Page 6
2—
Science Foundations and Basic Processes
Predictability and Variability of Regional and Global Water Cycles
The cycling of water in its various states—solid, liquid, and gaseous—is a primary process within the Earth's climate system. Information on variability of states and fluxes over time is crucial for the understanding of the sustainability of local, regional, national, and international economies and ecosystems. It is essential to establish rates of cycling, changes in these resulting from human intervention, and the consequences of those changes for regional water availability. It is important to estimate the magnitude of potential changes in water reservoirs at the land surface (e.g., lakes, seasonal snow-packs, soil moisture, groundwater, glaciers, and ice sheets), changes in fluxes of water (e.g., precipitation, evaporation, runoff, and groundwater recharge), and changes in atmospheric water storage and transport, all of which have profound influences on the Earth's energy cycle and global change processes. Toward this end, a better understanding is needed of what causes both short-term and long-term variability in these fluxes that couple the land surface reservoirs of water with each other as well as with the oceans and atmosphere. It is a priority scientific objective to establish how much of the variability in the water cycle is predictable over a range of time and space scales.
Large-scale seasonal-to-interannual oscillations in climate have been shown to contribute significantly to the total variability in precipitation and temperature in some regions. These predictable patterns of variability create excellent opportunities for improving long-lead hydrologic forecasts based on measurements of climate indicators such as sea-surface temperature or snow-cover patterns.
An understanding of mechanisms linking large-scale climate variability with regional conditions also forms the basis for reducing the uncertainty associated with assessing regional impacts of global change over decadal-to-centennial periods. A region-specific ability to project the consequences of global change is now required, for example, by decision-makers concerned with long-term fixed capital investments in infrastructure such as dams, water diversion systems, and flood damage mitigation systems that are vulnerable to shifts in hydroclimatic regime. It is also required by policy makers debating whether possible shifts in hydroclimatic regime warrant increased measures to reduce greenhouse gas emissions (USGCRP, 1999).
Page 7
The continental hydrologic cycle is the link between climate forcing at the large scale and surface impacts at the regional scale. Understanding how variabilities in regional hydrologic cycles are linked to large-scale climate is a research priority in both the hydrologic and climate sciences (NRC, 1998b).
Key challenges for hydrologic science are to define spatial and temporal regimes of hydrologic systems in which predictability is high and to characterize and understand the nature of variability in hydrologic systems. These two challenges are interrelated.
Predictability is the extent to which the future state of a system can be estimated based upon the (theoretical) availability of a comprehensive set of observations characterizing the system's initial condition. Once a hydrologic system has been shown to possess a substantial degree of predictability, useful prediction schemes can be devised to estimate the response of the system to external forcing (e.g., land use changes) or to variability of the climate system. The estimation of this response requires an in-depth understanding of (1) the relative contributions of local and remote forcing mechanisms to the total variability in hydrologic systems, (2) the changes in the response and the characteristics of hydrologic systems as they are monitored or modeled at different spatial and temporal resolutions, and (3) the nature of local and regional feedback mechanisms that affect the response of hydrologic systems to local and external forcing factors.
In this respect, the ability to make useful predictions—e.g., weather forecasting at a local scale (about 1–10 Km) for hazards mitigation, preparation of water supply outlooks at the basin scale, and water availability projections under global change at a regional scale (about 102 to 103 km)—requires that the cycling pathways, the storage of water, and the nature of variability for physical processes be measured and understood.
Predictable Patterns of Seasonal-to-Interannual Variability
Issues of predictability form the essence of the scientific challenges associated with seasonal-to-interannual climate variability. The land, biosphere, atmosphere, and oceans are coupled together in an Earth system that has a wide range of time and space scales in its variability. For example, there are "slow" (e.g., deep groundwater and oceans) and "fast" (e.g., atmospheric water vapor and surface moisture) components in this system, whose rate of reaction is controlled in part by reservoir size, by the intrinsic rates of the processes involved, and by interactions with other processes.
The superposed variations with a variety of time scales means that there is potentially some predictability if the "slow" components can be isolated and monitored. For example, coupled air-sea interactions across the large tropical Pacific basin result in large-scale changes in climate that appear with some regularity (quasiperiodic) on a 2- to 6-year time scale. This phenomenon, known as the El Niño-Southern Oscillation (ENSO), has strong implications for hydroclimatic anomalies across the tropics, and it also affects conditions in the extratropics (NRC, 1998b). If the ENSO signal is linked to regional precipitation and temperature, then seasons-ahead predictions of the phenomenon can provide a means of narrowing the uncertainty associated with long-lead hydrologic predictions. The implications of such predictability for managing water resources, agriculture, electric power, and other climate-sensitive sectors are far-reaching (see Box A). ENSO is just one example; there are other phenomena in the Earth system that cause variations on inter-annual and longer time scales. For example, feedback between the atmosphere and snow cover, land moisture, vegetation cover, and regional water bodies other than the Pacific also induces systematic oscillations and excursions in regional climate.
Page 8
Three key scientific questions in hydrologic science related to variability in the global water cycle and its regional seasonal-to-interannual predictability are as follows:
• The effects of large-scale seasonal-to-interannual oscillations in climate caused by interactions between land, oceans, and the atmosphere are apparent in many hydrologic records. The physical processes and cause-effect relations, however, are not understood well enough to support the design and implementation of forecasts that are sufficiently robust to be relied upon for regional water-related decision-making. To what extent is regional-scale hydrology predictable? What types and locations of measurements will most enhance predictions?
• Predictability of regional hydrologic systems is limited to large-scale climate predictability if forcing is unidirectional, i.e., large-scale climate affects regional hydrologic systems. If there is two-way land-atmosphere coupling or modulation of climate by local hydrologic processes, then there is potentially enhanced predictability gained through coupled hydrologic modeling. Across which regions and seasons can predictions of regional water cycling be enhanced by robust coupled land-atmosphere modeling?
• Hydrologic extremes (e.g., droughts and floods) may be influenced by large-scale climate variability and local land-atmosphere exchanges in ways that are quite distinct from how those influences affect regional hydroclimate under near-normal conditions. Current hydrologic forecasts (e.g., water supply outlooks or flow forecasts) are better at predicting near-normal. conditions than they are at predicting extremes. Yet in terms of economic and environmental impact, predicting extremes is of much greater regional importance. What special physical and statistical features (e.g., process pathways, influences across scales) can be used to link large-scale climate and regional-scale hydrology in the case of extreme events, and how are these features different for the case of floods and the case of persistent droughts?
Sources of Long-Term Variability
Variability in hydrologic records includes contributions from both natural variability and human-induced changes of the landscape and climate. There are at least two critical gaps that research must address to separate the total variability into natural and human-induced contributions. First, the causes and spatial patterns of natural variability from both measured and paleoclimate records need to be understood. For example, tree-ring records in the southwestern United States show evidence of prolonged droughts that greatly exceed the magnitude and duration of droughts in the measured record, principally during the past 100 years. The extent and seventy of these events within the region and the link to large-scale climate changes are not known. Second, it needs to be ensured that in future decades, scientists have the long-term hydrologic measurements needed to detect changes and that they have the understanding to link those changes to alterations in the landscape and in the broader-scale climate system. For example, observations are currently lacking to detect a change in the mass of the Earth's largest freshwater reservoirs, the polar ice sheets, to within 25 percent of the annual accumulation (Houghton et al., 1996). Further, the amounts and spatial patterns of groundwater recharge to critical water supply aquifers in the United States and other parts of the world have never been measured, in part because of a lack of understanding of how to make accurate measurements of recharge rates (Simmers, 1988).
Two key questions in hydrologic science related to understanding long-term sources of variability are the following:
Page 9
|
• Accurate and long-term measurements, both for understanding past variability and detecting future changes, are critical. From a water-resources standpoint, the long-term events of greatest concern are sustained droughts and increases in the magnitude and frequency of floods. However, this issue goes well beyond water resources, concerning essentially all of the Earth's hydrologic systems. What combination of remote and in situ observations and paleohydrologic records are required to identify shifts in regional and local hydrologic properties resulting from both natural and human-induced factors?
• Shifts in regional hydroclimate are linked to large-scale patterns that change over long time scales. Understanding linkages between regional and large-scale processes is essential to interpreting the record of past variability and to using that record as an analog of the expected shifts in regional water availability under global change conditions. Are there spatial patterns in the variability of the hydrologic record that may serve as reliable predictors of the impacts of global change?
Linking Measurements and Understanding across Scales
Each process contributing to a hydrologic response, such as evaporation, snowmelt, surface runoff, or subsurface flow, has its own set of characteristic spatial and temporal scales. For example, the spatial resolution at which atmospheric systems are modeled is from 100 to 10,000 times the scale of heterogeneities in land surface topography, vegetation, soil texture, or snow cover. Similarly, variables like temperature or soil moisture can change significantly in a matter of hours, whereas vegetation changes occur seasonally. Measurements and models need to be designed so that they can aggregate processes at disparate scales.
Page 10
Besides the variability in the original processes, some new scales of variability emerge when processes are coupled together. The critical challenge for hydrologic science is to understand and describe the ways in which heterogeneous processes interact with one another at different scales to produce the variability found in hydrologic systems. That understanding must then be translated into enhanced predictions. Two key questions follow:
• Advances in measurement technologies allow topographic, soil, vegetation and snow properties, and other parameters to be specified at finer and finer scales in predictive models. There are computational limitations, however, on the ability of hydroclimatic models to incorporate all processes and scales. Thus, a key question is to what extent the spatial structure and magnitude of the (fine-scale) variability in a particular parameter or variable is required for realistic modeling at the large scale (upscaling). The answer to this question depends on the scale at which a given process or variable interacts with other processes. For example, a prediction of snowmelt, runoff, and groundwater recharge over days to weeks may possibly benefit from a much finer description of soil and vegetation properties than would land surface feedback into a climate model for seasonal-to-interannual forecasts. For land-atmosphere modeling where the atmospheric boundary layer is the link in the exchanges, there is a natural spatial integration associated with turbulent mixing. At what scales and for which processes should the spatial structure of surface heterogeneity be incorporated into the upscaling strategy for hydrologic models?
• The coupling between the land, biosphere, and atmosphere imposes strong constraints on some hydrologic processes in some instances. For example, the diurnal cycle of surface moisture and energy flux at local scale is strongly affected by entrainment processes associated with the growth and collapse of the atmospheric boundary layer. Similarly, the seasonal cycle of these same fluxes at the regional scale is largely determined by the radiative and energy balance of the overlying atmosphere. Of the physical constraints that come about because of the coupling of water and energy cycles, which may be used to bound the estimates of local and regional hydrologic fluxes?
Coupling of Hydrologic Systems and Ecosystems through Chemical Cycles
The supply of water and its distribution over the landscape are primary determinants of nutrient inputs to ecosystems and to ecosystem productivity. Evaporation of water through ecosystems is a major control on the terrestrial branch of the hydrologic cycle. Cycles of water, energy, nutrients, and carbon have been identified consistently as priority cross-cutting themes in understanding environmental change (NRC, 1998e; USGCRP, 1999). Scientists currently lack the detailed knowledge of these coupled cycles required to make rational choices regarding the global issue of greenhouse gas emissions, and they lack detailed knowledge of the local to regional issues associated with human environmental disturbances such as land-use changes and their consequent perturbations to water quality and flow. To make the necessary environmental choices, which will have tremendous socioeconomic implications, a fundamental understanding is needed of how water exerts a controlling influence on ecosystems, and vice versa, with special emphasis on areas where human activity is having the greatest impact.
Page 11
Characterization of Water and Chemical Pathways
Ecosystems and humans are affected by and rely on discharges of water from groundwater and surface water systems. These waters carry chemicals, including nutrients needed for survival and contaminants such as organic solvents or pesticides. Adverse effects include the possibility of chemical accumulation (e.g., selenium enrichment and increases in basin-scale salinity). Changes in climate or land use that affect groundwater recharge can cause changes in groundwater storage and discharge that could have adverse effects on humans and ecosystems. To protect human welfare and the integrity of the ecosystems on which humans depend, understanding is needed concerning the pathways that water and chemicals follow as they move across the landscape and through the subsurface, as well as concerning the physical, chemical, and biological transformations that occur along these paths.
Research has demonstrated that water and chemicals are transported most rapidly along preferential pathways (see Box B). The location and configuration of these pathways are influenced by the arrangement of soils and geological units having high permeability, as well as by cracks and fractures. Delineation of pathways is complicated by difficulties involved in characterizing the land surface (e.g., mapping vegetation and microtopography) and the inability to see into the subsurface. Pathway patterns can be postulated through the use of numerical models that track imaginary particles, which represent water molecules and/or chemicals. Results of such modeling studies allow scientists to quantify not only the particle pathways themselves, but also travel times and fluxes. The accuracy of these models depends on the characterization of subsurface media and the specification of boundary conditions (e.g., hydrologic fluxes).
Understanding interactions and pathways of water and chemical exchange between surface and subsurface hydrologic systems is impeded because there are many gaps in knowledge (e.g., measurement and understanding of key flux variables such as evaporation and groundwater recharge). If human welfare and the integrity of the ecosystems on which humans depend are to be protected, it will be necessary to address a number of priority science questions including the following:
• Lack of knowledge of the detailed distribution of subsurface materials is impeding the ability to track and predict the movement of solutes, including contaminants. Subsurface imaging has been an important tool for oil exploration for many years, and it has been used in groundwater investigations as well, but improved techniques are needed for addressing problems in both the petroleum industry and for hydrogeological site characterization. Can geophysical techniques (e.g., ground-penetrating radar and nuclear magnetic resonance) be refined to provide ways of imaging the subsurface to provide information on the distribution of geologic units and preferential flow paths in a variety of complex geological settings?
• Groundwater recharge to the water table and groundwater discharge to oceans, lakes, wetlands, and rivers are important to ecosystems and to human survival. Yet scientists are still struggling to measure the spatial and temporal distribution of groundwater recharge and discharge. Can a general methodology be developed (e.g., using innovations in chemical, isotopic, and thermal measurements) to measure or otherwise estimate the spatial and temporal distribution of groundwater recharge and discharge (and fluxes of associated chemicals) over a basin? Can preferential paths of water and chemicals through the vadose zone be incorporated into basin-scale recharge theories?
• Addressing large-scale problems such as basin-scale recharge, subsurface flow paths, or salinity sources will require an understanding of a number of fundamental issues that have been largely neglected over the past two decades. It also will require combining several powerful approaches that have largely
Page 12
|
been used independently in the past. These include (1) traditional geologic mapping and conceptual methodologies, (2) subsurface imaging, (3) numerical modeling, and (4) innovations in tracer techniques. What combinations methods will enable identifying the important flow paths for water and solutes in the vadose zone, and in the saturated zone, at scales from smaller than a hillslope up to regional aquifers?
Interactions between Hydrologic Systems and Ecosystems
In aquatic ecosystems, residence times, nutrient fluxes, and many other factors are determined by the rate at which water moves through the hydrologic cycle, yet little is known about how most terrestrial
Page 13
and freshwater ecosystems and the plants and animals that make up the ecosystem respond to changes in the hydrologic cycle. Being able to predict with confidence responses to climate variability and to anthropogenic perturbations requires great improvement in understanding the interactions and feedback mechanisms between hydrologic systems and ecosystems. Several aspects of these interactions deserve special attention:
• The next level of improvement in estimating regional evaporation from plants is likely to result from a combination of new observations (e.g., remote sensing) and models of vegetation that include photosynthesis and the assimilation of nutrients. Can coupling between the cycling of nitrogen—a limiting factor for growth in many diverse aquatic and terrestrial ecosystems—and the cycling of water and carbon in models of land surface—atmosphere interactions improve the ability to estimate regional fluxes of water in terrestrial ecosystems?
• The water-ecosystem linkage may be strongest but least understood in riparian systems. Riparian systems are under great stress because of human activities. These systems harbor a large majority of the regional biodiversity. Together with their associated groundwater systems, riparian regions also sustain human habitation and agriculture. Furthermore, these terrain features convey a large fraction of the exchange between the subsurface, surface, and atmosphere. Reliable tools for managing riparian communities do not exist, largely because of uncertainties in (1) riparian plant-water-nutrient relations, (2) basin boundary conditions, (3) physical hydrologic processes, such as riparian evaporation, over large areas, and (4) hydrological flow paths and residence tunes. Can advances in basic knowledge of macro/micronutrient constraints vs. other physical/climatic constraints (e.g., water, energy) as controls on ecosystems, and especially new approaches for understanding and describing groundwater-surface water interactions and associated processes, provide the necessary understanding of riparian areas?
• Long-term resource management of aquatic ecosystems must be based on a better fundamental understanding of how biogeochemical processes respond to the combined effects of climate variability, e.g., changes in the flux of organic matter, acid deposition, mineral weathering and of changes in the hydrologic cycle related to climate warning. Lakes and streams are largely buffered against changes in acidity by mineral weathering, the rate of which depends on the supply of acids from both precipitation and a basin's terrestrial ecosystem. Can sufficient knowledge of biogeochemical cycles be developed (e.g., knowledge of long-term mineral-weathering rates and of coupled water, carbon and nitrogen cycling) at small catchment to lager basin scales to enable scientists to make confident, multidecadal forecasts of basin and ecosystem response to perturbations, particularly where future anthropogenic effects may perturb sensitive headwater catchments beyond what they have previously experienced?
Human Disturbances of Hydrologic Systems and Ecosystems
Human-induced perturbations now range from local to global in scale and may cause shifts that affect humans and ecosystems well beyond the source of the perturbation. Hydrologic shifts appear as changes in atmospheric water, precipitation, runoff, evaporation, groundwater recharge, or chemical cycles. A major challenge facing scientists lies in separating the effects of human influences from climate variability in the hydrologic record. Five issues of special concern regarding the hydrologic cycle, nutrient cycles, and ecosystems follow. The first two stem largely from global environmental issues; the latter three are related to land-use changes.
Page 14
• Climate warming in response to increases in greenhouse gas concentrations is expected to accelerate the rate at which water moves through the global water cycle, with far-reaching influences on global ecosystems and economies. It is essential to establish current rates and possible changes in precipitation, evaporation, runoff, recharge, and atmospheric water vapor transports. Any assessment of climate change and of its causes and impacts must be based on significantly better observations of the water cycle. What are the regional changes in the rate of water cycling and, with consequent changes in associated nutrient and contaminant cycles, how are the structure and function of the world's ecosystems influenced?
• Knowledge of the rate of sequestration of carbon in lakes, reservoirs, peatlands (see Box C) and oceans, and knowledge of controls on those rates, is essential if the United States is to intelligently undertake measures to mitigate greenhouse gas emissions. The total amount of carbon in the combined pools, and the relative balance among these pools, is likely to be strongly influenced by global warming and particularly by changes in the water cycle. How strongly will delivery of carbon to freshwater ecosystems and export of carbon from them be affected by changes in the hydrologic cycle, specifically the supply of water to these systems and in residence times within them?
• Contamination of rivers and lakes from point discharges are often traced to the source, effectively monitored, and strongly regulated. Contamination from nonpoint discharges remains poorly understood and is thus virtually unregulated. Although nonpoint source contamination occurs in both urban and rural environments, the latter presents the greater scientific challenge. Rural nonpoint source contamination is largely due to the erosion and transport of soil particles, to which nutrients (largely associated with fertilizers) and pesticides are attached. Riverine inputs of nutrients result in extensive zones of low oxygen concentration in coastal waters. Although the ability to describe soil erosion is relatively good, there has been very little success in predicting the transport of soil particles and associated contaminants over the land surface to streams and rivers. What are the erosion rates, fluxes, and residence times for sediments transporting non-point-source contaminants and nutrients to downstream ecosystems?
• Alteration of land use as a result of urban growth is inevitable, causing shifts in hydroclimate and generally increasing flooding and pollutant loading. There is also increased groundwater pumping and loss of groundwater recharge associated with urbanization. This decrease in groundwater recharge at least partially explains the degradation of aquatic systems in urbanizing areas. Understanding the nature of the dependence of various aquatic ecosystems on groundwater discharge is needed. Understanding is also needed concerning how to manage water in urban areas to maintain adequate groundwater flows to critical aquatic systems. What are the changes in the hydrologic cycle, and in nutrient and contaminant cycles, caused by alterations in land use owing to urban land use (i.e., the conversion from natural landscape and agriculture to suburban developments)?
• Wetland ecosystems are active in the cycling of water, nutrients, atmospheric trace gases, and carbon. Human activity has altered major wetland regions. Restoration of wetlands is seldom carried out based on a scientific approach. Wetland restoration is in particular need of attention, because in most situations there is little effort to establish, or to reconstruct, the prior condition, and follow-up is often absent or too brief for the degree of success to be evaluated. Can the ability to use hydrologic science be improved substantially so that the damage to aquatic ecosystems caused by human disturbances can be repaired and, where possible, the ecosystem can be restored to their prior conditions?
Page 15
|
