Evaluating inter-model variability among CMIP6 projections of future Köppen–Geiger climate zones in Kazakhstan under SSP5-8.5

Abstract

The focus of our study is the uncertainty of the inter-model Köppen–Geiger climate classification map for the future under the SSP5-8.5 scenario. We created classification predictions using 32 models from the CMIP6 ensemble for the end of the century across Kazakhstan. We created a modal map to display the most frequently predicted climate zone for each pixel. To indicate the uncertainties of the predictions, additional maps were created that also show the secondary zones where their probability is higher than a specified threshold. To evaluate classification confidence, we applied three thresholds to visualize zones of high, moderate, and low certainty. This approach highlights regions where the climate classification is less reliable and where further modelling or data collection may be needed to enhance confidence. Given the significant variability among model outputs, our findings emphasize the need for cautious interpretation.