A multimodel ensemble using the entropy-TOPSIS method for projecting temperature for the Lower Tapi River Basin

Abstract

General Circulation Models (GCMs) developed by scientist and researchers, are often used for predicting future trends in climate parameters need downscaling to regional scale and bias corrections. This study aims to use Entropy-TOPSIS approach to rank 13 bias corrected CMIP6 models for Lower Tapi River Basin (LRTB). Using five performance indicators, R2 Score, KGE, Spearman's Correlation ρ, MAE and Standard Error. The entropy method gives weights to these performance indicators and TOPSIS method was used to rank the models based on performance indicator. A Sensitivity analysis was performed using 5000 iterations on differential weights to check the robustness of the ranking. The study found EC-Earth3, CanESM5, EC-Earth3-Veg, INM-CM4-8, INM-CM5-0 as top performer for modeling Tmax and ACCESS-ESM1-5, BCC-CSM2-MR, ACCESS-CM2, EC-Earth3-Veg, EC-Earth3 as top performer for Tmin for the given study area. An ensemble model was developed using REA weights to study the change in temperature. The study found increase in Tmax from 0.9 °C to 2.0 °C and Tmin as 1.34 °C to 2.7 °C for period of 2015–2099 under various scenarios. The study will help researchers in climate studies over Lower Tapi River basin such as Water Quality Studies, Evapotranspiration, Effect of temperature on Agriculture, Soil moisture content and other hydro-climatic studies.