• TABLE OF CONTENTS
HIDE
 Cover
 Table of Contents
 List of Figures
 Front Matter
 Main






Title: Hydrologic modeling of watersheds discharging adjacent to the Mesoamerican Reef : analysis summary, December 1, 2006
ALL VOLUMES CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00095899/00001
 Material Information
Title: Hydrologic modeling of watersheds discharging adjacent to the Mesoamerican Reef : analysis summary, December 1, 2006
Physical Description: Book
Language: English
Creator: Burke, Lauretta
Sugg, Zachary
Publisher: World Resources Institute
Place of Publication: Washington, D. C.
Publication Date: 2006
Copyright Date: 2006
 Record Information
Bibliographic ID: UF00095899
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.

Downloads

This item has the following downloads:

PDF ( 1 MBs ) ( PDF )


Table of Contents
    Cover
        Cover
    Table of Contents
        Page i
    List of Figures
        Page ii
    Front Matter
        Page iii
        Page iv
    Main
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
Full Text


WORLD
RESOURCES
INSTITUTE


Hydrologic Modeling of Watersheds Discharging

Adjacent to the Mesoamerican Reef

Analysis Summary December 1, 2006

By: Lauretta Burke and Zachary Sugg

With analytical contributions from:
Will Heyman and Shin Kobara (Texas A& M University)
Laurent Cherubin, Christopher Kuchinke, Claire Paris, and Johnathan Kool (University of Miami)

Watersheds of the Mesoamerican Reef Region


Funded by:

H SAID

UNITED NATIONS
,,+'OUNDATI ON





Part of the
ICRAN MAR
Alliance:

ICRAN



iudt^


In partnership with:




WWF


UNEP WCMC


The ICRAN Mesoamerican Reef project was supported by the U.S. Agency for International Development
(USAID) and the United Nations Foundation (UNF). The project is executed in conjunction with the United
Nations Environment Program Caribbean Environment Program (UNEP-CEP).

Collaborating on the watershed component of the ICRAN MAR partnership are WRI, UNEP World
Conservation Monitoring Center (WCMC) and the World Wildlife Fund (WWF).











Contents

A ck n ow led gm en ts .......... ......................................................... .. ...... .. .. ........... iv

Project Sum m ary ............................................................. .............. 1

Project Background ........... ................... ................. .................. ...... 2

K ey Findings ................ ............ .... ...... ..... ............ .............. 3

O v erv iew of M eth od ology ................................................................................................ 6
W atershed D elineation ................................................. .. ................ .. ............ 6
Hydrologic M odeling.......................... ......... 8
GIS Data Sets ............... .................... ........ ........................... 9
Modeling Erosion and Sediment Delivery...... ..................... .............. 11
Precipitation and R ainfall Erosivity................................... ............................. ....... 13
V vulnerability of Land to Erosion ..................................... .......................... ......... 14
M odeling Runoff and Pollutant Delivery .............. .................................... ........ 14
Circulation M modeling ......... .... .......... .............. .......... .. 15
M odel C alibration and V alidation ........................................... .................... ...... 16
Lim stations of the A nalysis......................................................... .......................... 19
Limitations of Hydrologic M odeling .................. ....... ..... ....................... .............. 20
Lim stations of Circulation M odeling ............. ............. .................. .............. 20

A analysis R results ............................... .......... .... .................................... 21
1. Sediment and Pollutant Delivery given Current Land Cover (2003/04) ............... 21
2. Comparison of Results for Current Land Cover (2003/04) to Hypothetical Natural
Land Cover ........................ ... .. .. ..... ...... . .. ......... ............ 22
3. Comparison of Current Land Cover to Three Development Scenarios in 2025........ 23
4. Extrem e E vents ......... .......... ...... .......................................... ...... ........ 25
5. Buoyant Matter Transport along the MAR........................................... 26
6. Vulnerability of the Land to Erosion ................ ..... .... ............. .............. 29
7. Local Sources of Sedim ent and N utrients............................................ .... .. .............. 31

Conclusion .................................. ......................... 33

Additional Technical Notes .. ............. ........................ 35









List of Figures


Figure A. Annual Sediment Delivery from MAR Watersheds................ .................... 4
Figure 1. Watersheds of the Mesoamerican Reef Region.......................................... 7
F igure 2. Soil E rosivity (K -factor).......................................................... ... ... .............. 12
Figure 3. Annual Precipitation Distribution................ ............ ................. ........... ... 13
Figure 4. R ainfall Erosivity (R -factor)......................................................... .................. 13
Figure 5. Temporal Relationship Between N-SPECT River Discharge Estimates and the
Mean Colored Detrital Matter (CDM) Absorption Coefficient from Imagery for the North
of H on du ras......................... ................. ....... ......... ... ......... ........ .... 19
Figure 6. Sediment and Nitrogen Delivery by Basin (for current land cover) ............... 21
Figure 7. Modeled Sediment Delivery from "Current" and Hypothetical "Natural" Land
C over.................................. ............. . ............ .......... 23
Figure 8. Land Cover Distribution within each Scenario.............................................. 24
Figure 9. CDM Plume Extent on November 14, 2000 ..................................... ....... 26
Figure 10 Estimated Monthly River Discharge from N-SPECT ..................................... 27
Figure 11. Simulated Buoyant Matter Concentration Along the MAR for December
(Current Land Cover and Sustainability First Scenarios)......................................... 28
Figure 12. Maximum Annual Simulated Buoyant Matter Concentration Along the MAR
(Current Land Cover and Sustainability First Scenarios).............................................. 28
Figure 13. Mesoamerican Reef Mapped by Buoyant Matter Concentration (Current Land
Cover and Sustainability First Scenarios)................................... ................................... 29
Figure 14. Vulnerability of Land to Erosion............... ......... .. .................... ... 30
Figure 15. Average Contribution of Sediment, Nitrogen, and Phosphorous by Sub-basin
............................................................................................................. . . . 3 1

List of Tables

Table 1. N -SPECT C-factor Coefficients ............................................................... 12
Table 2. N-SPECT Runoff Curve Numbers (Coefficients) by Land Cover Type and Soil
H ydrologic G group ............................... ... .... .. .. .... .. ....... ..... .. .................. 14
Table 3. Pollutant Coefficients for Phosphorous, Nitrogen, and Total Suspended Solids
(T S S ) ................ ..... ....... ............ ...... .... .... .. ....... . .......... ........... .. ........ 1 5
Table 4. Comparison of Natural and Current Land Cover (percent in each cover type).. 22
Table 5. Comparison of Regional Results for Annual Model Runs for Current and Natural
L an d C ov er ............... ........... .. ... ............... .... ... ....................... . . ...... . 22
Table 6. Comparison of Land Cover Scenarios (percent in each cover type) .................. 24
Table 7. Comparison of Regional Results for Annual Model Runs for Current Land
Cover and Three Scenarios in 2025 ...... ................. ...... .. ................ 25
Table 8. Percentage of Erosion, Nitrogen, and Phosphorous Sources by Country within
M AR D drainage ......... .. ....... ..................................................................... 32









List of Acronyms


CDM: Colored Detrital Matter
DEM: Digital Elevation Model
GEO: Global Environment Outlook
ICRAN: International Coral Reef Action Network
MAR: Mesoamerican Reef
N-SPECT: Nonpoint-Source Pollution and Erosion Comparison Tool
ROMS: Regional Ocean Modeling System
RUSLE: Revised Universal Soil Loss Equation
SeaWiFS: Sea-viewing Wide Field-of-view Sensor
SOA: Spectral Optimization Algorithm
SOTERLAC: Soil and Terrain Database for Latin America and the Caribbean
TSS: Total Suspended Solids
UNEP: United Nations Environment Programme
USAID: U.S. Agency for International Development
WCMC: World Conservation Monitoring Centre
WRI: World Resources Institute
WWF: World Wildlife Fund










Acknowledgments


This data CD was made possible through the generous contributions of the United
Nations Foundation and the US Agency for International Development under the
International Coral Reef Action Network (ICRAN) Mesoamerican Reef (MAR) project.

The World Resources Institute gratefully acknowledges the many partner organizations
and colleagues who contributed to this project and the production of this data CD,
including core ICRAN partners: UNEP-World Conservation Monitoring Centre
(WCMC), and the World Wildlife Fund (WWF). Thanks to Emil Cherrington (Belize
Coastal Zone Management Institute and Water Center for the Humid Tropics of Latin
America and the Caribbean, CATHALAC) for data development and guidance on spatial
data issues; Joep Luijten (consultant to UNEP-WCMC) and Lera Miles (UNEP-WCMC)
for land-use scenario modeling; Eric van Praag (consultant to UNEP-WCMC) for
facilitating the project's policy and training workshops; Liza Agudelo and Armeid
Thompson (ICRAN MAR project); and Kristian Teleki and Nic Barnard (ICRAN) for
encouragement and support throughout. Thanks to Will Heyman and Shin Kobara (Texas
A&M University) for assisting with calibration of the hydrological model, and to Jamie
Carter and Dave Eslinger (US National Oceanographic and Atmospheric Administration)
for providing guidance on implementation of the N-SPECT model. Thanks to Claire
Paris, Laurent Cherubin, Christopher Kuchinke, and Johnathan Kool (University of
Miami) for adapting a circulation model to examine sediment dispersion along the
Mesoamerican Reef; and to Nestor Windevoxhel and Alejandro Arrivillaga (The Nature
Conservancy) for jointly supporting the circulation modeling. Thanks to Ramon Frutos
(Belize Meteorological Department) for data on hurricane events and Jan Meerman
(Belize Tropical Studies) for data on land cover and watersheds. Thanks to the World
Bank / GEF Mesoamerican Barrier Reef System (MBRS) project for identifying N-
SPECT as a useful tool. Thanks to Melanie McField (WWF and Smithsonian Institution)
for constructive comments and guidance on impacts to the MAR; and Jose Vasquez,
Sylvia Marin, Shalini Cawich and Gina DeFerrari (WWF) for their helpful collaboration
under the ICRAN partnership. Thanks to Serge Andrefouet for provision of coral reef
data for the MAR. Thanks to Suzie Greenhalgh, Daniel Prager, Janet Ranganathan, Isabel
Munilla, and David Jhirad (WRI) for their valuable review and comments; and to Jenny
Guiling (WRI) for provision of data for calibration of model results. Thanks to Ron
Valenzuela (WRI) for editing the document and special thanks to Carolina de Rosas and
Ruth Nogueron (WRI) for translating this paper into Spanish.

This data CD was made possible through support provided by the office of Guatemala-
Central American Programs, U.S. Agency for International Development, under the
terms of Grant no. 596-G-00-03-00163-00. The opinions expressed herein are those of
the authors and do not necessarily reflect the views of the U.S. Agencyfor International
Development or of UNEP.









Hydrologic Modeling of Watersheds Discharging
Adjacent to the Mesoamerican Reef



Project Summary

This paper describes the methodology and results of a hydrologic analysis implemented
by the World Resources Institute (WRI) as part of the International Coral Reef Action
Network (ICRAN) Mesoamerican Reef (MAR) partnership. The objective of the analysis
was to quantify the impact of human alteration of the landscape on land-based threats to
the MAR to inform land-use planning, agricultural policy and practice, conservation
priority setting, and risk mitigation efforts.

Over a two year period, WRI collaborated with many partners in the region to evaluate
sediment and nutrients coming from land in over 400 watersheds that discharge adjacent
to the MAR. The analysis evaluates the amount of sediment and nutrients (nitrogen and
phosphorous) coming from each plot of land; the amount of eroded sediment and
nutrients reaching the river mouth (coastal discharge point); and the amount of sediment
reaching the reef. In addition, the analysis provides estimates of the increase in sediment
and nutrient delivery resulting from human activities, and predictions of future sediment
and nutrient delivery (in 2025) given varying land-use scenarios. This analysis is the first
of this scope and level of detail for the MAR region.

The results provide a preliminary overview of regional patterns of sediment and nutrient
runoff and delivery, and indicate how human alteration of the landscape can influence
these patterns. To ensure that the project's results and analytical methods support action
in the region, WRI makes the underlying data, analytical method, and modeling tools
publicly available, and has conducted training sessions with users in the region. Based on
this training, regional users can implement more detailed, focused analyses for smaller
areas, calibrating them to local conditions.

All data used in the analysis and all model results, accompanied by metadata, are
provided on the data CD, Watershed Analysis for the Mesoamerican Reef WRI/ICRAN
MAR project, 2006. This paper begins with background on the ICRAN MAR partnership
and a summary of key findings, followed by a description of the analytical methodology
and a summary of analysis results.










Project Background


Shared by Mexico, Belize, Honduras, and Guatemala, the Mesoamerican Reef (MAR),
stretches over 1,000 km, and is the largest continuous reef in the Western Hemisphere.
Alteration of the natural landscape for development, road construction, or agriculture can
have adverse impacts on coral reefs through increased delivery of sediment, nutrients,
and other pollutants to coastal waters. Threats from land clearing are higher in areas of
steep slope, intense precipitation, and erosive soils.
Appropriate land-use practices in erosion-prone areas are essential for the management of
watersheds to ensure that the transport of sediment, nutrients, and other pollutants to
coral reefs is minimized. In the Mesoamerican region, over 300,000 hectares of land is
allocated to the production of banana, oil palm, sugar cane, citrus, and pineapple crops.
Eroded sediments as well as the residues of fertilizer and pesticides used in these
industries drain through the rivers and streams and enter coastal waters along the
Mesoamerican reef.
As part of the International Coral Reef Action Network (ICRAN) Mesoamerican Reef
(MAR) project, the World Resources Institute (WRI) partnered with UNEP-World
Conservation Monitoring Centre (WCMC) and the World Wildlife Fund (WWF) to
provide comprehensive watershed analysis to complement the ICRAN MAR project's
activities on Sustainable Fisheries and Sustainable Tourism.


The ICRAN MAR watershed analysis was developed to produce information and tools
for examining the potential impact of different land use and development options in the
region and the associated impacts on water quality on the MAR. The project aims to:
Link patterns of land use within watersheds to the impacts at coral reefs, and
identify reefs at greatest risk of degradation;
Identify watersheds most vulnerable to erosion and those which contribute the
most sediment and pollution to coastal waters;
Adapt tools to forecast potential trends, evaluate different policy or development
options, and facilitate improved land management within the region;
Use the results of the models and diagnostic tools to help educate and encourage
key stakeholders to adopt a suite of "better management practices" to reduce
impacts on the coastal and marine resources.

The ICRAN MAR watershed project includes analytical components looking at land
cover change and the associated impacts on runoff, erosion, and sediment and pollutant
delivery to and transport within coastal waters. The project also includes on-the-ground
activities with agricultural businesses to implement better management practices. Many
local partners were consulted on modeling methods, for data input and evaluation, and on
agricultural management practices. Three ICRAN partners collaborated on this effort:
UNEP-WCMC developed scenarios of land cover change and provided land
cover data sets as input to the hydrologic modeling;









WRI implemented the watershed delineation and hydrologic analysis for the
MAR region, performed analysis of vulnerability to erosion, and coordinated the
circulation modeling along the MAR;
WWF led the work with agri-business to implement better management
practices.

WWF's work with agricultural businesses focuses on reducing the presence of priority
pesticides in the MAR marine environment and controlling soil erosion from major
commercial agricultural sectors. WWF is working with business partners on banana,
pineapple, citrus and sugar cultivation. The analysis performed under this project helps
WWF and others to set priorities for targeting better land-management practices as well
as to guide land-use planning.



Key Findings

This hydrologic analysis serves to integrate a wide range of data, and adapt modeling
tools for an innovative, region-wide analysis for the MAR. The region-wide results
presented in this paper should be considered preliminary and indicative of the overall
pattern and magnitude of erosion and nutrient and sediment delivery across the region.
An important aspect of the project is to provide these modeling tools to partners in the
MAR region so that they might apply them at higher resolution to produce more detailed
results for smaller areas within the MAR region. This approach will allow for refinement
and better calibration of the model to local circumstances within the region.

The origin of sediments and nutrients reaching the MAR:

o Most of the sediment and nutrients delivered by watersheds along the MAR originate
in Honduras. It is estimated that over 80% of sediment and over half of all nutrients
(both nitrogen and phosphorous) originate in Honduras.

o Guatemala was identified as a source of about one-sixth of all sediments and about
one-quarter of all nitrogen and phosphorous entering coastal waters along the MAR.

o The modeling suggests that compared to the other countries, relatively minor
percentages of the regional sediment load come from Belize and Mexico. Belize
contributes between 10-15% of nutrients and Mexico is estimated to contribute about
5%of the nutrients from all modeled watersheds. The estimate for Mexico is probably
an underestimate, as the contribution of underground rivers is not included in this
analysis.

o Of the 400 watersheds in the MAR region, the Ulua watershed in Honduras was
found to be the largest contributor of sediment, nitrogen, and phosphorous. Other
large rivers found to be significant contributors of sediment and nutrients are the












Patuca (in Honduras), Motagua (in Guatemala and Honduras), Aguan (in Honduras),
Dulce (in Guatemala), Belize River (in Belize), and Tinto o Negro (in Honduras).



Figure A. Annual Sediment Delivery from MAR Watersheds


i. 2006.





Runoff, erosion, and nutrient delivery to coastal waters are increasing:


o As a result of human alteration of the landscape, runoff and associated river discharge
at river mouths has nearly doubled; sediment delivery at river mouths has increased
by a factor of 20; nitrogen delivery has increased by a factor of 3, and phosphorous
delivery by a factor of 7. (Ratios are based on model results for current (2003/04)
land cover and on hypothetical natural (unaltered) land cover.)


SCoral Reefs
Basins
i Lakes
Rivers
[ National Borders
Sediment delivery by basin (metric tonne)
0-10,000
I5500 O0 j 000 OCJ.
3.,10 COCi 5O COO C-liQ
I 0 00 OcK) 1000 0,0'j 00
100 0 .l-, i OiA.


Nole lModeled eminle .l' Jinmieni d.ben
are m ail.al, of palT adhl nlmagu but
uc no, .bo[lte Icl-il' e Thr. rrnflr Ic l
etimlnlart ?Tqmw i ihe J n 3 wlr CeP
emtmile !' cdwenl Jelr%\









The potential impacts of development and land-use paths are varied:

o Under land-use scenarios which favor free markets and little policy regarding the
environment, nutrient delivery is likely to increase by about 10% by 2025, while
sediment delivery might increase by 13% or more.

o If environmental policies that favor sustainable development are implemented,
nutrient and sediment delivery are likely to be reduced by at least 5% from current
levels, promoting recovery of degraded corals.

o Implementation of better agricultural management practices will yield additional
reductions in sediment and nutrient delivery beyond those evaluated in this study,
which has focused on the effect of changes in land cover.










Overview of Methodology


In collaboration with partners in the MAR region, WRI:
implemented a watershed delineation for all land areas draining along the
Mesoamerican reef;
implemented a hydrologic analysis to examine sources of sediment and nutrients
from this entire drainage area, as well as the delivery of sediment and nutrients to
coastal waters;
applied this hydrologic analysis tool to examine sediment and nutrient delivery
for several land cover scenarios (current land cover, original or "natural" land
cover, and three scenarios of land cover in 2025);
provided outputs of the hydrologic analysis as inputs to a circulation model to
examine sediment transport along the MAR; and
collaborated with partners on calibration and validation of model results.



Watershed Delineation
Watersheds are an essential unit for this hydrologic analysis, as they serve to link a plot
of land with its stream and river network, and its point of discharge to the sea.
Watersheds were delineated from a hydrologically corrected 250m digital elevation
model (DEM). This DEM is based on 90m resolution NASA Shuttle Radar Topography
Mission (SRTM) data, which were projected and resampled to 250m resolution for use in
this project's land cover and hydrologic modeling. As elevation data from radar has
inherent inaccuracies, such as capturing the tops of trees and buildings as part of the
elevation, the elevation data needed some "correction" as part of the delineation process.
Some mapped rivers and lakes were superimposed on the DEM through a process called
"burning" the DEM. Rivers (as lines) from La Comision Centroamericana de Ambiente y
Desarrollo (CCAD) and lake and lagoon data (polygons) from WWF were used to
develop a data set for correction of the DEM.1 This grid was superimposed (as -20m
elevation) into the DEM to ensure that water was forced to flow in these depressions.
This "burned" version of the data set was used for basin delineation in the Environmental
Systems Research Institute's (ESRI) ArcMap software. For each 250m grid cell in the
DEM, the direction of water flow (FlowDirection) and the number of cells flowing into
each cell (FlowAccumulation) was identified, as well as which cells comprise a basin
(area draining to a single coastal point.) The 250m grid cell with the maximum flow
accumulation in each basin was identified as the "pour point," or point of discharge to the
sea. Basins of less than 5 km2 area were excluded. Over 430 basins of at least 5 km2 area
were identified. (Some additional technical notes on watershed delineation can be found
at the end of this report.)

1 See metadata on WatershedAnalysis for the Mesoamerican Reef data CD for data sources, as well as the
GIS data set used for correction.











This final set of watershed delineations for the MAR region benefited from several
rounds of review and the provision of additional data by several project partners.
Delineations were again reviewed at the ICRAN MAR watershed workshop held in
Belize in August 2006. The watershed delineations are reliable for most of the MAR
region, but are inaccurate for the Yucatan, due to widespread karst topography with
underground rivers and lack of perennial surface waters.


Figure 1. Watersheds of the Mesoamerican Reef Region


Basins
National Borders
Coral Reefs
Rivers
Lakes
Elevation in m (Srtm_250)
0 140
140 370
370-610
I. 610 850
850-1120
1120-1460
1460-1900
1900-3290





r a ^

- er% Op


t -.Source WRI, 2006









Hydrologic Modeling
Threat from land-based sources of sediment and nutrients was evaluated using the
Nonpoint Source Pollution and Erosion Comparison Tool (N-SPECT), developed by the
U.S. National Oceanographic and Atmospheric Administration (NOAA). N-SPECT is a
public-domain software which runs as an optional extension within ESRI's ArcMap
software. N-SPECT combines information on the physical environment (elevation, slope,
soils, precipitation, and land cover) to derive estimates of runoff, erosion, and pollutant
sources (nitrogen, phosphorous and total suspended solids) from across the landscape, as
well as estimates of sediment and pollutant accumulation and concentration in stream and
river networks. N-SPECT was implemented at 250m resolution for the MAR region.
Hence, N-SPECT evaluates each 250m resolution grid cell (6.25 ha area) for its
contribution to runoff, sediment, and pollutant delivery within the watershed. The results
of N-SPECT analyses are intended to be used as screening tools to help understand and
predict the impacts of management decisions on water quality and, ultimately, on near
shore coral reef health.
N-SPECT can be run to evaluate annual or event-based runoff, erosion, sediment, and
pollutant delivery. It can also be modified to evaluate these outputs on a monthly basis.
N-SPECT runs can either calculate the accumulation of runoff, sediment, and pollutants
across the landscape (this is the standard run) or "local effects," meaning how much
sediment, pollutant, or runoff comes from each individual grid cell. All of these options
were employed in the analysis of land-based threat along the MAR.
a. Annual runs For the MAR region, N-SPECT was applied to evaluate annual
runoff, sediment, and nutrient delivery associated with several different land
cover scenarios current land cover (circa 2003/04), hypothetical "natural" land
cover, and three scenarios of potential land cover in 2025. In these "annual" runs,
N-SPECT was run with a consistent elevation, slope, soils, and annual
precipitation. Only the land cover data set was varied so that the influence of land
cover change on sediment and nutrient delivery could be evaluated.
b. Monthly runs N-SPECT was also run using monthly precipitation data in order
to derive monthly estimates of runoff and sediment delivery at river mouths,
which are essential inputs to the circulation modeling for the region, which is
implemented on a monthly time frame.
c. Local effects / sources of pollution N-SPECT was run in "local effects" mode
on current land cover (circa 2003/04) in order to evaluate how much sediment and
nutrients (nitrogen and phosphorous) originate from each individual 250m grid
cell independent of contributions from adjacent / upstream grid cells.
d. Storm Events N-SPECT was also applied to examine the sediment and nutrient
runoff and delivery associated with hurricanes in the region. This application
using the "storm event" feature of N-SPECT, used the same elevation, slope, and
soils data as the previous runs, as well as current land cover (circa 2003/04), but
used local precipitation data for the multi-day storm events.










GIS Data Sets
All GIS data sets used in the hydrologic analysis are in Universal Transverse Mercator
(UTM) projection (NAD1927 for Central America datum). All data inputs and model
results are available on the data CD, Watershed Analysis for the Mesoamerican Reef
WRI / ICRAN MAR project, 2006.


Input Data sets used in the hydrologic analysis using N-SPECT were:
1. Elevation a 250m resolution digital elevation model was derived from 90m
resolution NASA Shuttle Radar Topography Mission (SRTM) data.
2. Soils Soils come from a vector (polygon) data set acquired from the Soil and
Terrain Database for Latin America and the Caribbean (SOTERLAC).
3. Precipitation monthly precipitation data come from a 1 km resolution global
raster precipitation data set called WorldClim2, which reflects long-term average
precipitation. Monthly precipitation data were summed to produce annual rainfall
estimates. All precipitation data were converted to inches for use in N-SPECT.
4. Land Cover five representations of land cover were used to allow for
evaluation of the effect changes in land cover have on sediment and nutrient
delivery to river mouths along the MAR.
o "Current" Land Cover National "Ecosystem Maps" for Mexico,
Guatemala, and Honduras (2003) and Belize (2004) were used as the basis
of the current land cover map. Data were merged for the four countries
and gridded at 250m resolution.3
o "Natural" Land Cover "Original" or natural land cover for Belize,
Guatemala, and the Yucatan comes from a vector data set by Pronatura on
"Original land cover." 4

Land Cover in 2025. Three land cover change maps from 2005 to 2025 were
developed for the MAR region, based on the Markets First, Policy First, and
Sustainability First scenarios from the UNEP Global Environment Outlook (GEO)



See 2005. Very High Resolution Interpolated Climate Surfaces for Global Land Areas. (14 pages)
Description of the development of the WorldClim global climate data set. In International Journal of
( lnIt,,,. .'i. Robert J Hijmans, et al. on WatershedAnalysis for the Mesoamerican Reef data CD, WRI /
ICRAN MAR project, 2006.

3Some "nodata" cells for land cover along the coast were filled in based upon the land cover class of the
nearest land cover cell. In addition, some locations were recorded as "water" based on data from Pronatura
on waterbody locations in Belize, Guatemala, and the Yucatan.

Land classes have been reclassified to match the classes used in the ICRAN MAR watershed analysis,
and gridded at 250m resolution. For Honduras, the Ecosystem map of Honduras (2003) was modified as
follows "Cultivated" and "Developed" land cover was reclassified based on the adjacent land cover type,
with the exception of water (not allowed to expand) and mangrove (only allowed to expand if below 20m
elevation).









process.5 The scenarios envisage differing social, political, and economic trajectories,
emphasizing outcomes for the environment and human well-being.
o "Markets First" scenario of land cover in 2025 In this market-oriented
development scenario based on the GEO Market First Scenario for Latin
America, most of the world adopts the values and expectations prevailing
in today's industrialized countries. The wealth of nations and the optimal
play of market forces dominate social and political agendas. Public policy
is geared toward supporting commercial interests and promoting the open
exchange of goods and services. Social and environmental policies receive
little attention or financial support, for it is assumed that economic growth
is in itself a sufficient route to progress. This scenario sees the greatest rate
of agricultural expansion. The 250m resolution land cover data reflecting
this scenario were developed at UNEP-WCMC.
o "Policy First" scenario of land cover in 2025 In this scenario based on
the GEO Policy First Scenario for Latin America, decisive initiatives are
taken by governments in an attempt to reach specific social and
environmental goals. A coordinated pro-environment and anti-poverty
drive balances the momentum for economic development at any cost.
Environmental and social costs and gains are factored into policy
measures, regulatory frameworks, and planning processes. All these are
reinforced by fiscal levers or incentives such as carbon taxes and tax
breaks. Land use becomes better regulated, especially around riverine
corridors. Associated 250m resolution land cover data were developed at
UNEP-WCMC.
o "Sustainability First" scenario of land cover in 2025 In this scenario
based on the GEO Sustainability First Scenario for Latin America, a new
environment and development paradigm emerges in response to the
challenge of sustainability, supported by new, more equitable values and
institutions. A more visionary state of affairs prevails, where radical shifts
in the way people interact with one another and with the world around
them stimulate and support sustainable policy measures and more
accountable corporate behavior. Efficiency in the use of energy, land, and
material resources is promoted. There are efforts to adopt an ecosystem
approach to land use planning, with particular attention to watershed
protection. Associated 250m resolution land cover data were developed at
UNEP-WCMC.


A quantification of land cover change was developed from preliminary inputs to GEO4 from the
International Futures and IMAGE modeling teams, and a set of accompanying narratives were adapted
from the Latin America and Caribbean group's input to GEO4. Protected area scenario maps were
developed on the basis of international policy targets for the coverage of biomes and endangered species,
and implemented differently in each scenario. The CLUE-S model was then used to allocate land cover
change. Further information can be found in Luijten, J., Miles, L., Cherrington, E. (2006) Land use change
modellingfor three scenariosfor the A4R region. Technical Report to ICRAN MAR project. UNEP-
WCMC.









Modeling Erosion and Sediment Delivery
N-SPECT evaluates annual erosion for each grid cell based on the Revised Universal Soil
Loss Equation (RUSLE) developed by the U.S. Dept. of Agriculture6. RUSLE combines
aspects of slope, rainfall, soil, and land cover to estimate annual soil loss for that location.

Equation 1: RUSLE

Average Annual Soil Loss (tons/acre) = R K L S C P

R -Rainfall-runoff erosivity factor
K -Soil erodibilityfactor
L *S Slope steepness and length factors
C -Cover-management factor
P -Supporting practices factor

The input data sets listed above are elevation, soils, precipitation, and land cover. N-
SPECT uses these data sets as the basis for deriving most of the factors that are used in
the RUSLE calculation. Some of the derivations are complicated and are described below
with additional detail in the technical notes at the back of this report. In summary:
LS-factor The LS factor adjusts erosion rates based on topography, assigning
higher rates to longer or steeper slopes and lower rates to shorter or flatter ones7.
Slope steepness (S) and slope length (L) are derived from the DEM by N-SPECT.
These are combined into one "LS-factor" grid.
K-factor The SOTERLAC soils database includes the soil-erodibility factor (K-
factor) attribute, which represents a soil's susceptibility to erosion by rainstorms.
It is an integrated average parameter based on several different erosion and
hydrologic processes. A low K-factor (about 0.05 to 0.2) indicates a high
resistance to erosion and a high K-factor (about 0.4 or greater) indicates easily
eroded soil. N-SPECT develops a GRID based on the K-factor for RUSLE
calculations.
R-factor A grid of the rainfall and runoff erosivity factor (R) for the study area
is an input to the N-SPECT model and must be acquired or developed by the user.
R-factor represents the average annual erosive effect of storms and is based on an
erosivity index (EI) calculated from the kinetic energy of storms and their
maximum 30-minute rainfall intensities. These data were not available for the
MAR region so an alternate method was used to approximate R-factor.
Development of this factor is described in the next section.
C-factor Each land cover type, such as forest, grassland, or cultivated land has
an associated cover-factor or C-factor. This is a relative erosion rate for the given
land cover. (See Table 1.) Cultivated land, with a C-factor of .240 is rated as
being 60 times as erosive as forest, which has a C-factor of .004. These C-factors

6USDA Agriculture Handbook No. 703, found on WatershedAnalysis for the Mesoamerican Reef data
CD, WRI / ICRAN MAR project, 2006.
7 Ward, Andrew D. and Stanley W. Trimble. Environmental Hydrology, 2nd Ed. CRC Press, LLC. 2004.










for land cover in the MAR region were adapted from C-factors provided by
NOAA in N-SPECT. Use of locally derived relative erosion rates would be a
valuable refinement of the model.
*P-factor RUSLE includes a supporting practices factor or P-factor, which
allows inclusion of the influence of conservation practices (such as strip-cropping
and terracing) that control and mitigate erosion. The "Supporting Practices" factor
module is not available in the current version of N-SPECT. For this reason, we
have focused our analysis on the effect of change in land cover on sediment and
pollutant delivery to river mouths adjacent to the MAR. In addition, information
on specific management practices across the MAR region is fairly limited.


Table 1. N-SPECT C-factor Coefficients
Code Land Cover Category Cover-Factor
3 Low Intensity Developed 0.030
4 Cultivated Land 0.240
5 Grassland 0.050
7 Forest 0.004
9 Scrub/Shrub 0.014
10 Palustrine Forested Wetland 0.003
17 Bare Land 0.700
18 Water 0.000
Source: USDA Agriculture Handbook No. 703


Fiaure 2. Soil Erosivitv (K-factor)


j E Si n% CO LM_ a Itao g

WO 0 42
o2,O~


Source: Soil and Terrain Database for Latin America and the Caribbean











Precipitation and Rainfall Erosivity


Figure 3. Annual Precipitation Distribution
Precipitation Scenario: N-SPECT requires a
raster (grid) precipitation data set, specification
of the rainfall type in the region (from four -'
synthetic 24-hr rainfall distributions), and the
average number of "raining days," which is
complicated by the large area and range of
precipitation across the region. All annual'
model runs were conducted using a long-term
average yearly rainfall grid (which needs to be "
specified in inches for N-SPECT), and number
of rainy days per year set to 40.8 This number of
rainy days was selected based on calibration of
the model to the MAR region by Will Heyman
and Shin Kobara at Texas A&M University.9


Source: WorldClim database


Figure 4. Rainfall Erosivity (R-factor)
R-Factor: R-factor was empirically derived
based on annual precipitation and elevation using
an equation for rainfall erosivity in Costa Rica.
... ;;,:'......... Project partners at Texas A&M University
gJ. performed statistical validation of the resulting
e ."',' R-factor for the MAR region and found this
s" equation to be a statistically valid method for use
in estimating erosion with the RUSLE for the
MAR region.10 The equation used for R-factor is
included in the technical notes at the end of this
report.



Source: WRI, 2006


8 Note: Precipitation must be in inches for N-SPECT. A single number of rainy days must be selected for
the entire study area, defined as the "average number of storms in a location in a year." The model was run
with "Type II rainfall" selected, as this best represents areas with intense rainfall events.
9 See document on "Hydrologic Model Calibration" on the Watershed Analysis for the Mesoamerican Reef
data CD.
10 Ibid.










Vulnerability of Land to Erosion
The N-SPECT model evaluates erosion and pollutants coming off the land for a given
land cover. The project also evaluated the inherent vulnerability of the landscape to
erosion based on slope, soil erodibility, and annual precipitation. This simplification of
the RUSLE used in N-SPECT excludes land cover. It serves to identify vulnerable areas
where conversion to an erosive land cover type should be avoided or, where converted,
better management practices should be targeted.


Equation 2: Vulnerability of Land to Erosion -


Vulnerability = R K S 0.6

R -Rainfall-runoff erosivity factor
K -Soil erodibilityfactor
S -Slope (in degrees)


Modeling Runoff and Pollutant Delivery
N-SPECT evaluates runoff based on soil characteristics, land cover, topography, and
precipitation. Runoff calculations are based on curve numbers developed by the U.S.
Department of Agriculture that reflect the general permeability of a given soil type.11
Curve numbers are associated with a soil's hydrologic group classification (A, B, C, or
D), which is a measure of the drainage potential or infiltration rate of the soil. Soils in
Group A are well drained, while soils in Group D drain poorly and tend to be water-
logged. Hydrologic group is an attribute contained in the SOTERLAC soils database.
Curve numbers also depend on land cover types. N-SPECT selects curve numbers for use
in runoff calculations based on the combination of hydrologic soil group and land cover
type at each grid cell. These curve numbers are presented in Table 2.

Table 2. N-SPECT Runoff Curve Numbers (Coefficients) by Land Cover Type and Soil
Hydrologic Group
Code Land Cover Category CN-A CN-B CN-C CN-D
3 Low Intensity Developed 0.61 0.75 0.83 0.87
4 Cultivated Land 0.67 0.78 0.85 0.89
5 Grassland 0.39 0.61 0.74 0.80
7 Forest 0.30 0.55 0.70 0.77
9 Scrub/Shrub 0.30 0.48 0.65 0.73
10 Palustrine Forested Wetland 0.00 0.00 0.00 0.00
17 Bare Land 0.77 0.86 0.91 0.94
18 Water 0.00 0.00 0.00 0.00
Source: Urban Hydrology for Small Watersheds TR-55, USDA-NRCS



" Refer to Urban Hydrologyfor Small Watersheds TR-55, USDA-Natural Resource Conservation Service
(NRCS) on Watershed Analysis for the Mesoamerican Reef data CD.









N-SPECT evaluates pollutant loads based on runoff, land cover, and topography.
Coefficients representing the contribution of each land cover class to runoff of pollutants
(nitrogen, phosphorous, and total suspended solids) are applied to land cover data sets to
approximate pollutant loads. These coefficients reflect the expected pollutant mean
concentration from each land cover type and were derived from published studies and
research by NOAA and are provided with the N-SPECT model. Modeling can be made
more accurate through the use of locally derived pollutant coefficients, but such data
were not available for the MAR region. The coefficient derivation process is described in
the N-SPECT Technical Guide on the Watershed Analysis for the Mesoamerican Reef
data CD.

Table 3. Pollutant Coefficients for Phosphorous, Nitrogen, and Total Suspended Solids
(TSS)
Class Coefficients
Value Name Phosphorous Nitrogen TSS
3 Low Intensity Developed 0.18 1.77 19.1
4 Cultivated Land 0.42 2.68 55.3
5 Grassland 0.48 2.48 55.3
7 Evergreen Forest 0.05 1.25 11.1
9 Scrub/Shrub 0.05 1.25 11.1
10 Palustrine Forested Wetland 0.20 1.10 19.0
17 Bare Land 0.12 0.97 70.0
18 Water 0.00 0.00 0.0
Source: Urban Hydrology for Small Watersheds TR-55, USDA-NRCS



Circulation Modeling


The University of Miami adapted a circulation model to examine the transport of buoyant
matter along the MAR. (Buoyant matter includes suspended sediments, organic detritus,
and dissolved nutrients.) The Regional Ocean Modeling System (ROMS) provides spatial
and temporal modeling of ocean circulation and transport by currents of river discharge
along the MAR. The model includes the barrier reef, reef lagoon, and adjacent oceanic
waters, as well as bottom topography (bathymetry) at 1km resolution. The horizontal
resolution of the simulation is 2km (grid cell size).12 Both the state of the ocean
(temperature, salinity, currents, and tides) and the surface fluxes (wind, rain, solar, and
radiative heat fluxes) are accounted for in the model simulation of oceanic and coastal
waters. They were taken from the Levitus ocean and atmospheric climatology,13 which
provides long-term monthly averages for a year. Monthly river discharge and sediment
delivery (used as a proxy for buoyant matter load) were provided from the N-SPECT



12 The vertical resolution of the ROMS model varies with distance from shore.
13 For more information on Levitus climatology see
http://ingrid.ldeo.columbia.edu/SOURCES/.LEVITUS94/ (ocean climatology) and
http://icoads.noaa.gov/status.html (atmospheric climatology).









model.14 Within the ROMS model, rivers are defined as point sources. The tracking of
buoyant matter dispersal is computed in ROMS through hydrodynamic passive
transport.15

From these methods, the model produces a climatology (reflecting long-term average
conditions) of the circulation and buoyant matter transport in the MAR region. Four
numerical simulations were done to understand the response of the ocean to the river
runoff and sediment delivery associated with different land cover scenarios and storm
events. The model was initiated with data reflecting "current" conditions (land cover for
2003/4 and mean monthly precipitation). The model was run for two model years to get
to a valid equilibrium representation of accumulated river discharge and sediment
delivery. The model reached its buoyant matter equilibrium in winter of the second year.
All simulations for comparison were started at the end of that second year. Simulations
were run to capture river discharge and sediment delivery associated with (a) the "current
land cover" scenario, (b) the "sustainability first" scenario, (c) the Hurricane Keith storm
event, and (d) the Hurricane Mitch storm event.



Model Calibration and Validation


Complex, multi-stage modeling should be validated at every possible stage of the
analysis. This analysis serves to integrate a wide range of data, and adapt modeling tools
for an innovative, region wide analysis. Where possible, data from published sources or
proxy indicators derived from remote sensing were used to calibrate and validate model
results, and these are described below. An important aspect of the project, however, is to
provide these modeling tools to partners in the MAR region, so that they might apply
them at higher resolution and then use local data to initially calibrate, and later validate
model results. The region-wide results presented in this paper should be considered
preliminary and indicative of overall patterns of erosion and nutrient and sediment
delivery across the region.


1. Evaluation of input data sets. The best available region-wide data sets were
used in this analysis. All input data sets (elevation, rivers, precipitation, soils, and
land cover) were evaluated for spatial accuracy and attribute consistency. No



14 Monthly estimates of sediment delivery (total erosion within the watershed) are provided as input to the
ROMS model. These estimates overstate sediment delivery, but are indicative of relative distribution,
seasonal patterns and the overall magnitude of sediment delivery. Within the ROMS model, the sediment
was used as a proxy for buoyant matter, which includes suspended sediments, organic detritus, and
dissolved nutrients.
15 The hydrodynamicc passive tracer transport" of the ROMS model uses an "advection-diffusion" scheme
which captures both transport by current advectionn) and dispersion by turbulence (diffusion.) Using
outputs of the ROMS model, the sources of the buoyant matter (river mouths) and destination (reef
locations) are linked through a connectivity matrix.










input data was identified as having limitations that would significantly affect
model results.16
2. Calibration of runoff and rainfall parameters. Runoff and erosion estimates
are very dependent on two input parameters the rainfall erosivity factor (R-
factor) and number of rainy days per year. Project collaborators at Texas A&M
University tested model input parameters (both R-factor and number of rain days)
to calibrate the runoff and sediment delivery components of the model. The
selected parameters achieved good correlations with discharge predictions from a
water balance model, and erosivity estimates.17 This calibration is described in
detail in "Hydrologic Model Calibration" on the Watershed Analysis for the
Mesoamerican Reef data CD.
3. Validation of river discharge. River discharge values from N-SPECT were
compared with values calculated using a water balance model"1 which has
compared well with actual runoff data in six watersheds in the Gulf of Honduras.
Discharge estimates had comparable patterns and were generally within 25
percent of the published estimates.
4. Local rates of erosion and pollutant runoff. Within N-SPECT, erosion source
estimates rely on application of the RUSLE, while pollutant runoff estimates rely
on application of runoff curve numbers developed by USDA. Both of these
estimation techniques are widely used. These equations could be made more
accurate in the future through the use of locally derived erosion and pollutant
coefficients, but such data were not available for the MAR region. Although only
limited local data were available for validation of these estimates, the overall
estimates of local erosion, N and P runoff (sources) were found to be within
reasonable bounds compared to estimates from other areas.19
5. Sediment and pollutant delivery to river mouths. Few data from the MAR
region are available for validation of model outputs. Sediment delivery results
have been compared with survey data from the Watershed Reef Interconnectivity
Scientific Study (WRIScS)20 and modeled estimates from Thattai et al.21 In


16 SRTM data, radar-derived elevation data, have inherent inaccuracies due to detection of natural and
manmade features such as trees and buildings. These errors will effect some slope calculations, but will not
effect overall model results significantly.
17 Water balance equation from Thattai, Deeptha, Bj6m Kjerfve, W.D. Heyman, 2003. Hydrometeorology
and variability of water discharge and sediment load in the Inner Gulf of Honduras, Western Caribbean.
Journal ofHydrometeorology 4: 985-995. Erosivity estimates from Mikhailova, E.A., R.B. Bryant, S.J.
Schwager, and S.D. Smith. 1997. Predicting rainfall erosivity in Honduras. Soil Science Society ofAmerica
Journal. 273-279.
18 Thattai, Deeptha, Bj6m Kjerfve, W.D. Heyman, 2003. Hydrometeorology and variability of water
discharge and sediment load in the Inner Gulf of Honduras, Western Caribbean. Journal of
Hydrometeorology 4: 985-995.
19 Ranges of reasonable values (low, typical, and high) were established for erosion (by applying the
RUSLE), for nitrogen runoff (by applying a nitrogen balance equation) and for phosphorous runoff (by
coupling RUSLE results with estimates of phosphorous concentration in the soil).
20 Nunny, Rob, M. Santana, P. Stone, D. Tillet, and Prof. D. Walling, 2001. An Investigation of the Impact
on Reef Environments of Changing Land Use in the Stann Creek District in Belize. Technical Report
Module 3. The Watershed Reef Interconnectivity Study (WRIScS) 1997-2000.









addition, sediment delivery and sediment yield (per unit area) were compared
with estimates for similar size and gradient watersheds from around the world.22
Estimates of sediment delivery from N-SPECT were found to be higher than
expected in the largest watersheds in the region. A key limitation of the N-SPECT
model is that it does not adequately account for sediment and nutrient attenuation
(loss or redeposition) within the watershed en route to the river mouth.23
Overestimation is likely to be greater for sediments than nutrients, as large
sediments are more affected by redeposition than nitrogen (in solution) or
phosphorous (attached to smaller soil particles.) N-SPECT results, therefore,
indicate a high end, or "worst case scenario" of nutrient and sediment delivery.
Overall yields (per unit area) of N and P seem reasonable, given the expected
ranges of N and P in runoff, described above. The estimates of sediment yield and
sediment delivery by basin seem high, particularly for the largest watersheds. In
the largest watersheds, sediment delivery might be overestimated by a factor of
two to four. Estimates of sediment and nutrient delivery at river mouths,
therefore, indicate the relative patterns and order of magnitude, but should not be
regarded as accurate absolute values.
6. Sediment Transport to Reef. Modeling of the extent of buoyant matter reaching
the MAR estimated in the ROMS model is being compared to Colored Detrital
Matter24 (CDM) maps derived from Sea-viewing Wide Field-of-view Sensor
(SeaWiFS) satellite imagery. The Spectral Optimization Algorithm (SOA)25 was
used to process SeaWiFS data in the optically complex26 waters of the Caribbean
Sea along the MAR. The output product, acDM(443) (called CDM) is an
absorption coefficient which is a good proxy for the buoyant matter transport.
Figure 5 reflects the lag between mean monthly river discharge and mean CDM in
processed SeaWiFS images.









21 Thattai, Deeptha, Bjrn Kjerfve, W.D. Heyman, 2003. Hydrometeorology and variability of water
discharge and sediment load in the Inner Gulf of Honduras, Western Caribbean. Journal of
Hydrometeorology 4: 985-995.
22 John D. Milliman and James P. M. Stvitski, 1992. Geomorphic / Tectonic Control of Sediment Discharge
to the Ocean: The Importance of Small Mountainous Rivers. In The Journal of Geology, 1992, volume 100,
p. 525-544.
23 N-SPECT applies a sediment delivery ratio (SDR) within each individual grid cell to adjust sediment
load, but not across the basin to account for redeposition en route to the river mouth.
24 CDM includes dissolved organic carbon from soil and plants, often referred to as Colored Dissolved
Organic Matter (CDOM), plus detrital particles (dead organic matter).
25 Kuchinke, C.P., H.R. Gordon, L.W. Harding, Jr., and K.J Voss, A coupled oceanic and atmospheric
spectral optimization algorithm for ocean color imagery in Case 2 waters: a validation for Chesapeake Bay,
SUBMITTED FOR PUBLICATION, Remote sensing of Environment (2006).
26 Optically complex waters include multiple constituents such as colored dissolved organic matter as well
as phytoplankton.










Figure 5. Temporal Relationship Between N-SPECT River Discharge Estimates and the
Mean Colored Detrital Matter (CDM) Absorption Coefficient from Imagery for the North of
Honduras
10
7 x10

6



4-

3



1 4 --River Discharge
CDM mean
0 1 1
Mar Jun Sep Jan Apr Jul Oct Jan
Months
Source: University of Miami, 2006.


CDM classifications from SeaWiFS images are used to evaluate two aspects of
the ROMS circulation model results (1) the seasonal variation of the buoyant
matter plume dispersion patterns, and (2) the seasonal variation of the total
buoyant matter concentration. Buoyant matter plumes from the ROMS circulation
simulation were compared to the CDM classifications along transects
perpendicular to the coast.

Seasonal patterns are similar between the estimated buoyant matter plumes from
the ROMS simulations and the seasonal trends observed in the SeaWiFS data,
though there is considerable year-to-year variation in the latter. Overall, the model
agrees very well with the observations both in time and location. Some
discrepancies exist, however, in the structure of the gradient of the plume. If we
compare the position of the edge of the average CDM plume from SeaWiFS and
the modeled buoyant matter plumes, the CDM plumes varies seasonally between
10 and 40km while in the ROMS model the plume is almost always 40km wide.
Such a difference could be due to a diffusion coefficient used in the ROMS
model, which is too small. Validation of the circulation model is described in
greater detail in "Dynamics of Buoyant Matter in the MAR Region" on the
Watershed Analysisfor the Mesoamerican Reef data CD.



Limitations of the Analysis


Any multi-stage modeling process will have inevitable inaccuracies, so it is vital to be
aware of model limitations and only use results appropriately.









Limitations of Hydrologic Modeling
A key limitation of the N-SPECT model is that the model does not adequately account
for sediment and nutrient attenuation (loss or redeposition) within the watershed en route
to the river mouth. As a result, estimates of sediment and nutrient delivery at river
mouths are exaggerated. The absolute numbers are not accurate, but are indicative of the
overall magnitude and patterns of sediment and nutrient delivery across the region. These
estimates are still useful for examining relative patterns and the implications of different
policy scenarios, because the estimates of percentage change are valid.
A second limitation of the modeling is the focus on the role of land cover change without
considering the effect of specific land management practices. This emphasis arises from
two issues. For the MAR region, information was only available on land cover type and
not on the location of specific management interventions. In addition, detailed
information on reductions in erosion and nutrient runoff associated with each of the
agricultural management practices is not currently available. Once this information is
developed, however, it should be possible to evaluate benefits in N-SPECT by treating
each management intervention on each land cover type as a new, unique land cover
category with a specific land cover factor (C-factor) and pollutant coefficients for the
given category (i.e. citrus crops with erosion control). This would be a valuable extension
to the current analysis.
Another limitation of the model is that dams, which serve to trap sediment, have not been
included in the model. It should be technically possible to develop a function to include
dams, provided that information on sediment retention rates by dams is available. This
feature could be considered in local applications of the N-SPECT model.

Limitations of Circulation Modeling
Regarding the circulation modeling of the MAR, the first limitation is that the result is a
climatology. Namely, the circulation is representative of the most common conditions
that one would encounter each year if every extreme or unusual weather or ocean state
event were removed. It reflects long-term average conditions. Therefore, day-to-day
comparison between the model circulation and current observations is irrelevant. But if
compared with observations during the same month over several years, then common
patterns can be found and compared to the model results.
The second limitation is the sparseness of the CDM interpretations because of cloud
cover in the SeaWiFS images they were derived from. For example, there are few clear
SeaWiFS images of the MAR region during the summer months. However, there are
enough observations to recover the global trend of the CDM concentration over a year as
shown in Figure 5.
The third limitation is that there is no direct relationship between the CDM absorption
coefficient and the CDM concentration. Direct measurements from water samples, which
are not available yet, are necessary to get such information. Therefore, only the
circulation model can be used to estimate the buoyant matter load to the reefs.
Finally, in the numerical model, buoyant matter will not accumulate or deposit, but will
be constantly transported until flushed out of the domain. However, it is possible to
estimate how much buoyant matter would accumulate by counting how much passes by









any location. The model and the observations reflect the maximum extent of the plume
and its behavior over time.

Analysis Results
The presentation of results first focuses on basin-level delivery of sediment and pollution
to more than 400 river mouths along the MAR. We initially examine sediment and
pollutant delivery given current land cover (2003/04) and subsequently compare these to
results which use other land cover scenarios ("natural" land cover and three scenarios of
land cover in 2025). Next, we examine sediment transport in coastal waters along the
reef. Finally, we present an analysis of the vulnerability of the land to erosion and an
evaluation of the local origin of sediment and pollution.

1. Sediment and Pollutant Delivery given Current Land Cover
(2003/04)

N-SPECT was used to evaluate accumulation of sediment, nitrogen (N), phosphorous (P),
and total suspended solids (TSS) in more than 400 watersheds across the MAR region.
The maps in Figure 6 reflect the relative accumulation of sediment and N at river mouths
across the region. The Ulua watershed in Honduras was found to be the largest
contributor of sediment, N, P and TSS. Other rivers identified as large contributors of
sediment and nutrients are the Patuca (in Honduras), Motagua (in Guatemala and
Honduras), Aguan (in Honduras), Dulce (in Guatemala), Belize River (in Belize), and
Tinto o Negro (in Honduras). (Watershed names are provided on the map in Figure A in
the Key Findings.)


Figure 6. Sediment and Nitro
I: :











2. Comparison of Results for Current Land Cover (2003/04) to
Hypothetical Natural Land Cover

To evaluate the impact of human alteration of the landscape on sediment and pollutant
delivery to river mouths along the MAR, the N-SPECT model was run on both current
(2003/04) land cover and on hypothetical natural (unaltered) land cover. Table 4 provides
a comparison of the land cover distribution for these two time periods. River discharge,
sediment delivery, and pollutant delivery (nitrogen, phosphorous, and total suspended
solids) to over 400 river mouths is summarized in Table 5. Human alteration of the
landscape has nearly doubled runoff and associated discharge at the river mouths. The N-
SPECT model suggests that sediment delivery has increased by a factor of 20, while
nitrogen has increased by a factor of 3, phosphorous by a factor of 7, and TSS by a factor
of 5.


Table 4. Comparison of Natural and Current Land Cover (percent in each cover type)
Land Category Natural Current
Developed / Urban 0.0% 0.3%
Cultivated 0.0% 32.4%
Savanna / Grassland 1.8% 1.6%
Forest 82.4% 50.6%
Scrub / Shrub 8.3% 10.4%
Wetland / Mangrove 6.0% 3.2%
Bare / beach / unknown 0.4% 0.3%
Water 1.1% 1.2%
Source: ICRAN MAR, 2006


Table 5. Comparison of Regional Results for Annual Model Runs for Current and Natural
Land Cover
Discharge Sediment Nitrogen Phosphorous TSS
Scenario (x 109 m3) (x 109 mt) (x 10 mt) (x 103 mt) (x 103 mt)
Current (2003/2004) 60 370 130 17 2,400
Natural (no
development) 34 17 45 2 470
Ratio of Current /
Natural 2 X 22 X 3 X 7 X 5 X
Note: Sediment, N, P and TSS delivery represent upper bound estimates for the region, as loss of sediment
or nutrients due to redeposition or other processes is not accounted for. The values reflect overall erosion
and pollutant rutt, within the region, and are indicative of the overall magnitude ofsediment and nutrient
delivery, but should not be regarded as absolute values. The relative relationship between the scenarios is
valid.
Source: WRI, 2006


Figure 7 presents both land cover and sediment delivery results for current land cover and
hypothetical natural land cover.












Figure 7. Modeled Sediment Delivery from "Current" and Hypothetical "Natural" Land
Cover


Modeled Sediment
Delivery for 'Current'


"Current' land
cover (200314)
I B a s in s
CURPENIT Lard Cov 2003/4)

( -) Grasiana
_.. J.eCliD L hrub
F-inted WVdand
-- Ba'r Lanl
' rNO Dala
I Coral Rfm


S Coral Rees,
Sedmenl oelivery b basin (mt)
0 10.00
10 w00 500 D0
I 500 000 2 w 000
/ i o ooo o soDO oo o
M 3 000 000 50 D0 000
3 m 1000O00 DD
9p -~awo


Hypothetical Natural"
land cover



I- Basins
"NaturafrLand Cover
Gril siand
ForaIt Evwergreen
Scrub I shrluh
Foienea VJelland
Bdre Lard
*. J m 1Wa er
'I No Data
Coal R-of

' v,' r


F Modeled Sedirment
5I Delivery from
hypolhelical' Natural
land cover



Q Coral Reefs
Sediment delivery by basin (mt)
0 10 000
10 000 500 Oi
m 300000 50, O000
3000000 50,000.000


Source: WRI, 2006.


3. Comparison of Current Land Cover to Three Development

Scenarios in 2025


Scenarios of land cover change in the MAR region though 2025 are used to evaluate the
impact of land cover change on river discharge, sediment, and pollutant delivery. Table 6
and Figure 8 reflect the distribution of land cover in these scenarios. The N-SPECT
model was run on these three land cover scenarios and results are summarized in Table 7.
The N-SPECT model suggests that the Markets First scenario would result in a 13%
increase in sediment delivery relative to that of current land cover, while sediment would
only increase by 5% under the Policy First Scenario and would decrease by 5% under the
Sustainability First Scenario. Nutrient and TSS delivery would also increase significantly


,-









from current land use under the Market First Scenario (8 11% increases in nitrogen,
phosphorous, and TSS delivery are projected). At the other extreme is the Sustainability
First Scenario which could result in a 4 5% percent decline in nutrients and TSS, based
solely on changes in land cover. Additional reductions in sediment and nutrient delivery
can be achieved through the implementation of better agricultural management practices,
which were not considered in these scenarios.


Table 6. Comparison of Land Cover Scenarios (percent in each cover type)
Markets Policy Sustainability
Land Category Current First First First
Developed / Urban 0.3% 0.6% 0.6% 0.5%
Cultivated (Ag) 32.4% 36.7% 34.2% 31.3%
Savanna / Grassland 1.6% 1.5% 1.5% 1.7%
Forest 50.6% 45.8% 48.9% 49.8%
Scrub / Shrub 10.4% 10.6% 9.9% 11.9%
Wetland / Mangrove 3.2% 3.1% 3.2% 3.1%
Bare / beach / unknown 0.3% 0.3% 0.3% 0.3%
Water 1.2% 1.4% 1.4% 1.4%
Source: ICRAN MAR, 2006


Figure 8. Land Cover Distribution Within Each Scenario


Natural Current Mrkt. 1st Policy
1 st

Source: ICRAN MAR, 2006


" Water
" Bare
" Wetland
" Scrub
0 Forest
O Grassland
O Cultivated
* Urban


Sust.
1 st


100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%











Table 7. Comparison of Regional Results for Annual Model Runs for Current Land Cover
and Three Scenarios in 2025
Discharge Sediment Nitro en Phosphorous TSS
Scenario (x 109 m ) (x 109 mt) ( ) (x 103 mt) (x 103 mt)
Current (2003/2004) 60 370 130 17 2,400


Change from
Current 5% 13% 8% 11% 10%
Policy First 61 390 135 18 2,480
Change from
Current 2% 5% 3% 5% 4%
Sustainability First 59 355 125 16 2,300
Change from
Current -2% -5% -4% -5% -4%
Note: Sediment, N, P, and TSS delivery represent upper bound estimates for the region, as loss ofsediment
or nutrients due to redeposition or other processes is not accounted for. The values reflect overall erosion
and pollutant rumorfwithin the region, and are indicative of the overall magnitude ofsediment and nutrient
delivery, but should not be regarded as absolute values. The relative relationship between the scenarios is
valid.
Source: WRI, 2006


4. Extreme Events

Using the N-SPECT model, WRI has modeled runoff, erosion, and sediment delivery in
the MAR region during extreme rainfall events, such as hurricanes. An analysis of the
Hurricane Keith storm event in Belize was implemented for an area that covered 90
watersheds in Belize.27 During the five-day storm event (Sept. 29 Oct. 3, 2000),
discharge from these 90 watersheds was estimated at approximately 30% of the normal
(modeled) annual total discharge, while sediment delivery was approximately 50% of the
normal annual. Impact of rainfall during Hurricane Mitch (Oct. 27-30, 1998) was
evaluated for over 100 watersheds in Honduras.28 Discharge was estimated at
approximately one-third of the normal (modeled) annual total and estimated sediment
delivery was over two-thirds of that predicted for an average year. This analysis
highlights the significant impact large storm events have on sediment delivery in the
coastal zone. Figure 9 shows the extent of the CDM plume on November 10th, 1998,
approximately two weeks after Hurricane Mitch.










27 Precipitation data for Hurricane Keith was provided by the Belize Meteorological Department.
28 Precipitation data for Hurricane Mitch comes from the US Geological Survey (USGS).









Figure 9. CDM Plume Extent on November 10, 1998


Source: Colored detrital matter (CDM) plume extent interpreted from SeaWiFS
imagery by Christopher Kuchinke, University of Miami. Note: the Spectral
Optimization Algorithm (SOA) replaces the entire 'standard' SeaWiFS
atmospheric correction and bio-optical algorithm.


5. Buoyant Matter Transport along the MAR


WRI applied the N-SPECT model on a monthly basis to develop estimates of average
monthly river discharge and sediment delivery for both the "current" land cover and
"sustainability first" land cover scenarios. These were used as input to the ROMS ocean
circulation model run at the University of Miami to predict buoyant matter transport
along the MAR. The ROMS model results reflect seasonal variation of buoyant matter
plume extent and concentration. Full year animations of buoyant matter circulation for
both scenarios are available on the Watershed Analysis for the Mesoamerican Reef data
CD. A more detailed description of the ROMS circulation model application and results,
as well as observations of circulation patterns from SeaWiFS is included in "Dynamics of
Buoyant Matter in the MAR Region" on the same data CD29. Figure 10 reflects the
seasonality of river discharge estimated by the N-SPECT model.









29 Cherubin, L.M., C. Kuchinke, C.B. Paris, and J.Kool, 2006. Dynamics of Buoyant Matter in the Meso-
American Region reefs from SeaWiFS data and from a high resolution numerical simulation. University of
Miami. Final report to World Resources Institute.


Log-scale CDM (m )
I High 07


Mid. 0.06


I Low 0001


,sp










Figure 10. Estimated Monthly River Discharge from N-SPECT


Average Monthly River Discharge from all MAR Basins

10,000
9,000
8,000
7,000
E 6,000
5,000
4,000
3,000
2,000
1,000






Source: WRI, 2006


Results from the two scenarios can be compared to explore the potential impact of
changes in land use on the modeled buoyant matter delivery to the MAR. The plume
extents and concentration from the two scenarios can be compared on a month-by-month
basis, or annual summary statistics can be developed to reflect the mean annual buoyant
matter concentration at each reef location, the maximum annual concentration, or the
number of months where the concentration exceeds some threshold.

Figure 11 reflects the simulated buoyant matter plume for December for both current land
cover and the Sustainability First scenario. The ROMS model predicts a less extensive
and less concentrated buoyant matter plume off of the coast of Honduras during
December, but predicts increased concentrations in two semi-enclosed bays of the
Yucatan Baha de la Ascension and Bahia del Espiritu Santo. Much of this change is due
to decreased sediment delivery in the Sustainability First scenario. The increase in
buoyant matter along the Yucatan, however, results from a change in the circulation
predicted by ROMS along the MAR. Much like variability in weather, there is variability
in ocean circulation. During winter in the Sustainability First scenario, the usual
anticlockwise rotating gyre north of Honduras is replaced by the Caribbean current. The
current flows straight from the eastern tip of Honduras to the coast of the Yucatan
carrying sediments from Honduras.













Figure 11. Simulated Buoyant Matter Concentration Along the MAR for December (Current
Land Cover and Sustainability First Scenarios)


\ Buoyant Matter Plume Buoyant Matter Plume
Simulation for Dcember / Simulation for December
Current Land Cover Sustalnabillty First Land
/ Scenarlo V Cover Scenario

Land
Buoyant Matter (gIm3)
0
1-2
p 2.,
S-- 26 75
201 400
,.?1 9 ,.f401 800
d 801 max
No Data








Figure 12 reflects the maximum buoyant matter plume extent and concentration for the
year for the same two land cover scenarios. ROMS simulations suggest that under the
Sustainability First scenario the buoyant matter plume extent and concentration will be
reduced along the coast of Honduras, and less buoyant matter will reach the barrier reef
in southern Belize. The simulations again predict increased concentrations in the semi-
enclosed bays of the Yucatan. These reductions in buoyant matter plume extent could
result from changes in sediment delivery resulting from land cover change (reductions in
cultivated land) as well as due to variability in the ROMS simulations.



Figure 12. Maximum Annual Simulated Buoyant Matter Concentration Along the MAR
(Current Land Cover and Sustainability First Scenarios)


\Buoyant Matter Plume -
Maximum Extent for Year
Current Land Cover
/ Scenario


Buoyant Matter Plume -
SMaximum Extent for Year
/ stalnabllity First Land
/( Cover Scenario








0


R1^--


MAR
Land
3u'vdanil Madnil r 'n3
1-2
3-8
3 25
76 200
201- 400
S401 800
801 max
No Data










Figure 13 depicts coral reefs of the MAR, mapped whether or not the buoyant matter
concentration exceeds 3 grams per cubic meter (g / m3) during any month of the year.
Under the ROMS simulation for current land cover, 12% of the coral reefs of the MAR
exceed this threshold, while under the Sustainability First scenario, this proportion drops
to 10%. This improvement points to the significant effect that improved land
management can have on sediment and pollutant delivery to the MAR.


Figure 13. Mesoamerican Reef Mapped by Buoyant Matter Concentration (Current Land
Cover and Sustainability First Scenarios)


/ Currnt Land Cover
4Pe n aflo



I9





16'rer


Startlnablilty First
Land GCovr Senalio









Joi _'W


6. Vulnerability of the Land to Erosion

Development of land management priorities requires detailed local-level information on
sources of sediment, and the vulnerability of areas to erosion. WRI has developed a
landscape-wide indicator of the vulnerability of land to soil erosion. It incorporates the
slope of the land, soil erodibility, and annual precipitation into a 1 km resolution indicator
of the relative vulnerability of the land to erosion. This indicator does not consider the
current land cover or land use. Rather, it provides an overall indicator of erosion-prone
areas, and therefore, a guide to areas where restrictions on development, or the
implementation of best agricultural management practices should be encouraged.
Vulnerability is high in many areas in Guatemala and Honduras as well as some
mountainous areas in Belize. (See Figure 14.)


Land
Maximum Annual
Buoyant Matter
Concentration
* 0-3g m3
* >3glm3











Figure 14. Vulnerability of Land to Erosion


SBasins
M Lakes
Rivers
Vulnerability of land to erosion
M Very Low
Low
I Medium
High
I Very High


e,
















7. Local Sources of Sediment and Nutrients


N-SPECT was run in "local effects" mode to
evaluate the amount of eroded sediment and nutrients
originating from within each 250m resolution grid
cell, independent of contributions from adjacent or
upstream grid cells. The evaluation was done using
current land cover (2003/04). These results were
summarized by sub-basin, as an aid to priority-setting
and the targeting of better management practices.
The following maps reflect the average erosion,
Nitrogen runoff, and Phosphorous runoff per grid cell
within each sub-basin. Overall watershed boundaries
are also shown both for reference and to allow a
linking of sources within the sub-basins with the
previous maps of sediment and nutrient delivery by
basin.





Figure 15. Average Contribution of Sediment,
Nitrogen, and Phosphorous by Sub-basin


Average Annual Soil Erosion
by Sub-basin
',




S- --


I,..5,.''


Average Annual Nitrogen
Source by Sub-basin




if


4U ) ~ 1 W09 5Om C
(OO BP
(4 o 00 o
I 4 0000
( *jpr~ tlo *BO-IX
.T s -rt'r io-eoE


*A1
Average Annual Phosphorous
Source by Sub-basin



/


I- I


;r
,.. ~
~tF
~: ~~ ~e.~~-
i~,""' 7


C ~.


U.,
~ .
.. Im
na
~ I~
s
I ' r


~""tl~,`z`
\SiZISl~iiiii~
1~1~ I-


*-










Local sources of sediment, N, and P were also summarized by country. Most of the
sediment and nutrients delivered by watersheds along the MAR originate in Honduras.
The model suggests that over 80% of sediment originates in Honduras, while 17% of
sediment originates in Guatemala. Relatively minor percentages come from Belize and
Mexico. Honduras is also the largest source of nutrients (55% of N and 60% of P), while
Guatemala contributes about one-quarter of all N and P in these watersheds. Belize
contributes about 12-13% of both N and P, while Mexico is estimated to contribute about
5% of the total of these nutrients from all modeled watersheds.


Table 8. Percentage of Erosion, Nitrogen, and Phosphorous Sources by Country within
MAR Drainage
Erosion Nitrogen Phosphorous
(Percent of (Percent of (Percent of
Country total) total) Total)
Honduras 83% 55% 60%
Guatemala 17% 26% 25%
Belize 1% 13% 12%
Mexico 0% 6% 4%
Source: WRI, 2006










Conclusion


The analysis presented in this paper provides a regional overview of the magnitude and
pattern of sediment and nutrient delivery to coastal waters of the MAR. Based on this
analysis, we conclude that:

Policy action is needed to address the contributions from agricultural lands.
Most of the sediment and nutrients delivered to the MAR from watersheds in the region
come from agricultural lands in Honduras and Guatemala. The contributions of Belize
and Mexico are substantially less, but still pose a threat along their coasts. Many
promising initiatives to decrease pollution within the region are underway. These include
sustainable forestry management and integrated watershed management in Guatemala;
improved land use planning, reforestation and soil conservation programs in Honduras;
and similar initiatives in Belize and Mexico. These important efforts need public support,
recognition, and continued investment.

Results can help identify areas in need of better agricultural management.
This analysis identified vulnerable areas where conversion to an erosive land use should
be avoided, or where converted conservation practices should be implemented. It also
identified areas with high erosion and nutrient runoff, where better agricultural
management practices should be targeted.

Policies that support sustainable development can reduce sediment and nutrient
delivery.
As evidenced by our findings, land-use planning, integrated watershed management, and
other policies that support sustainable development can help to lessen erosion and
pollution runoff, thereby decreasing sediment and nutrients reaching the MAR.

More detailed modeling is needed to create more accurate information at higher
resolutions.
Regional-scale analyses are useful for providing an overview and for prioritizing areas in
which action is needed. However, local analyses provide more detailed and accurate
information that policymakers need in order to target their interventions. The tools
provided on the data CD, Watershed Analysis for the Mesoamerican Reef(WRI/ICRAN
MAR project, 2006), allow users to perform more detailed analyses of sediment and
nutrient delivery within smaller areas in the MAR region, such as at the watershed level.

More specifically, the N-SPECT model can be applied to individual watersheds or groups
of watersheds using the 90m elevation data provided or with the user's own data. More
detailed local modeling will improve the accuracy of the results, by using higher
resolution data on slopes and land cover, and by calibrating the model to local soils and
precipitation regimes.

It would be valuable to extend the current analysis to include the effect of improved
agricultural management practices on erosion and pollutant runoff Such an extension









would require detailed information on how each practice influences erosion rates and
pollutant runoff coefficients. Once such information is available, it should be possible to
use N-SPECT to evaluate reductions by treating each management intervention on each
land cover type as a unique category with specific erosion and pollutant runoff
characteristics. For example, citrus groves with cover crops planted to reduce erosion
might be treated as a separate category.

Enhancements to the model are needed to improve the accuracy of sediment and
nutrient delivery estimates.
We recommend that the N-SPECT model developers make investments to enhance the
model to account for sediment redeposition and nutrient loss during transport within
watersheds. This would result in a modeling tool that is more capable of estimating actual
sediment and nutrient delivery at river mouths. The accuracy of these estimates could
then be evaluated with field measurements.

Analyses such as these can help to evaluate progress in reducing land-based sources
of threat.
A number of national initiatives, as well as donor-funded regional initiatives, seek to
reduce or mitigate threats to the MAR. This analysis can help these initiatives to estimate
their progress by giving them the information they need to ensure they are moving in the
right direction.

Transnational natural resource management can be strongly supported by analyses
such as these.
To mitigate and reduce the land-based threats to the MAR, constructive regional
cooperation among a variety of stakeholders is necessary. Examples include the
multilateral cooperation agreements among the four countries involved in this analysis,
and agreements between the agriculture and tourism sectors and civil society groups. This
tool works across borders and sectors, creating information that allows productive
discussion on threat origins and potential mitigation measures.



The International Coral Reef Action Network (ICRAN) collaboration will continue to
support the application of analysis results and modeling tools in the region. For more
information on ICRAN, please visit: www.icran.org.


For questions or comments about this analysis please contact:
Lauretta Burke (lauretta)wri.org) and Zachary Sugg (zsu2gg(wri.org)
World Resources Institute
10 G St. NE,
Washington, DC 20002
+1 (202) 729-7600
On the web at: reefsatrisk.wri.org










Additional Technical Notes


1) Watershed Delineation We delineated watersheds at 250m resolution at WRI
using both N-SPECT and ArcMap. We needed to run the delineation in N-SPECT
so that it would accept the DEM, and be able to detect river locations (where flow
accumulates) and watershed boundaries. N-SPECT, however, lumps coastal
watersheds, which is not useful for examining specific river discharge or sediment
delivery. To get around this, we extracted the flow direction output from N-
SPECT and used this as the basis for deriving more detailed basins in ArcMap
(using the BASINS command in ArcTools\Spatial Analyst). We also ran Flow
Accumulation in ArcMap to identify rivers and streams. We combined flow
accumulation and basins to identify the point of high flow in each basin, and
assigned this point as the "pour point" or river mouth.

2) Slope Length in N-SPECT. We did two watershed delineations in N-SPECT.
One was based on the "burned," or hydrologically corrected DEM and resulted in
a good watershed delineation. The other was based on a "raw" or unburned DEM
and resulted in a poor watershed delineation, but more accurate slope and "slope
length" calculations. (A burned DEM has an artificially steep slope along the
burned rivers.) As such, we ran both watershed delineations, and copied the slope-
length grid (called LSgrid in N-SPECT) from the "raw" delineation to the
"burned" delineation. We then ran all future processes on the "burned" accurate
delineation, which now includes a more accurate slope-length grid.

3) R-factor the equation used for R-factor :

R = 3786.6 + 1.5679 (Precip in mm) 1.9809 (Elevation in m)

R is in metric units (MJ mm ha-1 h-1 y-1) or (megajoule mm per hectare
per hour per year.) However, N-SPECT requires US units: (hundreds of feet tonf *
inch acre-1 hour-1 year-1). One can convert from metric to US units by dividing
by 17.02.

Reference: USDA-ARS Agriculture Handbook No. 703




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs