Quantifying Water Vapor Age: A Dynamical Constraint on the Response of the Global Water Cycle to Climate Change

Abstract. One important uncertainty of climate change is the rate of acceleration of the global water cycle. While the Clausius-Clapeyron equation predicts a 7 % rise in atmospheric moisture per degree of warming, climate models suggest globally averaged precipitation only changes by 1 % to 3 % (the hydrologic sensitivity). This discrepancy can be explained by a 4 % to 6 % increase in water vapor (WV) age per degree of warming. Typically, changes in precipitation are used to calculate changes in WV age by assuming a well-mixed atmosphere to calculate a global value, which obscures the effects of regional dynamics on changes to the WV age spectrum. In this work, we have developed novel passive tracers for calculating the WV age spectrum in an Eulerian model. By directly simulating the moments of the WV age spectrum, our method resolves the temporal and spatial variability of the WV age distribution, and allows us to calculate the age spectrum of the precipitating water as well. We have demonstrated our technique by implementing these tracers in the Isca modelling framework using a quasi-realistic configuration which we then perturbed with global sea surface temperature anomaly experiments. In our control experiment, the globally averaged mean WV age and mean age of precipitation was 7.06 and 7.33 days respectively, and compared well to previous estimates using Lagrangian models. With climate change, WV age generally increased in the troposphere and decreased in the stratosphere. Globally, the mean WV age increased by 4.93 % K^-1 and the age of precipitation increased by 3.16 % K^-1. With our temporal and spatial resolution of WV age, we also resolve the effects of the Hadley cell strengthening, as well as the poleward shift and intensification of the midlatitude eddies and storm tracks, demonstrating how the regional changes in WV age differed from the regional changes in precipitation age.