Assessing the TMDL Approach to Water Quality Management (2001)

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.

Executive Summary Over the last 30 years, water quality management in the United States has been driven by the control of point sources of pollution and the use of effluent-based water quality standards. Under this paradigm, the quality of the nation’s lakes, rivers, reservoirs, groundwater, and coastal waters has generally improved as wastewater treatment plants and industrial discharg- ers (point sources) have responded to regulations promulgated under authority of the 1972 Clean Water Act. These regulations have required dischargers to comply with effluent-based standards for criteria pollutants, as specified in National Pollutant Discharge Elimination System (NPDES) permits issued by the states and approved by the U.S. Environmental Pro- tection Agency (EPA). Although successful, the NPDES program has not achieved the nation’s water quality goals of “fishable and swimmable” waters largely because discharges from other unregulated nonpoint sources of pollution have not been as successfully controlled. Today, pollutants such as nutrients and sediment, which are often associated with nonpoint sources and were not considered criteria pollutants in the Clean Water Act, are jeopardizing water quality, as are habitat destruction, changes in flow regimes, and introduction of exotic species. This array of challenges has shifted the focus of water quality management from effluent-based to am- bient-based water quality standards. This is the context in which EPA is obligated to implement the Total Maximum Daily Load (TMDL) program, the objective of which is attain- ment of ambient water quality standards through the control of both point and nonpoint sources of pollution. Although the TMDL program origi- nated from Section 303d of the Clean Water Act, it was largely overlooked during the 1970s and 1980s as states focused on bringing point sources of pollution into compliance with NPDES permits. Citizen lawsuits during the 1980s forced EPA to develop guidance for the TMDL program, which is now considered to be pivotal in securing the nation’s water quality goals. 1 

2 Assessing the TMDL Approach to Water Quality Management Under TMDL regulations promulgated in 1992, EPA requires states to list waters that are not meeting water quality criteria set for specific designated uses. For each impaired water, the state must identify the amount by which point and nonpoint sources of pollution must be reduced in order for the waterbody to meet its stated water quality standards. Meeting these re- quirements, many of which have been imposed by court order or consent decree, has become the most pressing and significant regulatory water quality challenge for the states since passage of the Clean Water Act. Given the most recent lists of impaired waters submitted to EPA, there are about 21,000 polluted river segments, lakes, and estuaries making up over 300,000 river and shore miles and 5 million lake acres. The number of TMDLs required for these impaired waters is greater than 40,000. Un- der the 1992 EPA guidance or the terms of lawsuit settlements, most states are required to meet an 8- to 13-year deadline for completion of TMDLs. Budget requirements for the program are staggering as well, with most states claiming that they do not have the personnel and financial resources necessary to assess the condition of their waters, to list waters on 303d, and to develop TMDLs. A March 2000 report of the General Accounting Of- fice (GAO) highlighted the pervasive lack of data at the state level avail- able to set water quality standards, to determine what waters are impaired, and to develop TMDLs. Subsequent to the GAO report and following issuance by EPA of up- dated TMDL regulations, Congress requested that the National Research Council (NRC) assess the scientific basis of the TMDL program, includ- ing: • the information required to identify sources of pollutant loadings and their respective contributions to water quality impairment, • the information required to allocate reductions in pollutant load- ings among sources, • whether such information is available for use by the states and whether such information, if available, is reliable, and • if such information is not available or is not reliable, what method- ologies should be used to obtain such information. Of concern to the nation’s lawmakers was the paucity of data and informa- tion available to the states to comply with program requirements and meet water quality standards. Indeed, as the TMDL program proceeds, the best available science, especially with regard to nonpoint sources of pollution, will be needed for regulatory and nonregulatory actions to be equitable and 

Executive Summary 3 effective. Report recommendations are targeted (1) at those issues where science can and should make a significant contribution and (2) at barriers (regulatory and otherwise) to the use of science in the TMDL program. Chapters 2, 3, and 4 discuss the information required to set water quality standards, to list waters as impaired, and to develop TMDLs (including the identification of pollution sources), while Chapter 5 discusses the role of science in allocating pollutant loading among sources. Chapters 3 and 4 go into considerable detail about the monitoring, modeling, and statistical analysis methods needed to collect data and convert it to information, and to assess and reduce uncertainty. This report represents the consensus opinion of the eight-member NRC committee assembled to complete this task. The committee met three times during a three-month period and heard the testimony of over 40 in- terested organizations and stakeholder groups. The NRC committee feels that the data and science have progressed sufficiently over the past 35 years to support the nation’s return to ambient-based water quality management. Given reasonable expectations for data availability and the inevitable limits on our conceptual understanding of complex systems, statements about the science behind water quality management must be made with acknowl- edgment of uncertainties. The committee has concluded that there are creative ways to accommodate this uncertainty while moving forward in addressing the nation’s water quality challenges. These broad conclusions are elaborated upon below. TMDL PROGRAM GOALS The TMDL program should focus first and foremost on improving the condition of waterbodies as measured by attainment of designated uses. Work on meeting the strict time demands within the budget con- straints cited by most states has focused on administrative outcomes as measures of success for the TMDL program. However, the success of the nation’s premier water quality program should not be measured by the number of TMDL plans completed and approved, nor by the number of NPDES permits issued or cost share dollars spent. Success is achieved when the condition of a waterbody supports its designated use. Adequate monitoring and assessment must be used to improve the listing of impaired waterbodies and to characterize the effectiveness of the actions taken to meet the designated use. 

4 Assessing the TMDL Approach to Water Quality Management The program should encompass all stressors, both pollutants and pollution, that determine the condition of the waterbody1. Proposed regulations may limit the applicability of the program to only those water quality problems caused by chemical and physical pollutants. Given their demonstrated effectiveness, activities that can overcome the effects of “pollution” and bring about waterbody restoration—such as habitat resto- ration and channel modification—should not be excluded from considera- tion during TMDL plan implementation. Scientific uncertainty is a reality within all water quality pro- grams, including the TMDL program, that cannot be entirely elimi- nated. The states and EPA should move forward with decision-making and implementation of the TMDL program in the face of this uncertainty while making substantial efforts to reduce uncertainty. Securing desig- nated uses is limited not only by a focus on administrative rather than wa- ter quality outcomes in the TMDL process, but also by unreasonable ex- pectations for predictive certainty among regulators, affected sources, and stakeholders. CHANGES TO THE TMDL PROCESS This report focuses on how scientific data and information should be used within the TMDL program. Science plays a crucial role in the stan- dards-setting process, in the decision to add waters to the 303d list, in the development of the TMDL plan, and in the allocation of pollutant loads among various sources (although its importance relative to the role of pol- icy decisions varies). The committee finds that although the state of the science is sufficient to develop TMDLs to meet ambient water quality goals in many situations, programmatic issues substantially hinder the use of the best available science. Thus, the following changes in the TMDL process are recommended, with an understanding that without such changes, the TMDL program will be unable to incorporate and improve upon the best available scientific information. States should develop appropriate use designations for waterbod- ies in advance of assessment and refine these use designations prior to TMDL development. Clean Water Act goals of fishable and swimmable waters are too broad to be operational as statements of designated uses. 1 This refers to the legal definitions of “pollutant” and “pollution,” which are given in Box 1-1 of Chapter 1. 

Executive Summary 5 Thus, there should be greater stratification of designated uses at the state level (such as primary and secondary contact recreation). The appropriate designated use may not be the use that would be realized in the water’s predisturbance condition. Sufficient science and examples exist for all states to inject this level of detail into their water quality standards. To ensure that designated uses are appropriate, use attainability analysis should be considered for all waterbodies before a TMDL is developed. EPA should approve the use of both a preliminary list and an ac- tion list instead of one 303d list. Many waters now on state 303d lists were placed there without the benefit of adequate water quality standards, data, or waterbody assessment. These potentially erroneous listings con- tribute to a very large backlog of TMDL segments and foster the percep- tion of a problem that is larger than it may actually be. States should be allowed to move those waters for which there is a lack of adequate water quality standards or data and analysis from the 303d list back to a prelimi- nary list, as shown in Figure ES-1. This would provide the assurance that listed waters are indeed legitimate and merit the resources required to complete a TMDL. If no legal mechanism exists to bring this about, one should be created by Congress. The data requirements and other criteria that should be used to differentiate the preliminary list from the action list are discussed in the report. No waterbody should remain on the prelimi- nary list for more than one rotating basin cycle. TMDL plans should employ adaptive implementation. As shown in Figure ES-2, adaptive implementation is a cyclical process in which TMDL plans are periodically assessed for their achievement of water qual- ity standards including designated uses. If the implementation of the TMDL plan is not achieving attainment of the designated use, scientific data and information should be used to revise the plan. Adaptive imple- mentation is needed to ensure that the TMDL program is not halted be- cause of a lack of data and information, but rather progresses while better data are collected and analyzed with the intent of improving upon initial TMDL plans. Congress and EPA need to address the policy barriers that inhibit adoption of an adaptive implementation approach to the TMDL program, including the issues of future growth, the equitable distribution of cost and responsibility among sources of pollution, and EPA oversight. 

6 Assessing the TMDL Approach to Water Quality Management All Waters Determine Designated Use/ Standard Screening Assessment “Preliminary” List Full Review Use/ Assessment Standard “Action” List (303d) TMDL Planning Adaptive Implementation FIGURE ES-1 Framework for water quality management. 

Executive Summary 7 TMDL Plan Immediate and Experiments Model Long-term Actions/ Refinement Monitoring no Meeting Designated Use? yes Back to initial list of all waters for continuing assessment in the rotating basin process FIGURE ES-2 Adaptive implementation flowchart. USE OF SCIENCE IN THE TMDL PROGRAM This report suggests changes in the data used and analytical methods employed that will support the revisions to the TMDL process recom- mended above. The following sections highlight the use of science in the TMDL program steps as illustrated in Figure ES-1. Additional recommen- dations about the scientific basis of the program not included in this execu- tive summary are found throughout the report. Water Quality Standards The TMDL process is primarily a measurement process and as such is significantly impacted by the setting of water quality standards. Water 

8 Assessing the TMDL Approach to Water Quality Management quality standards consist of two parts: a specific desired use appropriate to the waterbody, termed a designated use, and a criterion that can be meas- ured to establish whether the designated use is being achieved. The criterion used to measure whether the condition of a water- body supports its designated use can be positioned at different points along the causal chain connecting stressors (such as land use activities) to biological responses in a waterbody. Positioning the criterion involves a trade-off between forecast error for the stressor–criterion relationship and the adequacy of the criterion as a measure (surrogate) for the designated use. Model results that forecast the impact of the stressor on the criterion are likely to be more uncertain as the criterion is positioned farther from the stressor and closer to the designated use. On the other hand, position- ing the criterion closer to the stressor and farther from the designated use is likely to mean that the criterion is a poorer measure or surrogate for the designated use. Biological criteria should be used in conjunction with physical and chemical criteria to determine whether a waterbody is meeting its designated use. In general, biological criteria are more closely related to the designated uses of waterbodies than are physical or chemical measure- ments. However, guiding management actions to achieve water quality goals based on biological criteria also depends on appropriate modeling efforts. All chemical criteria and some biological criteria should be defined in terms of magnitude, frequency, and duration. The frequency com- ponent should be expressed in terms of a number of allowed excursions in a specified period. Establishing these three dimensions of the criterion is crucial for successfully developing water quality standards and subse- quently TMDLs. Water quality standards must be measurable by reasonably ob- tainable monitoring data. In many states, there is a fundamental discrep- ancy between the criteria that have been chosen to determine whether a waterbody is achieving its designated use and the frequency with which water quality data are collected. This report gives examples of this phe- nomenon and makes suggestions for improvement. Waterbody Assessment and Listing Ambient monitoring and assessment programs should form the basis for determining whether waters are placed on the preliminary list or the action list. 

Executive Summary 9 EPA needs to develop a uniform, consistent approach to ambient monitoring and data collection across the states. The rotating basin ap- proach used by several states is an excellent example of a framework than can be used to conduct waterbody assessments of varying levels of com- plexity, for example to support 305b reports, to place impaired waters on a preliminary list or action list, and to develop TMDLs. In that regard, EPA should set the TMDL calendar in concert with each state’s ro- tating basin program. Evidence suggests that limited budgets are preventing the states from monitoring for a full suite of indicators to assess the condition of their waters and from embracing a rotating basin approach to water quality management. Currently, EPA is assessing the sufficiency of state resources to develop and implement TMDLs. Depending on the results of that assessment, Congress might consider aiding the states, for example through matching grants to improve data collection and analysis. Evaluated data and evidence of violation of narrative standards should not be exclusively used for placement of a waterbody on the action list, but is useful for placement on the preliminary list. EPA should develop guidance to help states translate narrative standards to nu- meric criteria for the purposes of 303d listing and TMDL calculation and implementation. EPA should endorse statistical approaches to defining all waters, proper monitoring design, data analysis, and impairment assessment. For chemical parameters, these statistical approaches might include the binomial hypothesis test or other methods that can be more effective than the raw score approach in making use of the data collected to determine water quality impairment. For biological parameters, they might focus on improvement of sampling designs, more careful identification of the com- ponents of biology used as indicators, and analytical procedures that ex- plore biological data as well as integrate biological information with other relevant data. TMDL Development The scientific basis of the latter half of the TMDL process revolves around a wide variety of models of varying complexity that are used to re- late waterbody conditions to different land uses and other factors. Models are a required element of developing TMDLs because water quality stan- dards are probabilistic in nature. However, although models can aid in the 

10 Assessing the TMDL Approach to Water Quality Management decision-making process, they do not eliminate the need for informed deci- sion-making. Uncertainty must be explicitly acknowledged both in the models selected to develop TMDLs and in the results generated by those mod- els. Prediction uncertainty must be estimated in a rigorous way, models must be selected and rejected on the basis of a prediction error criterion, and guidance/software needs to be developed to support uncertainty analy- sis. The TMDL program currently accounts for the uncertainty em- bedded in the modeling exercise by applying a margin of safety (MOS); EPA should end the practice of arbitrary selection of the MOS and instead require uncertainty analysis as the basis for MOS deter- mination. Because reduction of the MOS can potentially lead to a signifi- cant reduction in TMDL implementation cost, EPA should place a high priority on selecting and developing TMDL models with minimal forecast error. EPA should selectively target some postimplementation TMDL compliance monitoring for verification data collection so that model prediction error can be assessed. TMDL model choice is currently ham- pered by the fact that relatively few models have undergone thorough un- certainty analysis. Postimplementation monitoring at selected sites can yield valuable data sets to assess the ability of models to reliably forecast response. EPA should promote the development of models that can more ef- fectively link environmental stressors (and control actions) to biologi- cal responses. A first step will be the development of conceptual models that account for known system dynamics. Eventually, these should be strengthened with both mechanistic and empirical models, although em- pirical models are more likely to fill short-term needs. Such models are needed to promote the wider use of biocriteria. Monitoring and data collection programs need to be coordinated with anticipated water quality and TMDL modeling requirements. For many parameters, there are insufficient data to have confidence in the results generated by some of the complex models used in practice today. Thus, EPA should not advocate detailed mechanistic models for TMDL development in data-poor situations. Either simpler, possibly judgmental, models should be used or, preferably, data needs should be anticipated so that these situations are avoided. In order to carry out adaptive implementation, EPA needs to fos- ter the use of strategies that combine monitoring and modeling and expedite TMDL development. This should involve the use of Bayesian 

Executive Summary 11 techniques that can combine different types of information. Although the modeling framework proposed in this report calls for improvements in models, there are existing models that can be applied rapidly and effec- tively within an adaptive implementation framework. FINAL THOUGHTS Through the adoption and use of the preliminary list/action list ap- proach, adequate monitoring and assessment approaches, sound selection of appropriate models, and adaptive implementation described in this re- port, the TMDL program will be capable of utilizing the best available sci- entific information. It is worth noting that the success of these approaches is directly related to the provision of adequate personnel and financial re- sources for data collection, management, and interpretation and for the de- velopment of sufficiently detailed and stratified water quality standards.