3.4.2 Groundwater
The demand for groundwater is likely to increase in the future, the main reason being increased water use globally. Another reason may be the need to offset declining surface water availability due to increasing precipitation variability in general and reduced summer low flows in snow-dominated basins (see Section 3.4.3).
Climate change will affect groundwater recharge rates, i.e., the renewable groundwater resource, and groundwater levels. However, even knowledge of current recharge and levels in both developed and developing countries is poor. There has been very little research on the impact of climate change on groundwater, including the question of how climate change will affect the relationship between surface waters and aquifers that are hydraulically connected (Alley, 2001). Under certain circumstances (good hydraulic connection of river and aquifer, low groundwater recharge rates), changes in river level influence groundwater levels much more than changes in groundwater recharge (Allen et al., 2003). As a result of climate change, in many aquifers of the world the spring recharge shifts towards winter, and summer recharge declines. In high latitudes, thawing of permafrost will cause changes in groundwater level and quality. Climate change may lead to vegetation changes which also affect groundwater recharge. Also, with increased frequency and magnitude of floods, groundwater recharge may increase, in particular in semi-arid and arid areas where heavy rainfalls and floods are the major sources of groundwater recharge. Bedrock aquifers in semi-arid regions are replenished by direct infiltration of precipitation into fractures and dissolution channels, and alluvial aquifers are mainly recharged by floods (Al-Sefry et al., 2004). Accordingly, an assessment of climate change impact on groundwater recharge should include the effects of changed precipitation variability and inundation areas (Khiyami et al., 2005).
According to the results of a global hydrological model, groundwater recharge (when averaged globally) increases less than total runoff (Döll and Flörke, 2005). While total runoff (groundwater recharge plus fast surface and sub-surface runoff) was computed to increase by 9% between the reference climate normal 1961 to 1990 and the 2050s (for the ECHAM4 interpretation of the SRES A2 scenario), groundwater recharge increases by only 2%. For the four climate scenarios investigated, computed groundwater recharge decreases dramatically by more than 70% in north-eastern Brazil, south-west Africa and along the southern rim of the Mediterranean Sea (Figure 3.5). In these areas of decreasing total runoff, the percentage decrease of groundwater recharge is higher than that of total runoff, which is due to the model assumption that in semi-arid areas groundwater recharge only occurs if daily precipitation exceeds a certain threshold. However, increased variability of daily precipitation was not taken into account in this study. Regions with groundwater recharge increases of more than 30% by the 2050s include the Sahel, the Near East, northern China, Siberia, and the western USA. Although rising watertables in dry areas are usually beneficial, they might cause problems, e.g., in towns or agricultural areas (soil salinisation, wet soils). A comparison of the four scenarios in Figure 3.5 shows that lower emissions do not lead to significant changes in groundwater recharge, and that in some regions, e.g., Spain and Australia, the differences due to the two climate models are larger than the differences due to the two emissions scenarios.
The few studies of climate impacts on groundwater for various aquifers show very site-specific results. Future decreases of groundwater recharge and groundwater levels were projected for various climate scenarios which predict less summer and more winter precipitation, using a coupled groundwater and soil model for a groundwater basin in Belgium (Brouyere et al., 2004). The impacts of climate change on a chalk aquifer in eastern England appear to be similar. In summer, groundwater recharge and streamflow are projected to decrease by as much as 50%, potentially leading to water quality problems and groundwater withdrawal restrictions (Eckhardt and Ulbrich, 2003). Based on a historical analysis of precipitation, temperature and groundwater levels in a confined chalk aquifer in southern Canada, the correlation of groundwater levels with precipitation was found to be stronger than the correlation with temperature. However, with increasing temperature, the sensitivity of groundwater levels to temperature increases (Chen et al., 2004), particularly where the confining layer is thin. In higher latitudes, the sensitivity of groundwater and runoff to increasing temperature is greater because of increasing biomass and leaf area index (improved growth conditions and increased evapotranspiration). For an unconfined aquifer located in humid north-eastern USA, climate change was computed to lead by 2030 and 2100 to a variety of impacts on groundwater recharge and levels, wetlands, water supply potential, and low flows, the sign and magnitude of which strongly depend on the climate model used to compute the groundwater model input (Kirshen, 2002).
Climate change is likely to have a strong impact on saltwater intrusion into aquifers as well as on the salinisation of groundwater due to increased evapotranspiration. Sea level rise leads to intrusion of saline water into the fresh groundwater in coastal aquifers and thus adversely affects groundwater resources. For two small, flat coral islands off the coast of India, the thickness of the freshwater lens was computed to decrease from 25 m to 10 m and from 36 m to 28 m for a sea-level rise of only 0.1 m (Bobba et al., 2000). Any decrease in groundwater recharge will exacerbate the effect of sea-level rise. In inland aquifers, a decrease in groundwater recharge can lead to saltwater intrusion of neighbouring saline aquifers (Chen et al., 2004), and increased evapotranspiration in semi-arid and arid regions may lead to the salinisation of shallow aquifers.