A Review of the USGS National Water Quality Assessment Pilot Program (1990)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 NAWQA Design Evaluation INTRODUCTION As discussed in the previous chapter, the scope of a truly na- tional water quality assessment is enormous, both in scale and complexity. Hence, it is essential that the scope of NAWQA be defined judiciously in recognition of the overall budgetary con- straints, the organizational strength of the USGS, and the salient characteristics of the physical and cultural systems that define water quality. This chapter contains an evaluation of the design of the NAWQA program components, including integration of the surface water and ground water study units, how study units were selected, the general exclusion of lakes and estuaries from the study plan, the sampling design program, data collection methods, the choice of chemical and biological constituents, quality as- surance/quality control aspects, data management, the analytical framework of the program, and products of the NAWQA program. Integration of Ground Water and Surface Water Description The original concept specified 123 separate surface water and ground water study units: 69 surface water and 54 ground water. A major difficulty with the "separate" approach is that it had the potential for missing the important linkages between surface water and ground water systems--linkages that can have profound effects on the water quality of both systems. As the pilot programs pro- ceeded, it became apparent to the committee and to the USGS that the integrated approach was more beneficial. Some of the pilot programs provide good examples of the importance of incorpora- ting surface water and ground water interactions into water qual- ity studies. 26

NAWQA Design Evaluation 27 For example, in the Delmarva Peninsula pilot project--a ground water investigation--pesticides have caused some contamination of surface water and shallow ground waters. The pilot project used transects of shallow wells installed adjacent to streams to track the infiltration of pesticides from agricultural lands down to the water table and from there into streams. The pilot project in the lower Kansas River basin--a surface water investigation--perhaps best exemplifies the importance of the integrated approach. Along the Kansas River, considerable interchange of water occurs between the river and its alluvial aquifer. During high flows, the river re- charges the aquifer, during low flows, base flow from the aquifer contributes an estimated 1 to 4 cubic feet per second of flow per river mile (Fader, 1974~. The exchange of water probably has a significant effect on ground water and surface water quality, although the effects were not known at the inception of the study. In the Blue River basin near the Kansas-Nebraska state line, streamflow depletion by wells is about two orders of magnitude greater than that predicted by an analog model in the early 1960s (Alley and Emery, MY. in the nor~nwes~ern per o~ `~; Buy unit (the upper Big Blue River and Little Blue River in Nebraska), the interaction between the streams and the aquifers is great. In dry periods, discharge from the High Plains aquifer sustains these streams. Irrigation return flow from the extensive croplands undoubtedly contributes to the relatively high specific conductance found downstream in the Kansas River; the sulfate ion is a major offender. Similarly, the use of nitrogen fertilizers may contribute to the upward trend in nitrite and nitrate levels in the Big Blue River; a major pathway could be base flow to the river. Ground water and surface water interaction is also important in some portions of the Central Oklahoma aquifer study unit; this was known prior to the inception of the pilot project. However, the spatial distribution of this interaction was different from that which was previously believed. Prior to the onset of the pilot project, the USGS team thought that most interaction occurred near the major streams, the Canadian and Cimarron Rivers. The pilot project team discovered that the major interaction occurred along two smaller streams, the Deep Fork and Little Rivers, which served as the major ground water sinks instead of the aforementioned larger streams. This surprising result could have important ramifi- cations for water quality, especially for the four streams involved. Certain other pilot projects might also benefit from an in- tegrated approach. The Yakima River basin pilot project--a sur- face water investigation--is one such example. Interchange between surface water and ground water is probably important in the

28 NAWQA Pilot Program upper, forested reaches of the watershed. In the lower portion of the basin, over 400,000 acres are under irrigation. Subsurface irrigation return flow could affect stream water quality, as the potential exists for water quality degradation arising from irriga- tion (NRC, 1989~. Critique The original NAWQA plan to keep surface water and ground water study units separate had the potential for missing or mini- mizing important water quality linkages between these two systems. Therefore, in recognition of the important water quality implica- tions of surface water and ground water interactions and the committee's concern that the original NAWQA concept downplayed these implications, the USGS decided in March 1989 to redefine the study units. Instead of the original 123 separate study units, 60 integrated surface water/ground water study units now exist; selection of these new units will be discussed in another section. The study units are shown in Figure 2.1 and listed in Table 2.1. Summary The emphasis on integrated study units will result in an im- proved program since it will now examine the water quality of an integrates! hydrologic system. Flow paths between aquifers and streams will be delineated so that the investigators will be able to quantify the effects of one component (i.e., surface water or ground water) of the system on the water quality of the other component. The integrated approach will also allow the investi- gators to ascertain the physics of the system to a greater extent than before, thus enabling them to understand the cause and effect relationships, which could be extended to other parts of the coun- try. One of the important aspects of the NAWQA program is an understanding of water quality cause and effect relationships, and the committee believes that the integrated approach will enhance such understanding. The USGS plans to use teams of surface water and ground water specialists to review study unit findings in an effort to develop a basic understanding of the interrelationships among surface waters, ground waters, and the water quality constituents carried by those waters. To the extent that these reviews are successful, they should enable better plan- ning of new projects as they come on-line.

NAWQA Design Evaluation :__ · . . r 29 FIGURE 2.1 Proposed NAWQA study units in the United States. SOURCE: Leahy et al, 1990.

30 NAPTHA Pilot Program TABLE 2.1 Proposed Study Units for a Full-Scale National Water Quality Annulment Program USGS Northeastern Region 1. New Hampshire-Southern Maine Basins 2. Southeastern New England 3. Connecticut Valley Drainage 4. Hudson River Basin 6. Long Island 6. Delaware River Basin 7. Lower Su~quehanna River Basin 8. Delman~a Peninsula 9. Potomac River Basin 10. Allegheny and Monongahela River Basins 11. Kanawha River Basin 12. Lake Ene-Lake Saint Claire Drainage 13. Great and Little Miami River Basins 14. White River Basin 15. Upper Illinois River Basin 16. Lower Illinois River Basin 17. Westem Lake Michigan Drainage 18. Minneapolis-St. Paul Basin 19. Red River of the North USGS Southeastern Region 20. Albemarle-Pamlico Drainage 21. Upper Tennessee River Basin 22. Santee Basin and Coastal Drainage 23. Apalachicola-Chattahoochee Basin 24. Georgia-Florida Coastal Plain 25. Southern Florida 26. Kentucky River Basin 27. Mobile River and Tributaries 28. Mi~i~ippi Delta 29. Chicot-Evangeline 30. Lower Tennessee River Basin USGS Central Region 31. Eastern Iowa Basins 32. Ozark Plateau 33. Central Oklahoma 84. Trinity River Basin 35. Balcones Fault Zone 36. Central Nebraska 37. Kansas River Basin 38. Upper Arkansas River Basin 39. Central High Plains 40. Southern High Plains 41. South Platte River Basin 42. North Platte River Basin 43. Cheyenne and Belle Fourche Basins 44. Yellowstone Basin 45. Upper Colorado Basin 46. Rio Grande Valley 47. Northern Rockies Intermontane Basins 48. Great Salt Lake Basin USGS Western Region 49. Upper Snake River Basin 50. Southern Arizona 51. Mid-Columbia Basin 52. Yakima River Basin 5S. Puget Sound Drainages 54. Willamette Basin 55. Sacramento Basin 56. Western Great Basin 57. San Josquin-Tulare 58. Santa Ana Basin 59. Oahu 60. Cook Inlet Basin

NAWQA Design Evaluation 31 Lakes and Estuaries Description In the NAWQA program the term "surface water" is virtually synonymous with "stream." Estuaries will not be considered in the program. Lakes (including impoundments) will only be considered if they significantly affect downstream water quality. However, the extent of this consideration is not clear, nor is the manner in which the significance of individual lakes will be evaluated. The exclusion of lakes has been justified by the USGS on the basis of cost. Critique Estuaries and lakes (including impoundments) are critical components of the nations water resources. Much of the surface water supply of the nation is drawn from lakes (either natural or engineered). Lakes and estuaries support rich and diverse ecosys- tems, which provide the basis of an extensive seafood industry and abundant recreational opportunities. Lakes also interact with other freshwater components of the hydrologic cycle, in that they act as both sources and sinks of various water quality constituents. Thus they can significantly affect downstream surface water or down- gradient ground water quality. It is clearly in the nation's interest to monitor, assess, and understand the water quality of both lakes and estuaries. With respect to the major estuaries and lakes in the nation, extensive assessment programs already exist. Notable examples include the Chesapeake Bay, San Francisco Bay, and the Great Lakes. In general, the level of activity in these large systems is much greater than could be supported under the NAWQA program. At best, NAWQA can provide additional information regarding loadings of various water quality constituents. However, NAWQA can and should benefit from the modeling techniques that have been developed during the intensive investigations of these major lake and estuary systems. In the aggregate, the smaller lakes and estuaries in the nation constitute a very important resource. While the committee believes that this resource deserves attention, it supports the decision of the USGS at this time to exclude estuaries from NAWQA and to consider lakes only insofar as they affect downstream surface water quality and downgradient ground water quality. This support is based on our recognition that the USGS, while

32 NAWQA Pilot Program traditionally strong in mathematical modeling of estuaries and lakes, has relatively little experience or capability in biological ant} chemical modeling. The committee, however, believes strongly that the USGS must include in NAWQA consideration of those lakes that significantly affect downstream or downgraclient water quality. This considera- tion should involve more than just upstream ant! downstream monitoring of water quality, and should include some degree of mathematical modeling. If developed in the early stages of a study unit, a preliminary model can be used to establish the potential significance of existing lakes, and to guide the monitoring pro- gram. Once collected, monitoring data can be user! to evaluate the moclel, and if warranted, to guide its improvement. This in turn may lead to adciitional monitoring requirements. This iterative use of models is an essential element of a process-oriented assessment, and is discussed further in the Analytical Framework section later in this report. Summary The committee believes that an assessment of the long-term trends in the water quality of lakes (including impoundments) and estuaries should be part of any long-term national water quality assessment. At some point, the water quality of lakes should receive the same level of attention in the NAWQA program as that of streams and ground water. However, given the present lack of personnel in the USGS with expertise in the biological and chemi- cal modeling of lakes and estuaries, we believe that the NAWQA program should not be expanded at this juncture to include es- tuaries. However, we recommend that initially lakes should be considered but only as they affect downstream surface water quality or downgradient ground water quality. The committee further recommends that the first set of inves- tigations in the NAWQA program include one or more study units in which lakes are likely to be significant contributors to down- stream and/or downgradient water quality. This will give the USGS an opportunity to enhance its capabilities with respect to lakes. Additionally, we recommend that mathematical models be developed at the initial stages of each study unit investigation involving lakes.

NAWQA Design Evaluation Introduction Selection of Study Units The selection of study units is described in Hirsch et al., 198S, which is the source for much of the discussion in this section. Some changes in the study unit selection process, as described in Hirsch et al., 198S, were dictated by the decision to designate combined surface water-ground water study units; these changes are included in the discussion herein. Since the selection of study units is based largely upon the USGS's hydrologic unit classifica- tion system, a brief description of that system, taken from Seaber and others (1986), is warranted. Description The USGS divides the United States into 21 major geographic regions and 222 hydrologic subregions. The regions contain either the drainage area of a major river, such as the Missouri region, or the combined drainage area of a series of rivers, such as the Texas- Gulf region, which includes a number of rivers draining into the Gulf of Mexico. Eighteen regions comprise the conterminous United States, with individual regions specified for Alaska, the Hawaiian Islands, and Puerto Rico and other Caribbean areas. A hydrologic subregion includes the area drained by a river system, a reach of a river and its tributaries in that reach, a closed basils), or a group of streams forming a coastal drainage area. These subregions are further subdivided into 352 accounting units that nest within, or are equivalent to, the subregions. The accounting units are used by the USGS in designing and managing the National Water Data Network. A set of 200 surface water candidate study units (CSUs) were identified, based primarily on the hydrologic subregions described in Seaber and others (1986~. When a given subregion was too large, it was subdivided to form a number of CSUs. In other cases, subregions were combined into a single CSU or a particular ac- counting unit from one subregion was added to another. In the selection of the surface water CSUs, attempts were macle to select ones having similar areas. Modifications were sometimes made to ensure that all sources to a major lake or estuary were contained within the same CSU. An attempt was also made to select CSUs that had relatively homogeneous land use and environmental characteristics. A set of 116 ground water CSUs were also iden- tified. The boundaries of these CSUs were less well defined than ~ _ _ 33

34 NAWQA Pilot Program those of the surface water CSUs; in some cases, ground water CSU boundaries coincided with those of surface water CSUs. Each set of CSUs was screened using a modified linear pro- gramming model (Fox and Scudder, 1986~. The objective function of the linear program minimized the number of study units se- lected, given certain constraints: (1) the study units account for about 60 percent of the nation's water use, as measured by popula- tion served by public water supplies and by total withdrawals excluding thermal and hydropower uses; and (2) each state contains at least one study unit having at least 30 percent of its area within that state. This exercise yielded two maps: one of surface water study units and one of ground water study units. Next, a few study units that were downstream of major river basins (e.g., the Colorado and Ohio rivers) were deleted; the rationale for these deletions was that these basins integrate water quality conditions for many upstream basins (Hirsch et al., 1988~. The two sets of study units were then merged and reviewed by all USGS district and regional offices to provide study unit boundaries that ac- counted for both surface water and ground water boundaries. Some study unit boundaries may be refined once the full-scale program gets under way. Finally, study units were specified for Hawaii and Alaska. The 60 study units encompass about 40 per- cent of the land area of the conterminous United States and incor- porate about 60 to 70 percent of the nation's water use (fresh surface water withdrawals excluding thermal and hydropower uses). It should be noted that in selecting the study units, emphasis was placed on "key river basins" and ground water units that corresponded somewhat to the USGS Regional Aquifer Systems Analysis (RASA) program. Essentially, the purpose of the RASA program is to emphasize the physical hydrogeology of certain regional aquifers. By attempting to coordinate NAWQA study units with RASA efforts, the USGS will take advantage of infor- mation already being generated. The RASA program emphasizes the physics of the ground water systems, something that must be understood to undertake a water quality assessment that seeks to identify cause and effect relationships. The coordination of RASA and NAWQA, where possible, will result in an improved NAWQA product. Critique The geographic coverage of the study units is good, considering that only 40 percent of the continental United States is included.

NAWQA Design Evaluation 35 The variation in size is substantial: the smallest study unit, Long island (number 5 in Figure 2.1 and Table 2.1), is about 1,300 square miles, and the largest is the Ozark Plateau (number 32), which is over 60,000 square miles. The pilot project study units are rela- tively small--3,000 to 15,000 square miles. Given the large size of approximately eight study units, the combined ground water-sur- face water nature of these study units, and the USGS's lack of experience in a large study unit, the committee is concerned that NAWQA objectives may not be met in some of the larger units. Summary Study unit selection is generally good; the coverage is broad and represents a variety of hydrologic regions. Coordination of study units with RASA study areas is an approach that should result in some degree of synergism and aid in delineating cause and effect relationships. Some of the study units are quite large, much larger than any of the seven pilot project study units. To see how well the NAWQA program can accommodate a large study unit, the committee recommends that early in the full-scale NAWQA program, the U.S. Geological Survey select two large study units for investigation. SAMPLING DESIGN Sampling Approach Introduction The NAWQA pilot program originally consisted of separate surface- and ground-water study units with separate sampling designs. The committee's evaluation therefore considers the sur- face- and ground-water sampling designs separately. The full- scale NAWQA will be based on combined surface- and ground- water study units; however, a combined sampling design has not yet been developed. Hence, the committee also points out some issues that need to be considered in designing a sampling program for the combined surface- and ground-water study units. Ground Water Description The major objectives of the ground water investiga- tions (Hirsch et al., 1988) are to (1) describe ground water quality

36 NAUSEA Pilot Program conditions for major hydrogeologic settings; (2) describe the geo- graphic distributions within the study units of selected water quality constituents and problem areas; (3) define long-term trends in ground water quality; and (4) identify, describe, and explain, if possible, the major factors that affect observed current conditions and trends in ground water quality. To attain these objectives, NAWQA projects will conduct three different types of ground water sampling activities: (1) regional, (2) targeted, and (3) long term. Regional sampling will be conducted throughout each major hydrogeologic setting within a study unit. All of the national target variables for chemistry and physics will be measured in an effort to provide descriptive statistics and to form an initial basis for describing the geographic distribution of water quality vari- ables within the study unit. Sampling will be done in three dimen- sions and will be designed to be unbiased with respect to particular "problem areas." Targeted sampling will occur in selected locations for specific groups of water quality constituents; it will focus on "problem areas" but will not be a "plume chasing" exercise. As an example, in the Central Oklahoma aquifer pilot project, targeted sampling variables are (1) naturally occurring trace substances (NOTS) such as arsenic, selenium, chromium, uranium and gross-alpha radio- activity and (2) organics (especially pesticides) beneath central Oklahoma City. Problem areas for targeted sampling will be identified from the regional sampling results, knowledge of the hydrogeologic and land-use factors that contribute to water quality, and consultation with local agencies and individuals involved with ground water quality. Three targeted sampling approaches will be used: (1) search-oriented sampling, (2) statistical hypothesis testing, and (3) local-scale transects (Hirsch et al., 1988~. Local-scale transects will involve high-density sampling of one or more wells per square mile to characterize representative set- tings that commonly occur throughout large parts of the study unit. These local-scale studies should be useful for examining the inter- relationship between surface water and ground water quality. Transects provide very good information about a study unit's physical hydrogeology. The Delmarva Peninsula pilot project has used transects of shallow wells installed adjacent to streams to follow the infiltration of pesticides from agricultural lands down to the shallow aquifer and from there into streams. Local-scale transects have also been used in the Carson basin pilot project. Search-oriented sampling will involve sampling for constituents within particular settings in which they are most expected. The

NAWQA Design Evaluation 37 NOTS sampling in the Central Oklahoma aquifer study unit is an example. Statistical hypothesis testing will be used to test hy- potheses on the regional distribution of contaminants. For exam- ple, a program could be designed to test the hypothesis that the ground water quality beneath areas of differing land use varies. Both search-oriented sampling and statistical hypothesis-testing will be used at scales ranging from several hundred square miles to areas approaching the study unit size. The final sampling activity, long-term sampling, will identify and examine temporal trends in ground water quality. Sampling will be conducted in wells representing ground water of different ages and hydrogeologic settings. The local networks established in the targeted samplings will also be used. Sampling frequencies or target constituents will not be rigid; as the results of the long-term sampling are obtained, certain sampling frequencies may be changed and locations or constituents dropped with the appropriate safeguards so as not to jeopardize long-term analysis. This flexible approach is beneficial in that it will conserve limited resources and concentrate on constituents and locations that will produce meaningful results. Critigue One of the committeets concerns is whether adequate coverage can be provided, especially in the larger study units. For example, current regional sampling in the three ground water pilot projects calls for an initial batch of 150 to 200 samples for each study unit during the first 2 years. This initial batch will provide the preliminary statistical summary. The three ground water pilot project study units are small, 3,000 to 6,050 square miles. Under the revised plan, study units not only will combine both ground water and surface water but also will be 10 to 20 times larger, as large as 60,000 square miles. It is unclear whether such a small number of samples will be sufficient for the much larger, com- bined surface- and ground-water study units. Long-term sampling is beneficial, and the committee is pleased to see its inclusion in the NAWQA program. It is one of the unique aspects of the program. Currently, there is no national systematic examination of the long-term trends (or lack thereof) in ground water quality. Surface Water Description This description of the sampling approach to be used in the surface water component of NAWQA is based on the docu-

38 NAWQA Pilot Program ment describing concepts for a NAWQA program (Hirsch et al., 1988~. The major objectives of the surface water investigations are to (1) describe the occurrence and spatial distribution of a broad array of water quality constituents, (2) provide information on the seasonal variation and frequency of occurrence of selected water quality constituents at key locations, (3) estimate loads of selected water-quality constituents at key locations, (4) define long-term trends in the concentrations and/or loads of selected constituents at key locations, and (5) identify, describe, and explain, if possible, the major factors that affect observed conditions and trends in surface water quality. To attain these objectives, NAWQA projects will conduct three different types of surface water sampling: (1) fixed-station sampling, (2) synoptic sampling, and (3) studies of selected reaches. Fixed-station sampling in the surface waters of each study unit will be conducted in order to describe the seasonal variations and frequency of occurrence of selected water quality constituents, to estimate loads, and to define long-term trends in water quality. Likely fixed-station sampling sites include (1) sites at the mouths of major tributaries or other points that account for a large por- tion of the total basin runoff; (2) sites up- and downstream from reservoirs or other areas likely to affect water quality; (3) sites on streams with homogeneous land use in the basin; (4) sites near important water uses in the basin (e.g., public water-supply in- takes); and (5) existing stations from other programs (e.g., NASQAN). These stations will be sampled monthly during the active phase with an additional 3 to 6 high-flow samples taken annually. Synoptic surveys will be conducted in each study unit to pro- vide a finer degree of spatial resolution than is attainable with fixed-station samples and to examine the relationship between water quality conditions and land- or waste-management practices. These surveys will consist of measurements on selected parameters taken at many sites during a brief period representative of a particular hydrologic condition. The number and timing of synop- tic surveys will vary among study units based on the questions being addressed. Studies of selected reaches will be conducted to understand the sources, distribution, and fate of particular water quality con- stituents in greater detail than is available from either the fixed or synoptic sampling. The committee agrees that mathematical modeling will be useful in these studies. The USGS views selected reach studies as likely candidates for funding from the USGS cooperative program.

NAWQA Design Evaluation 39 Critigue Typical water quality sampling programs include fixed- interval sampling at fixed stations. Unique and valuable additions to this in NAWQA include fixed-station sampling at high-flow conditions, synoptic sampling, and selected reach studies. Fixed- station sampling at high-flow conditions will be essential for load and transport calculations. Synoptic sampling will be a valuable tool for examining relations between land use and water quality conditions. The committee is concerned whether the land-use data will be: · available in a timely manner, · classified in adequate detail, · of the proper scale and resolution, and · updated frequently enough to be useful for documenting cause and effect relationships in long-term trends. It is clear that NAWQA's ability to ascertain cause and effect relationships lies primarily with the selected reach studies. The committee knows of no selected reach studies funded during the pilot studies. There is also concern that modeling is discussed only in the context of selected reach studies. However, it is this com- ponent that is likely to vary most between study units in the amount of financial support provided. Our concerns with respect to modeling and cause and effect analysis in NAWQA are pre- sented in greater detail in the Analytical Framework section of this report. Combined Surface- and Ground-Water Study Units The sampling programs described in the previous sections were developed for separate surface water and ground water study units. The study units are now combined, and the USGS needs to re- examine the sampling programs in light of this new organization. The following three issues are among those that should be ad- dressed in this reexamination: 1. Ground water and surface water sampling protocols have been developed independently. With the new emphasis on ground water and surface water interaction, it may be desirable to have some protocols developed in tandem. For example, if models are going to be developed to describe some aspect of this interaction, comparable procedures, estimation techniques, and precision levels may be needed to ensure that sensible and meaningful model interpretations can be drawn.

40 NAWQA Pilot Program 2. Because the integration of ground water with surface water samples has been a relatively recent design modification, it is still unclear how samples will be allocated between ground and surface water. Surface water systems are visible and, hence, easier to delineate than ground water systems. Ground water flow is inher- ently three-dimensional and more areally extensive than surface water flow, which dictates that sampling is more difficult and expensive than for surface water flow. Since ground water flows more slowly than surface water, sampling over long periods of time may be required to provide information on water quality changes. Therefore, more ground water samples may be needed to provide the same level of precision provided by the surface water pa- rameter estimation. The determination of the optimum allocation of sample points should be based in part on the level of desired precision in the parameters to be estimated as well as on the amount of inherent variation likely to be encountered in each type of sampling. 3. The need to integrate surface- and ground-water informa- tion may influence the choice of sampling location or timing of synoptic sampling. . ~ a. Summary of Sampling Design The ground water sampling program will consist of regional, targeted, and long-term sampling. The sampling approach generally seems adequate, although the com- mittee is concerned about the adequacy of regional ground water sampling in the larger study units. The long-term sampling pro- gram is a valuable and unique component of NAWQA. Surface water sampling in NAWQA will consist of fixed-station sampling, synoptic sampling, and selected reach studies. The high- flow samples and synoptic sampling are viewed as another unique and worthwhile aspect of NAWQA that is not part of ongoing water quality programs. Acquisition of adequate land-use data is an essential feature of NAWQA. However, the land-use data must be of adequate detail, available in a timely manner, and updated frequently. Most importantly, the integration of ground water with surface water will require a new mindset for the USGS operational field teams. In the past, the surface water, ground water, and water quality personnel appear to have operated independently from one another. The new combined approach, which the committee endorses, poses a new challenge to the national headquarters staff to make sure that the degree of communication and integration necessary for a national assessment of surface water, ground water and water quality actually occurs.

NAPTHA Design Evaluation 41 Now that each study unit represents a hydrologic system encom- passing both surface water and ground water, the committee recom- mends that the USGS review the sampling strategy to ensure that comparable procedures and precision levels are used. The review may not require a major redesign of the sampling program, but it should consider three major items: 1. the distribution and location of sampling sites for surface water and ground water; 2. quantification of the interaction between surface water and ground water; and 3. the water quality implications of these interactions. Rotational Versus Continual Data Collection Description As currently planned, NAWQA assessment activities will be conducted on a rotational or cyclical basis (i.e., with high and low activity periods for each study unit), rather than on a continual basis, for data collection and analysis (Hirsch et al., 1988~. (See Figure 2.2.) Initially, this rotational structure will consist of five phases: 1. Project planning 2. Analysis of existing data (retrospective) 3. Intensive data collection and interpretation 4. Report completion 5. Low-level sampling and analysis 1 year 1 year 3 years 1 year 4 years The project planning and retrospective phases actually overlap during the first 2 years of an initial assessment cycle. Some reconnaissance sampling also will occur during these first 2 years. After completion of the initial cycle in a particular study unit, the project planning and retrospective phases would require only one year. Thus, after the initial cycle, subsequent assessment cycles would consist of 4 years of high activity (intensive data collection, data analyses, and report preparation), followed by 4 years of low- leve! activity. The program does provide for some flexibility, e.g., some study units may encounter conditions that justify extension of the

42 on - o - - - a - 11 ~ .s' ' tori 7 He Is I Ili ~ i! Ma a_ Ma 5 4~ an o ret 8 Cl) an an an a_ Jo o aft en. (~a U.a to Can

NAWQA Design Evaluation normal 3-year intensive data collection period. It is also expected that federal or state cooperative programs or local agencies may conduct additional sampling and analysis activities during low- activity periods if circumstances warrant. The 60 study units will be divided into three groups of 20 selected to provide broad geo- graphic coverage and a range of hydrologic conditions. The pro- gram will be staged so that only one group of 20 study units is in the intensive data collection and analysis phase at any given time. Critique 43 There are several reasons for adopting a rotational, rather than a continual, data collection and analysis approach. First, using a rotational scheme, the USGS is able to cover a larger number of study units for a given staff size and budget. Second, the condi- tions and trends with which NAWQA is concerned develop rela- tively slowly. Third, as the USGS has emphasized, NAWQA is not designed to be only a long-term monitoring program (for which continual data collection and interpretation might be appropriate), but also an assessment of water quality trends. While NAWQA is intended to assess water quality conditions and trends, it is also designed to develop a scientific understanding of the factors that affect those conditions and trends. Thus, the emphasis of NAWQA is on producing timely interpretive reports focused on specific topics of regional and national interest. Summary While the committee agrees with this rationale for the rota- tional assessment approach, it recommends that the duration and staging aspects be reevaluated by the USGS and an external scien- tific advisory committee after completion of the first cycle of 20 study units. The committee recommends that extensions of inten- sive study be granted only for issues that are truly regional or national in scope. In addition, the committee recommends that the mathematical model for each study unit be selected in the first year of the cycle as part of the planning effort. Choice of Chemical Constituents Chemical and Physical Characteristics Description One of the key elements of the NAWQA design is the selection of a set of chemical and physical constituents to form the

44 NAWQA Pilot Program group of target variables on which the program will focus (Appen- dix B. #22~. The chemical and physical data base emphasizes water quality degradation, such as that which occurs from non- point sources of contamination or from many point sources. National target variables consist of a common set of physical measurements, inorganic constituents and organic compounds to be measured by specified analytical methods. Study unit variables will be selected for each study unit by the project team and are a supplement to the national list. This flexibility of adding con- stituents will meet the needs of local and regional water quality issues. The NAWQA program has selected a wide range of chemical and physical measurements to monitor in a nationally consistent manner. This broad-based list of water quality constituents will provide the data to assess a number of potentially national water quality issues, such as sedimentation, dissolved oxygen, nutrients (nitrogen and phosphorus), toxic substances, and bacteria levels. Table 2.3 summaries the general classes of compounds that will be measured in each of the various environmental compartments, i.e., ground water, surface water, bed material, tissues, and suspended sediment. The individual constituents to be measured in the surface and ground water samples are listed in Tables 2.4 to 2.6. Table 2.4 indicates those "core" constituents to be measured at all surface water sites, while Table 2.5 indicates the additional con- stituents to be measured seasonally during the first 12 months of the sampling period. These tables also indicate the "reporting level" for each constituent which is slightly higher than the detection limit of the method. Table 2.6 lists the constituents to be analyzed in ground water and shows the high degree of consistency between surface and ground water samples. The extensive list of pesticides, herbicides, and other volatile compounds to be measured is pro- vided in Appendix C. In addition to the chemical data described above, consistent records of ancillary information will be recorded on stream flow and basin characteristics, well and aquifer characteristics, and land use and other measures of human activity. Table 2.7 outlines the data to be gathered for wells which are sampled and is an example of the level of detail which will be documented by NAWQA. The compounds and ancillary information to be monitored as well as the analytical methods to be used in NAWQA were selected by a group of scientists from both inside and outside the USGS. The inorganic and physical constituents were selected primarily on the basis of their effects on human health, ecosystems, and agricul- ture and on the relevance to water quality issues. Some con- stituents, such as many of the major ions in ground water, were

NAWQA Design Evaluation TABLE 2.5 Chem~cal and Physical Measurements to be Made in Variou~ En~nronmental Matnces by National Water Quality A~t Program 45 Ground Surface Bed Swpended Water Water Material Ti~ues Sediment Field Measurements (Temp, pH, D.O.) X X Solids, dimohred and suspended X Major ions X X Nutrients X X Trace metale X X X X X Organic carbon X X X VOC's X X Serru-,rolatiles X* X X Acid extractables X* Org P pesticides X* PCB's X* X X X* Org C1 insecticides X* X* X X X* Carbamates X X Triazines, Org N pest. X X Chlorophenoxy herbicides X X R ad i o n u c l i d es X X ~Only detennined in those study units where reconna~ance sa~npling finds occurrence.

46 NAUSEA Pilot Program TABLE 2.4 -- Field measurements and constituents, to be detenruned in all surface-water samples collected from Fred stations as part of the pilot National Water-Quality Averment Program Compound WATSTORE Reporting code level FIELD MEASUREMENTS ALKALINITY (mg/L as CaC03) DISSOLVED OXYGEN PH (standard units) SPECIFIC CONDUCTANCE (umho/cm at 25) TEMPERATURE (degrees Celsius) MAJOR CONSTITUENTS 00410 00400 00095 00010 CALCIUM, DIS. (mg/L as Ca) 00915 .02 MAGNESIUM, DIS. (mg/L ~ Mg) 00925 .01 SODIUM, DIS. (mg/L as Na) 00930 .20 POTASSIUM, DIS. (mg/L as K) 00935 .10 CHLORIDE, DIS. (mg/L as Cl) 00940 .10 SOLIDS, R.O.E. AT 180'C (mg/L0 70300 1.00 SULFATE, DIS. (mg/L as S04) 00945 .20 NUTRIENTS NITROGEN, DIS., NH4 + ORG (mg/L as N) 00623 .20 NITROGEN., NH4 + ORG (mg/L as N) 00625 .20 NITROGEN, DIS., NH4 (mg/1 as N) 00608 .01 NITROGEN, DIS., NO2 + NO3 (mg/L as P) 00631 .10 PHOSPHORUS, DIS., ORTHO (rng/L as P) 00671 .01 PHOSPHORUS, TOT., (mg/L as p) 00665 .01 TRACE ELEMENTS ARSENIC, DIS. (ug/L AS As) 01000 1.0 ARSENIC, SUS., TOT. (ug/g as As)* 1.0 CADMIUM, DIS. (ug/L as Cd) 01025 0.1 CADMIUM, SUS., TOT. (ug/g as Cd)* 0.2 CADMIUM CHROMIUM, DIS. (ug/L as Cr) 01030 0.5 CHROMIUM, SUS., TOT. (ug/g as Cr)* 1.0 CHROMIUM COPPER, DIS. (ug/L as Cu) 01040 0.5 COPPER, SUS., TOT. (ug/g as Cu)* 0.7 COPPER LEAD, DIS. (ug/1 as Pb) 01049 0.5 LEAD, SUS., TOT. (ug/g as Pb)* 7.0 LEAD MERCURY, DIS. (ug/L as Hg) 71890 0.1 MERCURY, SUS., TOT. (ug/g as Hg)** 1.0

NAUSEA Design Evaluation TABLE 2.4 -- continued Compound SEDIMENT SUSPENDED SEDIMENT (ug/L) PERCENT FINER THAN 0.062 mm ORGANIC CARBON WATSTORE Reporting code level 8~0154 70831 CARBON, ORG., DIS. (ug/L as C) 00681 0.1 CARBON, ORG., SUS. (ug/L as C) 00689 0.1 *Suspended, Total = Total digestion being made on the suspended sediment. **Mercury, SUS, TOT = Digestion with HNO3 and Na, Cr2, O7. Source: Wilber, W., personal communication to M. Conditt of Procter and Gamble, 1990. 47

48 NAUSEA Pilot Program TABLE 2.5 -- Field measurements and constituents, to be determined seasonally at fixed stations during the first 12 months of operation as part of the National Water-Quality A - emment Program Compound WATSTORE Reporting code level FIELD MEASUREMENTS ACIDITY (mg/L as H) MAJOR CONSTITUENTS 71825 BROMIDE, DIS. (mg/L as Br) 71870 0.01 FLUORIDE, DIS. (mg/L as Et) 0~50 0.10 MAJOR METALS AND TRACE ELEMENTS ALUMINUM, DIS. (ug/L as Al) 01106 10-20 ALUMINUM, SUS., TOT. (ug/g as Al)* 20.00 ANTIMONY, DIS. (ug/L as Sb) 01095 1.00 ANTIMONY, SUS., TOT. (ug/g as Sb)* 1.00 BARIUM, DIS. (ug/L as Ba) 01005 2.00 BARIUM, SUS., TOT. (ug/g as Ba)* 1.00 BERYLLIUM, DIS. (ug/L as Be) 01010 .50 BERYLLIUM, SUS., TOT. (ug/g as Be)* .70 BORON, DIS. (ug/L as B) 01020 1.00 BORON, SUS., TOT. (ug/g as B)* 2.00 MOLYBDENUM, DIS. (ug/L as Mo) 01060 10.00 MOLYBDENUM, SUS., (ug/g as Ma)* 2.00 NICKEL, DIS. (ug/l ~ Ni) 01065 5-10 NICKEL, SUS., TOT. (ug/g as Ni)* 0.70 SELENIUM, DIS. (ug/L as Se) 01145 1.00 SELENIUM, SUS., TOT. (ug/g as Se)* 5.00 SILVER, DIS. (ug/L as Ag) 01075 2-10 SILVER, SUS., TOT. (ug/g as Ag)* 0.20 VANADIUM, DIS. (ug/L as V) 01085 6.00 VANADIUM, SUS., TOT. (ug/g as V)* 1.00 ZINC, DIS. (ug/L as Zn) 01090 3.00 ZINC, SUS., TOT. (ug/g as Zn)* 0.01 IRON, DIS. (ug/L as Fe) 01046 10.00 IRON, SUS., TOT. (ug/g as Fe)* 3.00 MANGANESE, DIS. (ug/L as Mn) 01056 1.00 MANGANESE, SUS., TOT. (ug/g as Mn)* 0.70 OTHER CYANIDE, DIS. (mg/L as ON) 00723 GROSS alpha RAD, DIS. (mg/L as U. natural) 80030 GROSS ALPHA as U natural FILTRATION, laboratory GROSS beta RAD, DIS. (pCi/L as SR-90/Y-90 natural) 0.01 0.40 0050 0.40 0.40

NAWQA Design Evaluation TABLE 2.5 -- continued 49 Compound GROSS BETA pCi/L as Cs-137 GROSS alpha MAD, BUS. (mg/g as U. natural) GROSS beta BAD, SUB. (pCi/g as SR-90/Y-90 natural) RADON-222 (50 pCi/L) WATSTORE Reporting code level 80040 80060 0.40 0.40 *Suspended, Total = Total digestion being made on the suspended sediments. Source: Wilber, W. personal communication to M. Conditt, Procter and Gamble, 1990.

50 NAPTHA Pilot Program TABLE 2.6 -- Water-quality constituents to be included in sample analysm for all three ground water pilot projects, National Water-Quality Assessment Program Constituent(~) Laboratory code (LC) or schedule number (SH) Properties and major constituents pH Specific conductance Alkalinity Calcium Magnesium Potassium Sodium Chloride Fluoride Sulfate Silica Nutrients Ammonia Nitrite Nitrite and nitrate Kjeldahl nitrogen (ammonia plus organic nitrogen) Orthophosphate phosphorus Major metals and trace elements Antimony Arsenic Barium Beryllium Boron Cadmium Chromium Cobalt Copper Iron Lead Lithium Manganese Mercury Molybdenum Nickel Selenium Silver Strontium Vanadium Zinc LC0068 LC~69 LC~70 SH1043 SH1043 LC0054 SHi043 LC1213 SH1043 LC1200 SH1Q43 LC0301 LC0160 LC0228 LC0268 LC0162 LC~77 LC0112 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 SH1043 LC0225 SH1043 SH1043 LC0087 SH1043 SH1043 SH1043 SH1043

NAPTHA Design Evaluation TABLE 2.6 continued 51 Constituent(~) Laboratory code (LC) or schedule number (SH) Radionuclides Grow alpha Grow beta Radon-~22 Tritium Discolored organic carbon Volatile organic compounds Carbamate insecticides Chlorophenox r-acid herbicides Nitrogen-contaunng pesticides (largely tnasine herbicides) Deutenum/protiurn Oxygen-18/o~rgen-16 SH0456 or SH045822 SH0456 or SH0458 LC1369 LC0624 or LC10433 Organic compounds Stable isotope ratios LC0113 SH1380 SH1359 SH0079 SH1389 LC0300 LC0489 1Samples should be collected and stored at project office for possible analyze later. 2Use SH046 if estimated concentration of dissolved solids is 1~ than 250 mg/L and SH0458 Otherwise. Appropriate laboratory method depends on use of date (Robert Michel, U.S. Geological Sunrey, written communication, 1988~. Source: Wilber, W. personal communication to M. Conditt, Procter and Gamble, 1990.

52 NAUSEA Pilot Program TABLE 2.7 Site charactenstice for sampled wells in the National Water Quality A_ment Program Unique site identification number Type of site (well! drain, and others) Data reliability Project identification number District, state, and county codes Latitude and longitude of site Altitude of land surface Topographic setting Use of site (observation well, withdrawal well, and 80 forth) Primary we of water Aquifer name code Aquifer type code (unconfined, confined, mixed) Depth of well Water level Depth to top and bottom of each open integral Depth to top and bottom of each geohydrologic unit Rated pump capacity Type of lift Date of well construction Method of construction Type of finish Type of surface seal Casing material Sampling method Primary reason for well selection (with or without regard to known or suspected local problem areas) Occurrence of various land uses and local features such as gas stations and septic tanks within a 100 ft and 1/4 mi radius of the sampling well Predominant land use within 100 ft radius of the sampling well Predominant land use within 1/4 mi radius of the sampling well Percent of total area within a 1/4 mi radius of the well that consists of predominant land use Known occurrences of major changes in land use near the well within the last decade Local agricultural practices Source: Hirsch et al., 1988.

NAWQA Design Evaluation 53 selected to provide information on the geochemical environment associated with each sample. The list of organic constituents was selected by examining several broad-spectrum methods that were capable of detecting several target compounds at once and by reviewing the EPA priority pollutant list as well as the Safe Drinking Water Act. A comparison was made of the list of chemi- cal constituents measured in NAWQA with those measured in EPA the National Oceanic and Atmospheric Administration (NOAA), and the Fish and Wildlife Service programs (Tables 3.2 and 3.3, Appendix C), and attention was given to complementing the other programs rather than duplicating their analyses. In general, the suites of synthetic organic compounds to be measured in surface water will be different than those for ground water. Also, recon- naissance level sampling will be done to determine whether a particular group of organic compounds needs to be measured more extensively. Importantly, as knowledge and technology in trace organic analysis improves with time, the list of organic targets may be modified. Critique The NAWQA program has selected a broad-based list of physical and chemical constituents, which will allow a number of potential water quality issues to be addressed. The list of con- stituents is thorough and well thought out. A group of scientists from both inside and outside the USGS studied and selected the constituents and methods while considering the efforts of other environmental agencies ant! striving to complement the other programs (Rickert, personal communication, 1989~. NAWQA has properly selected hydrophilic compounds to measure in the water samples and selected hydrophobic compounds to measure in tissues based upon their octanol/water partition coefficients. One decision by NAWQA program personnel was to not add dioxin and furans (Table 3.3, Appendix C) to the NAWQA list of constituents because of the prohibitive cost and the small number of labs that perform analysis. The committee agrees that this is a wise choice since the EPA already gathers extensive dioxin data; however, the EPA's dioxin data should be included in NAWQA's retrospective reports. The committee urges a strong coordination between biologists and chemists at the study unit level role so that the choice of sites for ecological sampling should be well coordinated with sites chosen for physical and chemical monitoring in NAWQA. Analysis of the same toxic trace metals and organics in water and tissue samples will help correlate exposure to bioaccumulation. Also, in

54 NAWQA Pilot Program the determination of pesticides in fish, the same constituent should be measured in both water and sediment to get a distribution of the pesticide into the various environmental compartments. If a regional study issue is believed to be a national issue, then a USGS oversight committee should determine whether it is appro- priate to pursue that issue on a national scale. This flexibility will also enable the NAWQA program to detect emerging water quality issues rather than focus on the classical or current water quality constituents. In order to evaluate those emerging water quality problems, the USGS should consider a library archive of spectroscopic data, such as ultraviolet (UV), infrared (IR), nuclear magnetic resonance (NMR), or gas chromatograph/mass spectrometer (GC/ME;) scans, from sample extracts so that years from now, when a new chemical issue emerges, the fingerprints will provide some clue as to the background levels or magnitude of the problem. We cannot know today what the water quality issues of the year 2000 are going to be; thus archived spectroscopic information would be very useful. A real strength of NAWQA lies in the nationally consistent data gathered from sensitive, consistent analytical methods. This will allow water quality data to be compared across the United States for many years. This consistency is the key to making long- term assessments. Summary The committee agrees with the choice of chemical and physical constituents selected for the NAWQA pilot program. The consis- tent water quality data base across the nation will be a real strength of the program. Also, the flexibility of NAWQA to add constituents of local or regional interest will facilitate the iden- tification of emerging water quality issues. Therefore, the committee recommends the following: · The EPA's dioxin data, as well as other pertinent water quality data, should be included in the USGS retrospective reports. · A strong coordination role should be created to manage the integration of ecological data with chemical and physical data collection. · An oversight committee should review the constituents that are added at the regional level to determine if a regional water quality problem is really a national problem. · The USGS should consider establishing a library archive of spectroscopic fingerprints of all water samples.

NAWQA Design Evaluation Description 55 Choice of Biological Constituents To accomplish the stated goals of NAWQA, it will be essential to combine the results of physical, chemical, and biological studies. Biological studies in NAWQA will attempt to meet the following goals: 1. Determine the occurrence and distribution of fecal con- tamination through synoptic surveys for fecal indicator bacteria. 2. Determine the occurrence (and bioavailability) of potentially toxic trace elements and synthetic organic compounds through animal and plant tissue analysis. 3. Assess the relations between physical and chemical charac- teristics of streams and functional or structural aspects of the biological community through ecological surveys. 4. Define and quantify biological processes that affect physical and chemical aspects of water quality (Appendix B. =29~. Fecal contamination will be assessed in NAWQA with analyses of available information and by measuring concentrations of Escherichia colt (E. ~ during each 'of the 3 years of the active phase as part or annual synoptic surveys. The objectives of these surveys are to describe the occurrence and distribution of E. cold within surface waters of the study unit and to relate them to pat- terns of land use and waste management practices. The techniques for doing this have been described in a' protocol (Appendix B. #30) and used in surface water pilot projects. The synoptic survey conducted during the pilot phase of the 'Yakima project (Appendix B. #69) provides an example of the results of such a survey. Sampling for contaminants in tissues in NAWQA will be done in three phases: reconnaissance, synoptic, and nonintensive. During the first phase, samples from relatively few sites will be tested for a large number of contaminants. In the second phase (intensive sampling phase) samples from more sites will be tested for a limited set of contaminants. The tissue analysis protocol (Appendix B. #31) sets out a decision tree to be utilized in deter- mining which species to sample at a site. This will result in a suite of target species for the national survey. The introduced and widely distributed Asiatic clam, Corbicula fluminea' will be the target species where it is present. Other mollusc species will be second choice, then fish species, aquatic insects, and last, sub- mersed macrophytes.

56 NAWQA Pilot Program The objective of the ecological survey in NAWQA is to docu- ment the current status and long-term trends in biological com- munities and to provide background ecological information to complement other NAWQA measurements. The protocol (Appendix B. =33) proposes a core set of a few measurements using similar approaches in all study units at sites similarly selected. Stage I sampling will be done in year 1 of the active phase and will include reconnaissance sampling of 30 to 50 sites with qualitative sampling of benthic invertebrates. Stage II sampling will be done at about 15 sites during years 2 and 3 with sampling for fish abundance and diversity once per year. In addition to the sites used for synoptic sampling in other parts of NAWQA, sites used in the ecological survey will also include low (first through third) order sites selected in pairs so that one member represents a refer- ence site, and the second is representative of a particular type of land use. Quantitative description of habitats (e.g. descriptions of riparian vegetation, bank and substrate characteristics, photo documentation) will be assessed at each site. Critique Fecal Contamination The presence of fecal contamination in surface and ground water has routinely been estimated using either the total or, more commonly in the recent past, the fecal coliform test (American Public Health Association, 1989~. Thus the vast majority of historical sanitary data is reported as total or fecal coliform. However, the NAWQA program proposes to assess the sanitary quality of surface water by testing specifically for the presence of the bacterium E. coli. This bacterium is associated with the feces of warm blooded animals, including humans, and makes up an important subset of coliforms which are measured in the fecal conform test. This method has been chosen by NAWQA because the EPA has found a better statistical correlation between suspected gastrointes- tina1 diseases among swimmers and the E. cold test results than with the number of fecal coliforms found in the nonmarine bathing waters tested (Dufour, 1984~. On this basis, the use of this test for the NAWQA program is justified. Because historical data as well as data currently being collected by other agencies are reported in terms of total or fecal coliforms there is concern that the use of the E. Coli test alone will provide information which cannot be compared to prior bacteriological water quality analyses. If such comparisons are warranted then it

NAWQA Design Evaluation 57 will be necessary to perform the more traditional fecal coliform tests as well as the E. Coli test. The protocol developed for the detection of fecal contamination (Appendix B. #30) provides a clear description of the techniques to be used. If all projects follow that protocol, the data should be comparable within the program. For synoptic surveys to be of maximum value, fecal contamination should be assessed at the same time as synoptic sampling for nutrients and dissolved oxygen. The frequency of monitoring for fecal contamination should be such that (1) statistical variations in the collected data can be considered in evaluating differences between stations, (2) at critical sites where fecal contamination and human exposure vary sea- sonally, such variation can be evaluated, and (3) the impact of changes in wastewater discharge practices (such as a change in chlorination practices) can be assessed. Clearly only one sample per year per station will be of limited value in addressing these issues. Monthly sampling at selected fixed stations in addition to that done during synoptic surveys is recommended. To date NAWQA has collected data concerning fecal con- tamination in only one study unit during one synoptic survey (Appendix B. #69~. Therefore, it is too early for the committee to evaluate the ability of the program to provide regional or national assessments or to evaluate long-term trends. Tissue Analyses Biological tissues provide the best information about the availability of contaminants to the biota, provide a time- averaged assessment of contaminants, and increase the probability of detecting trace contaminants because of higher concentrations in tissues (Appendix B. #29 and 31~. The protocol describing how tissue analysis will be used in NAWQA (Appendix B. #31) provides an excellent summary of existing tissue monitoring, a clear ratio- nale for choice of contaminants to be analyzed, a decision tree to use in selecting target species, and a detailed protocol for sampling tissue removal and sample preservation. A fully implemented NAWQA could fill an existing void in tissue monitoring programs because of the broad coverage of freshwaters, consistency in target species sampled, and its ability to relate contaminants in tissue samples to other physical, chemical, and biological data at the site. Field testing of this protocol in the Yakima and Illinois River pilot projects began only recently (summer 1989~. Samples located were sent to laboratories under contract to the U.S. Fish and Wildlife Service (USFWS), which is also responsible for quality control. Since the results of this work will not be available until

58 NAWQA Pilot Program summer 1990, the committee cannot evaluate this component of the effort, although it seems to be a wise decision to use the USFWS labs so that the data will be compatible and complementary to those being collected as part of the National Contaminant Biomoni- toring Program (NCBP). The U.S. Geological Survey has also begun discussions with the National Institute of Standards and Technology (NIST) regarding long-term cryogenic storage of tissue samples. The archiving of carefully selected samples is a critical component of any long-term monitoring program. It is impossible to determine what substances will be contaminant problems in the future, and hence it is essen- tial that considerable attention be paid to archiving samples that will be useful for future analyses. The USGS should adopt proto- cols for procedures to use for archiving samples and maintaining the archives. Equally important is devising protocols to guide study units in selecting how many and which samples are to be ar- chived. Archiving samples is costly; hence researchers at sampling sites must be judicious in their choices of samples to archive. Samples should be archived from currently contaminated as well as reference sites. It may be wise to make the number of samples to be stored a function of the variability observed at the site rather than simply archiving a fixed percentage at all sites. Based on a careful assessment of the literature (Appendix B. =32), the USGS has decided not to include toxicity testing as a routine part of NAWQA. Toxicity testing is not appropriate for a long-term monitoring program at this time because: tests results are procedure and species dependent, experimental systems do not adequately duplicate the com- plexity or variability of natural systems, and · tests are often not adequately sensitive. Ecological SurveYs The inclusion of an ecological survey in NAWQA is an important addition to the more traditional physical and chemical measures of water quality. The biota provide an integrated picture of water quality because they are exposed to the range of stresses at a site and may respond to a short-term change in water quality (e.g., a spill) that would be missed by the routine sampling effort. In addition, sensitivity to different stresses varies across taxonomic groups. Hence, an assessment of the biotic com- munity offers a level of interpretation not available with only physical or chemical measures of water quality. Furthermore, much of the public concern over water quality stems from a desire for the protection of public health and aquatic life, and it is

NAWQA Design Evaluation 59 valuable to include these aspects in a water quality monitoring program. It is clear that the ecological survey portion of the program is still in the methods development phase. Preliminary field sampling (Appendix B. #55) has provided some useful insights on using fish as indicators of water quality. However, there are no data with which to judge the appropriateness of methods proposed for inver- tebrates, algae, or habitat documentation. In the material reviewed by the committee, there has been no discussion of how the biologi- cal data collected will be incorporated into a national or even regional synthesis. Despite considerable recent progress, the bio- logical component of NAWQA lags behind other aspects. Effort has been put into designing appropriate sampling protocols, and it is now time to provide examples of how these data are to be interpreted to provide a regional and national synthesis. Future NAWQA review committees will need to consider how well this sampling program meets the described objectives. This committee did not have the data necessary to make this assessment. Biological Processes Affecting Water Oualitv Measurements of the rates of biogeochemical processes (e.g., Vitrification and denitrifica- tion) controlling concentrations of elements in water will not be a routine part of NAWQA (Appendix B. #29~. They may be in- cluded as a part of selected reach studies if the problem warrants their inclusion. The decision not to include these measurements is a consequence of fiscal constraints and lack of agreement among scientists as to the best techniques to use. The absence of these measures will limit the ability to develop mechanistic models of transformations at NAWQA study sites and will limit the ability of the NAWQA program to understand some of the aspects of cause and effect relationships. The extent to which modeling and under- standing will be limited will depend on local conditions and could be anticipated by the early application of models in the study unit. Summary Fecal contamination will be assessed using the E. cold test. Because the use of this method-will limit the ability of NAWQA teams to relate their data either to historical patterns or to concur- rent measurements made by other agencies, the committee recom- mends that parallel standard tests for the presence of fecal coli- form also be made. To provide a measure of temporal variation in

60 NAWQA Pilo! Program fecal contamination, the committee recommends including monthly E. cold and fecal coliform tests at a limited number of fixed-station sites in study units where fecal contamination is a potential pro- blem. The procedures for analysis of contaminants in tissues appear adequate to provide an assessment of tissue contamination that fills a void in existing tissue monitoring programs. The committee recommends that USGS adopt protocols for archiving and maintaining tissue samples, and that they devise protocols to guide site personnel in selecting how many and which samples to archive. The ecological survey will use a core set of measurements (e.g., algal biomass, invertebrate and fish abundance and diversity, and habitat characterization) to document the cur- rent status and long-term trends in biological communities. The inclusion of an ecological survey is a valuable addition to NAWQA. Because there are few data from the pilot projects with which to evaluate the ability of the tissue monitoring and ecological survey efforts to meet the stated objectives, the committee recommends that another scientific advisory group review this component of the program when more data are available. Quality Assurance/Quality Control Description The NAWQA's quality assurance (QA) program consists of a matrix of record keeping, methods and prototype validation, and standardized field and laboratory activities which will be consis- tent across the NAWQA network (Mattraw et al., 1989~. The QA program requires that: 1. data collection activities follow USGS approved methods, 2. data analysis and interpretation procedures are documented and capable of being verified, and 3. reports are technically and editorially sound and consistent with USGS policies. Quality assurance will employ the following: 1. field blanks to deal with contamination, 2. field spikes to deal with degradation of sample constituents, 3. multiple samples at the same time and location to deal with sampling precision,

NAWQA Design Evaluation 61 4. splits of collected samples sent for lab analyses to account for lab precision, and 5. standard samples sent as unknown samples to lab to account for lab accuracy. In terms of personnel, one full-time person will be responsible for QA on the headquarters staff, and one person per regional NAWQA team will be assigned to oversee QA on all the projects within that region (D. Rickert, personal communication 1989~. Most of the laboratory analyses for the NAWQA pilot program have been provided by the National Water Quality Laboratory (NWQL). The Branch of Quality Assurance (BQA) has two respon- sibilities: developing, implementing, monitoring, and updating a water resources division-wide QA program that addresses sample collec- tion, field measurements, laboratory analysis, and data manage- ment; and 2. providing technical administrative guidance to field, labora- tory, and data base QA programs. The BQA had responsibility for providing an unbiased, external review of the QA program associated with the pilot NAWQA program. As of this writing 2/90--confirmed with W. Wilber via telephone conversation), a decision has not yet been made to use either a government owned, contractor operated (GOCO) or an in- house operation for lab analyses. Several of the USGS personnel would like to retain the in-house operation. Based on an economic analysis (W. Wilber and W. Alley, personal communication 1989), it would be less expensive to use in-house lab analyses because a GOCO operation would require an additional 5 percent of current laboratory support. The pilot projects underwent periodic review to ensure that the quality of the data, interpretations, and reports met the standard of the USGS. Periodic reviews provided an objective assessment of the effectiveness of project work activities, procedures, and docu- mentation. These reviews were conducted by a review team con- sisting of members from the regions, headquarters, BQA, National Research Program (NRP), and other districts. Each team prepared a report summarizing their findings from the review. The reviews were thorough and followed the quality assurance checklist for surface-water projects or the checklist for ground water projects. Each major element of the study unit plan was examined for adherence to the QA/QC procedures. The data were examined for

62 NAWQA Pilot Program completeness and conformance to information management guide- lines. Laboratory QA samples were checked against NAWQA minimum standards and for proper review and (if needed) action. Critique The USGS appears to be committed to ~ sound practice of quality assurance and quality control. The QA program assures that technically sound procedures will be used in the NAWQA program. The reviews demonstrated that the proposed QA/QC program is working in the study units. The review process provided a com- plete examination of the QA/QC program in the study unit. It identified deficiencies and actions to correct them. In addition, the reviews identified areas where additional national guidance should be considered to aid the study units. The following were items mentioned in more than one study unit review: 1. concerns with aspects of low-level trace organic analyses; 2. the need for national archival guidelines; and 3. more guidance on how to evaluate the QA laboratory samples. The site review program is an effective process for assuring the integrity of the NAWQA quality assurance program. Summary The proposed QA plan is a sound approach for managing the NAWQA data bases to ensure utility and longevity beyond the first rotation of sites. In the full NAWQA program, periodic data base reviews will need to be continued to ensure completeness and accuracy of data. The results of all the pilot project audits should be reviewed for recurring areas of concern--particularly with monitoring for low-level organic compounds. The committee be- lieves that the QA program is a strong component of the NAWQA program. For the program's stated purposes it is desirable to have one agency, preferably a nonregulatory agency, responsible for a program of this scale. Consistency of sampling protocols and data collection will aid in analysis and be essential for timely and meaningful interpretation and synthesis. It is important that the headquarters person in charge of QA be in regular communication with the four regional NAWQA QA personnel to coordinate their activities closely, thereby ensuring that comparable standards and practices are being employed across the full NAWQA network of sites.

NAWQA Design Evaluation 63 There is a lack of consistency in terms of water quality data collected by federal agencies. In the EPA 305(b) program, each state defines its water quality issues (USEPA, 1989~. The com- mittee would like to see more coordination of the NAWQA program with other agencies, especially in such areas as uniformity of water quality parameters monitored and more consistent sampling and analytical techniques. The NAWQA program has the advantage over EPA's 305(b) program owing to its coordinated effort among the four USGS regions in terms of data collection protocols and methods. While local issues are important, the committee believes that the national synthesis planned in NAWQA will be a contribu- tion. In time, after a full NAWQA has been operational, it would be desirable for the USGS to determine how much overlap exists between the two programs, NAWQA and 305(b) to avoid costly duplication of effort. At the very least, both agencies should have access to the data from both programs. Data Management Description For data management (i.e., data documentation, collection, archiving, and retrieval), the pilot studies have used the USGS National Water Data Storage Retrieval System (WATSTORE) and the National Water Information System (NWIS) data bases, which are periodically transferred to EPA's STORET system. For the full NAWQA program, a new system, NWIS-2, is being designed and is targeted for operation in FY 1992. The development of this system is currently under way as part of the upgrading of computer resources for the Water Resources Division of the USGS. Critique WATSTORE is on a 15-year-old Amdah} Computer, which uses outdated computer technology from the 1960s and 70s (W. Wilber, personal communication 1989~. People unfamiliar with WATSTORE have great difficulty using the system. As part of the computer upgrade mentioned above, the USGS is developing speci- fications for a new water quality information system. Eventually, NWIS-2 will replace WATSTORE (W. Wilber, personal communica- tion 1989~. Currently, state data bases are on Prime computers that are uploaded to the national WATSTORE system and are periodi- cally transferred to EPA's STORET system. It is not clear whether

64 NAWQA Pilot Program the new system will handle the Geographical Information System (GIS) data any differently from what was done in the pilot studies. The committee endorses the revamping of WATSTORE and en- courages the USGS to work quickly to develop and implement NWIS-2. The NAWQA criterion for acceptable storage and documenta- tion is to ensure that the data records and documentation will be useful to a succeeding project team 6 or more years after comple- tion of the initial investigative phase (i.e., beyond the first cycle of activity). In addition to the sanitary water quality data, the ecological survey is documenting the current status of the biologi- cal community (fish and invertebrate distribution data and tissue sampling for bioaccumulation) at each NAWQA site to provide a basis of the spatial variation and assess long-term changes in the biological community. The data management plan for the Ecologi- cal Survey is currently being revised. Sharing information and data files with the USGS within the scientific community is not easy because very few of the USGS personnel are connected to the Internet, a collection of local net- works (campus, governmental, and industrial) linked together by regional networks and attached to a national backbone (NSF Net). The purpose of the Internet is to provide high-speed communica- tion among members of the scientific community and to provide access to remote computing resources. Linking the USGS scientists to the scientific community via Internet would facilitate greater collaboration and coordination. Summary The committee believes it would be much easier to communicate with the USGS if their personnel and their computers had access to the Internet system. This would facilitate better coordination with universities and other research facilities and agencies. The committee endorses the revamping of WATSTORE and encourages the USGS to work quickly to develop and implement their new national water information system. ANALYTICAL FRAMEWORK Description For reasons discussed in Chapter 1, the committee firmly believes that a national assessment of our nation's waters must take a strong process-oriented approach. Such an approach improves

NAWQA Design Evaluation 65 basic understanding of the physical, chemical, and biological processes that contribute to water quality and is essential to intel- ligent water-resource management. The committee further believes that the basic design of NAWQA is strongly process-oriented, and that the USGS is well suited to execute such a design. Much of the NAWQA design reflects an emphasis on hydrologi- cal processes. Study units are based on hydrological boundaries. Monitoring locations will be determined based on understanding of the relevant flow system. Synoptic surface water surveys provide opportunities to sample distinct hydrological events. Reach studies ~ ~ . . ~ ~ ~ . ~ . ~ ~ and local-scale transects enable aetallea analysis OI lmpOrlalll subsystems. The integration of surface water and ground water studies provides sorely needed opportunities to understand the important interactions between surface and ground waters. The USGS is ideally suited to conduct a process-oriented assess- ment of our nation's waters. The overall mission of the USGS is to provide information that will assist resource managers and policy- makers at the federal, state, and local levels in making sound decisions. This information is generated through assessments of the quantity and quality of the nation's natural resources, in- cluding minerals, energy resources, and water. In the case of water resources, the USGS has collected and interpreted data on water quantity and quality for more than 100 years. While many of these investigations have been relatively small in scale, a signifi- cant number have been comparable to a NAWQA study unit. A notable example is the ongoing study of selenium in the San Joaquin Valley in California (Gilliom, et al., 1989.) The committee reviewed the activities of the USGS in this study and was favor- ably impressed with the scientific understanding that has resulted. Furthermore, it is clear to the committee that without such under- standing, management strategies in the region would be poorly grounded. The strength of the USGS in conducting a process-oriented assessment of our nation's waters lies in the fact that it has well- trained, personnel representing a variety of disciplines dispersed throughout the nation. This means that a particular study-unit investigation will not be starting from scratch, since it will have access to personnel who understand the local and regional hydrol- ogy. Given the enormous variation of the hydrologic processes through the nation, such expertise is critical. Therefore, the committee believes that the NAWQA design has a strong process orientation, and that the USGS is well suited to execute the design. However, in the course of evaluating the NAWQA program, the committee also concluded that the ability of NAWQA to elucidate hydrologic processes should be strengthened

66 NAWQA Pilot Program by a greater reliance on mathematical modeling. This issue is discussed in the subsequent sections. Critique An important feature that appears to be lacking from the overall NAWQA program, particularly the regional pilot studies, is the quantitative and analytical framework that identifies the principal cause and effect relationships. While this issue has been qualitatively addressed in NAWQA documents and in comments by USGS personnel, its quantitative definition, which is essential for decisionmaking, appears to be inadequate at this stage. The link between sources, both point and distributed, and the water quality response of a particular system, can be provided by a mathematical model that defines the spatial and temporal distribution of the concentration of relevant water quality constituents. The formula- tion of such a model should be an integral part of the individual studies. Although it has been stated or suggested that some kind of modeling would follow the data collection, there remains a reason- able question that this task may not be effectively realized. More importantly, the structuring of a quantitative model should be incorporated in the early phase of each project, to whatever degree is possible given the nature of the problem and available data. The use of models can serve many purposes in NAWQA. First, a model provides a tool for organizing what is known about a given system, at the scale of interest. Consequently, a model also serves to reveal gaps in knowledge. A model can be used to pre- dict what should be the critical factors affecting water quality (e.g., point versus nonpoint sources). This information can be used in designing the sampling strategy. Once data have been collected, they in turn can be used to refine the model. A model provides a way to account explicitly for meteorologic and hydrologic varia- tions. This makes it much easier to resolve water quality impacts caused by other factors, such as human activities. It also makes it possible to extrapolate for conditions other than those observed during the sampling period. A model can be used to simulate water quality for different scenarios of human activity. Such a capability is essential for evaluating hypothetical management options and is also useful in evaluating the impact of past manage- ment practices. Thus, water quality modeling can be used in all phases of water quality assessment, including design of the monitoring program, analysis of data, and evaluation of past and potential management

NAWQA Design Evaluation options. Of particular importance is the interactive nature of modeling and data collection. These activities are mutually sup- portive, and each should evolve throughout the assessment process. This will lead to a better assessment, as well as to an improved understanding of the relevant physical, chemical, and biological processes. 67 Just as we recognize the necessity for using models in NAWQA, we also realize that the state-of-the-art in water quality modeling may have to be improved if models are to be used effectively in support of NAWQA. While a number of excellent surface water quality models are widely used for management purposes, most notably those supported by EPA, the applicability of these models to large spatial scale water quality assessment is unclear. Existing surface water models were designed to be used at relatively small scales and are most commonly applied to evaluate compliance with standards under hypothetical flow conditions (such as the 7-day, 10-year low flow). In comparison, NAWQA study-unit investiga- tions will be conducted! at enormous scales and will involve a very large suite of water quality constituents over a wide range of flow conditions. With respect to ground water, the USGS has played a leading role in developing models of both flow and quality. Of particular importance is the large-scale flow modeling that was undertaken by the USGS in the RASA program. These models should provide a framework for ground water modeling in many of the NAWQA study units (and in fact, have been used this way to some extent in the pilot studies). What remains, of course, is the incorporation of water quality into these models. A more difficult challenge is the integration of surface water and ground water _.__ ^— and_ arm_. ~—In,. . _^ ~ modeling. Given the lack of appropriate existing models in some cases, the USGSts inexperience in most cases, and the considerable difficulty of the task, some models will prove inadequate or infeasible. Therefore, a midcourse model evaluation should be built into the assessment cycle so that corrections can be made in sufficient time. Summary Water quality modeling should be an essential component of NAWQA. In the short term, the USGS should make use of existing models. Because the USGS has relatively little experience with existing surface water quality models, it will need to begin ~ :~ A ~ ~ : :- ~ ~ There is also a critical need to determine the data required by these models, since the current NAWQA design may not include collection of immenlatelv to nrov~de Or nDtaln the necessary tralDlnE.

68 NAWQA Pilot Program necessary data. (For example, there is no plan to collect meteoro- logical data, which will be essential for watershed models of water quality.) During the first round of intensive monitoring, the leading water quality models should be applied over a wide range of situations. This would not only enhance the assessment effort, but would also provide valuable information on the capabilities of the models. A midcourse mode' evaluation should be built into the assessment cycle so that corrections can be made in sufficient time. Over the long term, the USGS should be developing a coherent strategy for water quality modeling in the context of a large spatial scale assessment, addressing such issues as scale, model complexity, and surface and ground water interactions. To the fullest extent possible, this should be done in cooperation with other federal agencies, particularly the Environmental Protection Agency. PRODUCTS Description The goals of NAWQA are to "describe the status and trends in the quality of the nation's surface and ground water resources and to provide a sound, scientific understanding of the primary natural and human factors affecting the quality of these resources" (Ap- pendix B. #36) in order to "provide an improved scientific basis for evaluating the effectiveness of water quality management programs and for predicting the likely effects of contemplated changes in land and water management practices" (Appendix B. #38~. Essential to achieving this goal of improving the evaluation and prediction of the consequences of management practices is the dissemination of the results of NAWQA to the appropriate audi- ences. The USGS has proposed three broad classes of report topics to be prepared under NAWQA (Appendix B. #38~: 1. statistical descriptions of water quality conditions and changes over time, 2. information on the geographic distribution of contaminants across the United States, and 3. information on key factors that affect water quality. It is the committee's understanding that this information will be provided in USGS water-supply papers, open-file reports, water resources investigations reports, yearbook articles, and other publi- cations, as well as in scholarly journal articles. For example, each

NAWQA Design Evaluation 69 pilot project has been assigned a Water-Supply Paper number. The "A" chapter of each will be a project summary, the "Be chapter will be a retrospective report (see below), and later chapters will be specific to each study unit (Appendix B. #37~. Presentation methods in these reports will include tables, graphs, maps, and supporting text. Raw data will be available via a single, consistent computer data base. Similar products are to be prepared for both the study-unit investigations and the issue-based and national synthesis team studies. Unfortunately, because of the timing of the committee's review, no NAWQA reports presenting results at either the study-unit or synthesis level were available for review, with the exception of retrospective reports for 5 of the pilot projects (Appendix B. #62, 70, 79, 83, 89~. Therefore, our evaluation of NAWQA products is necessarily broad and general for the most part. The retrospective reports available for review focus on compiling available existing data from federal, state, and local agencies, evaluating the suit- ability of these data for NAWQA study unit activities, and pre- senting a preliminary spatial analysis of water quality conditions in the study unit via mapping, summary statistics, and descriptive text. They also contain background material of the study unit, its natural and cultural features, and its hydrologic systems. Critique The retrospective reports available for review have been pre- pared and designed well, and should serve the purpose for which they were intended. During their visits to the pilot projects, committee members were frequently told that the compilation of all available data into one documented source and data base was by itself a significant contribution to local decisionmaking. The reports do exhibit some unevenness in the quality of their analyses, with some study-unit teams apparently being more capable and creative than others in extracting and presenting useful informa- tion from the available data. After examining the more general aspects of the planned products of NAWQA, the committee first noted that the maps and reports at the USGS are in general highly respected among the potential users of NAWQA. Potential users interviewed by the committee were practically unanimous in that opinion. In fact, some local and state users made the point that data and informa- tion provided by the USGS will be far more helpful to them in their water quality management responsibilities than data and information acquired by the users themselves because local and

70 NAWQA Pilot Program state governmental authorities have such high regard for the integrity of the USGS. One concern brought to the attention of the committee during its evaluation is the timeliness of USGS reporting. A number of potential users expressed frustration at the time required for publication of USGS reports, especially those containing data. The committee commends the USGS for its decision to publish the Water-Supply Papers for the pilot projects initially as open-file reports in order to accelerate the availability of this information. However, the committee would like to encourage the USGS to be vigilant and creative in seeking out ways to minimize the time required to publish NAWQA findings. In considering the three broad classes of report topics en- visioned by the USGS for NAWQA, the committee noted that the first two classes will be of greatest interest in support of policy formulation and resource allocation at fairly high levels of deci- sionmaking, while the third class will be of greatest interest to water quality management decisionmaking at the implementation level. It was also noted that the first two classes of reports will in many ways be the easiest to produce. The committee urges the USGS to strike an appropriate balance and assure that adequate resources are directed toward reports describing the results of cause and effect analyses, even if those are more difficult to produce. Many of the potential users of NAWQA with whom the committee spoke, especially on the local and state levels, em- phasized the value they would place on reports bettering their understanding of processes. The committee is most concerned, however, with the relatively narrow scope of the retrospective studies. While the study-unit teams appear to have been thorough in identifying and acquiring available data, the extent of their investigations was quite limited. In particular, the investigations focused almost exclusively on ambient water quality data in streams and aquifers. Data on sources of contamination were not collected and analyzed, in spite of the availability of extensive data bases (e.g., EPA's Permit Compliance System and Industrial Facility Reports). In addition. the retrospective reports make no attempt to identify and assess available studies of water quality processes in the study units or available water quality models that have been developed for all portions of the study units. If the goal of NAWQA is indeed to improve our understanding of water quality processes on the study- unit, regional, and national scales, it is imperative that information on sources and previous investigations of processes be included within the concept of a retrospective report.

NAWQA Design Evaluation Summary The committee recommends that the concept of a retrospective report be expanded to include data on known sources of water quality constituents, existing studies of water quality processes within the study unit, and previously developed conceptual and mathomat;ral mn~l~lc of the nhvsical. chemical. and biological 71 .——~ . ., _ ~ ~ .~ ~ , ~ , _ processes influencing water quality in the study unit. The committee encourages the USGS to be vigilant and creative in seeking ways to minimize the time required to publish the findings from the National Water Quality Assessment program. In addition, the committee urges the USGS to ensure that adequate resources~are directed toward reports describing the results of cause and effect analyses, even if those are more difficult to produce. Committee-Initiated Case Study Description In order to obtain a sharper focus on how a cause and effect analysis would be conducted under the NAWQA program, the committee recommended that the USGS conduct a special study of the effect of changes in wastewater treatment on trends in the downstream water quality of the Illinois River. Wastewater treat- ment was selected as the problem area owing to the large federal investment in these facilities The Illinois River Basin was selected as the study area because ~t Is relatively simple from a hydrologic point of view and an extensive long-term data base already exists. As a result of this request, NAWQA personnel conducted a detailed inventory of past and present wastewater treatment data in this area. Their study focused on the ava~lan~ty and su~t- ability of municipal wastewater treatment information. Their results, indicating several areas where significant improvements are needed, are being featured in a series of three articles in Water Environment and Technology, the journal of the water Pollution Control Federation (Appendix B. #46) and in a USGS Open File Report (Appendix B. #57~. . · . . · . ~ . . · . Critique Although the NAWQA personnel did a very thorough evalua- tion of quality control problems with the existing data, they have not yet completed the originally requested analysis of cause and

72 NAWQA Pilot Program effect relationships for this test study area. Thus the committee cannot evaluate how cause-and effect analysis is to be ac- complished under the NAWQA program. While it is understand- able that the USGS is more comfortable monitoring since it is their traclitional forte, NAWQA needs to be a first-class assessment activity if it is to serve its intended! purpose. Summary The USGS should immediately intensify its cause and effect assessment activities and provide complete case studies and a coherent national methodology for doing cause and effect assess- ments. This initial effort should utilize all available data and not be restricted to data generated by the NAWQA monitoring pro- gram. The results of this effort should be reviewed by an external science advisory committee.