Drivers of Dry Day Sensitivity to Increased CO2
While it is well established that global precipitation increases with climate warming, there is less certainty about regional precipitation changes, including extremes. Increases in regional extreme precipitation can often be accompanied by decreases in the frequency of lighter rain rates, and therefore an increase in the number of dry days. This work focuses on the regional responses to CO2 forcing through the lens of dry days and dry spells. The CMIP6 ensemble is used to quantify the changes in annual dry days and the annual maximum dry spell length. A subset of 4 models that include additional experiments are analyzed to decompose the contributions of the response into three additive components: the fast adjustment, the slow response associated with uniform SST warming, and the slow response associated with a patterned SST response.
All models project a strong drying around the Mediterranean as well as over most subtropical Mediterranean-like regional climates (e.g., Southern Africa, Australia, Chile), Amazonia and Southeast Asia. The spatial distribution of the precipitation response is mostly shaped by changes in the wet day frequency, and so the number of dry days is a critical diagnostic of the regional hydrologic response. While the models agree on many areas where the number of dry days increase, individual models can deviate regionally, leading to significant uncertainty. In the subset of models, the uniform warming component of the response is associated with widespread drying. One model shows a much weaker signal than the others, and it is instead dominated by the patterned SST warming. That suggests a larger role of large-scale circulation changes in that model than the others. The fast response is smaller than the slow response but is relatively robust across the models. Closely related to the number of dry days, the length of dry spells consistently increases in most subtropical regions, including the Mediterranean Basin, California, Chile, Southern Africa, and Australia.
The CMIP6 models simulate an overall increase in the number of dry days and the length of dry spells over land in response to an abrupt CO2 quadrupling. This widespread meteorological drying is particularly strong in the subtropics but is also found in some tropical (e.g., Amazonia) and extratropical (northern mid-latitudes) areas. Globally averaged over land, the annual mean precipitation increase is mostly explained by an increase in daily precipitation intensity. At the regional scale, the balance between precipitation intensity and frequency changes and the relative influences of the drivers of dry day changes is model-dependent, but the fast effect and the uniform warming effect show the most robust impacts. The SST pattern effect is less robust and likely more uncertain given the key role of model-dependent changes in large-scale circulation and the inability of models to capture the recent patterns of observed SST changes. Importantly, the changes in dry days and dry spells seem only weakly related to global warming, signally the importance of regional effects and structural uncertainties in models that can impact land-atmosphere feedbacks.