Which weather systems are projected to cause future changes in mean and extreme precipitation in CMIP5 simulations?

Future changes in precipitation due to climate change are of great concern to society. However, questions such as Which weather systems will cause which changes? and Is the relative importance of these weather systems likely to change in the future? have not been addressed fully yet. Here we present the first global estimates of the relative contributions of different weather systems (i.e., tropical cyclones, extratropical cyclones including fronts, and others) to changes in annual mean and extreme precipitation in the late 21st century using multimodel projections of the Coupled Model Intercomparison Project Phase 5. Although the models present biases in tropical cyclones over southern hemisphere, in particular, the representations of global weather system patterns are comparable to the reanalysis data. Total precipitation from tropical cyclones decreases (increases) in the tropics (subtropics) and that from extratropical cyclones including fronts decreases (increases) on the equatorial (poleward) side of the storm tracks. In addition, the mean intensity and frequency of system-wise precipitation can change significantly even without considerable changes in annual amounts. We found that the subtropics, particularly in the Pacific and North Atlantic, are the regions where the proportions of precipitation by weather systems in annual mean and extreme precipitation display notable changes, suggesting distinct shifts in climate regimes. These regions have a common feature: they undergo the influence of several distinct weather systems in the present climate. In regions where climate regime shifts are projected, even the weather systems that have a minor contribution to precipitation in the present climate may cause considerable changes in annual and extreme precipitation.