Floods and Droughts in a Changing Climate – Now and the Future

EarthzineArticles, Climate, Extreme Weather Theme, Original

Map showing Change in the likelihood of summer season drought (defined as a once in ten years event for the late 20th century) by the end of the 21st century

Paul A. Dirmeyer
Center for Ocean-Land-Atmosphere Studies
Calverton, Maryland


Floods and droughts are among the most dangerous and costly of all natural disasters. According to statistics from the United Nations [1], during 1970-2005 over 30% of natural disasters were floods and nearly 15% were droughts or drought-related (wild fires and extreme high temperatures). During the 30-year period 1980-2009, floods accounted for more deaths in the United States than hurricanes, tornados or lightning, ranking first among weather fatalities [2]. Droughts are the main cause of agricultural distress, accounting for over $11 billion in damage in the United States during the first decade of this century [3].

Floods and droughts are not equal-but-opposite phenomena. Droughts are the absence of rain, and thus grow and retreat in severity at rates paced by the climatological (“normal”) precipitation in an area. Droughts strongly impact agriculture, disrupting the annual harvest cycle, affecting prices for agricultural commodities in real time through market speculation, and, through lingering scarcities, rippling through regional and global economies for many months after the drought has ended. In some regions with decadal time scale climate variations, such as the southwestern United States or the Sahel in Africa, droughts can persist for many years or even decades. Floods, on the other hand, are caused by extreme excesses of precipitation or the sudden release of a surfeit of water from storage, such as a reservoir or snowpack. They tend to be more localized than droughts and rather short-lived, lasting hours to days, although large-scale floods can last weeks or months.

There is great concern that a changing climate brought about by the unprecedented human-driven changes to the composition of the atmosphere could increase the frequency or severity of droughts and floods. The latest report from the Intergovernmental Panel on Climate Change (IPCC) states that an increase in the frequency of extreme precipitation events is “very likely” to expand and the area of the globe affected by increased drought is “likely” to expand [4]. In this article, we discuss historical trends in floods and drought, future projections, and the requirements of observations to monitor changes.

Historical Trends

There are numerous ways to define or categorize both floods and droughts. Meteorological floods and droughts are classified based on anomalies in precipitation. Hydrologic floods and droughts are measured in terms of deviations of streamflow or river water depth from historical norms. Agricultural floods and droughts are defined by their impact on crops and livestock. Other more specialized stake holders such as dam and reservoir operators, hydroelectric power concerns, river transportation networks and municipal water managers, to name a few, have their own criteria for defining floods and droughts. We will speak here primarily in terms of meteorological extremes based on anomalies from climatological annual cycles of precipitation.

Reliable raingauge data exist for much of the world beginning from the late 19th to mid-20th centuries, depending on the location. Satellite estimates can help fill gaps in the gauge network and provide estimates over oceans, but these estimates extend back only to the late 1970s. Records can be extended back in time in many areas using biological and geological proxies such as tree ring and pollen data, lake sediment samples and erosion patterns. These proxy methods are not as precise as instrumental measurements.

There exists in the Earth’s climate a degree of natural variability on all time scales, and when speaking of variability in precipitation, the extremes are precisely what we call floods and droughts. These can be random chance events that reflect the chaotic aspects of the climate system. However, there are slowly-varying elements of the climate system that can enhance the likelihood of droughts or floods in any given year. Chief among these is the cycle of El Niño / La Niña events, a quasi-periodic oscillation of sea surface temperatures and oceanic heat content that is part of a coupled ocean-atmosphere mode of variability in the tropical central and eastern Pacific Ocean. El Niño affects precipitation variability on a 2-6 year cycle across much of the globe exacerbating drought or flood, depending on its phase, over Australia, Indonesia, much of North and South America, East and South Africa [5]. It also interacts with the Asian monsoon system in a complicated fashion [6]. On longer timescales there are variations such as the Pacific Decadal Oscillation and the Atlantic Multidecadal Oscillation that appear to be instrumental in driving long-term droughts such as the 1930’s Dust Bowl over the western Great Plains of the United States, and the prolonged drought over the Sahel during the late 20th century [7]. There is also growing evidence that the state of the land surface (e.g., anomalies in soil moisture) can act as a positive feedback, exacerbating both droughts [8; 9] and floods [10; 11].

Although extremes in precipitation exist in the climate record on a variety of timescales, evidence is mounting that we are already witnessing the signature of human-influenced global warming in precipitation observations. An upward trend in both the frequency and intensity of heavy precipitation events has been found over the United States during the 20th century [12]. A subsequent study of global precipitation gauge records confirmed these positive trends in wet extremes over much of the world [13]. Recently, a careful comparison to observations of multi-model projections of changes in extreme rainfall during the last half of the 20th century suggests that observed increases across more than half of the monitored areas of the Northern Hemisphere can be attributed to increases in greenhouse gases [14]. Drought also appears to be on the increase over the last half century, driven not only by regional downward trends in precipitation, but also by the drying effect on soils of increasing temperatures [15] and the change in timing of spring snowmelt [16].

Map showing Change in the likelihood of summer season drought (defined as a once in ten years event for the late 20th century) by the end of the 21st century

Figure 1. Change in the likelihood of summer season drought (defined as a once in ten years event for the late 20th century) by the end of the 21st century, according to multi-decade simulations with the operational model of the European Centre for Medium-range Weather Forecasts (ECMWF) based on the A1b scenario of the IPCC.

Projections for the Future
A synthesis of climate model projections presented by the IPCC indicates an increase in both the occurrence of intense precipitation events (floods) and the number of dry days (droughts) over large portions of the globe, including many areas that may see significant increases in both extremes [17]. To illustrate this, we show here recent findings using the global atmospheric model of the European Centre for Medium-Range Weather Forecasts (ECMWF) [18]. This model was run at a much higher spatial resolution than any of the climate models that contributed to the IPCC conclusions. Unlike those climate models, these simulations were made with specified observed ocean surface temperatures for the late 20th / early 21st century, and to simulate the projected warming those same ocean temperatures were added to the climate change signal from one of the IPCC model simulations in order to simulate late 21st century conditions (based on the A1b scenario for greenhouse gas emissions) [19].

Figure 1 shows the projection for the change in the likelihood of summertime (June through August) precipitation falling in the lowest 10%, or a once-in-ten-year’s drought based on the climate of the late 20th century simulation (1961-2007). Using that threshold as the definition of drought, we then compare to all summer seasons of the future scenario (2071-2117). “Double” means what was a once-in-ten-years drought is projected to double in frequency to once-in-five years. Likewise, “triple” and “quadruple” mean three and four times more likely. The large-scale features of the ECMWF model are in very close agreement with the official IPCC projections [4; 17]. Both projections indicate Southern Europe, North Africa, Mexico, the Caribbean, and a large portion of the United States and Canada will likely suffer dramatic increases in the incidence of drought. One should not read too much into the details of the small-scale structure apparent in this simulation (run at a spatial resolution of 16km or 10 miles). Rather the take-home message should be that there is a great deal of local-scale structure, and it is quite likely that over several decades, a specific location in the Midwest U.S., for instance, could experience much more drastic or mild effects than a locality 50 or 100 km away. The ECMWF model projections also indicate that over most land areas there will be a significant increase in the number of days with no rain.

Figure showing the Change from the late 20th to the late 21st century in the average amount of precipitation occurring during the five wettest six-hour intervals (equivalent to the worst flash floods) in reference 47-year periods

Figure 2. Change from the late 20th to the late 21st century in the average amount of precipitation occurring during the five wettest six-hour intervals (equivalent to the worst flash floods) in reference 47-year periods.

At the other end of the spectrum, there is also projected to be an increase in the number of days with heavy precipitation. Figure 2 shows the change in total precipitation in the five strongest flash floods – here defined as the five wettest 6-hour periods in the 47-year simulations of current versus future climate. Very few locations are projected to experience a decrease in extreme rainfall events, and large portions of the world are forecast to have 30% or more precipitation falling in those extreme events. Across much of the tropics and subtropics, increases of 100% or greater are common. These results are also very consistent with the conclusions of the IPCC for extreme precipitation events [17].

What Are the Mechanisms?

It is a fair question to ask, what are the mechanisms behind the changes in precipitation that all of these models project? First of all, the amount of water vapor that air can hold is a direct function of its temperature. That amount grows exponentially as air becomes warmer. This is illustrated in Figure 3, which shows the percentage of the air, by mass, composed of water vapor for a range of relative humidities from 10% to 100% as a function of air temperature. For a 1°C increase from the global mean surface air temperature (currently about 15°C), the capacity of the air to hold water vapor increases nearly 7%. The more moisture there is in the air, the more moisture can condense and fall as rain or snow. Global models and theoretical calculations indicate that the atmosphere conserves its overall relative humidity as the global mean temperature changes, so an increase in the amount of moisture in the air seems likely. Other factors may also contribute toward more vigorous precipitation, such as the fact that the lower atmosphere is warming faster than the middle or upper atmosphere over most of the globe. This causes the atmosphere to be more thermally unstable, and more conducive to triggering thunderstorms and convective precipitation. Factors such as these appear to be behind the tendency toward more floods.

Graph showing The percentage of air, by mass, composed of water vapor, plotted as a function of temperature (bottom axis) and relative humidity (various curves).

Figure 3. The percentage of air, by mass, composed of water vapor, plotted as a function of temperature (bottom axis) and relative humidity (various curves).

On the other end of the spectrum, the tendency toward more frequent and prolonged drought also has several causes. Even with no change in mean precipitation, a shift towards greater percentages of rainfall coming in intense events would cause greater storm runoff and less infiltration of water into the ground. Coupling that with higher mean air temperatures leads to less moisture in the soil, which could have potentially disastrous consequences for agriculture. As mentioned earlier the state of the land surface, particularly soil moisture, seems to act as a reinforcing feedback to local and regional precipitation [20]. Dry soils are accompanied by reduced evaporation and plant transpiration, leading to drier and more stable air that further inhibits convection and rainfall.

Conclusions and Observational Implications

Recent observations are corroborating the projections of climate models in terms of changes to the Earth’s climate in response to the changing composition of the atmosphere driven by human industrial and agricultural practices. Most locations can expect an increase in the frequency and intensity of heavy precipitation events, and many locations can expect more frequent episodes of drought. “Gentle rains” are becoming less common. The greatest projected changes, however, are yet to come.

Many facets of the climate system remain poorly observed or not monitored at all. For example, soil moisture is currently monitored on a routine basis in at most a few hundred locations worldwide. Vast areas of the continents are hundreds if not more than a thousand kilometers from the nearest routine measurement site. Rain gauge networks in some countries are very dense, but in others they are unmaintained or nonexistent. Satellites offer the promise of routine global measurements of key aspects of the global water cycle, but all satellite instruments measure radiances in specific bands of the electromagnetic spectrum that must be processed and interpreted in terms of physical properties of the land and atmosphere. This is a complex and imprecise process. Furthermore, the average environmental satellite is in service for approximately 10 years, often replaced by a platform carrying different instrumentation that cannot be seamlessly cross-calibrated with previous data.

The result of such problems is uneven monitoring of the evolving climate of the Earth, and added uncertainty as to what is actually occurring and why. Long-term planning and support for global monitoring systems is needed – with perspectives well beyond specific field campaigns, satellite missions, or national networks. The design-build-operate cycle for instrumentation needs to occur with the time scales of climate change in mind – at least 50 years. Consistency in measurements is crucial for long-term monitoring. The more accurately we can measure changes in the Earth’s hydrologic cycle, the better we can plan and prepare for mitigation and adaptation.


[1] ISDR (International Strategy for Disaster Reduction), “Disaster Statistics” http://www.unisdr.org/disaster-statistics/pdf/isdr-disaster-statistics-occurrence.pdf, 2006.
[2] National Weather Service, “Weather Fatality, Injury and Damage Statistics” http://www.weather.gov/os/hazstats.shtml, 2010.
[3] National Oceanic and Atmospheric Administration, “Economics of Drought Data and Products” http://www.economics.noaa.gov/?goal=weather&file=events/drought&view=costs, 2011.
[4] Intergovernmental Panel on Climate Change, “Climate Change 2007: Synthesis Report,” IPCC, Geneva, Switzerland, 2007.
[5] Ropelewski, C. F., and M. S. Halpert, “Global and Regional Scale Precipitation Patterns Associated with the El Niño/Southern Oscillation,” Monthly Weather Review, vol. 115, 1987, pp. 1606-1626.
[6] Torrence, C., and P. J. Webster, “Interdecadal Changes in the ENSO–Monsoon System,” J. Climate, vol. 12, 1999, pp. 2679–2690.
[7] Schubert, S., and 32 co-authors, “A U.S. CLIVAR Project to Assess and Compare the Responses of Global Climate Models to Drought-Related SST Forcing Patterns: Overview and Results,” J. Climate, vol. 22, 2009, pp. 5251–5272.
[8] Fischer, E. M., S. I. Seneviratne, D. Lüthi, and C. Schär, “Contribution of land-atmosphere coupling to recent European summer heat waves,” Geophys. Res. Lett., vol. 34, 2007, L06707, doi:10.1029/2006GL029068.
[9] Fennessy, M. J., and J. Shukla, “Impact of initial soil wetness on seasonal atmospheric prediction,” J. Climate, vol. 12, 1999, pp. 3167-3180.
[10] Beljaars, A. C., P. Viterbo, M. J. Miller, and A. K. Betts, “The anomalous rainfall over the United States during July 1993: Sensitivity to land surface parameterization and soil moisture anomalies,” Mon. Wea. Rev., vol. 124, 1996, pp. 362-383.
[11] Dirmeyer, P. A., and J. L. Kinter III, “The Maya Express – Late spring floods in the US Midwest,” Eos – Transactions of the American Geophysical Union, vol. 90, 2009, pp. 101-102.
[12]Karl, T. R., and R. W. Knight, “Secular Trends of Precipitation Amount, Frequency, and Intensity in the United States,” Bull. Amer. Meteor. Soc., vol. 79, 1998, pp. 231–241.
[13] Groisman, P. Ya., R. W. Knight, D. R. Easterling, T. R. Karl, G. C. Hegerl, and V. N. Razuvaev. “Trends in Intense Precipitation in the Climate Record,” J. Climate, vol. 18, 2005, pp. 1326-1350.
[14] Min, S.-K., X. Zhang, F. W. Zwiers and G. C. Hegerl, “Human contribution to more-intense precipitation extremes,” Nature, vol. 470, 2011, 378–381.
[15] Dai, A., K. E. Trenberth, T. Qian, “A Global Dataset of Palmer Drought Severity Index for 1870–2002: Relationship with Soil Moisture and Effects of Surface Warming,” J. Hydrometeor, vol. 5, 2004, pp. 1117–1130.
[16] Barnett, T. P., J. C. Adam, and D. P. Lettenmaier, “Potential impacts of a warming climate on water availability in snow-dominated regions,” Nature, vol. 438, 2005, pp. 303-309.
[17] Meehl, G.A., T.F. Stocker, W.D. Collins, P. Friedlingstein, A.T. Gaye, J.M. Gregory, A. Kitoh, R. Knutti, J.M. Murphy, A. Noda, S.C.B. Raper, I.G. Watterson, A.J. Weaver and Z.-C. Zhao “Global Climate Projections” In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2007, pp. 747-846.
[18] European Centre for Medium-range Weather Forecasts, “IFS Documentation – Cy33r1: Operational Implementation 3 June 2008,” http://www.ecmwf.int/research/ifsdocs/CY33r1/index.html, 2009.
[19] Kinter III, J. L., D. Achuthavarier, J. Adams, E. Altshuler, B. Cash, P. Dirmeyer, B. Huang, E. Jin, L. Marx, J. Manganello, C. Stan, T. Wakefield, T. Palmer, M. Hamrud, T. Jung, M. Miller, P. Towers, N. Wedi, M. Satoh, H. Tomita, C. Kodama, Y. Yamada, P. Andrews, T. Baer, M. Ezell, C. Halloy, D. John, B. Loftis, and K. Wong, “Revolutionizing climate modeling – Project Athena: A multi-institutional, international collaboration,” Bull. Amer. Meteor. Soc., 2011, submitted.
[20] Koster, R. D., P. A. Dirmeyer, Z. Guo, G. Bonan, E. Chan, P. Cox, H. Davies, T. Gordon, S. Kanae, E. Kowalczyk, D. Lawrence, P. Liu, S. Lu, S. Malyshev, B. McAvaney, K. Mitchell, T. Oki, K. Oleson, A. Pitman, Y. Sud, C. Taylor, D. Verseghy, R. Vasic, Y. Xue, and T. Yamada, “Regions of strong coupling between soil moisture and precipitation,” Science, vol. 305, 2004, pp. 1138-1140.