Abstract Climate model selection is crucial for enhancing the reliability of climate projections, which are widely employed for policy development, particularly in water resources management. This study proposes a novel methodology that incorporates large-scale circulation drivers and extreme climate indices for evaluating and ranking climate models’ performance. As a case study, the methodology was applied to assess the performance of selected CMIP6 climate models in Victoria, Australia, focusing on key climatic variables including precipitation, temperature, solar radiation, sea surface temperature, geopotential height and specific humidity. Five precipitation-based indices R10, R20, R40, CDD and PRCPTOT, and two temperature-based indices SU25 and SU35 were used. Seven CMIP6 models were evaluated over the period 1979–2013 at a daily step using statistical indicators, including Root Mean Square Error (RMSE), Spatial Correlation (SC), Percentage Bias (Pbias) and Nash-Sutcliffe Efficiency (NSE). The models were ranked using Compromise Programming, individually as well as in groups. GFDL-ESM4 consistently ranked high across most groupings and emerged as the overall top-performer model. In contrast, MIROC6 and CNRM-CM6-1 consistently ranked lower across multiple groupings. The proposed methodology is dynamic and versatile, as it integrates large-scale circulation variables and extreme climate indices in the evaluation process, offering wide applicability in climate change impact assessment studies.
Adhikari et al. (Thu,) studied this question.