Spatial decision support system for assessing agricultural nonpoint source pollution using GIS
Abstract
A spatial decision support system was developed to assist in assessing nonpoint source (NPS) pollution using a NPS pollution model and geographic information system (GIS). The tools developed were evaluated in two watershed located in two different regions and the significance of effects due to the extraction of model inputs from a GIS data base on the NPS pollution model were highlighted. Using the GRASS (Geographical Resources Analysis Support System) GIS and AGNPS (AGricultural Non-Point Source), a distributed parameter model, the spatial decision support system (SDSS) was developed. The SDSS assists with extracting the 22 input parameters for each cell of the AGNPS model from only 7 GIS layers and with minimum user interaction. Thereby, significant amounts of time, labor and expertise can be saved. Further, the SDSS assists with visualizing and analyzing the output of the AGNPS model simulations. Various components of this tool include displays of sediment, nutrient, and runoff movement from a watershed. The SDSS was evaluated for the Animal Science (ANSI) watershed located in Indiana and the Henrietta watershed located in Texas. Topographic attributes such as slope steepness and slope length are important factors in predicting soil loss and chemical movement using hydrologic simulation models. The effects of four slope estimation algorithms (neighborhood, quadratic, best fit plane and maximum slope method) in deriving slope steepnesses, slope lengths and erosion estimates using the SDSS were explored. Significant differences were found among the four slope prediction methods in topographic attributes and erosion estimates at the outlet of the watershed and within the watershed. Aggregation of data, required by the NPS pollution model, has significant effects on the NPS pollution model results. The SDSS was used to study the effects of cell resolution on the NPS pollution model for the two watersheds. Aggregation reduced the slope steepness and increased the USLE slope length for a cell. The methods presented allow the measurement of the spatial variability of several important inputs and their effects on model results.
Degree
Ph.D.
Advisors
Engel, Purdue University.
Subject Area
Agricultural engineering|Environmental science
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.