Abstract
Reliable projections of extremes by climate models are becoming increasingly important in the context of climate change and associated societal impacts. Extremes are by definition rare events, characterized by a small sample associated with large uncertainties. The evaluation of extreme events in model simulations thus requires performance measures that compare full distributions rather than simple summaries. This paper proposes the use of the integrated quadratic distance (IQD) for this purpose. The IQD is applied to evaluate CMIP5 and CMIP6 simulations of monthly maximum and minimum near-surface air temperature over Europe and North America against both observation-based data and reanalyses. Several climate models perform well to the extent that these models’ performance is competitive with the performance of another data product in simulating the evaluation set. While the model rankings vary with region, season and index, the model evaluation is robust against changes in the grid resolution considered in the analysis. When the model simulations are ranked based on their similarity with the ERA5 reanalysis, more CMIP6 than CMIP5 models appear at the top of the ranking. When evaluated against the HadEX2 data product, the overall performance of the two model ensembles is similar.
Highlights
Current climate projections indicate a significant warming of the hottest days and the coldest nights in all land areas of the world already under low emission scenarios (Hoegh-Guldberg et al 2018), and even more severe increases are projected for higher emission scenarios (Sillmann et al 2013b)
Climate model evaluation has traditionally been performed by comparing summary statistics that are derived from simulated model output and corresponding observed quantities using, for instance, the root mean squared error (RMSE) or mean bias (Flato et al 2014)
HadEX3 is very similar to HadEX2 and the reanalyses ERA5 and ERA-Interim are quite similar, while the NCEP-DOE Reanalysis 2 (NCEP-2) reanalysis performs poorly; for winter TNn only 7 out of 48 climate model simulations perform worse than NCEP-2
Summary
Thordis L Thorarinsdottir , Jana Sillmann, Marion Haugen, Nadine Gissibl and Marit Sandstad.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
