Analyses of stream flow records for the United States over the past several decades show primarily increases, which is consistent with trends in precipitation (Lins and Slack, 2005). However, these observed changes in stream flow are due in large part to the aggregate effects of many human influences, of which climate change is only one (Gerten et al., 2008). Of the world's 200 largest rivers, 22.5 percent showed downward trends over the period 1948 to 2004, and 9.5 percent showed upward trends, both mostly as a result of climate variations (Dai et al., 2009). While projections of runoff changes generally mimic precipitation trends, such projections are uncertain in part because runoff is influenced by rates of evapotranspiration—the sum of evaporation of water from the surface and transpiration of water though the leaves of plants. The effects of temperature change and changes in CO2 on plant processes can in turn affect evapotranspiration, and thus the magnitude of runoff (Gedney et al., 2006; Piao et al., 2007; Wolock and Hornberger, 1991).
Extreme conditions, namely floods and droughts, are generally of greatest concern to water managers. In addition to climate change, these events and can be magnified by human-influenced factors such as urbanization, streambed alterations, and deforestation. It is not clear whether the frequency of extreme runoff events has increased during the last several decades. Milly et al. (2002) reported a measurable increase in large floods, but Kundzewicz et al. (2005) found 27 increases, 31 decreases, and 137 with no significant trend in 195 catchments worldwide. These differences reflect both the regional nature of precipitation shifts as well as the multiple changes occurring in any individual region. For example, catchment-specific land use changes and streambed modifications may have occurred over the period of record and may mask or enhance the climate change signal. Such challenges suggest that adaptive water management decisions will require regional climate information and may differ in their specific application from one river basin to another. Given the observed increases in heavy precipitation events and the expectation that this intensification will continue, assessments indicate that generally, the risk from floods will increase in the future. However, local water, land use, and flood risk-management decisions can modify the actual flood vulnerability of communities and built infrastructure (Kundzewicz et al., 2007). Flood-control measures themselves can be a primary reason for changes in intensity of flooding (Pinter et al., 2008).
Long-term records do not exist for evapotranspiration. Trends in pan evaporation, a standard measurement of water loss to the atmosphere from an exposed pan of water at some meteorological stations, are actually negative for the past several decades in the United States (Golubev et al., 2001), which is the opposite of what would be expected under a warming climate. Several explanations are possible. Brutsaert and Parlange (1998) argue that pan evaporation reflects potential rather than actual evapotranspiration and that actual evapotranspiration and pan evaporation should have opposite signs due to feedbacks caused by the heat transferred during the transformation of water from liquid to vapor. An alternate explanation is that net surface radiation actually decreased in the United States during the past several decades due to increased cloudiness, and hence actual evapotranspiration decreased (Huntington, 2004). Discerning trends and making projections for evapotranspiration is complicated further by the indirect effect of increased CO2 concentrations, which can alter plants' water-use efficiency (Betts et al., 2007). Thus, although evapotranspiration is a critically important process in the water cycle, our ability to understand trends and to predict the impacts of climate change on it is limited (Fu et al., 2009; Kingston et al., 2009).
Was this article helpful?