Abstract
This study conducts a comprehensive evaluation of climate models, assessing their ability to simulate climate indices and applying the Mann-Kendall test across different Shared Socioeconomic Pathways (SSPs). The analysis identifies nuanced strengths and weaknesses in various models, with CanESM5 demonstrating overall accuracy, Earth3-veg excelling in Diurnal Temperature Range (DTR), and limitations observed in Fogals-G3 and Miroc6. MRI-ESM2-0 exhibits lower overall accuracy, emphasizing inherent variability in modeling. The complexity of climate modeling is evident as none of the models consistently replicates observed values across all variables. The evaluation of vulnerability indices under SSP 585 conditions highlights diverse model projections, especially in temperature extremes like TX10p, TX90p, and TN90p. Mann-Kendall test results reinforce model-specific trends under different SSPs, emphasizing the need for cautious interpretation. Ongoing refinement is crucial for improved accuracy, stressing the necessity of a comprehensive understanding of model-specific outcomes to assess potential climate impacts under different scenarios. Furthermore, a dedicated examination of the Guwahati region focuses on temperature-based indices from 2015 to 2050, utilizing GCMs, downscaling models, and the Mann-Kendall test. This region-specific analysis, vital for water resource management and disaster preparedness, anticipates a rise in temperature and temperature indices, providing valuable insights for local planning and environmental protection.