Climate change impacts on soil erosion and nutrient losses in the Great Lakes region

Wang Lili, Purdue University


Pollutants from non-point sources (NPS) have become the primary reason for water quality degradation in the Great Lakes region after great progress has been made in reducing point source pollution through the enforcement of total maximum daily loads (TMDL). Climate change may impact NPS pollutant transportation processes through influencing runoff generation, soil hydrological conditions, and evapotranspiration with changes to precipitation quantity, intensity, frequency, and to air temperature. As there have been few studies focused on future climate change impacts on the NPS pollutants in the Great Lakes region, this study was conducted to fill a gap in our knowledge by exploring how changes in regional climate may impact sediment, nitrogen, and phosphorus losses by the end of 21st century. A macroscale soil erosion model (a coupling of the Variable Infiltration Capacity model and the Water Erosion Prediction Project model, the VIC-WEPP model) was applied to three states in the Great Lakes region: Michigan, Wisconsin and Minnesota under three climate change scenarios (A2, A1B, and B1). Over three future periods (2030s, 2060s, and 2090s), predicted annual soil loss decreased by 0.4 to 1.8 ton ha-1 with soil loss increases in the northern and central study domain due to precipitation increases and soil loss decreases in the southern study domain as a result of air temperature increases. Seasonally, soil loss was projected to decrease by 0.3 to 1.1 ton ha-1 in summer due to decreasing precipitation and increase in fall and winter by 0.9 to 2.0 ton ha-1 as a result of increasing precipitation. In the second part of this project, a new coupled soil erosion and water quality model, the WEPP-WQ model, was developed to address deficiencies in existing water quality models that rely on the empirical soil erosion model (Universal Soil Loss Equation model, USLE) and its derivatives. The new coupled model was evaluated for several point sites using both simulated single event storms, and continuous multi-year observations. The model performed quite well in simulating nutrient losses for single storm events with that R2 is higher than 0.8, Nash-Sutcliffe efficiency (NSE) is higher than 0.65, and percent bias (PBIAS) is less than 31% for runoff, sediment loss, nitrate nitrogen loss, total nitrogen loss, soluble phosphorus loss, total phosphorus loss. In predicting time series nutrient loss, WEPP-WQ model could simulated nitrate nitrogen decently with the ratio of the root mean square error to the standard deviation of measured data (RSR) higher than 0.65, NSE higher than 0.51, and PBIAS greater than -35%. The calibrated model was then applied to watersheds in the Great Lakes region, and used to quantify changes in nutrient losses under the same future climate scenarios used to evaluate soil losses. Due to the increase of precipitation quantity and intensity and frequency of extreme storm events, total phosphorus loss was projected to increase by 28% to 72% for the Green Lake watershed and 31% to 108% for the Walworth watershed in the future periods. Nitratenitrogen losses were projected to increase by 5% to 38% for the Green Lake watershed and 8% to 95% for the Walworth watershed as a combined result of increase in precipitation quantity and intensity and frequency of extreme storm events, with the major influencing factors being different in each future period.




Cherkauer, Purdue University.

Subject Area

Climate Change|Soil sciences|Agricultural engineering|Environmental engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server