Development of a coupled soil erosion and large-scale hydrology modeling system

Authors: MAO, DAZHI - Purdue University; CHERKAUER, KEITH - Purdue University; Flanagan, Dennis

Submitted to: Water Resources Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/18/2010
Publication Date: 8/31/2010
Citation: Mao, D., Cherkauer, K.A., Flanagan, D.C. 2010. Development of a coupled soil erosion and large-scale hydrology modeling system. Water Resources Research. 46:W08543.

Interpretive Summary: Soil erosion is a major concern in the U.S. and the world. Computer simulation models are often created to represent mathematically the physical processes that occur in nature and cause soil erosion, and to estimate and effectively manage the land to minimize erosion problems. In this study, a large scale hydrology model called the Variable Infiltration Capacity (VIC) model was used to estimate processes such as water infiltration into the soil, surface runoff water moving off the soil, and especially winter processes such as freezing and thawing soil, and snow melting. VIC was linked with a hillslope soil erosion model using components from WEPP (the Water Erosion Prediction Project model) to estimate soil detachment, sediment transport, sediment deposition and delivery. The coupled models were tested by applying them to two small watersheds in Minnesota, and did a good job of estimating observed snow depths and changes in observed soil temperatures. The frozen soil calculations in the model predicted more erosion in the spring months, during thawing and snow melt occurrence, with greater effects at more northern locations. This research impacts other scientists working with models at large river basin scales and in areas impacted by cold weather processes, providing a tool for soil erosion estimation.

Technical Abstract: The impact of frozen soil on soil erosion is becoming increasingly important for sustainable management of soil resources, especially in regions where agricultural land use is dominant. A newly developed coupled modeling system that integrates the Variable Infiltration Capacity (VIC) model and the physically-based Water Erosion Prediction Project (WEPP) model, is capable of predicting soil loss at large-scales. To evaluate the impact of seasonal soil frost on the modeling system, simulations were conducted in two sub-watersheds in the Minnesota River basin. The frozen soil algorithm of the coupled model was first evaluated in this study by comparing coupled model results with observational data. Multiple regression analysis was then used to identify hydrologic variables most important to the prediction of soil erosion. The coupled model system was also used to evaluate soil loss with and without the representation of soil frost to quantify the effects of frozen soil on soil erosion. Precipitation and runoff, not unexpectedly, were the two most significant prediction variables for soil erosion, but soil rill erodibility and critical shear stress adjustments were also found to be important. Soil erodibility adjustments were higher when a frozen soil is thawing than when the soil frost algorithm was not used, which contributed to higher soil loss rates. Watershed simulations showed that the frozen soil algorithm produced more soil erosion due to improved runoff estimation than simulations without the frozen soil algorithms in the spring months, suggesting frozen soil algorithms needs to be included in the coupled model simulation. On an average annual basis, frozen soil related soil losses accounted for 9.1 % and 4.0% of the total annual soil loss in the two evaluation watersheds. Frozen soil was shown to have a greater impact on soil loss in northern regions that are prone to soil frost than southern regions.