作者: Thordis L Thorarinsdottir , Jana Sillmann , Marion Haugen , Nadine Gissibl , Marit Sandstad
关键词: Grid 、 Set (abstract data type) 、 Rare events 、 Climate model 、 Ranking 、 Econometrics 、 Index (economics) 、 Context (language use) 、 Climate change 、 Computer science
摘要: Reliable projections of extremes by climate models are becoming increasingly important in the context change and associated societal impacts. Extremes definition rare events, characterized a small sample with large uncertainties. The evaluation extreme events model simulations thus requires performance measures that compare full distributions rather than simple summaries. This paper proposes use integrated quadratic distance (IQD) for this purpose. IQD is applied to evaluate CMIP5 CMIP6 monthly maximum minimum near-surface air temperature over Europe North America against both observation-based data reanalyses. Several perform well extent these models' competitive another product simulating set. While rankings vary region, season index, robust changes grid resolution considered analysis. When ranked based on their similarity ERA5 reanalysis, more appear at top ranking. evaluated HadEX2 product, overall two ensembles similar.