Simulating accumulation and melt of snow for RUSLE2 databases

Authors: McCool, Donald; QIU, HANXUE - WASHINGTON STATE UNIV; ANDERSON, TODD - CORNELL UNIVERSITY

Submitted to: Proceedings of the American Society of Agricultural and Biological Engineers International (ASABE)
Publication Type: Proceedings
Publication Acceptance Date: 2/6/2007
Publication Date: N/A
Citation: N/A

Interpretive Summary: Lack of a means to include snowmelt runoff and erosion in the Revised Universal Soil Loss Equation Version 2 (RUSLE2) hampers its use in areas where there is a period of melting of snow accumulated during the winter. Performance of other models that use RUSLE2 as the driver for hill slope erosion is also affected. Because RUSLE2 relies on input data sets to drive the erosion process, we sought models that could effectively estimate daily snow accumulation and melt and provide data from which snowmelt erosivity databases could be developed using long-term daily disaggregated weather data from the RUSLE2 climate databases. The Soil-Plant-Air-Water (SPAW) model was selected and tested for ability to simulate accumulation and melt of snow. Tests were based on daily weather data collected over a forty-one-year period from eight selected weather stations in cold or high elevation cropland areas of the western U S, the Northern Great Plains, and the northeastern U S. Depth of new snow predicted with SPAW default parameters was generally less than observed and the snow pack was predicted to melt before observed records indicate. Adjusting the accumulation and melt temperatures allowed for a better match to the observed data. The RUSLE2 daily disaggregated weather data were then used in SPAW to develop daily snow water equivalent accumulation and melt values for the eight selected locations. When RUSLE2 average daily climate data was used in SPAW, the model was less sensitive to the input accumulation and melting temperature for stations with low winter temperature and heavy snow fall. However, the model was sensitive to these input parameters for stations where winter temperature is high and snow fall is light. For most cold or heavy snow areas, reasonable estimates of the timing of the accumulation and melt of snow can be obtained using the SPAW model with the daily temperature and precipitation values in RUSLE2 databases.

Technical Abstract: Lack of a method to properly account for snowmelt runoff and erosion in the Revised Universal Soil Loss Equation Version 2 (RUSLE2) hampers its use in areas where there is a period of melting of snow accumulated during the winter. This also adversely affects the performance of other models that use RUSLE2 as the driver for hill slope erosion. Because RUSLE2 relies on input data sets to drive the erosion process, we sought models that could effectively estimate snow accumulation and melt and provide data from which snowmelt erosivity databases could be developed using the disaggregated daily weather data from the RUSLE2 climate databases. The Soil-Plant-Air-Water (SPAW) model was tested for ability to simulate accumulation and melt of snow. Model tests were based on daily climate data collected over a forty-one-year period from eight selected weather stations in cold or high elevation cropland areas of the western U S, the Northern Great Plains, and the northeastern U S. Depths of new snow predicted with SPAW default parameters were generally less than observed and the snow pack was then predicted to melt before observed records indicate. Adjusting the accumulation and melt temperatures allowed for a better match to the observed data. The model performance also indicated that considering ET and air temperature would achieve better prediction. When RUSLE2 average climate data was used in SPAW, the model was insensitive to the input accumulation and melting temperature for stations with low winter temperature and heavy snow. However, the model was sensitive to these input temperatures for stations where winter temperature is high and snow fall is light. Here, snow accumulation is less likely to be a winter long event; rather, snow accumulates and melts frequently during the winter, creating challenges for simple modeling approaches. For heavy snow or cold areas, reasonable estimates of the average timing of the accumulation and melt of snow can be obtained using the SPAW model and the disaggregated daily temperature and precipitation values in RUSLE2 databases.