The fractional increase in global mean precipitation ([Formula: see text]) is a first-order measure of the hydrological cycle intensification under anthropogenic warming. However, [Formula: see text] varies by a factor of more than three among model projections, hindering credible assessments of the associated climate impacts. The uncertainty in [Formula: see text] stems from uncertainty in both hydrological sensitivity (global mean precipitation increase per unit warming) and climate sensitivity (global mean temperature increase per forcing). Here, by investigating hydrological and climate sensitivities in a unified surface-energy-balance perspective, we find that both sensitivities are significantly correlated with surface shortwave cloud feedback, which is further linked to the climatological pattern of cloud shortwave effect. The observed pattern of cloud effect thus constrains both sensitivities and consequently constrains [Formula: see text]. The 5%-95% uncertainty range of [Formula: see text] from 1979-2005 to 2080-2100 under the high-emission (moderate-emission) scenario is constrained from 6.34[Formula: see text]3.53% (4.19[Formula: see text]2.28%) in the raw ensemble-model projection to 7.03[Formula: see text]2.59% (4.63[Formula: see text]1.71%). The constraint thus suggests a higher most-likely [Formula: see text] and reduces the uncertainty by ~25%, providing valuable information for impact assessments.
© 2023. Springer Nature Limited.