Short Link
Full Link

By Dr. Aiguo Dai, Ph.D.
Professor, State University of New York, Albany

Click to enlarge.
Figure 1. Linear trends in (a) observed precipitation and (b) calculated self-calibrated Palmer Drought Severity Index (PDSI) with Penman-Monteith potential evapotranspiration during 1950-2010 (from Dai, 2013, Nature Climate Change), and (c) observed river runoff from 1948-2004 (from Dai et al. 2009, J. Climate). Red and orange colors indicate drying. Stippling in (a) and (b) indicate the trends are statistically significant. These physically-related but not identical measures of drought show consistent drying trends over many land areas.

The largest impacts of human-induced climate change in the coming decades will likely come from large percentage changes in weather and climate extremes, as these extreme events will change at much faster rates than the mean climate does. Among the projected changes of extreme events, increased risk of drought, heat waves and wild fires is a major concern, as climate models predict higher summer temperatures, increased evaporative demands, fewer rainy days, and drier soils over many land areas.

In fact, large drying trends may have already occurred since the 1950s over many land areas such as most of Africa, South and East Asia, Southern Europe, eastern Australia, and parts of South and Central America, as shown by historical records of streamflow, precipitation and drought index (Fig. 1). The rising temperature since the late 1970s has enhanced the drying over many land areas, although natural, decadal precipitation variations (e.g., associated with decadal variations in sea-surface temperatures in the Pacific) may have contributed considerably to the historical drying trend over some of these regions such as the U.S., South Asia, and eastern Australia.

As atmospheric CO2 and other greenhouse gases continue to rise, the impact of global warming will, however, become increasingly dominant over natural variations in the coming decades. The averaged soil moisture changes projected by climate models (Figure 2) suggest wide-spread drying in the Americas, Europe, Australia, and southern Africa. One might think 5-10% reduction in decadal-mean soil moisture content is not a big deal, as it can vary from year to year at much larger amplitudes. However, such a mean shift can lead to much larger increases (e.g., more than doubling by the 2090s) in global drought areas, as shown by Figure 3. The reason is that droughts are located on the left-side tail of the probability distribution of soil moisture at a given location, and the probability for drought can increase greatly for even just a small shift in the mean.

Click to enlarge.
Figure 2. Projected future changes (2080-2099 minus 1980-1999) in annual precipitation in mm/day (left column) and soil moisture content in percent in the top-10cm layer (right column.) Results from the climate models used to formulate the 2007 IPCC report (called CMIP3 models) are shown in the top row, and results from the models used to formulate forthcoming 2013 IPCC report (called CMIP5 models) are shown in the bottom row. The medium emissions scenario (called A1B for the CMIP3 models and RCP4.5 for the CMIP5 models) is shown for each set of models. Humanity is currently on pace to emit much more CO2 to the atmosphere than these medium emission scenarios suggest, and the predicted changes in precipitation and soil moisture are more more drastic than this for the higher emission pathway that we are currently on.

This future increased risk of drought is further illustrated (Figure 4) by the self-calibrated Palmer Drought Severity Index (PDSI) with potential evapotranspiration estimated using the Penman-Monteith equation and calculated using the surface meteorological data from the climate models. While the PDSI is designed to measure droughts under current climate conditions and thus its quantitative interpretation for future climates needs to be cautious, the combination of the soil moisture (Figures 2-3) and PDSI (Figure 4) changes does suggest a dire projection of increased risk of severe droughts.

Click to enlarge.
Figure 3. Time series of the percentage dry (red lines) and wet (green) areas of global land from 1950-2099 projected by CMIP3 (from the 2007 IPCC report) and CMIP5 climate models (from the 2013 IPCC models) under medium emissions scenarios. The dry and wet areas are defined, respectively as the bottom and top 20 percentiles of the soil moisture content in the climate models during 1950-1979.
Click to enlarge.
Figure 4. The self-calibrated Palmer drought severity index (PDSI) with the potential evapotranspiration estimated using the Penman-Monteith equation and calculated using the surface meteorological data from CMIP3 (left) and CMIP5 (right) models under medium emissions scenarios for selected decades. Currently, PDSI less than -3 is considered severe to extreme drought over the U.S.
Dr. Aiguo Dai received his PhD in Atmospheric Science from Columbia University in 1996. He is currently an Associate Professor at the University of Albany, SUNY. He has studied climate change, in particular, hydroclimate change, for more than 15 years. He has published over 80 peer-reviewed journal articles with a total of citations close to 8000 and an H-index of 44. He currently serves as an Editor of Journal of Climate and chairs the Committee on Climate Variability and Change of the American Meteorological Society.
Try Our BETA
We're updating
the WU experience.

Try it!