Precipitation and its extremes in changed climates

While the specific humidity of the atmosphere is expected to increase as the climate warms, roughly consistent with constant relative humidity in a global mean, precipitation increases do not keep pace with specific humidity increases. To investigate systematically how precipitation changes with climate, we conducted a series of simulations with an idealized general circulation model in which we varied the optical thickness of the longwave absorber. In the simulations, globalmean precipitation increases approximately linearly with global-mean surface air temperature over a wide range of climates and asymptotically approaches a maximum in warm climates, with the maximum estimable from the surface energy budget. The large-scale (grid-scale) precipitation, which is particularly important in the extratropics, exhibits more complex behavior, with a maximum in a moderately warm climate and smaller values in colder and warmer climates. Changes in precipitation extremes likewise do not scale with specific humidity changes but exhibit more complex behavior, with a wide range in which they increase approximately linearly with global-mean surface air temperature at a fractional rate approximately equal to that of the mean precipitation and, for some extreme-value statistics, with a maximum or plateau in warm climates.