干旱区生态环境知识资源中心

Farmland especially dry farmland managed in traditional ways has high wind erosion risk and contributes mainly to dust emission in arid area. Modeling predicting provides a general view to soil erosion susceptibility, and is very helpful for the understanding of potential spatial source of wind erosion. This study applied the Single-event Wind Erosion Evaluation Program (SWEEP) to predict soil wind erosion of farmland in the study area. SWEEP is a standalone version of the erosion sub-model from the Wind Erosion Prediction System (WEPS). It needs fewer calculation parameters than WEPS and is often used for single erosion events of limited size. The objective of this study was to test the feasibility of using SWEEP to estimate annual wind erosion of farmland over large areas (downwind distance >1600 m) with limited wind data (2005-2011) from weather stations. We validated the simulation results by comparing them with field measurements and wind tunnel data for the same soils. The soil material eroded by wind included PM10, suspension, saltation, and creep particles. Suspension particles were the main component involved in the soil loss (averaged 61.8%). However, saltation and creep particles dominate the particle clouds for fields with a downwind length of less than 550 m (averaged 56.6%), and the mass flux dominated by suspension particles stabilizes when this length is longer than 1000 m (averaged 83.4%). PM 10 always has very low proportion (<2.3%). Our validation results suggest that it is feasible to use SWEEP for large areas with limited wind data. However, SWEEP could not simulate the small soil losses that occur especially at low wind velocities well. Many factors contribute to this problem, but the main one is overestimation of the threshold wind velocity. Previous research suggest that it will be difficult to replace SWEEP’s calculation algorithms for the threshold wind velocity, but both these algorithms and some SWEEP parameters must be improved to provide accurate predictions of soil erosion. (C) 2016 Elsevier B.V. All rights reserved.