Possibility of spatial estimation of soil erosion using Revised Universal Soil Loss Equation model and generalized additive model in post‐hard coal mining spoil heap

Abstract Revegetation and plant cover establishment on coal mining heap areas are effective methods for reducing runoff and soil erosion. They are essential components for reestablishing ecosystems in ecologically vulnerable areas. Our study area covered the Upper Silesia coal region. The research focused on active and passive reclamation plots of the study area with different land covers, that is, woodland, grassland, forbland, and control plots. This study aimed to evaluate the spatial allotment of soil erosion in various reclamation techniques and land covers of a mining area using a combination of the Revised Universal Soil Loss Equation (RUSLE) model, geographic information systems, and regression modeling with generalized additive models. We demonstrated that the combination of different data analysis methods can be used to accurately predict soil degradation, even if not all factors used in the RUSLE model are wall‐to‐wall available. However, only sheet and rill erosions can be predicted using this model. The results demonstrated the spread of soil erosion over the post‐coal mining heaps, varying from 0.03 Mg/ha/year to 583.47 Mg/ha/year. The highest soil erosion rate was observed at the control site (bare rock without vegetation). Approximately 62.6% of the region remained in the safe zone, whereas approximately 1.4% of the area was classified as a very severe zone for erosion.