Relating Sediment Yield Estimations to the Wet Front Term Using Rainfall Simulator Field Experiments

Abstract

Depth to wet front is generally considered as the amount of water that penetrates into soil and wets the internal soil layer. This is an important variable especially in applications such as runoff generation and sediment yield estimation. This variable in some cases is used in the hydrological science, in the form of a proxy for infiltration as an important factor in soil erosion processes. This work deals with estimating event-based suspended sediment yield in relation to depth to wet front to capture the significance of the depth to wet front in Modified Universal Soil Loss Equation (MUSLE). In this field study, a rainfall simulator equipped with drip systems installed over an experimental hillslope plot was used to generate rainfall with intensities of 45, 60, and 70 mm/h over three slopes 10, 20, and 30% with three repetitions. The coefficient of determination (R2) and the Nash–Sutcliffe efficiency (ECNS) were applied as performance metrics for the model. The results revealed that storm rainfall energy or Erosivity Index (EI30) itself was not suitable for estimating the sediment yield. On the other hand, incorporating both EI30 and rainfall-runoff resulted in higher model efficiency. Upon addition of the depth to wet factor into MUSLE model, a greater efficiency was obtained. Further, the results of event-based simulated and observed sediment indicated that they closely corresponded to a straight line (푅2 = 0.94 and NSE = 0.93). Thus, the depth to wet front was a useful variable in sediment yield simulation. The Nash efficiency coefficient and correlation factor for the estimation total sediment yield were obtained (R2 = 0.93; NSE : 0.92) for both calibration and (R2 = 0.93; NSE : 0.86) validation stages. The result suggests that consideration of the depth to wet front in MUSLE model may lead to improved hydrological and sediment applications.