Quantifying CMIP6 Data Accuracy: A Comprehensive Analysis Using Multiple Metrics for the Whole India Grid Region to Identify Optimal Climate Models

The efficacy of the climate model in forecasting temperature measurements iscarefully examined in this work, with a particular emphasis on the impact of biascorrection and model-specific characteristics. The Mean Bias Error (MBE) analysesshow a consistent pattern: for both highest (Tmax) and minimum (Tmin)temperatures, positive MBE values suggest overestimation and negative valuesimply underestimating of recorded temperatures. This shows that MBE valuesand model accuracy are positively correlated, especially when it comes totemperature extreme prediction. Examining the linear relationship fol- lowingthe bias adjustment reveals different effects on correlation strength. A modestpositive connection is maintained by certain models, such as IPSL-CM6ALR,whereas a significant positive association is maintained by others. Nuancedinsights are revealed by skill scores and Root Mean Square Error (RMSE)computations, which show consistent pre-correction skill ratings among models.After rectification, there is a slight improvement in skill ratings. EC-Earth3 andAWI-CM-1-1-MR regularly show the lowest RMSE values, indicating better performance,particularly when it comes to portraying Tmax. This refined studyconcludes by highlighting the complex relationship between bias correction andspecific model properties and the performance of climate models. Both EC-Earth3and AWI-CM-1-1-MR consistently show themselves to be reliable predictors oftemperature extremes.