Time-evolving sea-surface warming patterns modulate the climate change response of precipitation in Mediterranean-like regions

Regions with a Mediterranean-like climate, apart for California, are projected to receive less rainfall due to climate change, thus posing serious implications for future water availability for societal and agricultural needs. At a first order, it is often assumed that water availability is proportional to global mean warming. Yet, the mechanisms controlling the precipitation response in Mediterranean climates remain only partly understood, as shown by the substantial uncertainty that still characterises the climate model projections. Here, by analysing projections from the CMIP5 climate models, we show that the linear scaling with warming does not apply in three key Mediterranean-like regions, namely Chile, California and the Mediterranean proper. In particular, despite long-term warming, the models show that the projected precipitation reduction in Chile and the Mediterranean halts as soon as anthropogenic forcing is stabilised, while the precipitation increase in California accelerates. By examining the response to an abrupt quadrupling of CO2, we demonstrate that such non-linearity in the time-evolution of precipitation cannot be solely explained by the well-known rapid adjustment to radiative forcing, but it is instead due to distinct fast and slow patterns of atmospheric circulation change, that are themselves forced by the time-evolution in the spatial patterns of sea-surface temperature warming. In particular, while the fast warming is favourable to force a poleward shift of the mid-latitudes jets, hence drying the Mediterranean and Chile, the slow warming, including an el nino-like pattern in the tropical Pacific, inhibits such shifts and precipitation changes, while favouring the wetting of California. The results show that stabilising GHG concentrations will have an immediate benefit to the hydro-climate of these Mediterranean-like regions, while pointing to constraining uncertainty in the patterns of surface warming as an important step to increase confidence in the future projections.