Droughts are considerably more difficult to define in quantitative terms than floods. Droughts may be expressed in terms of rainfall deficits, soil moisture deficits, lack of flow in a river, low groundwater levels, or low reservoir levels; different definitions are used in different sectors. A hydrological drought occurs when river or groundwater levels are low, and a water resources drought occurs when low river, groundwater, or reservoir levels impact water use. Low river flows in summer may not necessarily create a water resources drought, for example, if reservoirs are full after winter; conversely, a short-lived summer flood may not end a water resources drought caused by a prolonged lack of reservoir inflows. Water resources droughts therefore depend not only on the climatic and hydrological inputs but critically on the characteristics of the water resource system and how droughts are managed. This section focuses on hydrological drought, particularly on low river flows. Different studies have used different indices of low river flows, including the magnitude of minimum flows, the frequency at which flows fall below some threshold, the duration of flow below a threshold, and the cumulative difference between actual flows and some defined threshold.
At the global scale, Arnell (1999b) explored the change in the minimum annual total runoff with a return period of 10 years under several scenarios, based on HadCM2 and HadCM3 GCMs. He shows that the pattern of this measure of low flow (which is relatively crude) changes in a similar way to average annual runoff (as shown in Figure 4-1) but that the percentage changes tend to be larger. Arnell (1999a) mapped a different index of low flow across Europethe average summed difference between streamflow and the flow exceeded 95% of the time, while flows are below this thresholdunder four scenarios. The results suggest a reduction in the magnitude of low flows under most scenarios across much of western Europe, as a result of lower flows during summer, but an amelioration of low flows in the east because of increased winter flows. In these regions, however, the season of lowest flows tends to shift from the current winter low-flow season toward summer.
Döll et al. (1999) also modeled global runoff at a spatial resolution of 0.5°x0.5°, not only for average climatic conditions but also for typical dry years. The annual runoff exceeded in 9 years out of 10 (the 10-year return period drought runoff) was derived for each of more than 1,000 river basins covering the whole globe. Then the impact of climate change on these runoff values was computed by scaling observed temperature and precipitation in the 1-in-10 dry years with climate scenarios of two different GCMs (Chapter 3), ECHAM4/OPYC3 and GFDL-R15. Climate variability was assumed to remain constant. For the same GHG emission scenario, IS92a, the two GCMs compute quite different temperature and more so precipitation changes. With the GFDL scenario, runoff in 2025 and 2075 is simulated to be higher in most river basins than with the ECHAM scenario. The 1-in-10 dry year runoff is computed to decrease between the present time (19611990 climate) and 2075 by more than 10% on 19% (ECHAM) or 13% (GFDL) of the global land area (Table 4-4) and to increase by more than 50% on 22% (ECHAM) or 49% (GFDL) of the global land area. These results underline the high sensitivity of computed future runoff changes to GCM calculations.
There have been several other studies into changes in low flow indicators at the catchment scale. Gellens and Roulin (1998), for example, simulated changes in low flows in several Belgian catchments under a range of GCM-based scenarios. They show how the same scenario could produce rather different changes in different catchments, depending largely on the catchment geological conditions. Catchments with large amounts of groundwater storage tend to have higher summer flows under the climate change scenarios considered because additional winter rainfall tends to lead to greater groundwater recharge (the extra rainfall offsets the shorter recharge season). Low flows in catchments with little storage tend to be reduced because these catchments do not feel the benefits of increased winter recharge. Arnell and Reynard (1996) found similar results in the UK. The effect of climate change on low flow magnitudes and frequency therefore can be considered to be very significantly affected by catchment geology (and, indeed, storage capacity in general). Dvorak et al. (1997) also showed how changes in low flow measures tend to be proportionately greater than changes in annual, seasonal, or monthly flows.
Other reports in this collection