A Bias-Corrected HighResMIP Dataset for Impact Assessment Studies

Le résumé fourni par la source

Climate impact assessments increasingly require high-resolution climate projections that capture fine-scale processes. While the High Resolution Model Intercomparison Project (HighResMIP) provides global climate simulations at 25-50 km resolution without statistical downscaling, systematic biases still limit their direct application for impact assessment studies. Here we present BC-HiRMIP, the first comprehensive globally bias-adjusted HighResMIP experiments at daily temporal and 0.5° spatial resolution, covering the period 1979-2050. Across four global climate models (MPI-ESM1-2-XR, EC-Earth3P-HR, CNRM-CM6-1-HR, HadGEM3-GC31-HM), BC-HiRMIP includes up to 11 essential meteorological variables (temperature, precipitation, humidity, radiation, wind, and pressure), spanning equilibrium climate sensitivities of 2.99-5.62 °C. The datasets were bias adjusted using the ISIMIP3BASD v3.0.1 methodology that preserves model-projected climate change signals across distribution quantiles with W5E5 v2.0 as reference. Comprehensive validation across diverse climate zones demonstrates substantial bias reduction, minor differences between raw and bias-adjusted climate change signals and distributional characteristics. This standardized, multi-variable, multi-model dataset bridges the gap between climate modeling capabilities and impact assessment needs, enabling applications in hydrology, agriculture, renewable energy, and climate service research.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.