A Dynamic Programming Approach to a Class of Nonpoint Source Pollution Control Problems

This paper presents a new approach to modeling, analyzing, and solving a class of environmental control problems dealing with sediment deposition. An efficient dynamic programming algorithm is designed to handle the spatial characteristics of soil movement through a watershed, and its ultimate impact on water channels and/or reservoirs. The model generates "sediment abatement cost frontiers" which summarize the trade-off information needed for watershed planning and management. This information can also be used to identify and target special-problem areas. The paper presents both results on the efficiency of the DP algorithm compared to other methods, and results on the application of the model to real world cases.