HYDROLOGIC MODELING OF LAND PROCESSES IN PUERTO-RICO USING REMOTELY-SENSED DATA

An integrated, multi-disciplinary effort to model land processes affecting Mayaguez Bay in western Puerto Rico is described. A modeling strategy was developed to take advantage of remotely sensed data. The spatial, interannual, and seasonal variability of sediment discharges to the bay were also evaluated. Classified images of remotely sensed data revealed the spatial distribution and quantities of land use classes in the region and aided in the discretization of the watershed into homogeneous regions. These regions were modeled using a geomorphic modeling technique based upon spatially averaged parameters. Simulation results from the modeling effort compared favorably with observations at two locations within the watershed. Results showed that runoff and sediment loads from the area exhibit a marked seasonal trend and that deforested areas located in the foothill regions of the watershed contribute a disproportionate share of the sediment load to the bay. In years when rainfall distributions are uniformly distributed over the area, the sediment yields may be up to 100 percent higher than years when the rainfall is concentrated in the heavily forested mountainous regions.