A modified CSLE for soil loss prediction under different vegetation patterns at slope scale in China

• Published in: International Soil and Water Conservation Research
• Main Authors: Shi, Wenhai, Bao, Jiachi, Wang, Miaomiao, Chen, Zhongming, Yu, Jinle, Chen, Hongjun, Song, Wenyi, Xin, Yan
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2025
• Subjects: Chinese soil loss equation; Flow path length; Soil loss prediction; Vegetation patterns; Flow path length; Chinese soil loss equation; Vegetation patterns; Soil loss prediction
• ISSN: 2095-6339
• DOI: 10.1016/j.iswcr.2025.06.003

Vegetation plays a fundamental role in reducing soil erosion by shielding the soil surface from raindrop impact and runoff erosion, promoting water infiltration to decrease runoff, and enhancing soil stability. Beyond the extent of vegetation cover, its spatial distribution is critical for optimizing erosion control and ensuring long-term ecological sustainability. Existing soil erosion models, such as the Chinese Soil Loss Equation (CSLE), predominantly focus on vegetation type and coverage while neglecting the spatial configuration of vegetation. This oversight can introduce uncertainties in predicting soil erosion. To overcome this shortcoming, this study proposes a new method by integrating vegetation spatial pattern indices, with particular emphasis on the mean flow path length index (MFLI), into the conventional CSLE framework. The MFLI effectively captures the positional distribution and spatial arrangement of vegetation, providing a more refined analysis of erosion dynamics at the slope scale. Using this index, a revised biological control factor (B) was developed. The proposed method was validated with data from 52 experimental plots across China and further tested with optimized parameters on five additional representative sites. Results demonstrated that the new approach substantially outperformed the conventional storm-based CSLE model, achieving model efficiencies of 0.686 and 0.636 during calibration and validation, respectively. In summary, the proposed method offers a more accurate and reliable prediction of soil erosion under diverse vegetation pattern conditions at the slope scale. By integrating spatial distribution characteristics of vegetation, it provides an improved tool for soil and water conservation, supporting more precise erosion prediction and mitigation strategies.