A performance evaluation was conducted for the Coupled Model Intercomparison Project Phase 5 (CMIP5) GCM ensemble over the south-central United States. A suite of temperature- and precipitation-based statistical metrics were calculated over the modeled historical period and compared to observation-based datasets. Models were ranked based on individual metrics as well as overall performances. No best model was found for all circumstances but general biases across models were identified.
This assessment may prove useful to regional downscaling and impacts researchers who may be restricted by the total number of GCMs they can utilize. Results from this analysis will be provided as well as a discussion on when it may be useful and appropriate to use.