Water resources

We examined four studies that assessed the potential impacts of climate change on water resources: Alcamo et al. (1997), Arnell (1999), Vorosmarty et al. (2000), and Doll (2002).

Arnell (1999) used a macro-scale hydrological model to simulate river flows across the globe, and then calculated changes in national water resource availability. These changes were then used with projections of future national water resource use to estimate the global effects of climate change on water stress, and to estimate the number of people living in countries that experience water stress or in counties that experience a change in water stress. Vorosmarty et al. (2000) used a water balance model that is forced offline with GCM output to estimate the number of people experiencing water stress. Alcamo et al. (1997) used a global water model that computes water use and availability in each of 1,162 watersheds, taking into account socio-economic factors that lead to domestic, industrial, and agricultural water use as well as physical factors that determine supply (runoff and ground water recharge). Some aspects of the model's design and data came from the IMAGE integrated model of global environmental change (Alcamo et al., 1994). The study relied on two GCMs for physical and climatic input. Alcamo et al. (1997) estimated the scarcity of water by means of a criticality index, which combines the criticality ratio (ratio of water use to water availability) and water availability per capita in a single indicator of water vulnerability. Doll (2002) used a global model of irrigation requirements, reporting changes in net irrigation requirements. Net irrigation was computed as a function of climate and crop type, with climatic input generated by two transient climate models.

The results from the water studies are far less consistent and conclusive than those of other sectors. Figure 3, based on Arnell's (1999) results, indicates the changes in the number of people living in countries experiencing water stress with increasing temperature. Arguably, it is impacts to this category of people that are most important. However, establishing what constitutes water stress is ultimately a rather subjective step. Nevertheless, there is not much change in water stress by this measure between the 2020s and the 2080s (increases in GMT of roughly 1°C and 3°C, respectively). As might be expected, the relatively wetter HadCM2 model predicts fewer people living in water stressed conditions. Figure 2e also shows the difference between the total population of countries where stress increases and the total population of countries where stress decreases. This measure gives a better sense of the total number of winners versus losers (though one could argue that the gains of winners do not really offset the losses of losers) with regard to changes in water stress, regardless of arbitrary thresholds. The trend is still ambiguous, since one model predicts net loss (HadCM2) and another predicts net gain (HadCM3). Counter to what one might expect, it is the drier model (HadCM3) that predicts a larger population of people in countries where water stress decreases. This is driven mainly by the fact that in the HadCM2 scenario, stress increases in the populous countries of India and Pakistan, while in the HadCM3 scenario, stress decreases in these countries. In both figures, the results are sensitive to large countries flipping from one situation to another. Regionally, the countries where climate change has the greatest adverse impact on water resource stress are located around the Mediterranean, in the Middle East, and in southern Africa. Significantly, these countries are generally least able to cope with changing resource pressures. Overall, these results indicate the importance of the regional distribution of precipitation changes to estimates of water resource impacts.

Vorosmarty et al.'s (2000) results indicate that climate change has little effect globally on water resource pressure. The effects of increased water demand due to population and economic growth eclipse changes due to climate. Here again it is important to note regional changes, which are masked by global aggregates. Vorosmarty et al. predicted significant water stress for parts of Africa and South America. This is offset by estimated decreases in water stress resulting from climate change in Europe and North America. In general, climate change produces a mixture of responses, both positive and negative, that are highly specific to individual regions. Of course, there is only a limited amount of climate change by 2025, the date at which the Vorosmarty et al. analysis ended.

Alcamo et al. (1997) presented results that highlight the impact of climate change on future water scarcity for only one point in time, 2075, and for one of the two GCMs that the study employed. The study suggested that, globally, overall annual runoff increases and water scarcity is somewhat less severe under climate change. In a world without climate change, 74% of the world's population is projected to live in water scarce watersheds by 2075. However, with climate change, this figure is reduced to 69%. These results are consistent with those of Vorosmarty et al. (2000), suggesting that climate change is not the most important driver of future water scarcity. Growth in water use due to population and economic growth is the decisive factor. Though Alcamo et al. (1997) suggested that climate change may ameliorate water scarcity globally, regionally the picture is quite different. Some 25% of the earth's land area experiences a decrease in runoff in the best guess scenario (which combines moderate estimates of future intensity and efficiency of water use) according to Alcamo et al. (1997), and some of this decrease is estimated to occur in countries that are currently facing severe water scarcity. The Alcamo et al. (1997) results also point to the possibility that industry will supersede agriculture as the world's largest user of water.

Figure 3. Water resources

Impacts on water resources as a function of temperature

1500 i 1000 -500

Source: Arnell, 1999. In both cases results are shown as averages for the decades of the 2020s, 2050s, and 2080s.

Note: Figure 3 shows two measures of the impact of climate change on users of water resources, both derived from Arnell (1999). Data represented by an "x" are changes in the number of people in countries using more than 20% of their water resources. This measure focuses on impacts on those people who live in or near a state of water stress. Data represented by a solid square are the difference between the total population in countries where water stress increases and countries where water stress decreases. This measure looks at the number of winners versus losers regardless of the whether they live in a state of water stress or not.

Doll's (2002) results mirror those of Vorosmarty et al. When cell-specific net irrigation requirements are summed over world regions, increases and decreases of cell values caused by climate change average out. Irrigation requirements, however, increase in 11 out of 17 of the world's regions by the 2020s, but not by more than 10%. By the 2070s, increases occur in 12 of these regions, 10 of which also show an increase in the 2020s.

The relationship between water resources and climate change appears to be inconclusive. A clear trend did not appear in the studies, perhaps because of the methods used and because of inconsistent changes in regional precipitation patterns across the climate models. Averaging world regions or even countries presents many problems. The water basin is the critical unit for analysis of water resources. Changes in one part of a basin, such as increased or decreased runoff, will affect other parts of the basin. Such changes have little effect outside the basin unless one basin feeds into another or is connected to another via water transport infrastructure. Since basins and transport infrastructure do not necessarily conform to national borders, an analysis based on estimating a uniform change for individual countries may not capture realistic impacts on water resources.

A second critical reason why we do not see a clear relationship between increases in GMT and effects on water resources appears to be inconsistent estimates of changes in regional precipitation. An increase in would increase global mean precipitation. However, the nature of regional changes in precipitation is quite uncertain and varies considerably across climate models. Differences in precipitation patterns from one climate model to another are probably more important than differences in mean temperature in terms of effect on estimates of impacts on water resources. Beyond this, the impacts on water resources are extremely complicated and can depend on such factors as how water is consumed, the ability to adjust uses, legal and institutional constraints, and the capacity to build or modify infrastructure.

Nevertheless, an argument can be made that adverse impacts to the water resources sector will probably increase with higher magnitudes of climate change.18 This argument is based on two considerations. One is that water resource infrastructure and management are optimised for current climate. The more future climate diverges from current conditions, the more likely it is that thresholds related to flood protection or drought tolerance will be exceeded with more frequency and with greater magnitude than they currently are. The second consideration is that more severe floods and droughts are expected to accompany higher magnitudes of climate change. Some regions might benefit from a more hydrologically favourable climate, but it seems unlikely that the majority of the world's population would see improved conditions, especially since systems are optimised for current climate.

Continue reading here: Human health

Was this article helpful?

0 0