GIS Based Soil Erosion Prediction Using RUSLE and SPI Models for Southern Ontario Watersheds
This study was completed in cooperation with The Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA) in order to aid with analysis of landscape erosion prediction models. The initial task used the Revised Universal Soil Loss Equation (RUSLE) to investigate parameters such as the impacts of high seasonal variability and water contents combined with the LS-Tool to calculate slope length and steepness (LS) values from a 0.5m digital terrain model (DTM) derived from LiDAR Point Data. Geographic information system (GIS) processes were used to map the risk of soil erosion for the Indian-McGregor Creek sub-watershed in Chatham-Kent, Ontario. Furthermore, the Stream Power Index (SPI) was calculated using a raster slope data file for the watershed in order to study the results of the flow and its erosive power. This was done with the intention of determining the validity of using SPI as a predictive model for landscape erosion risk. The outputs and data of both the RUSLE and SPI analysis were compared against aerial imagery to determine the accuracy of the predictive models when applied to the identification of sites with potentially high gully erosion. It was determined that although SPI consistently mimicked the predictions of the RUSLE model, it did so with far less resolution and specificity. Potential reasons for this related to the topography of the region being studied and are discussed in detail within the report.