Sea surface warming patterns drive hydrological sensitivity uncertainties

The increase in global-mean precipitation with global-mean temperature (hydrological sensitivity; $$\eta$$ η ) is constrained by the atmospheric energy budget, but its magnitude remains uncertain. Here we apply warming patch experiments to a climate model to demonstrate that the spatial pattern of sea surface warming can explain a wide range of $$\eta$$ η . Warming in tropical strongly ascending regions produces $$\eta$$ η values even larger than suggested by the Clausius–Clapeyron relationship (7% K−1), as the warming and moisture increases can propagate vertically and be transported globally through atmospheric dynamics. Differences in warming patterns are as important as different treatments of atmospheric physics in determining the spread of $$\eta$$ η in climate models. By accounting for the pattern effect, the global-mean precipitation over the past decades can be well reconstructed in terms of both magnitude and variability, indicating the vital role of the pattern effect in estimating future intensification of the hydrological cycle.