A circulation-based performance atlas of the CMIP5 and 6 models for regional climate studies in the Northern Hemisphere mid-to-high latitudes

Abstract. Global climate models are a keystone of modern research. In most applications relevant for decision making, they assumed to provide plausible range possible future states. However, these have not been originally developed reproduce the regional-scale climate, which is where information needed in practice. To overcome this dilemma, two general efforts made since their introduction late 1960s. First, themselves steadily improved terms physical and chemical processes, parametrization schemes, resolution implemented system components, giving rise term “Earth model”. Second, global models' output has refined at regional scale using limited area or statistical methods what known as dynamical downscaling. For both approaches, however, it difficult correct errors resulting from wrong representation large-scale circulation model. Dynamical downscaling also high computational demand thus cannot be applied all available On background, there an ongoing debate community on whether thrive away “model democracy” paradigm towards careful selection strategy based capacity key aspects observed climate. The present study attempts useful such by providing performance assessment historical model experiments CMIP5 6 recurring atmospheric patterns, defined Jenkinson–Collison approach. latest generation (CMIP6) found perform better average, can partly explained moderately strong relationship between horizontal atmosphere. A few rank favourably over almost entire Northern Hemisphere mid-to-high latitudes. Internal variability only small influence ranks. Reanalysis uncertainty issue Greenland surrounding seas, southwestern United States Gobi Desert but otherwise generally negligible. Along study, prescribed interactively simulated components identified each coupled configuration simple codification introduced describe complexity sense.