Performance Evaluation of CMIP6 Climate Models for Rainfall and Erosivity in the Thamirabharani Basin, India

Global Climate Models (GCMs) are an imperative component in water resource management for the identification of key factors of climate, simulation of climatic behavior, and prediction of future conditions. The performance evaluation of GCMs is another vital task for ensuring reliable climate projection. The study assesses 35 CMIP6 GCMs from the NASA Earth Exchange Global Daily Downscaled Projections (NASA NEX-GDDP) datasets against India Meteorological Department (IMD) observations (1950-2014) to evaluate their performance in simulating rainfall over the Thamirabharani River Basin. Model performances were assessed using six statistical indicators, such as Pearson Correlation Coefficient (CC), Nash–Sutcliffe Efficiency (NSE), Normalized Root Mean Square Error (NRMSE), Percent Bias, Kling–Gupta Efficiency (KGE), and Skill Score (SS). The relative importance of these indicators was considered through three weighting schemes: Equal weight, Entropy-based weight, and Principal Component Analysis (PCA)-derived weight. Additionally, five Multi-criteria decision making (MCDM) Techniques are PROMETHEE-II, TOPSIS, VIKOR, MOORA, and Compromise Programming (CP), were applied to comprehensively rank the GCMs. Based on the combined outcomes of the statistical analysis and ranking using Group Decision Method (GDM), the best-performing models, ACCESS-CM2, CanESM5, MIROC6, NorESM2-MM, and BCC-CSM2-MR, were identified. These selected models are used to project future rainfall and rainfall erosivity for two different shared socioeconomic pathways (SSPs), SSP4.5 and SSP8.5, respectively. Under the SSP 4.5 scenario, the rainfall and rainfall erosivity are projected to increase by 1.34 and 1.40 times respectively in the future. Under SSP 8.5 the increase are higher at 1.54 and 1.47 times, respectively. Rainfall and erosivity are going to be more intense toward the end of the century, especially under high-emission conditions, which implies higher risks for soil erosion and water resource challenges.