Machine learning-based identification for the main influencing factors of alluvial fan development in the Lhasa River Basin, Qinghai-Tibet Plateau

Alluvial fans are an important land resource in the Qinghai-Tibet Plateau with expansion of human activities. However, factors alluvial fan development poorly understood. According to our previous investigation and research, approximately 826 exist Lhasa River Basin (LRB). The main purpose this work is identify influencing by using machine learning. A index (Di) was created combining its area, perimeter, height gradient. 72% data, including Di, 11 types environmental parameters matching catchment 10 commonly used learning algorithms were train build models. 18% data validate remaining 10% test model accuracy. feature importance illustrate significance Di. primary modelling results showed that accuracy ensemble models, Gradient Boost Decision Tree, Random Forest XGBoost, not less than 0.5 (R2). Tree XGBoost improved after grid their R2 values 0.782 0.870, respectively. selected as final due optimal generalisation ability at sites closest LRB. Morphology development, a cumulative value relative 74.60% XGBoost. will have better adding training samples other regions.