Prioritizing the selection of CMIP6 model ensemble members for downscaling projections of CONUS temperature and precipitation

Abstract Given the mismatch between large volume of data archived for sixth phase Coupled Model Intercomparison Project (CMIP6) and limited personnel computational resources downscaling, only a small fraction CMIP6 archive can be downscaled. In this work, we develop an approach to robustly sample projected hydroclimate states in downscaling test whether selection single initial condition (IC) ensemble member from each model is sufficient span range modeled over conterminous United States (CONUS) CONUS sub-regions. We calculate pattern-centered root mean square difference IC anomalies relative model’s historical climatology shared socioeconomic pathway (SSP) projections 30-year time periods compare ratio inter-model intra-model variability metric. Regardless SSP, generally much greater than at scales as whole, well most However some variables scenarios, inter- are similar sub-CONUS scales, indicating that selecting per may 21st Century across CONUS, but specific regions variables, more careful members necessary. Regionally-resolved Taylor diagrams identify where efforts should focused if available do so. Our results suggest that, with parsimonious sampling, requisite expense temperature precipitation fields subsequent CMIP activities increase marginally despite great volumes successive phase.