How relevant are frequency changes of weather regimes for understanding climate change signals in surface precipitation in the North Atlantic-European sector? – a conceptual analysis with CESM1 large ensemble simulations

Abstract. Climate change affects the climatology of surface precipitation in spatially in-homogeneous ways and it is challenging to identify and quantify the contribution of atmospheric circulation changes to this pattern. Various methods have been developed to characterize the large-scale atmospheric circulation and assess its changes, e.g., by classifying the flow into so-called weather regimes or circulation types. Several studies have then related frequency changes of these regimes due to global warming to changes in surface weather parameters. However, even without regime frequency changes, the climatology of surface parameters may change due to so-called regime intensity changes (e.g., a particular regime becomes on average wetter or drier). In this study, the question of how relevant frequency changes of weather regimes are for understanding climate change signals in surface precipitation is addressed with a novel conceptual framework. For every regime i, a spatially varying parameter γ_i(P) is introduced, which corresponds to the ratio of the contributions from regime frequency vs. regime intensity changes to the climate change signal of precipitation P. Conceptual considerations show that γ_i(P) is (i) proportional to the relative change of regime frequency, (ii) proportional to the regime-specific anomaly of precipitation, and (iii) inversely proportional to the climate change effect on regime intensity. The combination of these independent and competing factors makes the study of γ_i(P) interesting and insightful. As a specific example application of this framework, we consider a 7-category weather regime classification in the North Atlantic-European sector and large ensemble simulations with the CESM1 climate model under the RCP8.5 emission scenario for the periods 1990–1999 and 2091–2100. Considering γ_i(P) for surface precipitation P in this simulation setup reveals that (1) γ values are typically less than 0.3 and therefore, to first order, frequency changes of WRs are of secondary importance for explaining climate change signals in P – in contrast, the intensity changes dominate, which are to a large degree, but not entirely, related to the so-called thermodynamic effects of global warming; (2) the main reason for the generally low values of γ are the comparatively small WR frequency changes and the limited regime-specific anomalies of P, in particular over continental Europe; and (3) γ values tend to be slightly larger for precipitation variables that are less constrained by thermodynamic arguments, i.e., γ for the number of wet days is larger than γ for the number of heavy precipitation days. In summary, this study provides a generally applicable framework to quantify climate change effects of regime frequency changes on surface parameters, it illustrates the key conditions that must be fulfilled such that these frequency changes can become relevant, and, at least in our application, it shows that these conditions are generally not fulfilled.