How Pseudo is Pseudo-Global-Warming?

Abstract:

The pseudo-global-warming (PGW) method has been widely used to simulate future changes in precipitation and hydrology at regional scales. In PGW simulations, lateral boundary conditions are supplied by reanalysis, and the mean-state temperature and wind fields are perturbed to approximate the response to global warming. As a result, PGW simulations omit aspects of the hydrologic response to global warming that arise from changes in transient winds and temperature. Here I quantify the errors introduced by this limitation, focusing specifically on the change in precipitation minus evaporation (P-E). Building on previous work by Seager et al. (2010), I analyze 6-hourly output from the CESM Large Ensemble, and demonstrate that the change in P-E in each season can be decomposed into distinct contributions from changes in i) mean-state temperature, ii) mean-state winds, iii) transient temperature, and iv) transient winds. The first two of these contributions represent aspects of the P-E response that are captured by the PGW method, while the last two represent aspects that the PGW method leaves out. Within the latter category, I find that the contribution from changes in transient temperature (iii) is small except at high latitudes, where temperature variance decreases due to polar-amplified warming. In contrast, changes in transient winds (iv) contribute to substantial drying over much of the midlatitudes, with especially large impacts during wintertime in the southwestern US and western Europe. These results reveal both strengths and limitations of using the PGW method to assess hydrologic impacts of global warming in various regions.