Google Earth Engine-based morphometric parameter evaluation and comparative analysis of soil erosion susceptibility using statistical and machine learning algorithms in large river basins

Morphometric analysis of a river basin, encompassing the assessment of shape, drainage, and terrain characteristics, is crucial for erosional, hydrological, and hydro-geological studies However morphometric analysis of large river basins conventionally poses computational challenges. In this study, Google Earth Engine (GEE), a cloud-based platform was utilized to conduct morphometric modeling and evaluation of geomorphology of river basins using JavaScript Application Programming Interfaces (APIs). Consecutively, soil erosion susceptibility mapping based on morphometric parameters using an ensemble of statistical and machine learning algorithms specifically Principal Component Analysis (PCA), Random Forest (RF), and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) was performed. These methods were applied to India's largest (Ganga: 834,933.65 Km²) and second-largest (Godavari: 310,000 Km²) river basins. Correlation-based comparative analysis of PCA and RF-derived erosional susceptibility rankings for the Ganga (GARB) and Godavari (GRB) river basins was also performed followed by the identification of highly susceptible zones. Results indicate that 28.50% (86,800 Km²) of GRB and 9.85% (81,823.49 Km²) of GARB are highly susceptible to geological erosion. Highly susceptible GARB sub-watersheds were 2.7 times more susceptible than those in GRB, while less susceptible GARB sub-watersheds were 1.96 times more susceptible than GRB counterparts. Machine learning with RF excelled in river basins with greater morphometric variation (GARB), while PCA proved more effective in homogenous river basins (GRB). This study introduces an alternative method for morphometric modeling based erosional analysis, emphasizing the suitability and constraints of statistical and machine learning algorithms and providing valuable guidance for soil conservation efforts.