Comparing surface erosion processes in four soils from the Loess Plateau under extreme rainfall events

This research aims to improve erosion control practice in the Loess Plateau,by studying the surface erosion processes,including splash,sheet/interrill and rill erosion in four contrasting soils under high rainfall intensity(120 mm h^(−1))with three-scale indoor artificial experiments.Four contrasting soils as sandy loam,sandy clay loam,clay loam and loamy clay were collected from different parts of the Loess Plateau.The results showed that sediment load was significantly impacted by soil properties in all three sub-processes.Splash rate(4.0-21.6 g m^(−2∙)min^(−1))was highest in sandy loam from the north part of the Loess Plateau and showed a negative power relation with the mean weight diameter of aggregates after 20 min of rainfall duration.The average sediment load by sheet/interrill erosion(6.94-42.86 g m^(−2∙)min^(−1))was highest in clay loam from middle part of the Loess Plateau,and the stable sediment load after 20 min showed a positive power relation with the silt content in soil.The average sediment load increased dramatically by rill and interrill erosion(21.03-432.16 g m^(−2∙)min^(−1)),which was highest in loamy clay from south part of the Loess Plateau.The average sediment load after the occurrence of rill showed a positive power relation with clay content and a negative power relation with soil organic matter content.The impacts of slope gradient on the runoff rate and sediment load also changed with soil properties.The critical factors varied for different processes,which were the aggregate size for splash erosion,the content of silt particles and slope gradient for sheet/interrill erosion,and the content of clay particles,soil organic matter and slope gradient for rill erosion.Based on the results of the experiments,specific erosion control practices were proposed by targeting certain erosion processes in areas with different soil texture and different distribution of slope gradient.The findings from this study should support the improvement of erosion prediction and cropland management in different regions of the Loess Plateau.