Kevin Trenberth from the National Center for Atmospheric Research began the workshop with a talk on how precipitation is expected to change in a warming climate and on the link between these changes and extreme hydrologic events. He discussed the physical basis for concurrent increases during extreme high precipitation and longer durations of dry periods, yet he noted that global and regional climate models are demonstrably poor at most aspects of forecasting the hydrological cycle. Pasha Groisman of the National Oceanic Atmospheric and Administration highlighted his past and ongoing observational studies (as well as findings by other foreign researchers) that strengthen the opinion that in the past several decades over most of the extratropics, or mid-latitude regions, precipitation has become more intense; however, he also acknowledged that the data presented could be analyzed in many ways. The presenters discussed the various approaches to categorizing the data and impacts that might have on the resulting conclusions. Richard Seager of Columbia University agreed with Groisman by articulating that it is a robust prediction of state-of-the-art climate models that greenhouse gas-induced global warming will cause the wet regions of the planet (in the deep tropics and the mid to high latitudes) to get wetter while the subtropical dry zones get drier. When recent climate observations are reviewed for evidence of these changes, trends are found to have been dominated by large-amplitude natural decadal atmosphere-ocean variability. He concluded that near-term hydroclimate prediction must account for both anthropogenic change and the evolution of natural modes of variability, but he noted that all models still have significant room for improvement. Seager presented an example of an ensemble analysis in which the underlying models were equally divided between wetter and drier results.
Aquifer-atmosphere interactions can be important in landscapes where the water table is shallow (<2m) and the watershed topography is gentle, stated Mark Person of New Mexico Tech. He found that it may soon become possible to incorporate physically based representations of aquifer hydrodynamics into Global Climate Models (GCMs) given recent advancements with supercomputers. This integration may help to improve predictions of the long-term consequences of droughts on water resources and climate dynamics. Siegfried Schubert of NASA’s Goddard Space Flight Center offered thoughts on the nature of drought—he argued that the possibility of predicting long-term drought rests largely on the strength of sea-surface temperature linkages to the land component of the hydrological cycle, and on the ability to predict sea-surface temperature changes.
Tom Huntington from the U.S. Geological Survey assessed the published record of trend analysis in various components of the hydrologic cycle and associated variables. Huntington concluded that the evidence is mixed for an increase in the frequency, intensity, and duration of extreme weather events like hurricanes and floods. He noted that intensification of extreme
weather events may be related to great recycling of water through the hydrologic cycle and may not be directly proportional to more extreme hydrologic events. Harry Lins of the U.S. Geological Survey posed the question of what the likely effects of an accelerated hydrological cycle might be on streamflow in general, and on floods in particular. Data and published literature indicate that the relative precipitation sensitivity (elasticity) of mean streamflow with respect to precipitation is much greater than that of peak streamflow, and that precipitation sensitivity decreases as flood return period increases. Hence, while flood peaks are quite likely to increase if precipitation increases, their fractional change relative to a given fractional change in the mean precipitation is less than the fractional increase in the mean flow.
Richard Vogel of Tufts University noted that multiple sources of uncertainty and non-stationarity are now inherent in nearly all water-resource planning problems, and it is important that the water resources engineering community shift away from the stationarity paradigm on which it has presumed for many decades. He described work that has analyzed several thousand flood peak records from across the United States. Although not yet definitive, his work suggests that, where non-stationarity is evident, it is more likely to be associated with changing land use (especially urbanization) than with climate change. Katie Hirschboeck of the University of Arizona argued for the necessity of moving beyond conventional methods for estimating the frequency of extreme hydrologic events. She described an approach based on parameterization of spatially and temporally varying hydroclimatic extremes (which she called synoptic hydroclimatology) as a starting place for making operationally useful decisions about the impacts of climate change on hydrologic extremes.
Gerald Galloway of the University of Maryland discussed the nature of guidance that can be given today to U.S. water management planners to deal with an uncertain hydrologic future. He acknowledged that floods are acts of nature, and flood consequences are a result of man, so while flood risk calculations are not precise, risk assessment provides insights and a basis for prioritization. U.S. water managers need to deal with present problems by using the Precautionary Principle in future planning with a newly developed national policy. Michael Hayes, of the University of Nebraska, Lincoln, spoke about the status of drought risk management in the United States. He highlighted several key issues that should be considered including the facts that drought is a local issue, monitoring is essential, mitigation and planning require innovative ideas, worse-case scenarios should be considered, and communication between scientists and the public is key.