The consequences of increasing climate variability are difficult to predict given the complex interactions of processes within the biogeochemical and hydrologic cycles. The Soil Water Assessment Tool (SWAT) model was calibrated and validated for the Upper Big Walnut Creek watershed in central Ohio to assess these effects of climate change with a focus on nutrient pollution in the area’s surface water. Due to its proximity to rapidly urbanizing areas near Columbus, the watershed in question contains a representative mixture of land uses, as well as tiled and untiled farmland, indicating results found here will be widely relevant. After SWAT model calibration and validation, future climate scenarios will be run using the IPCC 2007 report as a guideline. Scenarios will focus on increasing inter-storm periods and rainfall intensity and the examining the resulting changes in nitrogen and phosphorus concentrations in surface water bodies. Concurrent with the development of a SWAT model simulation, multivariate regression analysis is being performed on eight sub-watersheds within the Upper Big Walnut Creek in order to determine relationships between watershed physical characteristics (slope, land use, stream length, etc.), temperature, precipitation, and observed nutrient loading in surface waters. Due to the sensitivity of the biogeochemical cycle to highly variable conditions such as soil moisture and temperature, analysis performed at a sub-watershed scale will help better estimate nutrient concentrations in the stream. The results of this small-scale analysis will then be applied at the watershed level.
A possible use of this work is to enhance prediction methods for ungauged watersheds. Satellites are one of the best data sources for such watersheds and offer excellent spatial coverage; however, this comes at the cost of low resolution. To compensate, remotely sensed data will be referenced against other information and the resulting downscaled data will be used force the regression-based model. Furthermore, due to its well-established status, the SWAT model will be used to assess the performance of the empirical relationships developed during the regression analysis. Depending on these results, future directions of work may include modification of SWAT’s runoff and biogeochemical cycling equations.