literacy. Many
success stories abound.
We have introduced computing and analytical tools that enable us
to monitor, model, and predict the impact of hydrological changes
better than ever before. We can present our findings most
convincingly with our professional quality graphical, numerical,
and textual presentations. In some areas, our technical and
scientific staffs reach virtuoso levels of mastery. However, in
our excitement with the parade of impressive new tools, we have
neglected the more illusive and lackluster task of managing the
In previous decades, when computing facilities were centralized
and more difficult to use, most of the work involving computers
was delegated to professional programmers and analysts. 'These
professionals had to learn certain disciplines and work habits that
were part and parcel to the holistic problem of creating useful
and lasting computer applications. Documentation skills were as
important as programming skills for long term success. Methodical
planning, design, and testing were standard skills required on
every project. Documentation standards (both written and
internal to the programming code) had to be followed. Managers
in such environments had to enforce good work habits and assure
that the applications were designed well, documented for usage as
well as maintenance purposes, thoroughly tested, and properly
stored. They had to see that all source code and documentation
conformed to the appropriate standards.
History is abundant with evidence that not all computer managers
and programmers developed good work habits and professional
skills. Even in a relatively closed culture in which the benefits
of standards, conventions, and disciplines were painfully clear, it
was difficult to instill professional work habits in everyone.
In today's environment, the computing tools are much easier for
the non-computer professional to use. Anyone who desires can
learn to use a computer. There is no closed culture to create
and police its own standards. Responsibilities relating to the role
of information worker ranks well below (if at all) each employee's
primary job activities. Their managers are people who generally
have not led or groomed their people in computer lore and the
disciplines that support good information management. It is not
uncommon for some managers to publicly disdain the technology
and the gory details of managing a constant deluge of data.
Itr,,. inlon,'ol.on or, th.s she*.! I pi.,,..'. to QUAD DATA Corp Disclosufe f~o trnctions ore stoted in this document
Overall, our current middle and upper managers are generally
weak on information management skills. They are also preoccupied
with present crises and feel that information management matters
are not an effective use of their time. Project leaders tend to
be focused on the current products and feel they have neither
the time, nor the responsibility to do any documentation beyond
the completion of each project work product. While they try to
take care of their own needs, few people feel that there is any
coordinated and corresponding effort among their counterparts to
warrant their participation. Moreover, they are convinced that
such an effort will put them behind on more visible projects.
These observations are not intended in a negative sense but are
the realities of the day.
Given this overall mind set, who then, is to educate and groom
professional work habits with respect to information management?
Who is to raise everyone's level of consciousness to know that
they have inherent responsibilities as information workers? Who
can coordinate the development of organizational standards and
have the authority to police and enforce them? This is the
function of the Chief Information Officer and an assistant called
the Data Administrator. These roles will be discussed briefly
later in this paper.
While we are maturing well with respect to our acquisition and
use of the technology, we are making no significant progress in
cultivating a maturity in the way we manage the data and
information. Our management of information is somewhat hap-
hazard compared to all other vital resources within the organiza-
tion. At times we seem to be like the farmer who was so
enthralled with acquiring new tractors and equipment that he
neglected to anticipate the market and plan the most strategic
and profitable crops to grow. The problem doesn't show up right
away and may not be recognized when it does. A few of the
problems I routinely observe within state agencies include the
Conflicting data element definitions between project
leaders, departments, divisions, and other agencies.
No documentation to show how each data element was
No documentation to show how each data base is structured.
S!. (,, af ..o orn ttis sheet is pTop'.elry to QUAD DAIA Corp Disclosure restr'c ons ore stoled in this document
* No documentation to show the interrelationships between
data elements and data sets within data bases.
* No documentation regarding what data is available and over
what time period it was collected.
* Many similar, but incompatible data files on many different
computers within a given agency. Few groups have any
routine procedures to archive final data in any central data
* No easy method of locating and aggregating all the organi-
zation's appropriate data from distributed data sets for
larger more comprehensive studies.
* No standards or procedures established to make it easy to
transport data from one computer system to another.
* No master data structure or data element definitions for all
to use and follow.
* No long range plan to collect strategic data that can be
anticipated for tools and capabilities that will be available
in the future.
* No central authority within organizations to lead a compre-
hensive and unifying effort with respect to information
management standards and corresponding documentation
No electronic method of locating data and information at
the organizational level. Modelers and information analysts
frequently complain that the most time consuming part of a
project is "scrounging" through the organization for
existing data. It is reasonable to assume that they don't
always find all the relevant data.that the organization has
collected (at great expense), especially historical data.
No document storage and retrieval mechanism at the
No written strategies and policies for centralizing high use
data, archiving low use data, or destroying useless and
No format procedures to insure that important data and
documents (that will be significant over time) adheres to
data definition standards and is properly loaded into the
central archive data base on a routine basis.
Sc:.,7 04,'t.o on th sheet ecry o D A op Dcloure stc s e n thiocumen
tlrc tnf.' 1Dt on on this sheet i! propntaip ry to QUAD DAIA Corp Disclosure restictoons are stated in this document
Much of the data we would like is very expensive to collect and
process. Data collected consistently over a period of time is
critical to analyzing changes that have occurred and to try to
identify the source causing the change. Initial data collected
from a given site generally becomes a baseline for monitoring
subsequent changes. It is impossible to go back into time and
collect data that was omitted or lost. We have an inherent
responsibility as information workers to retain this irreplaceable
data in a form that can is easy to use, spontaneously, whenever
a need exists.
We are beginning to see word processing capabilities on almost
every desk. We have greatly increased our output and quality of
documents. Yet, we have no mechanism for storing and retrieving
from the organizations cumulative wealth of documents.
From a management prospective, data is one of the few resources
on which we spend substantial sums of money, yet we have no
institutional mechanisms with which to strategically plan, budget,
and preserve data on a corporate or organizational level. We
tend to be preoccupied with computer hardware and spend too
little thought on the long term management, stewardship, and
strategic use of irreplaceable information.
We routinely collect massive amounts of data from the field.
Yet, it is more the exception than the rule that we have a
central repository of uniformly defined and formatted data.
There is rarely a mechanism or procedure to insure that any two
project leaders use the same data definitions, formats, or data
structure. Continuity between higher units in the organization is
extremely rare. Compatibility with other agencies is merely by
coincidence, if at all. With all the advantages of desk top
computing, we pay the price of a multitude of disjoint data sets
in a highly distributed system.
Upper management appears to be unaware of the situation with
respect to information management. They are not being given an
accurate assessment of the entire situation. There is a tendency
among the staff to report only the showcase successes and to
omit certain areas that are receiving inadequate attention. Many
are silenced by the fear that any attention paid to information
management may result in restrictive rules and substantial
increases in an already saturated work load, without noticeable
benefit to their work.
In many cases data is collected faster that it is put into machine
readable form. Almost all functional units have file cabinets,
closets, attics, even warehouses that are crammed with large
volumes of data still in its original paper form. While it may not
be advantageous to put all data into a computer, it is not clear
r o is p QUAD DATA Corp D on e stated in this document
..=- .. ,. Inlo,n|i) ot on th s Srr.e! IS p,'.-ir,-tJ!r tL, QUAD DATA Corp DOsclo 'e I.-.:,c., on ore stOted In this document
that a conscious strategy, policy, or plan has been developed to
consistently manage this situation.
Almost all of the successful data base applications have resulted
from the persistence and determination of a lone champion. It is
the result of such champions, usually at the neglect of other
duties, who doggedly pursued the completion of a data base
project. Even so, the completed projects rarely are documented
so that the work can be continued or otherwise maintained by a
successor. Frequently, when the organization loses such an
employee, all data and other machine readable information
generated by that employee falls into disuse. The lack of
documentation, uniform data compatibility, and distributed comput-
ing generally discourages any successor in trying to understand
and reconstruct what a predecessor has done. It is usually
considered easier to ignore prior efforts or to start from scratch.
Frequently, this results in a great deal of lost brainpower,
manpower, and historical information. This is very expensive and
wasteful. It is also unnecessary.
SIMULATION, MODELING, FORECASTING, PREDICTION
In our attempts to better understand entire hydrological systems,
we have developed relatively sophisticated modeling software.
However, we rarely have the data required to make effective use
of our modeling potential. We are limited by the data we collect.
We are even more limited by the data kept from previous years
to study changes in conditions over time. As organizations at
large, we do not seem to be doing sufficient advanced planning
to anticipate information we will need in the future.
What we can learn from modeling is not completely foreseeable.
Yet the foremost constraint on where this can lead is not our
modeling tools or computing appliances, but on available data.
During the next decade, it is reasonably certain that modeling
will become much more sophisticated and important than it has
been during this past decade. Aside from techniques, models will
become more integrated to include more related data such as
rainfall, well logs, water chemistry, stage data, etc. over increas-
ingly larger geographical areas. It is also reasonable to conclude
that other application areas, such as expert systems and GIS
(Geographical Information System) analytical tools, will continue
to evolve and require a well planned base of date. The usefulness
of these activities will be governed largely by the data we plan,
collect, and manage today.
Ni- inrc, tonr on nthli sheet I propreiary to Q'JAD DATA Corp Disclosure retrC ctons ore stated in this document
If our sophistication in managing information is inadequate now,
it is going to get worse. The situation will deteriorate at an
accelerating pace the longer we delay in taking corrective action.
There are a number of trends that will contribute to enlarge the
problem as time passes:
The ratio of computers to information workers continues to
increase. In some areas, this ratio will surpass a 1:1 ratio
within the next two years, as a result of specialized systems.
The foreseeable economy of scale suggests that smaller
more individualized computers will be more cost effective
than large central computers, hence the trend away from
centralization will continue.
Information workers continue to aggregate data sets on the
smallest computer that will adequately do the job. This is
motivated by convenience, ease of use, better response
time, and the desire for more autonomy. This continues to
result in greater numbers of disjoint data sets in relatively
un-managed distributed environments. In other words, the
organization's base of data continues to become more
fragmented rather than centralized.
Technology enables u.s to generate more numeric data,
geographical data, and textual documents at an increasing
The sheer volume of data that is collected is getting
increasingly difficult to manage and store. A substantial
portion of the data collected is never put into machine
Much of the technology acquired in the early 1980's (or
earlier) is nearing the end of its market life and is soon to
Hardware life cycles continue to shorten. Some IBM
mainframe computers are now realizing a market life of
less than six years. One reputable forecasting company
has observed that product cycles (new releases of successor
products) is now averaging two year intervals and is still
dropping. Personal computers have much shorter market
lives than large mainframe computers.
1 Computer Economins Sourcebook, Computer Economics,
Inc., Carlsbad, California, 1988.
Itn. cri 'atlo n co It.: sneet is piopneloty to QUAD DATA Corp Disclosure restrictions ore stated in this document
Lack of management and standards enforcement is making
data base conversion to more modern tools increasingly
difficult for organizations.
Future growth in the sophistication of integrated modeling
will be increasingly undermined by the lack of uniform and
Continuous data samples in a readily accessible form.
Despite our progress, our tools, and our ability to use them, a
substantial amount of data and information that is already in
electronic form will not be sufficiently organized and accessible
to be functionally available to future water management workers.
Beyond the first time use, for all practical purposes, much of the
data collected during this past decade is not functionally available
for general use today. Locating the data and converting it to
common data definitions and formats is frequently very labor
intensive and costly.
RECOMMENDED ACTIONS AT THE WATER MANAGEMENT
A strategic plan for future informational needs must be dynamic.
It requires continuous adjustment driven by long and short term
priorities. Before recommended actions are discussed it is useful
to consider the following:
The primary person or project leader who collects, processes,
and analyzes the data feels a keen sense of ownership and
protectiveness of the data.
The sheer volume of information about the environment
together with the complexity of government operations at
many different levels and geographic areas has made
control, coordination, and communication of information
between the various entities of the State difficult and
It is impractical to try to develop a centralized clearing
house or repository of environmental information at the
State level at this time. Distributed data bases are the
mainstream trend for the foreseeable future. Efforts must
be focused in locating available information, accessing that
information, easily transporting that information to the
people who need it. Procedures should be developed for
this process. Data definition standards must be developed.
Data dictionaries must be accessible.
S **, ,r-:rrr,,r. orn this sheet : pr'pc'.tOu Ic QUAD DATA Co'p DSsclosu'e rest!-,tons ore stated in this document
Leave each data set owner in charge of their data set.
However, make, them conform to the standards and proce-
Accept the distributed data base concept at the operational
level, but create procedures to transfer copies of all
appropriate data and documents to the organization's
central data base.
At the organization level, establish standards, conventions,
and procedures for keeping data definitions and data bases
compatible and up-to-date.
*Develop search and retrieval facilities that contain informa-
tion about your information. This is necessary in order to
locate what relevant information exists.
Create a forum in which all effected parties can brainstorm
to anticipate future problems, informational needs, tools
that are likely to be available, and the data needed to
maximize the organization's resources and capabilities.
Develop long term strategies, plans, and cost estimates
Develop the infrastructure within the organization to deal
specifically with the planning and management of information.
This should specifically include the implementation of the
Chief Information Officer (CIO) and one or more information
planners that includes a Data Administrator (DA).
STRONG LEADERSHIP IS REQUIRED FOR ORGANIZATION-WIDE
Success stories almost always are the result of a lone champion
who persisted till the application was operational. In order for
the entire organization to achieve this level of success, such a
champion is required in the executive office. Leadership must
come from the top. The most ideal situation is when the Chief
Executive Officer aggressively assumes the role of Chief Informa-
It seems clear that in any organization having such functions and
positions as Chief Executive Officer (CEO), Chief Financial
Officer, Chief Operating Officer, and Chief Legal Officer, the
function and need exists for a position of Chief Informhation
It r nlornint.on on this sheet is piop'ielory to QUAD DA1A Corp Disciosu', icr-s.ctions ore staled in this document
The argument for the position of Chief Information Officer was
well presented in a paper in 1983 by two of QUAD Data Corpora-
tion's principles, Dr. J.H. Poore and A.P. Jensen.2 A brief
summary of the characteristics and role of the key information
executives is as follows:
The Chief Executive Officer (CBO):
Responsible for the proper use of corporate resources, be
they human, technical, or material.
Responsible for translating knowledge into action.
Chief decision maker, planner, economist, allocator of
people, materials, and time in the organization.
Chief strategist, counselor and arbitrator.
Visionary, philosopher, symbolic personality of the organiza-
tion, and primary maker of changes in the organization.
The Chief Information Officer (CIO):
Similar in rank and authority to the Chief Financial
Officer. The CIO must be a member of the executive
The CIO must aide, shape and promote the corporate
Overall information resource architect and planner.
Responsible for long range planning and implementation
with specific emphasis on information, decision support
systems, and information technology.
Reports directly to the CEO or is the CEO.
Understands the mission, the organization, information, and
decision support systems (non-technical).
Must be structurally unbiased regarding organizational units
2 Poore, J.H. and Jensen, A.P., "The Chief Information
Officer", QUAD Data Corporation, Tallahassee, Florida,
S ..-. irWrr,'otco n on th.s sheet is propr:e'or, to QUAD DATA CorIp Disclosure festrclions ore stoled in this documents
established, the activities begin shifting toward maintenance and
re-evaluation. This may lead to a degree of boredom in the CIO.
On the other hand, it can present an opportunity for the CEO to
resume the function of the CIO. Ultimately, the CIO role must
be assumed by the CEO. The Captain of the ship cannot navigate
the optimal course if all the resources to keep the ship and crew
functioning healthily are not integrated into the planning and
long range strategy. The CEO cannot optimize the organization's
position in the future if informational resources are not integrated
into the long range as well as present strategy. In water
management, future problems are likely to become more complex,
more emotional, and require faster decisions that must be backed
up with well documented science.
RECOMMENDED ACTIONS AT THE STATE LEVEL
Within the past few years there was an initiative at the State
level to create a centralized repository of environmental data.
Little tangible progress has been seen. Such an undertaking is
totally impractical during this period of time for a variety of
reasons. Even to attempt, such a task requires a number of
preliminary steps that would be substantial achievements in
themselves. The first is to merely identify the sources and
characteristics of the data being collected. An automatic directory
of data sets and contact people would be a useful first stage.
Second, it would be necessary to get each data set owner/manager
to create a data dictionary pertaining to each data set in their
domain. This would be a mutually valuable compilation of
information for all those involved. Next, a standard state-wide
definition of all data elements (a master data dictionary) would
have to be developed. Based upon that instrument, along with a
master data structure design, general purpose reformatting and
transformation software would have to be acquired or developed.
Only then could individual data sets be converted and transported
to a central environmental repository.
Getting each data set owner to fully participate would be extremely
difficult. Creating a sufficient computing facility would be
extremely expensive. The disk space alone would be staggering
for such a potentially voluminous data base. The return on this
investment would be dubious. Perhaps only permitting data would
be cost beneficial. Even so, timeliness of the data updates might
be the "Achilles heel" that would reduce the data base's usefulness.
With respect to plans at the state-wide level, a few comments
are presented below that were extracted from the Growth Manage-
l-.e ior'mo' cior, on th.. shO I p1OpnlCct'. to QUAD DAIA Corp Disclosure restrictions ore stated in this document.
Must cause changes that anticipate and facilitate future
Initiates the creation of instruments and services such as:A
Data Retention Plan/Priority/Hierarchy
A Data Base Dictionary
A Modeling Plan/Wish List
An Information cost/benefit study and plan
An electronic Information Locator Service
The Data Administrator (DA):
Creates detailed designs for the corporate data base.
Concerned with data structure, definitions, dictionary,
security, integrity, procedures.
Liaison with all constituents of the data base.
Stronger on diplomacy (social skills) and decision support
system design rather than technical skills.
Authority over data not people.
Reports directly to the CIO.
The Data Base Administrator (DBA):
Implements the data design.
Technical skills with the GDBMS (Generalized Data Base
Management System) software.
*Constructs data base applications.
The first one to three years of the CIO's mission are very
challenging. During the first few years of planning and building
the new information oriented mind set, a full time effort is
required. However, at some point, the mode and style of operation
is set and becomes more of a monitoring and maintenance
activity. While periodic reviews of the entire information
strategy must continue, the process should eventually be instilled
into every mind in the organization and the human part of the
system should function somewhat routinely.
Over time, the challenge for the CIO is likely to diminish. After
the fundamental strategies, mechanisms, and plans have been
I ,, Introrrr ior. on thi, shee' is prop'etlo'y to QUAD DATA Corp Disclosure feRtricltons ore stated in this document
Address the reluctance of state agencies and employees to
use value added services. There exist a wide-spread mind
set within most government workers that there is something
wrong with paying the private sector to perform services
involving the repackaging of government (public) data.
They fail to understand that they are not paying twice for
their own data, but are paying for time savings and
convenience. A companion problem is the misconception
that staff time does not cost anything, that it is already
paid for, that it is a non-expense and therefore has no
value. Taxpayers could get more bang for the buck in the
long run if state employees understood that it is more cost
effective to work smart and take advantage of labor saving
services that are cost/effective, while using their own time
to its best advantage for the state.
Recent progress in the use of information technology is evident
in many areas and the benefits are enjoyed daily. Basic facilities
have been put in place and many of the professionals in the
water management districts are using them to great advantage.
However, there are other areas, perhaps more important, where
the pay back will take longer and the benefits will be less
immediate and less tangible. -Information management is such an
Whether we do it by design or by default, now is the time when
the basis for water management during the next decade is being
established. Future water management activities will require data
bases designed and developed now as well as accompanying
decision support systems that must be started now and evolved
over the next several years. The technology must facilitate the
retention of information (geographical, numerical, and textual) in
an accessible and usable electronic form which exploits new
technologies and techniques for managing and using data.
i ', nf r 3t.on on this sheet is proptelot to QUAD DATA Corp Disclosure rel nctions are stated in this document