Future changes to north Pacific mean-state and the connections to the western US precipitation extremes from multiple GCM large-ensembles

Abstract:

Future changes to precipitation extremes are projected by coupled climate models. However, differences in these projections arise, due to both inter-model structural differences and internal climate variability. To date, it has been difficult to separate these two factors; the recently released Multi-Model Large Ensemble Archive now provides a valuable resource to disentangle internal variability from inter-model physical differences. We examine the mean and extreme precipitation changes over the North Pacific Ocean and the western US coast (WUS) to see how mean-state changes affect the behavior of extremes. Overall, the subtropical region (25 to 35°N) is consistently predicted to be drier, except over the eastern Pacific region in CanESM2. Over the northward midlatitude regions and higher latitudes, the precipitation shows a widespread increase. The Pacific Hadley cell exhibits a widening tendency in all of the models except CanESM2, which is consistent with the different precipitation changes exhibited. These different mean-state patterns are associated with distinct regional changes in precipitation extremes: from south to north (here, from California to Washington), models show different levels of increases. The contributions of dynamic and thermodynamic factors are quantified using a decomposition of future moisture flux changes, which shows that thermodynamic increases in moisture are dominating but with certain discrepancies among models and at different mid-latitude regions. Although the dynamic contribution is relatively moderate, it induces notable uncertainties among different models in terms of the signs of the contributions. This suggests that reducing uncertainty among future projections of extremes will require more accurate constraints on future mid-latitude circulation changes.