4
Scientific Opportunities for USGS

FOCUS AREAS AND ISSUES IN WATERSHED RESEARCH

Historically, the term "watershed research," as carried out by the U.S. Geological Survey (USGS), has referred to detailed studies of physical, chemical, and biological systems and processes occurring in watersheds ranging in area from 1 or 2 hectares to a few square kilometers. Scientific studies in such watersheds have focused on basic physical processes such as sediment yield and transport, streamflow generation, and rainfall-runoff relationships. As discussed in Chapter 3, investigators also have used such watersheds as natural environmental laboratories to study processes related to atmospheric deposition, acid rain, and the transport of trace constituents through the environment as well as for biological surveys and vegetation studies. The results of such work have formed the basis for much of our current understanding of hydrologic processes.

The relatively small scales of these historic research watersheds have made detailed data collection and process studies possible but sometimes have limited the transferability of the research to larger-scale problems. For example, understanding the transport and fate of pesticides in the environment is of great interest to state and federal agencies charged with monitoring and regulating pesticide use, but these agencies are commonly responsible for regions encompassing thousands of square kilometers and containing multiple watersheds. Scientific studies of pesticide transport and fate, in contrast, usually have been either site specific or confined to small watersheds.

The committee believes the USGS will need to focus its efforts on resolving important problems, even more strongly in the future than it has in the past. Thus, future watershed research programs at the USGS must focus on scientific topics having direct relevance to current and future water policy



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 33
Watershed Research in the U.S. Geological Survey 4 Scientific Opportunities for USGS FOCUS AREAS AND ISSUES IN WATERSHED RESEARCH Historically, the term "watershed research," as carried out by the U.S. Geological Survey (USGS), has referred to detailed studies of physical, chemical, and biological systems and processes occurring in watersheds ranging in area from 1 or 2 hectares to a few square kilometers. Scientific studies in such watersheds have focused on basic physical processes such as sediment yield and transport, streamflow generation, and rainfall-runoff relationships. As discussed in Chapter 3, investigators also have used such watersheds as natural environmental laboratories to study processes related to atmospheric deposition, acid rain, and the transport of trace constituents through the environment as well as for biological surveys and vegetation studies. The results of such work have formed the basis for much of our current understanding of hydrologic processes. The relatively small scales of these historic research watersheds have made detailed data collection and process studies possible but sometimes have limited the transferability of the research to larger-scale problems. For example, understanding the transport and fate of pesticides in the environment is of great interest to state and federal agencies charged with monitoring and regulating pesticide use, but these agencies are commonly responsible for regions encompassing thousands of square kilometers and containing multiple watersheds. Scientific studies of pesticide transport and fate, in contrast, usually have been either site specific or confined to small watersheds. The committee believes the USGS will need to focus its efforts on resolving important problems, even more strongly in the future than it has in the past. Thus, future watershed research programs at the USGS must focus on scientific topics having direct relevance to current and future water policy

OCR for page 33
Watershed Research in the U.S. Geological Survey issues. This study identified six general focus areas that provide a context for potential future watershed research at the USGS (Table 4.1). The topics range in scope from local to regional to national and even global and will require research at a variety of spatial scales. Some topics are purely scientific, in the context of traditional watershed studies (hydrology, hydrogeology, geochemistry, soil science, engineering, biology). Many others cross disciplinary lines and may involve such areas as economics, risk assessment, toxicology, and environmental policy. Several aspects of the focus areas and issues identified in Table 4.1 deserve mention. First, there are obvious interrelationships among focus areas, and no single focus area can be addressed in the absence of other issues. For example, most or all water quality issues have implications for water quantity and land use. Second, the range of disciplines involved is very broad, ranging from classical hydrology and hydrogeology to more general economic and public policy considerations. Third, there is a vast array of spatial scales, from site-specific studies to global issues of climatic change and carbon cycling. Finally, all the issues are relevant to the pursuit of sustainable development. The following discussion highlights some of these issues in the context of watershed research needs. Water Quality Water quality can have a direct impact on human health and on the health of ecosystems. Thus, protection and improvement of the quality of ground water and surface water in the United States, particularly the protection of drinking water supplies, continue to be high priorities. Protection of surface water quality has long been a concern of most federal and state regulatory agencies and involves all components of the surface water budget. Programs in the Black Earth Creek priority watershed in Wisconsin (GAO, 1995) are good examples of attempts to maintain and improve surface water quality through coordinated efforts of local landowners; concerned citizens; and local, state, and federal regulatory agencies. Most such watershed management projects have focused on agricultural areas and have a duration of several years. The projects include an assessment phase, in which the current water quality is characterized; an implementation phase, in which specific water quality improvement steps, usually in the form of best management practices, are carried out; and a postaudit phase, in which improvements in water quality are documented. Best management practices (BMPs) are activities designed to maintain or improve overall watershed quality. Typical best management practices implemented in

OCR for page 33
Watershed Research in the U.S. Geological Survey TABLE 4.1 Focus Areas and Example Issues in Watershed Science Focus Area Issues Water quality - Protection of ground water and surface water quality - Material transport and fate - Process scaling - Urban storm water - Sediment transport - Ground water flow in karst areas and fractured rocks - Model development and improvement Water availability and conservation - Water availability - Multiobjective water management - Optimization management - Ground water recharge - Ground water/surface water relationships - Consistent ground water supplies during periods of inconsistent precipitation Land use and land use change - Basin-wide water management - Effects of agricultural and industrial best management practices - Effects of urbanization - Effects on sediment and trace constituents - Contaminated sites and urban brownfields - Habitat protection - Susceptibility mapping Natural hazards - Flood hazards (erosion, deposition, structural damage, water quality impacts) - Flood forecasting - Drought - Slope stability Climatic variability and change - Hydrologic responses as indicators of climatic change - Hydrologic feedback to climatic change - Global carbon cycling Aquatic habitat alteration and restoration - Aquatic habitats in streams - Natural flow regimes - Wetlands function and restoration - Structural versus restorative approaches

OCR for page 33
Watershed Research in the U.S. Geological Survey agricultural watersheds include streambank erosion controls, improved storage and disposal of animal wastes, improved agricultural tillage practices, and water control structures; local landowners and regulatory agencies share the expenses of implementing these practices. A major problem with many such watershed management programs is that the water quality benefits resulting from the management practice changes are often difficult to predict or assess using current models and field techniques. Some parameters, such as sediment loading in small upland watersheds, respond rapidly to upstream management practices, producing measurable downstream sedimentation changes in only a few months or years and within relatively short distances. However, other parameters, such as nitrates, pesticides, and trace metals, often have much longer residence times in the hydrologic system, and downstream changes in these parameters resulting from BMP implementation may not become apparent for years or decades. Further, those changes may occur far downstream. A case in point is the Big Spring Basin in Iowa (Hallberg et al., 1983), where significant reductions in the amount of nitrogen fertilizer applied to agricultural fields have so far failed to have any statistically significant impact toward reducing nitrate concentrations in ground water in aquifers below the fields or in downstream surface waters. Investigators in other watersheds have made similar observations. To quote from a recent General Accounting Office report to Congress (GAO, 1995, p. 13), "... even given rigorous monitoring, demonstrating a link between changes in land use and diminished chemical pollution is difficult, if not impossible, especially within a short time frame." Obviously, there is a need for longer-term (10- or 20-year) surveillance and research on the effects of changing land use on watersheds. The USGS has the scientific staff, data management facilities, and long-term funding mechanisms necessary to undertake just such long-term studies. Therefore, long-term evaluation of the effects of land use changes and management practices on watersheds presents significant scientific opportunities for the USGS. Critical issues also exist for ground water quality at watershed scales. The 1996 reauthorization of the Safe Drinking Water Act included an earlier mandate that U.S. Environmental Protection Agency (EPA) provide guidelines for the establishment of wellhead protection areas (WHPAs) around public drinking water supply wells. A WHPA is the area of the land surface that corresponds generally to the hydrogeologic capture zone from which the well collects water. Local governments can, in theory, control land use within the WHPA to eliminate or reduce contamination sources, thereby protecting the quality of water produced by the well. Unfortunately, the delineation of well capture zones can be technically difficult, particularly in complex hydrogeo-

OCR for page 33
Watershed Research in the U.S. Geological Survey logic settings, and local municipalities rarely have the financial resources to conduct detailed hydrogeologic studies around each well to be protected. Many municipalities are therefore basing WHPAs on the results of simplified semianalytic or analytic element computer codes, such as those developed by Blandford and Huyakom (1990) and Haitjema et al. (1994), which account for only limited hydrogeologic complexity. The resulting capture zone estimates are therefore uncertain. Furthermore, the actual capture zone can be almost impossible to verify in the field. Clearly, more work is needed, both on field methods for assessing hydrogeologic conditions near water supply wells and on computer techniques for developing capture zone estimates. While techniques for measuring and estimating contaminant loads and runoff from small watersheds are plentiful, tracking the fate of nutrients, heavy metals, sediments, and other contaminants in water bodies subject to large variability remains an elusive task. The physical processes affecting these substances—deposition, resuspension, and various transport mechanisms—often take place in aquatic environments where streamflow, temperature, velocity, wind, and other forces acting on those environments are subject to large fluctuations over short periods of time. Typically, it is unclear whether the fate of material in a system is being controlled by reaction or transport processes. In general, then, an integrated understanding of the biogeochemistry of sediments and subsurface environments is lacking. Without this knowledge the responses of a system to changes in inputs cannot be predicted. The atrazine story demonstrates these challenges. Farmers have used the herbicide atrazine effectively for over 30 years in many parts of the Midwest to control invasive grassy weeds in corn and other crops. Unfortunately, atrazine has become a ground water contaminant and was shown recently to be present at trace (part per billion) quantities in over 50 percent of the drinking water wells in some parts of Wisconsin (LeMasters and Doyle, 1989). Detection of the pesticide at such low levels was not possible just a few years earlier because sufficiently sensitive analytical techniques were not yet available. The fate of this pesticide in ground water and surface water is a complex process), involving advective-dispersive transport, sorption-desorption to soil particles and degradation, as the parent atrazine compound is transformed into atleast three known metabolic products. Even though atrazine use in many parts of the Midwest currently is being banned or sharply reduced, it is likely that atrazine and its metabolites could persist at low concentrations in watersheds for many years. As a result, contaminant levels in ground water and surface water might increase as atrazine leaches downward from the soil zone. Presently there are no

OCR for page 33
Watershed Research in the U.S. Geological Survey BOX 4.1 ILWAS—The Small Watershed Approach The Integrated Lake-Watershed Acidification Study (ILWAS) illustrates the value of simplifying an environmental problem using the small watershed approach. ILWAS, conducted in the early 1980s by a consortium of academic, federal, and private researchers, asked a well-posed question: ''In lake-watershed systems receiving similar amounts of acidic deposition, why are some lakes acidified and others neutral?'' By studying small watersheds, this problem became tractable; the processes controlling water acidification and neutralization could be isolated and identified. ILWAS was conducted in the Adirondack Mountains of northern New York, an area that receives high levels of acidic deposition and is sensitive to acidification because of its crystalline bedrock. The study focused on two small (~2 square kilometers) lake-watershed systems 30 kilometers apart. Both watersheds were pristine, forested, and had similar compositions of bedrock and glacial till. Both watersheds also received nearly identical acidic deposition (mean annual pH of 4.2). Yet Panther Lake was neutral (typical pH near 7.0), whereas Woods Lake was acidic (pH 4.5 to 5.0). Woods Lake was too acidic to support fish. Many components of the two lake-watershed systems, including vegetation, soils, surficial geology, till and bedrock mineralogy, and in-lake features, were evaluated for four years. The chemical composition of water was monitored along its flowpath in each watershed from when it entered as precipitation until it left as lake outflow. Samples were collected of incident rain and snow, throughfall (canopy drip), soil water percolate from the organic and mineral horizons, ground water in the glacial till, inlet stream water, and lake water outflow. Changes in chemistry along a flowpath were related to the physical environment through which the water was moving (canopy, soil, till, etc.), providing the clues needed to understand the acid neutralization process. The key to the contrasting chemical response of the two systems is differences in the catchment hydrology. Peak flow at Panther Lake outlet is more attenuated, and base flow is more sustained, relative to Woods Lake. These differences, in turn, stem from differences in surficial geology, as revealed by geophysical surveys. At Woods Lake,

OCR for page 33
Watershed Research in the U.S. Geological Survey the till cover is relatively thin, and water infiltrates only a short distance before moving laterally downslope. At Panther Lake, till is much thicker and water follows longer flowpaths. Despite similar mineralogical compositions of the surficial material and bedrock in the two watersheds, the longer water residence time within the much deeper tills of the Panther Lake system is sufficient to allow weathering reactions to buffer atmospheric acidity. A lasting contribution of ILWAS is the recognition that hydrology, as controlled by surficial geology, is often an important factor in determining the sensitivity of a lake or watershed to acid deposition. verified integrated process models to predict the long-term transport, transformation, and fate of atrazine and related pesticides in ground water/surface water systems. The tracking of contaminants from one scale to another presents many research challenges. Problems associated with extrapolating laboratory results and models to small watersheds and small watershed results to large watersheds are well documented (NRC, 1991). Models that are appropriate at one scale are not necessarily useful or practical at larger or smaller scales. Inaccurate estimates of kinetic parameters relative to transport rates particularly handicap modeling efforts. Overall, very, few models have been formulated, calibrated, and verified at larger scales. Much fundamental process-oriented watershed research is tractable only at the scale of a small experimental watershed, yet actual environmental problems usually occur at larger scales. Investigative and analytical techniques to transfer knowledge gained at small scales to watershed management and problem solving at larger scales, both spatial and temporal, are clearly lacking. Many scientific opportunities exist for water quality monitoring and research in urban and suburban watersheds. Urban settings can deliver a broad mix of potential contaminants to surface water and ground water, and hydraulic residence times are commonly short, offering little opportunity for degradation or sorption of contaminants prior to downstream discharge. Two principal sources of rainfall-related water quality problems in urban and suburban watersheds are point sources, including both combined sewer overflows (CSOs) and sanitary sewer overflows (SSOs), and storm-water discharges (SWDs), including nonpoint sources of pollution. Both CSOs and SSOs violate water quality standards, while SWDs also contribute suspended solids, create bed loads, and exacerbate problems associated with accumula-

OCR for page 33
Watershed Research in the U.S. Geological Survey tion and transport of contaminated sediments. In either case, transient events can create serious adverse impacts on receiving waters that have tended to defy resolution. Very little is known about the ecotoxicological impact of these periodic urban discharges to downstream regions. Best management practices for urban watersheds are still evolving. Predictions of urban storm water quality based on models developed in rural or undeveloped areas are prone to failure. Because of its established record of achievement in data collection and assessing major point and nonpoint sources of nutrient flux in U.S. watersheds, the USGS is well suited to broaden its current activities to include complementary scientific evaluations of large urban and suburban watersheds affected by hydrometeorological events. Such an expanded focus would help identify and record the indicator parameters descriptive of the spatial and temporal impacts of discharges from the urban/suburban complex into contiguous water resources and contribute to development of policy and management strategies protective of human health and the environment. Sedimentation studies represent additional research opportunities. There is a long history of study of sediment erosion, transport, and deposition by the U.S. Department of Agriculture's (USDA) Agricultural Research Service and Natural Resources Conservation Service and by universities, but major scientific, issues related to sedimentation remain. Perhaps the most pressing of these is in the area of contaminant transport, fate, and impact, as sediments play an important and poorly understood role in the behavior of chemicals in both surface water and ground water (NRC, 1991). Because most hydrophobic pollutants are associated with particulate material, many have accumulated to high levels in sediments and evolved into "in place" pollutants (i.e., polychlorinated biphenyl (PCBs); see, for example, Harris et al., 1988; Larsson et al., 1992). Contaminated sediments may pose a continuous threat to a system due to resuspension events or remobilization caused by biological activity (e.g., bioaccumulation, biomagnification; Harris et al., 1990; Smith et al., 1988). Once again, major difficulties encountered in scaling up research findings at individual sites to large heterogeneous watersheds persist. Karst features and fractured rocks occur over large areas of the United States, yet methods for measuring and modeling ground water flow and ground water surface water interactions in such terrains are currently poor. Most hydrogeologic models are based on the physics of porous media flow, yet severe ground water and surface water problems occur in areas where ground water moves through underground cavities and solution channels (karst) or through interconnected fractures. Some fracture-flow models make the simplifying assumption that the fracture distribution is essentially two-

OCR for page 33
Watershed Research in the U.S. Geological Survey dimensional (Rouleau, 1988; Smith et al., 1989). Recently, several fracture-network models capable of characterizing three-dimensional fracture geometries have been developed (Dershowitz et al., 1994). These models require statistical characterization of the geometric and hydraulic properties of field-measured fracture sets. The models then simulate flow and transport through stochastically generated fracture networks. Few sets of field data exist with which to test and validate such models. Furthermore, integration of fractured rock investigations with larger-scale watershed studies has been attempted only rarely. The variety of hydrologic processes operating in watershed systems can best be studied using integrated models, such as the Modular Modeling System described by Leavesley et al. (1996). Such modeling systems link specific process models, such as precipitation models or rainfall runoff models, to environmental data sets through a geographic information system. Such models are powerful tools that can be used to study many watershed processes in a unified fashion and are particularly useful in identifying data gaps. Water Availability and Conservation Water supply and conservation continue to be critical issues over broad areas of the United States. As the nation's population continues to grow, demands for additional water for potable and industrial uses will also increase, and there will be increasing pressure to find and develop new sources of ground water and surface water while maintaining water quality and quantity. On a global basis, it is estimated that society's ability to appropriate runoff will increase by 10 percent in the next 30 years, while the population will increase by 45 percent in the same time period (Postel et al., 1996). With anticipated diversions from streams, together with the potential adverse effects of global change, aquatic species may become endangered by reduced instream flows over the coming decades. Thus, there is a continuing need for reliable, up-to-date water resource information to help water managers and regulatory officials make the best possible decisions about water supply alternatives. Such decisions are impossible without considering all the competing uses and costs involved in water supply. For example, even in humid areas of the United States, increasing ground water withdrawals in a particular area may have unwanted side effects, such as diminished base flow to streams and wetlands or land subsidence. Planning decisions cannot be made without a clear understanding of the physical interrelationships between ground water and surface water systems.

OCR for page 33
Watershed Research in the U.S. Geological Survey Meeting current and future demands for water supply while simultaneously minimizing the financial and environmental costs of water use and treatment is an example of multiobjective water management. Recently developed mathematical models and computer codes can link various water management options to the policy constraints of each option. For example, ground water policy evaluation and allocation models can be used to study the influence of regional institutional polices, such as taxes and quotas, on regional ground water use (Wagner and Gorelick, 1987). Hydraulic management models can help determine optimal locations of pumping wells and optimal pumping rates based on a variety of restrictions on local drawdown, hydraulic gradients, water quality, and production targets. Such optimization models can be powerful management tools, but the physical hydrologic and hydrogeologic data necessary for their use often are lacking. Future watershed management will require a better understanding of natural ground water recharge processes and of artificial recharge techniques (NRC, 1994a). Natural ground water recharge is the process by which water moves downward from the land surface to reach the saturated zone and becomes ground water. Natural recharge varies temporally and spatially and is notoriously difficult to measure (Mercer et al., 1982), yet accurate estimates of recharge rates and delineations of recharge areas are essential for most commonly used ground water models, such as MODFLOW. Field and modeling studies of recharge frequently involve several disciplines, including climatology, surface water hydrology, soil physics, geochemistry, and vaclose zone hydrology. Although several investigators (e.g., Stephens and Knowlton, 1986; Stoertz and Bradbury, 1989) have conducted detailed studies of recharge at specific sites, there have been few, if any, rigorous studies of the rates and spatial variations of ground water recharge at watershed scales. Artificial recharge projects have become common in areas of the United States experiencing ground water shortages and usually involve the construction of permanent or temporary surface impoundments with permeable beds (NRC, 1994a). Surface water held in such impoundments is allowed to seep slowly downward to recharge underlying aquifers. More research is needed to evaluate the long-term effectiveness of such projects in sustaining local and regional ground water withdrawals and their short-- and long-term effects on water quality, both locally and throughout watersheds. It is also important to evaluate water quality issues associated with artificial recharge practices. For instance, in parts of the arid southwest, low-quality surface waters are sometimes used to recharge ground water. The risks to future users of the ground water are largely unknown. Exchanges between ground water and surface water are key components of every watershed system, yet many past studies have focused on only the

OCR for page 33
Watershed Research in the U.S. Geological Survey surface water or ground water aspects of particular watersheds. Over the past 20 years USGS scientists and others have conducted significant studies of the interactions between lakes and ground water systems. The early work of Winter (1978) established a framework for describing the dynamics of ground water flow into and out of lakes and demonstrated how modern computer techniques can be used to guide the collection of appropriate field data through numerical simulation of lake systems. More recent studies (e.g., Krabbenhoft, 1992; Krabbenhoft et al., 1990) have added chemical and isotopic components to understanding of ground water/lake interactions, and such studies have implications for managing lakes to avoid or mitigate such problems as lake acidification due to acid rain, eutrophication due to nutrient-rich runoff, and accumulation of toxic substances such as mercury in lake sediments and biota. While contamination of ground water from landfills, hazardous waste sites, and underground storage tanks has come under special scrutiny during the past 15 years, much less attention has been given to the transport and fate of agricultural chemicals and nutrients in animal waste after they are applied to land surfaces. Important fractions of materials so applied infiltrate to ground water and move through shallow aquifers to nearby streams. Leakage from subsurface disposal systems also can follow similar pathways to streams. Knowledge of the fate of substances as they move through subsurface environments to surface waters is important not only in understanding the nature of contamination but also in designing management programs. Interactions between surface and ground waters are particularly important to understanding depletion and replenishment of aquifers during drought events. The science of these processes is poorly developed, and existing capabilities to predict aquifer responses to droughts and recharging surface water events are limited. Land Use and Land Use Change Watersheds at various scales respond as an integrated whole to the hydrologic changes imposed on them. Most local, national, and global environmental issues involve some aspect of land use or land use change. Such land use issues range from local agricultural practice to regional timber-harvesting methods to large-scale deforestation, and most of these issues can become emotionally and politically charged during public debate. In the interest of providing clear scientific guidance to decisionmakers, it is critical that we define the scientific questions involved in each issue and then collect

OCR for page 33
Watershed Research in the U.S. Geological Survey should organize its research efforts in watershed processes to address the need to understand hydrologic effects at the scale of watersheds hundreds to thousands of square kilometers in area. The first of the three general requirements for an effective research program in this area—an extensive measurement and monitoring network for basins of various sizes, including major rivers—is satisfied to a large extent by the National Water Quality Assessment (NAWQA) program. NAWQA is an ambitious program that seeks to evaluate the status of the nation's water quality (NRC, 1990). The program already has achieved many successes and the effort to synthesize results from around the country (the ''national synthesis'') is quite active (NRC, 1994c). By design, the NAWQA program has only a modest research component. The key to making the most out of the NAWQA results will be linkages to research efforts of other programs within the Water Resources Division (for example, see Box 4.3). To incorporate the second component of the "large-watershed research program—an integration of process studies on experimental watersheds—will require cooperation and collaboration between the USGS and other agencies. For example, the USGS can rely on the Agricultural Research Service (ARS) for research on processes related to agricultural practices and on the Forest Service for research on processes related to forestry practices. The USGS has much experience in monitoring and studying natural processes within relatively undeveloped watersheds. For the most part, manipulation experiments are not possible in these watersheds. The USGS should consider where gaps in watershed experimentation exist and keep its own watershed research program active by concentrating on these areas. For example, it might be determined that experimentation on hydrologic effects of urbanization is lacking. In this case the USGS could shift its efforts to implementation of such a program that would be designed to interface with information coming from the NAWQA program. The final component needed for the success of the "large-watershed' program is an active modeling effort. The USGS historically has had great success in developing models that have bridged the gap between detailed research and more regional studies. Examples include the ground water code MODFLOW and the geochemical code WATEQ. With the exception of the relatively recent work on the Modular Modeling System by George Leavesley and his colleagues, the USGS has not had similar successes in modeling for watershed science. The success of the USGS effort in watershed research demands that this aspect of the work be recognized as an ingredient of equal importance to the measurement and experimentation portions. The overall effort will entail the use of statistical analyses and mechanistic process modeling.

OCR for page 33
Watershed Research in the U.S. Geological Survey BOX 4.3 Sleepers River WEBB, NAWQA, and the Role of Small Watershed Research in Large Basin Studies Sleepers River in northeastern Vermont is the only WEBB site in an NAWQA basin. The Sleepers River WEBB study uses both a nested basin and a paired basin design to investigate hydrologic and biogeochemical processes in different land uses and at different basin scales. The outlet of the largest basin, the entire 111-square kilometers drainage of Sleepers River, was chosen as an indicator site by the Connecticut River NAWQA study. It is at this scale—approximately 100 square kilometers—that WEBB and NAWQA interface. WEBB and NAWQA have fundamentally different scientific missions. WEBB studies are process-oriented research projects, geographically restricted to one or two land uses in a single ecosystem. Sites are generally small undeveloped watersheds where processes are more effectively isolated. The goal of NAWQA is assessment of water quality status in large river basins, which encompass a broad range of land uses and often more than one ecosystem. Both programs have in common the goal of trend detection through long-term monitoring. The operational approaches of the two programs are likewise quite different. WEBB implements a spatially and temporally intensive sampling design to infer processes by linking closely spaced observations. Sampling in NAWQA, by contrast, is of necessity spatially and temporally sparse, limited to that deemed necessary to assess current water quality. NAWQA indicator sites each are characterized by one dominant land use and are an attempt to isolate water quality influences of that land use. Integrator sites reveal the combined influences on larger rivers. Process understanding gained by WEBB investigations is valuable to NAWQA. The frequent sampling in WEBB studies has shown that nutrient and contaminant fluxes are quite dynamic—information that can be used to guide the frequency of NAWQA sampling. For example, this approach was applied at an indicator site in the Hudson River NAWQA, where frequent sampling showed that pesticide transport occurs primarily during significant storms shortly after application. If streams are sampled only monthly, much of the annual export can go undetected. Sampling during snowmelt at Sleepers River

OCR for page 33
Watershed Research in the U.S. Geological Survey likewise revealed dynamic fluctuations in nitrate concentrations that were not captured by the NAWQA sampling at the same site. WEBB sites also serve as test sites for modeling efforts. For example, the NAWQA National Synthesis Team used TOPMODEL to map the susceptibility of pesticide runoff for the contiguous 48 U.S. states. TOPMODEL was applied to predict the percentage of overland flow in total flow, and the predictions were combined with pesticide application records to assess the risk of pesticide runoff to streams. The assumptions and generalizations needed to apply TOPMODEL at a national scale were tested directly in WEBB and other small watersheds. Issues such as appropriate digital elevation model (DEM) scale, topographic index calculation algorithm, and grid size for the final map were resolved in these research watersheds where model results could be evaluated in light of existing process understanding. Information from NAWQA studies also can point to areas where WEBB-like studies are needed. For example, within the Connecticut NAWQA, nutrient concentrations at indicator sites varied considerably. While some sites were similar to Sleepers River, other agricultural sites had much higher concentrations, and certain urban sites had the highest of all. Nutrient export from urban and intensive agricultural basins clearly dominates the loading to the main stem and the coastal zone. Process-level understanding of controls on nutrients would be of benefit to understanding current water quality and predicting its future trend. For example, isotopic analysis to identify point and nonpoint sources of nitrate would help to indicate where water quality improvement efforts should be targeted. As mentioned above, pieces of the ingredients for a program on "large-watershed" research already exist within the USGS. There are indications that integration of these ingredients is taking place on a selected basis. A few cases about which this committee is aware include the work of Leavesley and co-workers on the Gunnison Basin in Colorado (Battaglin et al., 1993); the work of Wolock on modeling nested basins (Wolock, 1995), which is being extended within the NAWQA umbrella and the research basin efforts at Sleeper's River (see Box 4.3); the work of Bencala and coworkers to interpret in-stream tracer studies throughout the Willamette NAWQA study on the basis of processes identified to be important in a series of small-basin experiments; and the work of Helsel (1994) to "break the scale barrier." If

OCR for page 33
Watershed Research in the U.S. Geological Survey a way can be found to organize and coordinate these efforts, the USGS may fill an important research niche in "large watersheds." A Knowledge Base for the Restoration of Watersheds Hydrologic alterations made to watersheds stem from stream channelization, impoundments, wetlands drainage, deforestation, and urbanization. Although many social and economic benefits have been realized from these human activities, some areas have experienced the unintentional consequences of exaggerated flood and drought, water quality degradation, reduced ground water recharge, and habitat impairment. In response, watershed restoration—a move to recreate some of the predisturbance hydrologic processes and landscape features—has been advocated as the best means to address problems of concern. One of the challenges to future water management is knowing the effectiveness and costs to restore the hydrologic regimes and ecological functions of watersheds, where such restoration is desired by society. Central to the success of watershed restoration is an understanding of, and ability to predict, how changes on the landscape and in water management affect hydrologic processes and ecological outcomes at different watershed scales. Some efforts to monitor and then judge the results of restoration efforts for particular landscape features, such as wetlands (Mitsch and Wilson, 1996), have been made. However, at present almost all assessment efforts are geared toward particular landscape features (e.g., wetlands or riparian zones) or to remediation of some contaminant as a target site. In evaluating the USGS program on hazardous materials science and technology, this committee concluded that the USGS should take an active role in helping to evaluate the effectiveness of remediation efforts for particular ground water areas (NRC, 1996). A similar recommendation to the USGS here with regard to restoration of whole watersheds is appropriate. The USGS has taken an active role in providing technical assessments of alternatives in some areas where whole watershed restoration efforts are under way (see Box 4.4). As this role expands, the USGS should commit to improving the science base supporting assessment protocols for watershed scale restoration. The strength of the USGS has been in areas of geoscience: in collecting data that allow assessment of the quality of water, in gaining a fundamental understanding of what natural processes are important in the flow of water and the transport of materials (including biogeochemical reactions), and in producing models that are useful in analyzing flow and transport in natural systems. The opportunity to derive ecological implica-

OCR for page 33
Watershed Research in the U.S. Geological Survey BOX 4.4 Redwood River Water Management Project The upper Mississippi River and its backwaters, wetlands, and floodplain forests are crucial habitat for many fish and wildlife species. The river is a major flyway for migratory birds, including up to 40 percent of North America's ducks, geese, swans, and wading birds. Many of these species breed in the prairie pothole country in the great river's watershed, including the Redwood River basin, which drains roughly 180,000 hectares of southwestern Minnesota into the Minnesota River. Flooding in the Redwood River basin has resulted in agricultural, urban, and residential damages, particularly at the town of Marshall, Minnesota. Wetlands drainage could be a major contributing factor to increased flood peaks and flood damages in the basin. Prior to agricultural drainage, roughly 43 percent of the basin was wetlands. Roughly 19 percent of these former wetlands areas are depressional and have potential value for stormwater storage. Over 82 percent of the watershed is in agricultural use, indicating extensive wetlands drainage for agriculture. Prior to drainage for agriculture, many of the wetlands in the Redwood River watershed were closed basins that stored water during rainfall events and did not contribute directly to flows in the Redwood River. Two flood control projects that were planned for the basin have not yet been constructed because of public opposition and anticipated adverse environmental impacts. One would divert water from the Redwood River to an adjacent river basin during periods of heavy flows. The residents of the other basin do not want to accept the water. The other project is a dam that would back water up on farm land and cause fluctuating water levels in a state-owned wildlife area. The Redwood and Minnesota rivers are also heavily polluted by suspended sediments, fertilizers, and pesticides that largely run off of agricultural land. The state of Minnesota has initiated a "Clean Water Action Partnership" to involve communities in the cleanup of the Minnesota River basin. Wetlands restoration and the installation of soil and water conservation practices could generate substantial benefits in the Redwood basin. Wetlands and soil and water conservation practices can significantly reduce nonpoint source pollution. The Redwood watershed is

OCR for page 33
Watershed Research in the U.S. Geological Survey also in the prairie pothole region of the upper Midwest, one of the most important waterfowl breeding areas in the United States, which makes it a high priority for wetlands restoration. Other potential benefits of a restoration approach to flood damage reduction include reduced soil erosion and increased groundwater recharge and water supply. A county-level joint powers board, the Redwood-Cottonwood Rivers Control Area, is taking the lead in developing a water management plan for the Redwood River watershed that would use wetlands restoration and soil and water conservation practices to reduce flooding, improve water quality, increase wildlife habitat, and provide other benefits. The board is made up of local farmers. A wide range of federal, state, and local agencies, as well as private-sector representatives are assisting with this initiative. The USGS has been involved in a hydrology work group to assess watershed models for use in predicting outcomes of various management scenarios. tions from this knowledge has been enhanced with the recent formation of the Biological Resources Division. Building on past strengths and this new opportunity, the USGS should advance the science of whole-watershed restoration in four critical areas (1) improvements in the ability to understand relationships among watershed hydrology, water quality, and habitat; (2) helping better understand conditions prior to disturbance; (3) relating the consequences of restoring damaged sites to watershed-scale outcomes; and (4) translating knowledge gained from data collection and experimental watershed studies into models that can be used to evaluate restoration actions. Developing the appropriate knowledge base for making informed decisions about watershed restoration will require a long-term commitment to research. Changes in water quantity and quality stemming from restoration efforts may occur over periods of years to decades. That is, interacting physical, chemical, and biological processes may take long times to equilibrate following alterations. Given the background variability in climate and weather, building the knowledge base for watershed restoration will require long-term monitoring and, possibly, long-term experiments on selected systems. Once again, the development of appropriate modeling tools to interpret and to generalize results must be recognized as being of equal importance to fieldwork.

OCR for page 33
Watershed Research in the U.S. Geological Survey Water chemistry sampling system in the Icacos watershed, Luquillo Forest WEBB site, Puerto Rico. Source: U.S. Geological Survey. A Knowledge Base for Urban-Suburban Hydrology The impacts of changes in land use on hydrology and, more recently, the combined impacts of climate change and land use, have been a focus for hydrologic research. Due to the efforts of a variety of individuals and organizations (including the ARS and the Forest Service), some quality information related to agricultural and forestry effects on watershed hydrology is available. Work on urban hydrology has been somewhat more scattered, perhaps because there is no agency responsible for determining the effects of urban and suburban land use changes. Because of this fragmentation, a concerted effort by the USGS to develop more systematic information related to hydrologic changes in response to suburban and urban development should be considered. Research needs in the area of urban hydrology are many and varied (e.g., Heaney, 1986). The USGS cannot hope to do all of the research that is needed. Rather, the consideration should be the organization of a research

OCR for page 33
Watershed Research in the U.S. Geological Survey thrust that would utilize existing strengths of the agency. The key is to integrate the efforts of researchers in the National Research Program with those of scientists working in the districts, either on NAWQA or other projects within the cooperative program. That is, there are efforts within NAWQA to investigate urban effects on water quality (e.g., Bruce and McMahon, 1994); there are efforts to understand recharge in urban areas (e.g., Michel et al., 1994); and there are efforts to apply simulation models to estimate the effects of urbanization on floods (e.g., Dinicola, 1994). What appears to be lacking is a coordinated program to accumulate extensive data for urban watersheds with the aim of adding to fundamental scientific understanding of processes in these watersheds and of extending and improving the ability to assess quantitatively the effects of land use changes in an urbanizing area. As with the previously discussed areas that this study has identified as possibilities for USGS emphases, the integration of monitoring, observations on small research watersheds, and modeling is critical for an effective program. A Knowledge Base for Erosion, Sediment Transport, and Sediment Deposition The USGS historically has been heavily involved in research on sediment transport in rivers. Over the past decade or so, there has been a decreasing emphasis within the agency on such research. The recognition that toxic chemicals often are transported in association with sediments has prompted renewed interest in work to understand sediment budgets and the processes by which sediments are removed from watersheds. The sediment budgets for large watersheds have large uncertainties (Parker, 1988). In conjunction with a concerted program on the hydrology of "large" watersheds (see above), the USGS should develop an integrated effort on sediment transport. This effort should include a measurement program for a hierarchy of basins around the United States nested so as to address issues of scaling from small watersheds to large watersheds and to record sediment inventories and processes that occur only on larger watersheds. The measurement program should be supplemented with a modeling effort to interpret the measurements and provide the framework for scaling process understanding from small to large watersheds. Special consideration should be given to urban watersheds where improved knowledge of sediment budgets may be of critical importance in understanding the effects of development on water quality and channel stability. An excellent example of what can be done is the USGS work in Puerto Rico (see Box 4.5), where basic research results in

OCR for page 33
Watershed Research in the U.S. Geological Survey BOX 4.5 Water, Energy, and Biogeochemical Budgets in and Adjacent to the Luquillo Experimental Forest of Eastern Puerto Rico The Water, Energy, and Biogeochemical Budgets (WEBB) program began funding research in eastern Puerto Rico during 1990. Sites are in and adjacent to the Luquillo Experimental Forest (LEF). Core funds come from the USGS Global Change Research Program. The Caribbean District Office of Water Resources Division (WRD) manages the program, with guidance from investigators of the National Research Program of WRD. Multiple-paired watersheds are used to compare geologically matched, natural and agriculturally developed environments. The USGS endeavors to characterize the processes that control the distribution and transport of carbon, major, important minor, and nutrient elements through soils, downslope, and out of watersheds. The core of this program is long-term, event-based chemical sampling and physical monitoring. A feature that distinguishes this WEBB site is a strong emphasis on geomorphic processes. Additional efforts include gas exchange studies and innovation of new biogeochemical approaches such as the development of equilibrium erosion theory and the design of techniques based on in situ-produced cosmogenic beryllium-10. Geographic information systems are used to extrapolate from site-specific studies to regional scales. Since its inception, work in the LEF has involved cooperation with the Long Term Ecological Research Program, with the International Institute for Tropical Forestry of the Forest Service, and with universities, Outside the LEF, most research has involved internal coordination with cooperator-based research programs developed by the USGS district office. The cooperators are agencies of the Puerto Rico's government concerned with hazards, such as floods and landslides, or with capacity loss in the Carraizo Reservoir, the principal water supply for San Juan. Puerto Rico is an excellent metaphor for future development in the tropics, having problems associated with deforestation and urbanization. The linkage between the research needs of cooperators and some of the basic scientific questions established a synergy. For example, researchers asked "do landslides ultimately control much of

OCR for page 33
Watershed Research in the U.S. Geological Survey the chemistry of tropical rivers in mountainous regions?", while cooperators asked "how can we assess landslide hazards?" Researchers asked "by what mechanisms does agricultural development accelerate erosion?''; cooperators asked "why is so much sediment being delivered to the Carraizo reservoir, so much so, that a 10-year drought provokes water rationing?" By melding the basic research envisioned in the WEBB program with cooperator-funded research, scientists came up with new and sometimes surprising answers. It was shown that denudation in the forested watersheds was at a near-steady state, whereas physical denudation in the agricultural watersheds was out of equilibrium and proceeding at an order-of-magnitude greater rate. Upland erosion in the agricultural watershed, driven by landslides, is resulting in vast deposits of colluvium and alluvium, which may indicate years of vexation for water providers. conjunction with monitoring efforts are providing the knowledge base that will be critical for informed decisions on watershed management practices (e.g., see Guzmán-Ríos, 1989; Larsen and Torres-Sánchez, 1996; Stallard, 1995). The basic data required for analyzing sediment transported from watersheds must be collected by using bedload samplers and flow-weighted suspended sediment sampling. The measurements are costly in terms of maintaining the gauging station and in terms of personnel. The number of sediment-sampling stations, like the number of stream-gauging stations, operated by the USGS has declined in recent years. The backbone of any scientific program on watershed erosion and related sediment transport and deposition is the collection of basic data. As with stream gauging (NRC, 1992b), careful attention should be given to supporting at least a minimal set of measurement stations for sediment transport studies. In the area of watershed management, effective utilization of research results associated with sediment transport requires that the USGS pursue coordination with agencies charged with managing water resources. Clearly, the USGS is aware of this need, as recently demonstrated in the controlled flood on the Colorado River below Glen Canyon Dam. USGS investigators had shown that the natural annual sediment scour and fill process maintained large sand bars along the river banks, kept sand bars clear of vegetation, and kept debris fans from constricting the river. Construction of the Glen Canyon Dam retarded the large annual spring floods and led to a reduction in the

OCR for page 33
Watershed Research in the U.S. Geological Survey size of sand bars, allowed vegetation to encroach on the channel and debris fans to build up, and caused filling of backwater areas used by native fish. As a solution to these problems, the USGS proposed to the Bureau of Reclamation a controlled flood from the Glen Canyon Dam as a way of reestablishing more natural river conditions, a proposal that was implemented successfully by the bureau in the spring of 1996. The USGS should exploit similar collaborative opportunities where practical. For example, the ARS is leading an effort to improve models for erosion prediction (Lane et al., 1988). Through minor extensions to its existing efforts in sediment transport mechanics and modeling, the potential exists for the USGS to enhance significantly the ARS program as well as other USDA programs.