3.4.3 Floods and droughts
A warmer climate, with its increased climate variability, will increase the risk of both floods and droughts (Wetherald and Manabe, 2002; Table SPM2 in IPCC, 2007). As there are a number of climatic and non-climatic drivers influencing flood and drought impacts, the realisation of risks depends on several factors. Floods include river floods, flash floods, urban floods and sewer floods, and can be caused by intense and/or long-lasting precipitation, snowmelt, dam break, or reduced conveyance due to ice jams or landslides. Floods depend on precipitation intensity, volume, timing, antecedent conditions of rivers and their drainage basins (e.g., presence of snow and ice, soil character, wetness, urbanisation, and existence of dikes, dams, or reservoirs). Human encroachment into flood plains and lack of flood response plans increase the damage potential.
The term drought may refer to meteorological drought (precipitation well below average), hydrological drought (low river flows and water levels in rivers, lakes and groundwater), agricultural drought (low soil moisture), and environmental drought (a combination of the above). The socio-economic impacts of droughts may arise from the interaction between natural conditions and human factors, such as changes in land use and land cover, water demand and use. Excessive water withdrawals can exacerbate the impact of drought.
A robust result, consistent across climate model projections, is that higher precipitation extremes in warmer climates are very likely to occur (see Section 3.3.1). Precipitation intensity increases almost everywhere, but particularly at mid- and high latitudes where mean precipitation also increases (Meehl et al., 2005, WGI AR4, Chapter 10, Section 10.3.6.1). This directly affects the risk of flash flooding and urban flooding. Storm drainage systems have to be adapted to accommodate increasing rainfall intensity resulting from climate change (Waters et al., 2003). An increase of droughts over low latitudes and mid-latitude continental interiors in summer is likely (WGI AR4, Summary for Policymakers, Table SPM.2), but sensitive to model land-surface formulation. Projections for the 2090s made by Burke et al. (2006), using the HadCM3 GCM and the SRES A2 scenario, show regions of strong wetting and drying with a net overall global drying trend. For example, the proportion of the land surface in extreme drought, globally, is predicted to increase by the a factor of 10 to 30; from 1-3 % for the present day to 30% by the 2090s. The number of extreme drought events per 100 years and mean drought duration are likely to increase by factors of two and six, respectively, by the 2090s (Burke et al., 2006). A decrease in summer precipitation in southern Europe, accompanied by rising temperatures, which enhance evaporative demand, would inevitably lead to reduced summer soil moisture (Douville et al., 2002) and more frequent and more intense droughts.
As temperatures rise, the likelihood of precipitation falling as rain rather than snow increases, especially in areas with temperatures near to 0°C in autumn and spring (WGI AR4, Summary for Policymakers). Snowmelt is projected to be earlier and less abundant in the melt period, and this may lead to an increased risk of droughts in snowmelt-fed basins in summer and autumn, when demand is highest (Barnett et al., 2005).
With more than one-sixth of the Earth’s population relying on melt water from glaciers and seasonal snow packs for their water supply, the consequences of projected changes for future water availability, predicted with high confidence and already diagnosed in some regions, will be adverse and severe. Drought problems are projected for regions which depend heavily on glacial melt water for their main dry-season water supply (Barnett et al., 2005). In the Andes, glacial melt water supports river flow and water supply for tens of millions of people during the long dry season. Many small glaciers, e.g., in Bolivia, Ecuador, and Peru (Coudrain et al., 2005), will disappear within the next few decades, adversely affecting people and ecosystems. Rapid melting of glaciers can lead to flooding of rivers and to the formation of glacial melt-water lakes, which may pose a serious threat of outburst floods (Coudrain et al., 2005). The entire Hindu Kush-Himalaya ice mass has decreased in the last two decades. Hence, water supply in areas fed by glacial melt water from the Hindu Kush and Himalayas, on which hundreds of millions of people in China and India depend, will be negatively affected (Barnett et al., 2005).
Under the IPCC IS92a emissions scenario (IPCC, 1992), which is similar to the SRES A1 scenario, significant changes in flood or drought risk are expected in many parts of Europe (Lehner et al., 2005b). The regions most prone to a rise in flood frequencies are northern and north-eastern Europe, while southern and south-eastern Europe show significant increases in drought frequencies. This is the case for climate change as computed by both the ECHAM4 and HadCM3 GCMs. Both models agree in their estimates that by the 2070s, a 100-year drought of today’s magnitude would return, on average, more frequently than every 10 years in parts of Spain and Portugal, western France, the Vistula Basin in Poland, and western Turkey (Figure 3.6). Studies indicate a decrease in peak snowmelt floods by the 2080s in parts of the UK (Kay et al., 2006b) despite an overall increase in rainfall.
Results of a recent study (Reynard et al., 2004) show that estimates of future changes in flood frequency across the UK are now noticeably different than in earlier (pre-TAR) assessments, when increasing frequencies under all scenarios were projected. Depending on which GCM is used, and on the importance of snowmelt contribution and catchment characteristics and location, the impact of climate change on the flood regime (magnitude and frequency) can be both positive or negative, highlighting the uncertainty still remaining in climate change impacts (Reynard et al., 2004).
A sensitivity study by Cunderlik and Simonovic (2005) for a catchment in Ontario, Canada, projected a decrease in snowmelt-induced floods, while an increase in rain-induced floods is anticipated. The variability of annual maximum flow is projected to increase.
Palmer and Räisänen (2002) analysed GCM-modelled differences in winter precipitation between the control run and around the time of CO2 doubling. A considerable increase in the risk of a very wet winter in Europe and a very wet monsoon season in Asia was found. The probability of total boreal winter precipitation exceeding two standard deviations above normal is projected to increase considerably (even five- to seven-fold) over large areas of Europe, with likely consequences for winter flood hazard.
Milly et al. (2002) demonstrated that, for fifteen out of sixteen large basins worldwide, the control 100-year peak volumes (at the monthly time-scale) are projected to be exceeded more frequently as a result of CO2 quadrupling. In some areas, what is given as a 100-year flood now (in the control run), is projected to occur much more frequently, even every 2 to 5 years, albeit with a large uncertainty in these projections. Yet, in many temperate regions, the snowmelt contribution to spring floods is likely to decline on average (Zhang et al., 2005). Future changes in the joint probability of extremes have been considered, such as soil moisture and flood risk (Sivapalan et al., 2005), and fluvial flooding and tidal surge (Svensson and Jones, 2005).
Impacts of extremes on human welfare are likely to occur disproportionately in countries with low adaptation capacity (Manabe et al., 2004a). The flooded area in Bangladesh is projected to increase at least by 23-29% with a global temperature rise of 2°C (Mirza, 2003). Up to 20% of the world’s population live in river basins that are likely to be affected by increased flood hazard by the 2080s in the course of global warming (Kleinen and Petschel-Held, 2007).