Runoff characteristics and soil erosion dynamic processes on four typical engineered landforms of coalfields: An in-situ simulated rainfall experimental study

Large engineered landform tracts formed during coalfield development and construction cause serious soil erosion and threaten the ecological security of the coalfield area. A series of in-situ simulated rainfall experiments with rainfall intensities ranging from 1.0 to 3.0 mm min(-1) were conducted to explore the runoff characteristics and erosion dynamic processes of four engineered landforms, namely, waste slag heap with more gravel (WSG), waste slag with more sand (WSS), waste soil heap (WSH) and disturbed land (DL), with undisturbed land (UL) as the control (CK). Compared with UL, the WSS, WSG and WSH runoff rates (RRs) increased by 38.83%, 71.39% and 63.16%, respectively, but decreased by 9.25% for DL. The slope flow of UL was in the subcritical laminar and subcritical transition flow regimes, while the four engineered landform flow regimes were mainly subcritical transition and subcritical turbulent flow regimes. The UL and DL soil loss rates (SLRs) followed a declining-fluctuating-levelling off trend, while the WSS, WSG and WSH SLRs followed a trend of multiple early-stage peaks and valleys before levelled off. The mean WSS, WSG, WSH, and DL SLRs were 11.19 times, 138.67 times, 73.47 times and 2.82 times that of UL, respectively. UL disturbance significantly increased the runoff ability to detach sediments by 2.48 94.90 times based on the relationships between runoff rate and soil loss rate. The shear stress (tau), stream power (omega), and unit stream power (U-p) critical values initiating DL soil erosion were 58.86%, 37.06% and 56.67% lower than those of UL, respectively, while higher critical hydrodynamic values were required to detach WSG, WSS and WSH particles. The rainfall intensity (RI), slope gradient, median particle size (D-50) fractal dimension (FD) and U-p were employed to develop an empirical model for predicting engineered landform SLRs. (C) 2019 Elsevier B.V. All rights reserved.