Author
 |
Hagen, Lawrence |
|
Submitted to: International Conference on Aeolian Research
Publication Type: Proceedings Publication Acceptance Date: 7/19/2002 Publication Date: 7/19/2002 Citation: In: J.A. Lee and T.M. Zobeck (eds.) ICAR5/GCTE-SEN Joint Conference; 2002; Center for Arid and Semiarid Lands Studies, Texas Tech University, Lubbock, TX. Interpretive Summary: Wind erosion models are widely used to design control practices and to estimate both on-site and off-site erosion impacts. But most have not had extensive validation. In this study, we compared measures soil loss with simulated soil loss for individual storms using the erosion submodel of the Wind Erosion Prediction System (WEPS). The measured data included 46 storms at seven locations in six different states. The WEPS model generally predicted less erosion than was measured for rare, large storms with more than 2 kg/m^2 (9 tons per acre) loss, but provided good agreement with measured values of other storms. Measured soil surface conditions were not available near some storm dates, so average conditions for the soil were used as model inputs. However, on dates with large storms the soil conditions were likely more erodible than the average condition input to the model. Overall WEPS predictions were in good agreement with measured soil losses, and thus, it should be a useful tool for those who design conservation practices and estimate erosion impacts. Technical Abstract: As part of the Global Change and Terrestrial Ecosystems Soil Erosion Network (GTE-SEN) wind erosion model validation project, we compared observed soil loss with simulated soil loss predictions for 46 individual storms using the erosion submodel of the Wind Erosion Prediction System (WEPS). The experimental sites were 2.5-ha., tilled circular areas located in larger fields that did not erode. Thirteen vertical clusters of sediment samplers were used measure horizontal soil discharge between the surface and one m height, along with a meteorological tower, associated weather transducers, and date logger. The average storm loss from the cluster measurements extrapolated to 180 m downwind was 0.82 kg/m^2, while the average predicted soil loss was 0.64 kg/m^2. The maximum differences between observed and predicted loss occurred during large erosion events where the predicted values were frequently less than those observed. Scatter in the cluster data along the wind direction suggested the initial field surfaces were not always uniform as assumed in the model. Also, some of the input field surface conditions were not measured close to the storm dates, so average values for the soil texture were input to the model. Linear regression of the storm data showed reasonable agreement between predicted and observed (R^2 = 0.71) with an intercept greater than zero. However, nonlinear regression using with a power equation showed that for storm losses less than 2 kg/m^2 the predictions were close to the 1:1 line, and the intercept was slightly less than zero. |
