Special Section on Climate Change and Water Resources: Climate Nonstationarity and Water Resources Management

Abstract

Over the past three decades, hydrologists and water resources specialists have been concerned with the issue of nonstationarity arising from several factors. First is the effect of human intervention on the landscape that may cause changes in the precipitation–runoff relationships at various temporal and spatial scales. Second is the occurrence of natural events such as volcanic explosions or forest fires that may cause changes in the composition of the air, the soil surface, and geomorphology. Third is the low-frequency component of oceanic–atmospheric phenomena that may have significant effects on the variability of hydrological processes such as annual runoff, peak flows, and droughts. Fourth is global warming, which may cause changes to oceanic and atmospheric processes, thereby affecting the hydrological cycle at various temporal and spatial scales. There has been a significant amount of literature on the subject and thousands of research and project articles and books published in recent decades. Examples of human intrusion on the landscape are the changes in land use resulting from agricultural developments in semiarid and arid lands (e.g., Pielke et al. 2007, 2011), changes caused by large-scale deforestation (e.g., Gash and Nobre 1997), changes resulting from open-pit mining operations (e.g., Salas et al. 2008), and changes from increasing urbanization in watersheds (e.g., Konrad and Booth 2002, Villarini et al. 2009). These intrusions change hydrologic response characteristics such as the magnitude and timing of floods. In many situations, current systems and management practices will be ill equipped to cope with such changes unless adjustments are made. Large-scale landscape changes such as deforestation in the tropical regions can potentially alter atmospheric circulation patterns, and consequently affect global weather and climate (e.g., Lee et al. 2008, 2009). Major natural events, such as the volcanic explosion of Mount St. Helens in 1980 or the El Chichon volcanic explosion of 1982 induce a shock to the climate system in the form of global cooling that continues for several years. These events can also affect global circulation. Low-frequency climate drivers of the oceanic– atmospheric system such as the El Nino/Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), Atlantic Multidecadal Oscillation (AMO), and Arctic Oscillation (AO) modulate global climate at interannual and multidecadal time scales. These drivers are the main sources of nonstationarity in global climate and hydrology. Large numbers of papers documenting the effect of these drivers on global hydroclimatology continue to emerge (e.g., Dilley and Heyman 1995; Mantua et al. 1997; Enfield et al. 2001; Akintug and Rasmussen 2005; Hamlet et al. 2005). In addition to climate variability and change due to the previously mentioned factors, anthropogenic warming of the oceans and atmosphere because of increased greenhouse gas concentrations and the ensuing changes to the hydrologic cycle are topics of serious pursuit. The international scientific community is making strides in understanding the potential warming and its effects on all aspects of climate variability [Intergovernmental Panel on Climate Change (IPCC) 2007], but the impacts on the hydrologic cycle remain debatable and inconclusive (e.g., Cohn and Lins 2005; Legates et al. 2005; Hirsch and Ryberg 2011). Based on analyses of the global mean CO2 (GMCO2) and annual flood records in the United States, no strong statistical evidence for flood magnitudes increasing with GMCO2 increases were found (Hirsch and Ryberg 2011). Although general circulation models have had success in the attribution of warming global temperatures to anthropogenic causes, their credibility and utility in reproducing variables that are relevant to hydrology and water resources applications is less clear. For example, the IPCC Report for Latin America acknowledges that “the current GCMs do not produce projections of changes in the hydrological cycle at regional scales with confidence. In particular the uncertainty of projections of precipitation remain high : : :That is a great limiting factor to the practical use of such projections for guiding active adaptation or mitigation policies” (Magrin et al. 2007; Boulanger et al. 2007). A variety of methods exist that address the concern of nonstationarity in hydrological processes and the topic remains an active research area. For example, in watersheds in which increasing urbanization has been documented causing significant effects in the flood response and magnitude, watershed modeling has been utilized to estimate the possible changes in the flood frequency and magnitude. Frequency analysis methods also have been applied when the parameters (or the moments such as the mean and variance) of a given model (e.g., the Gumbel model) may vary with time (e.g., Strupczewski et al. 2001; Clarke 2002). In addition, the role that low-frequency components of the oceanic– atmospheric system (represented, for example, by large-scale oscillations such as ENSO, PDO, and AMO) have on extreme events such as floods has been recognized. These large-scale forcing factors have been shown to exert in-phase and out-of-phase oscillations in the magnitude of floods, mean flows, and droughts