Role of land surface processes in increased aridity over land from global warming

Abstract:

Recent studies based on climate model projections suggest that aridity, defined as the ratio of annual-mean precipitation to potential evapotranspiration, P/PET, will generally increase over land in a warmer world, as PET increases faster with temperature than P. It has been argued that increasing aridity with temperature follows from relatively simple, robust atmospheric thermodynamics, independent of land surface processes. In particular, the arguments invoked involve enhanced warming over land compared to oceans and decreased terrestrial relative humidity, both of which increase PET. Still, it is unclear precisely how much coupling to the land surface may ultimately impact the aridity response to warming. Here we propose to analyze the role of land surface processes in the increase in aridity (i.e., the decrease in P/PET) over land with global warming. We use simulations from the Global Land Atmosphere Coupling Experiment (GLACE)-CMIP5 experiment, in which climate change simulations from several CMIP5 generation models were performed with and without long-term changes in soil moisture. The results indicate that, from a global-mean perspective, climate change-induced decreases in soil moisture and associated feedbacks are significant components of the projected increase in land aridity. The drying of soil with future warming leads to greater increase in PET, and a smaller increase in P, compared to the future projections with present day soil moisture. The greater increase in PET mostly arises from a larger reductions in relative humidity, which themselves stem from a combination of temperature and specific humidity changes, i.e., the land-ocean warming ratio and reduced moistening from reduced land evaporation. Spatial analysis of these changes highlights latitudinal differences between the two sets of simulations, with the largest signatures in from the subtropics into midlatitudes. While there is general agreement across the ensemble of GLACE-CMIP5 models, some sizeable differences in the strength of land surface effects do occur, highlighting the sensitivity to model framework. Still, these results point to land surface processes as actively contributing to warming-induced changes in aridity over land.