摘要：We present the global and regional hydrological sensitivity (HS) to surface temperature changes; for perturbations to CO2; CH4; sulfate and black carbon concentrations; and solar irradiance. Based on results from ten climate models; we show how modeled global mean precipitation increases by 2–3% per kelvin of global mean surface warming; independent of driver; when the effects of rapid adjustments are removed. Previously reported differences in response between drivers are therefore mainly ascribable to rapid atmospheric adjustment processes. All models show a sharp contrast in behavior over land and over ocean; with a strong surface temperature-driven (slow) ocean HS of 3–5%/K; while the slow land HS is only 0–2%/K. Separating the response into convective and large-scale cloud processes; we find larger inter-model differences; in particular over land regions. Large-scale precipitation changes are most relevant at high latitudes; while the equatorial HS is dominated by convective precipitation changes. Black carbon stands out as the driver with the largest inter-model slow HS variability; and also the strongest contrast between a weak land and strong sea response. We identify a particular need for model investigations and observational constraints on convective precipitation in the Arctic; and large-scale precipitation around the Equator.
