1 Journal of Geophysical Research: Atmospheres (1984–2012) 1990 Vol: 95(D7):9983-10004. DOI: 10.1029/JD095iD07p09983

Potential evapotranspiration and the likelihood of future drought

The likelihood of future drought is studied on the basis of two drought indices calculated from the Goddard Institute for Space Studies general circulation model (GISS GCM) transient and doubled CO2 climate changes. We use the Palmer drought severity index (PDSI) and a new supply-demand index (SDDI), the latter being the difference between the precipitation and potential evapotranspiration Ep, i.e., the difference between atmospheric supply of and demand for moisture. Both indices show increasing drought for the United States during the next century, with effects becoming apparent in the 1990s. If greenhouse gas emissions continue to increase rapidly, the model results suggest that severe drought (5% frequency today) will occur about 50% of the time by the 2050s. The results are driven by the large increase in EP, associated with the simulated climate warming. EP increases most where the temperature is highest, at low to mid-latitudes, while precipitation increases most where the air is coolest and easiest to saturate by the additional moisture, at higher latitudes. Thus atmospheric demand becomes greater than supply for most of the globe by the latter half of next century. We show that large EP changes can lead to soil moisture deficits, as in the PDSI hydrologic budget, and vegetation desiccation, as is implied by vegetation and climate models. We suggest that drought intensification has been understated in most GCM simulations, including the GISS GCM, because of their lack of realistic land surface models. Paleoclimate analogues in the Mesozoic and Cenozoic are reviewed which imply that arid conditions can result from either increased temperatures or decreased precipitation, consistent with our use of the SDDI. The results depend primarily on the temperature increase, in particular the model sensitivity of 4°C warming for doubled CO2. Global precipitation cannot keep pace with increased demand over land because the land surface warms more than the ocean surface; this effect, along with greater atmospheric opacity, reduces low level gradients and evaporation from the ocean. If the forecast temperature changes come to pass, these conclusions suggest that drought conditions will increase dramatically.