Evaluating nonpoint pollution using remotely sensed data in soil erosion models

IN defining nonpoint-source pollution problems, researchers and public officials must address the geographic extent and degree of seriousness of those problems. Evaluation of nonpoint-source pollution programs is especially important during these times of austere budgets. Conceivably, the greatest gains in soil conservation and pollution control efforts can be obtained by effectively targeting available resources. Methods of identifying those areas deserve urgent attention in order to develop future conservation and pollution control policies. Since G. W. Musgrave's effort to quantify soil erosion factors in 1947 ( 13 ), researchers have developed several models to predict soil loss or sediment yield, including the USLE (universal soil loss equation) ( 22 ), CREAMS (Chemicals, Runoff, and Erosion from Agricultural Management Systems) ( 8 ), SPUR (Simulation of Production and Utilization of Rangelands) ( 24 ), EPIC (Erosion-Productivity Impact Calculator) ( 20 ), SWRRB (Simulation for Water Resources in Rural Basins) ( 19 ), and ANSWERS (Areal Nonpoint-Source Watershed Environment Response Simulation) ( 1, 3 ). Results from these models can be used directly or indirectly to determine potential water quality problems from nonpoint sources of sediment. The models vary in type and extent of land use for which they are best suited. The USLE has been modified to address a wide range of land uses. CREAMS was adapted specifically for agricultural areas. SPUR was designed primarily for range-land use. Generally, these models use point measurements, such as runoff or land cover …