Scaling Potential Evapotranspiration with Greenhouse Warming

Potential evapotranspiration (PET) is a supply-independent measure of the evaporative demand of a terrestrial climateof basic importance in climatology, hydrology, and agriculture. Future increases in PET from greenhouse warming are often cited as key drivers of global trends toward drought and aridity. The present work computes recent and business as usual future Penman-Monteith PET fields at 3-hourly resolution in 13 modern global climate models. The percentage change in local annual-mean PET over the upcoming century is almost always positive, modally low double-digit in magnitude, usually increasing with latitude, yet quite divergent between models.These patterns are understood as follows. In every model, the global field of PET percentage change is found to be dominated by the direct, positive effects of constant-relative-humidity warming (via increasing vapor deficit and increasing Clausius-Clapeyron slope). This direct-warming term accurately scales as the PET-weighted (warm-season daytime) local warming, times 5%-6% degrees C-1 (related to the Clausius-Clapeyron equation), times an analytic factor ranging from about 0.25 in warm climates to 0.75 in cold climates, plus a small correction. With warming of several degrees, this product is of low double-digit magnitude, and the strong temperature dependence gives the latitude dependence. Similarly, the intermodel spread in the amount of warming gives most of the spread in this term. Additional spread in the total change comes from strong disagreement on radiation, relative humidity, and wind speed changes, which make smaller yet substantial contributions to the full PET percentage change fields.