MODELING SOIL EROSION WITHIN THE MILL CREEK WATERSHED, YOUNGSTOWN, OHIO

Mill Creek Park, established in 1896, is one of the largest metro-parks in the US, providing a variety of outdoor activities to the city of Youngstown, Ohio, and its surrounding areas. Popular attractions within the park include a number of man-made ponds and lakes, which have rapidly silted in since their construction, forcing several sediment-excavation projects. An investigation is underway to elucidate information from lake deposits to help develop an erosion model for contributing watersheds.

A preliminary erosion model based on the Universal Soil Loss Equation was established for a small subset of the Mill Creek watershed. This model incorporates existing datasets on soil characteristics, topography, and climate in an effort to constrain a parameter for land cover; information on sediment yield was used to solve for this unconstrained factor. Lily Pond is a 3-acre catch basin that receives sediment from surrounding forested hillslopes, typical terrain and land-cover types within the park. A 1974 excavation of the pond provided a baseline for assessing subsequent sediment sequestration, quantified from detailed bathymetric surveys and cores. Sediment-yield calculations were compared to the watershed-erosion model and provided a refined land-cover factor for future modeling within the park.

Current research is now underway to provide additional model-parameter constraints utilizing similar methodologies. Differences in sediment yield between watersheds of similar size, topography, and soil characteristics provide clues regarding the influence of land-cover types, which vary tremendously across the region. The Indian Run watershed, for example, is characterized by 21% urban land cover and 23% agricultural land cover whereas the Bears Den watershed is characterized by 66% and 0%, respectively. Improving individual parameters using comparison studies that incorporate empirical data on sediment yield are critical to the development of an erosion model that is regionally applicable.