902 F. 2d 962 - Natural Resources Defense Council Inc v. Administrator US Environmental Protection Agency

Separate opinion filed by Chief Judge WALD.

Concurring and dissenting opinion filed by Circuit Judge HARRY T. EDWARDS.

Concurring and dissenting opinion filed by Circuit Judge SILBERMAN.

In these consolidated cases, petitioners seek review of several aspects of the Revisions to the National Ambient Air Quality Standards for Particulate Matter, 52 Fed.Reg. 24,634 (1987) ("Revisions"), issued on July 1, 1987 by the Environmental Protection Agency ("EPA" or "agency"). For various reasons, we dismiss all of the challenges to the Revisions raised by the American Iron and Steel Institute and by the American Mining Congress; in addition, while we dismiss the challenge raised by the Natural Resources Defense Council and affiliated petitioners vis-a-vis a secondary ambient air quality standard protecting against visibility impairment, we order EPA to submit a statement of reasons within sixty days of the issuance of the mandate of this opinion explaining its decision not to initiate a rulemaking on a secondary standard protecting against acid deposition.

Before presenting a brief overview of the relevant portions of the Clean Air Act ("CAA" or "Act"), 42 U.S.C. Secs. 7401-7626 (1982), and of the regulatory developments culminating in the 1987 rulemaking, we describe the regulated pollutant--particulate matter. The term "particulate matter" includes the discrete particles captured and filtered by an air-sampling device meeting certain specifications. Particulate matter originates from stationary and mobile sources, both natural and man-made, such as the earth's crust, sea salt, tires and various industrial and agricultural processes.

The 1987 Revisions focus on particles with an aerodynamic diameter of 10 microns or less. "PM10 " is the abbreviation used to describe this group of particulate matter. Prior to the Revisions, particulate matter standards applied to a more inclusive subset of particulate matter, known as "total suspended particulates," or "TSP."

The Clean Air Act provides a comprehensive program for control of air pollution through a system of shared state and federal responsibility. Under Sec. 108 of the Act, EPA identifies air pollutants that are emitted from "numerous or diverse" sources and whose presence in the ambient air "may reasonably be anticipated to endanger public health or welfare." Id. Sec. 7408(a)(1). For each pollutant, EPA is required to issue a "criteria" document reflecting its health and welfare effects and a "control techniques" document discussing the costs and benefits of different types of emission controls. Id. Sec. 7408(a)(2), (b)(1).

Under Sec. 109, EPA must issue "primary" and "secondary" national ambient air quality standards ("NAAQS") for each pollutant identified under Sec. 108. See id. Sec. 7409(a)(2). The primary standards must protect the public health while allowing an adequate margin for safety; the secondary standards must protect the public welfare from any known or anticipated adverse effects. See id. Sec. 7409(b). The EPA Administrator is required to complete a thorough review of the standards, and revise them as appropriate, by December 31, 1980 and at five-year intervals thereafter. See id. Sec. 7409(d). Within nine months of the promulgation or revision of an NAAQS, each state must adopt or revise a plan for the attainment and maintenance of the standard. The state plan must be submitted for EPA's approval. See id. Sec. 7410(a).

In the 1977 Amendments to the Clean Air Act, Congress ordered the states to classify each of their air quality regions as in attainment, nonattainment, or unclassifiable with regard to the NAAQS then in effect. See id. Sec. 7407(d). For areas in attainment or unclassifiable, Congress prescribed measures to prevent significant deterioration (PSD) of air quality. See id. Secs. 7470-79. To this end, Congress specified the "maximum allowable increase" in concentrations of particulate matter. Id. Sec. 7473(b). States were required to ensure that new or modified emissions sources would not exceed these allowable increments. The 1977 Amendments also included the review and revision timetable of Sec. 109(d)(1), see id. Sec. 7409(d)(1), and directed the EPA Administrator to appoint an independent scientific review committee, see id. Sec. 7409(d)(2).

The 1987 Revisions to the ambient air quality standards for particulate matter were a long time coming. Drawing from a 1969 air quality Criteria Document for particulate matter, EPA first promulgated primary and secondary standards for particulate matter in April 1971. See National Primary and Secondary Ambient Air Quality Standards, 36 Fed.Reg. 8186 (1971). In the late 1970s, EPA began to revise the 1969 particulate matter Criteria Document to correct its scientific shortcomings. This process was completed when the Revised Criteria Document for particulate matter was published in December 1982. See Air Quality Criteria for Particulate Matter and Sulfur Oxides (1982).

On March 20, 1984, EPA proposed revisions to the particulate matter ambient air quality standards. See Proposed Revisions to the National Ambient Air Quality Standards for Particulate Matter, 49 Fed.Reg. 10,408 (1984). In the main, EPA proposed to replace the TSP indicator with PM10 for the primary standard, while retaining TSP for secondary standard, see id. at 10,412, 10,418-19; to select the levels of the revised 24-hour and annual primary standards from specified ranges, see id. at 10,415-17; to issue a secondary (welfare) standard aimed at the soiling and nuisance effects, while deferring a decision on a separate fine particle standard addressing visibility impairment and continuing to evaluate alternative approaches to reduce acid deposition, see id. 10,418-19; and to alter the form of the secondary standard, see id.

After a lengthy notice and comment period and the issuance of several supplemental proposals, EPA issued the final rule revising the particulate matter standards on July 1, 1987. See 52 Fed.Reg. 24,634. EPA replaced TSP with PM10 as the indicator for the primary standards, see id. at 24,638-39; selected annual and 24-hour primary standards from the specified ranges, see id. at 24,641-45; set a PM10 secondary standard for soiling and nuisance identical to the primary standard; and continued to defer on a fine particle standard to address visibility impairment while research proceeded on controlling acid deposition, see id. at 24,645-47.

Relevant details of the Revisions are discussed in the appropriate sections of this opinion.

In December 1988, EPA denied petitions for reconsideration of various aspects of the Revisions. See 53 Fed.Reg. 52,698, 52,705 (1988). This court then consolidated those petitions for review purposes in March 1989. After a joint motion by two parties to stay proceedings in one of the consolidated cases, we substituted No. 87-1438 as lead docket in August 1989.

II. THE AMERICAN IRON AND STEEL INSTITUTE'S CLAIMS

In this Part, we consider the petition from the American Iron and Steel Institute ("AISI"), challenging as arbitrary and capricious the Administrator's selection of 150 Sg/m3 and 50 Sg/m3 as, respectively, the twenty-four hour and annual national primary ambient air quality standards for particulate matter, measured in PM10 . In addition, AISI claims that the Administrator violated Sec. 109(b)(1) of the Clean Air Act, 42 U.S.C. Sec. 7409(b)(1) (1982), when he failed to discuss the degree of "safety" provided by the levels selected for the primary standards, that the Administrator erred in failing to consider the health consequences of unemployment in his assessment of the appropriate level for the primary standard, and that the control techniques information issued by EPA for particulate matter does not meet the requirements of Sec. 108(b)(1) and (c), 42 U.S.C. Sec. 7408(b)(1), (c) (1982). On the record before us, we find no basis to overturn or defer EPA's final rule instituting annual and twenty-four hour national primary ambient air quality standards for particulate matter, measured in PM10 s. We further conclude that AISI lacks standing to challenge the adequacy of the control techniques information issued by EPA. Accordingly, we deny AISI's petition for review of these matters.

In revising the primary standards for particulate matter pursuant to 42 U.S.C. Sec. 7409(d), the Administrator must set the primary standards at levels which in his judgment are "requisite to protect the public health." CAA Sec. 109(b)(1), (d), 42 U.S.C. Sec. 7409(b)(1), (d) (1982). Primary standards must be based on the air quality criteria issued under Sec. 108, and they also must "allow[ ] an adequate margin of safety." Id. The criteria document, which serves as the basis for establishing the pollutant levels, must "accurately reflect the latest scientific knowledge useful in indicating the kind and extent of all identifiable effects on public health ... which may be expected from the presence" of particulate matter "in the ambient air, in varying quantities." CAA Sec. 108(a)(2), 42 U.S.C. Sec. 7408(a)(2) (1982).

EPA completed a revised criteria document for particulate matter in 1982 ("Revised Criteria Document").1 In order to "bridge the gap" between the scientific review in the Revised Criteria Document and the judgments required by the Administrator in setting ambient air quality standards for particulate matter, EPA's Office of Air Quality Planning and Standards drafted a "Staff Paper."2 The Staff Paper recommended that the Administrator select a value from a range of concentrations from 150 to 350 micrograms per cubic meter (Sg/m3) as PM10 for the twenty-four hour particulate matter primary standard and a range from 55 to 110 Sg/m3 as PM10 for the annual particulate matter primary standard.³

In 1986, the Environmental Criteria and Assessment Office reviewed scientific studies on the health effects of particulate matter that had emerged since its review of relevant information in the 1982 Revised Criteria Document.4 On the basis of this updated analysis, EPA's Office of Air Quality Planning and Standards recommended that the Administrator consider twenty-four hour standards at levels from 140 to 250 Sg/m³ and annual standards at levels from 40 to 65 Sg/m³.5 The Administrator set the twenty-four hour particulate matter standard at 150 Sg/m³ and the annual standard at 50 S g/m³. See 52 Fed.Reg. at 24,643, 24,645. AISI contends that these levels are needlessly stringent. AISI asks this court to set aside the levels as arbitrary and contrary to law, and seeks remand to the Agency for further reconsideration and explanation.

B. EPA's Basis for the Twenty-Four Hour and Annual Particulate Matter Primary Standard Levels

AISI does not contest EPA's retention of both a short-term (twenty-four hour) and a long-term (annual) standard for particulate matter; nor does AISI contest the particle size fraction (PM10 ) used as an indicator. What AISI disputes are the numerical levels for the twenty-four hour and annual standards. AISI contends that the Administrator acted arbitrarily in setting the twenty-four hour and annual PM10 standards at 150 and 50 Sg/m³ respectively because he provided no basis for distinguishing the health effects associated with the levels selected from those associated with the levels rejected. AISI claims that the "only reliable scientific evidence shows that standards at the highest levels proposed would still protect the public health, including sensitive subgroups of the population, with an adequate margin of safety." Brief of the American Iron and Steel Institute at 20.

In reviewing the primary standards for particulate matter, and the "adequacy" of the margin of safety, we are reviewing "predictions within an agency's area of special expertise, at the frontiers of science," New York v. EPA, 852 F.2d 574, 580 (D.C.Cir.1988), cert. denied, --- U.S. ----, 109 S.Ct. 1338, 103 L.Ed.2d 809 (1989). In such circumstances, we must defer to the agency's interpretation of equivocal evidence, so long as it is reasonable. See id. And where, as here, the statute is "precautionary " in nature, the evidence "uncertain or conflicting" and the "regulations designed to protect the public health," the court "will not demand rigorous step-by-step proof of cause and effect." Ethyl Corp. v. EPA, 541 F.2d 1, 28 (D.C.Cir.) (en banc ), cert. denied, 426 U.S. 941, 96 S.Ct. 2662, 49 L.Ed.2d 394 (1976). "The Administrator may apply his expertise to draw conclusions from suspected, but not completely substantiated, relationships between facts, from trends among facts, from theoretical projections from imperfect data, from probative preliminary data not yet certifiable as 'fact,' and the like." Id. But we must, nevertheless, carefully review the record to ascertain that the agency has made a reasoned decision based on "reasonable extrapolations from some reliable evidence." Natural Resources Defense Council v. Thomas, 805 F.2d 410, 432 (D.C.Cir.1986); see also Marsh v. Oregon Natural Resources Council, --- U.S. ----, 109 S.Ct. 1851, 1861, 104 L.Ed.2d 377 (1989) (noting that courts must carefully review record and "satisfy[ ] themselves that the agency has made a reasoned decision based on its evaluation of the significance ... of the ... information"). It is against this standard of review that the court must consider EPA's adoption of the twenty-four hour and annual national primary ambient air quality standards for particulate matter.

AISI challenges the twenty-four hour standard on several grounds. First, AISI claims that "EPA has no reason for rejecting a standard at 250 Sg/m³ that

would not apply with equal force to a standard at 150 Sg/m³, and no reason for settling on a standard at 150 Sg/m³ that

would not firmly justify a standard at 250 Sg/m³." Brief of AISI at 20. According to AISI, "[t]he format of EPA's calculation gives the Agency so much latitude that it could, with artful assumptions, 'justify' virtually any number on the same basis as it did 150 Sg/m³" and EPA's selection of 150 Sg/m³ for the twenty-four hour standard is, therefore, "inherently arbitrary and must be set aside." Id. at 23. Second, and more specifically, AISI claims that EPA rested too little on the Lawther6 and London mortality7 studies. According to AISI, the Lawther study, taken alone, would support setting the twenty-four hour standard at 250 Sg/m³. See id. at 21. According to AISI, the London mortality data also indicate that a standard of 250 Sg/m³ would be "well below" the levels where a " 'scientific consensus' accepts 'pollution' as responsible for some unknown amount of 'life shortening' among 'the elderly and persons with pre-existing respiratory or cardiac disease.' " Id. (quoting 52 Fed.Reg. at 24,642-43. Third, AISI claims that EPA relied too much on the Six Cities Study, which is comprised of the Dockery study8 and the Ware study.9 See Brief of AISI at 10-15. AISI describes the Six Cities Study as the "linchpin" for setting the standards at the lower end of the proposed ranges: the Dockery study for the twenty-four hour standard and the Ware study for the annual standard.

We do not agree that the Administrator's selection of the twenty-four hour standard lacks the necessary reasoned analysis and supportive evidence. The Administrator concluded that a twenty-four hour PM10 standard greater than 150 S g/m³ would "present an unacceptable risk of premature mortality" and allow the possibility of significant lung function changes.10 After carefully reviewing the record, we find EPA's selection of the twenty-four hour standard reasonable in light of the divergent results in the studies and the agency's mandate to provide an adequate margin of safety. Studies contained in the record provided evidence of adverse health effects at levels below 250 Sg/m³.

As an initial matter, the Administrator noted that the data do not provide evidence of clear thresholds in exposed populations and, hence, the Administrator needed to select a level along a continuum of responses. See 52 Fed.Reg. at 24,642, 24,643 (citing studies by Mazumdar et al., 1982; Ostro, 1984; Shumway et al., 1983). In seeking the point along the continuum that would provide the requisite adequate margin of safety, EPA "was required to take into account all the relevant studies revealed in the record." American Petroleum Inst. v. Costle, 665 F.2d 1176, 1187 (D.C.Cir.1981), cert. denied, 455 U.S. 1034, 102 S.Ct. 1737, 72 L.Ed.2d 152 (1982).

The Administrator acknowledged that the Lawther study indicates that a PM10 standard of 250 Sg/m³ might contain some margin of safety11 and that the London mortality studies indicate that the upper portion of the proposed twenty-four hour range would be "well below" the pollution levels that produced "excess mortality" in London.12 But results from other studies indicating adverse health effects at lower concentrations of particulate matter, differences between the United States and London in particulate composition and difficulties in converting from British Smoke--the indicator for particulate matter used in the London and Lawther studies--to PM10 measurements, discouraged exclusive reliance on these results.13 The Administrator explained that more recent reanalyses of the London mortality data suggest that the risk of premature mortality to sensitive individuals extends to concentrations substantially lower than the 500 to 1000 Sg/m³ British Smoke levels of the historical London episodes in which the scientific consensus indicates that pollution was responsible for excess mortality.14 Indeed, the Administrator noted that one of the more recent studies, Ozkaynak & Spengler (1985), provides qualitative support for an association of daily mortality and particle concentrations in nearly contemporary United States atmospheres.15 These studies prompted the Administrator to consider levels that extended from 250 Sg/m³ down to the lower bound of the original staff range of interest--150 Sg/m³--and even lower. In light of these explanations and the results in other studies, the Administrator's decision not to rely more heavily on the Lawther and London mortality studies cannot be said to be arbitrary and capricious.

Nor can we conclude that the EPA's reliance on the Dockery study was unreasonable. The Dockery study reported statistically significant decreases in lung function in a group of children exposed to peak PM10 levels of 140-250 S g/m³.16 These decrements persisted for two to three weeks following the exposure and the study suggested the possibility of larger responses in a subset of the children.17 But EPA did not rely on the results of that study alone in selecting a standard at the low end of the range. A study by Dassen recorded similar effects in children exposed to PM10 levels of 200 to 250 Sg/m³ in the Netherlands, but no observable effects two days after exposure to PM10 levels estimated at 125 Sg/m³.18 EPA also considered several reanalyses of the London mortality data by Mazumdar et al. (1982), Ostro (1984) and Shumway et al. (1983), and qualitative data that suggested increased risks for sensitive groups, risks of potential effects not demonstrated in the quantitative epidemiological literature and "interactive responses" due to the presence of other pollutants.19

EPA explained that "[i]t is reasonable to expect that the effects observed (small reversible reductions in lung function in children) are, in most cases, more sensitive to air pollution than those observed in the London studies."20 "The staff assessment ... suggests that short-term lung function effects in children are possible across a range of 140-250 Sg/m³ or more as PM10 ."21 EPA noted that "[t]he standard is in the lower portion of the range where sensitive, reversible physiological responses of uncertain health significance are possibly, but not definitely, observed in children."22 And it concluded that a "substantial margin of safety below the levels at which there is a scientific consensus that particulate matter causes premature mortality and aggravation of bronchitis" is warranted "because of the seriousness of these effects and because of the recent analyses of daily mortality that suggest adverse effects may occur at particulate matter levels well below the consensus levels."23

In setting a standard under Sec. 109, the Administrator must "take into account all the relevant studies revealed in the record" and "make an informed judgment based on available evidence." American Petroleum Inst. v. Costle, 665 F.2d at 1187. The record shows that the Administrator did so. The Administrator relied on studies which showed adverse effects at and below the 250 Sg/m³ level. AISI essentially asks this court to give different weight to the studies than did the Administrator. We must decline. It is simply not the court's role to "second-guess the scientific judgments of the EPA." See New York v. EPA, 852 F.2d at 580. The Administrator articulated a satisfactory explanation, which is supported by the record evidence, that levels at the top end of the range might not adequately protect against adverse health effects. While there was some evidence that levels even at the low end of the range might produce adverse reactions in sensitive portions of the population, the Administrator did not act arbitrarily in drawing conclusions from the uncertain and conflicting data. The Administrator may reasonably apply his expertise to draw conclusions from "imperfect data," Ethyl Corp., 541 F.2d at 28, as he did here.

AISI next claims that the annual standard established by the Administrator at 50 Sg/m³ lacks a coherent scientific basis. See Brief of AISI at 24. In addition, AISI contends that there is no basis for distinguishing the level rejected from the level selected. AISI cites EPA's explanation that it would not set the annual level below 50 Sg/m³ on the ground that even if "some small risk of increased respiratory symptoms may exist at this concentration, the available data are currently inconclusive." See 52 Fed.Reg. at 24,645. According to AISI, EPA's rationale could as easily justify setting the annual standard at the top end of the range because the data do not show that people will suffer health risks at the top end of the range any more conclusively than that they will suffer from pollution levels at the bottom of the range. See Brief of AISI at 25. We cannot agree. The record is replete with support for the final standard.

The Administrator selected a standard near the middle of the "range of interest" of 40 Sg/m³ to 65 Sg/m³ recommended by the Office of Air Quality Planning and Standards. In support of this standard, the agency cites studies by Bouhuys, Hausman, Ostro and Ware.24 The Ware study "indicates the possibility of increased respiratory symptoms and illnesses in children at multi-year levels across a range of 40 to over 58 Sg/m³ as PM10 ."25 The Bouhuys study suggests some possibility of symptomatic responses in adults at long-term median levels at or below about 50 to 55 Sg/m³ as PM10 .26 Studies by Ostro (1987) and Hausman et al. (1984) suggest the possibility of respiratory effects in adults at comparable levels.27 Other studies provided evidence of no observed effects at or below 60 to 65 Sg/m³.28 The Administrator also noted that "the information available to support the existence of significant adverse effects at annual PM10 levels below 50 S g/m³--especially when 24-hour levels are maintained below 150 Sg/m³--is quite limited and uncertain."29

The Administrator acknowledged that these data are uncertain and limited in scope. Consequently, the Administrator considered it "particularly important to examine the results of qualitative data from a number of epidemiological, animal, and ambient particle composition studies when evaluating what constitutes an adequate margin of safety for an annual standard."30 In deciding to set the standard at the lower bound of the original proposed range, the Administrator reasoned that "[t]his standard provides a reasonable margin of safety against the serious effect of long-term degradation in lung function, which has been judged likely at estimated PM10 levels above 80-90 Sg/m³ and for which there is some evidence at PM10 levels above 60 to 65 Sg/m³."31

The Administrator admits that the 50 Sg/m³ did not spring from a bounty of definitive research as the clear and sole appropriate standard. But such is not required. The Administrator is required to provide an adequate margin of safety. And "[i]n setting margins of safety the Administrator need not regulate only the known dangers to health, but may 'err' on the side of overprotection by setting a fully adequate margin of safety." American Petroleum Inst. v. Costle, 665 F.2d at 1186. The level selected might not be the only one that would have been reasonable based on the Revised Criteria Document and the Revised Criteria Document Addendum. But we certainly cannot conclude that the level selected resulted from an unreasonable interpretation of the equivocal evidence, where some studies indicated threats to health even at the low end of the range. Cf. Public Citizen Health Research Group v. Tyson, 796 F.2d 1479, 1505 (D.C.Cir.1986) ("as long as Congress delegates power to an agency to regulate on the borders of the unknown, courts cannot interfere with reasonable interpretations of equivocal evidence"); New York v. EPA, 852 F.2d at 580 (upholding EPA's rejection of a particular air pollution model); American Petroleum Inst. v. Costle, 665 F.2d at 1186 (upholding EPA's standards for photochemical oxidants as reasonable "given the uncertain information"). "That the evidence in the record may also support other conclusions, even those that are inconsistent with the Administrator's, does not prevent us from concluding that his decisions were rational and supported by the record." Lead Indus. Ass'n v. EPA, 647 F.2d 1130, 1160 (D.C.Cir.), cert. denied, 449 U.S. 1042, 101 S.Ct. 621, 66 L.Ed.2d 503 (1980).

As to AISI's challenge to the Administrator's use of the Ware study, EPA agrees with AISI that the Ware study showed no evidence of reduced lung function in children at concentrations of 40 to 58 Sg/m³. See 52 Fed.Reg. at 24,644; Brief of AISI at 24. And EPA also noted that the study did not find gradients in symptoms and illness within some of the cities in the study that had somewhat smaller localized pollution gradients. See 52 Fed.Reg. at 24,644. But EPA also noted that the study did indicate the possibility of increased respiratory symptoms and illnesses in children at multi-year levels across a range of 40 to 58 Sg/m³ as PM10 , a finding AISI concedes. AISI and EPA differ only in the inferences drawn from the study "on balance." See Brief of AISI at 24. We cannot say, however, that the Administrator was unreasonable in his interpretation of the Ware results, especially in light of the other studies suggesting respiratory effects at these levels.

C. Consideration of the Effects of Unemployment

AISI next contends that EPA erred in refusing to consider the health conse quences of unemployment in determining the primary standards for particulate matter. This claim is entirely without merit. In three previous cases this court has emphatically stated that Sec. 109 does not permit EPA to consider such costs in promulgating national ambient air quality standards. See Natural Resources Defense Council v. EPA ("Vinyl Chloride "), 824 F.2d 1146, 1157, 1159 (D.C.Cir.1987) (en banc ); American Petroleum Inst. v. Costle, 665 F.2d at 1185; Lead Indus. Ass'n v. EPA, 647 F.2d at 1148-49. It is only health effects relating to pollutants in the air that EPA may consider. See 42 U.S.C. Sec. 7408(a)(2) ("Air quality criteria for an air pollutant shall accurately reflect the latest scientific knowledge useful in indicating the kind and extent of all identifiable effects on public health or welfare which may be expected from the presence of such pollutant in the ambient air, in varying quantities.") (emphasis added). Consideration of costs associated with alleged health risks from unemployment would be flatly inconsistent with the statute, legislative history and case law on this point.

AISI also challenges the agency's selection of the primary standard levels for particulate matter on the ground that EPA did not identify a "safe level" indicating what " 'risks are acceptable in the world in which we live.' " Brief of AISI at 26 (quoting Vinyl Chloride, 824 F.2d at 1165). AISI contends that both Sec. 109 and this court's decision in Vinyl Chloride require EPA to identify the risk to health it considers acceptable.

Neither the holding in Vinyl Chloride, nor the decision's rationale, is apposite to actions involving Sec. 109. The challenge in Vinyl Chloride related to the appropriate use of cost and technological feasibility factors in establishing an emission standard for a "hazardous air pollutant," pursuant to Sec. 112 of the Clean Air Act, 42 U.S.C. Sec. 7412 (1982). Under Sec. 112, the agency may consider cost and technological feasibility in selecting levels for hazardous pollutant standards. However, this court rejected the agency's argument that the Administrator "need only find that the costs of control are greater than the reduction in risk to health." Vinyl Chloride, 824 F.2d at 1164. In setting national emission standards for hazardous air pollutants pursuant to Sec. 112 of the Clean Air Act, "the congressional mandate to provide 'an ample margin of safety' 'to protect the public health' requires the Administrator to make an initial determination of what is 'safe.' " Id. This court held that the Administrator may not consider cost and technological feasibility in determining what is "safe"; such a determination "must be based solely upon the risk to health." Id. at 1166. Once EPA has determined a "safe" level of exposure, it may then consider cost and feasibility in providing for the "ample margin of safety" required by Sec. 112.

The two-step methodological requirement endorsed by Vinyl Chloride was necessary because of the need under Sec. 112 to sever determinations that must be based solely on health considerations from those that may include economic and technological considerations. Under Sec. 109, however, the Administrator may not consider cost and technological feasibility. Hence, the rationale for parsing the Administrator's determination into two steps is inapposite to actions under Sec. 109.

The foreclosure from consideration of cost and technological feasibility also persuades us that Sec. 109 itself does not mandate the methodological requirement we found necessary under Sec. 112. In facing substantive challenges to national ambient air quality standards, the Agency may sometimes need to articulate the level of threat to the population it considers tolerable; but there is no separate methodological requirement under Sec. 109 that the Administrator establish a measure of the risk to safety it considers adequate to protect public health every time it establishes a standard pursuant to Sec. 109.

Indeed, this court has previously declined to impose specific requirements on the Administrator's methodological approach to selecting an adequate margin of safety under Sec. 109. In Lead Industries, one petitioner argued that EPA must separately determine the maximum level of a pollutant that is protective of human health and the adequate margin of safety by which the level that is protective of health must be reduced. See 647 F.2d at 1161. The court ruled that the choice of approach to determining the margin of safety "is a policy choice of the type that Congress specifically left to the Administrator's judgment. This court must allow him the discretion to determine which approach will best fulfill the goals of the Act." Id. at 1162.

In the instant case, the Administrator determined the primary standards for particulate matter based solely upon the risk to health. Methodologically, that is all that is required under Sec. 109. We recognize that pursuant to a substantive challenge of the "safety" of the selected levels, EPA might be required to explain the risk it considered tolerable in meeting the requirements of providing a reasoned explanation for an administrative decision. But, partly because of the uncertainty of the data upon which the Administrator needed to rest his assessment, we do not find such further articulation necessary here. Consequently, AISI's complaint provides no basis for reversal or remand.

On December 7, 1989, AISI filed with the court a motion to hold the decision in abeyance pending final action by EPA on AISI's "Petition for Investigation and, If Appropriate, Reconsideration of the PM10 NAAQS," filed with EPA on the same date. AISI's Petition for Investigation asked EPA to commence an investigation of the "Six Cities" study. The motion and petition were prompted by a Wall Street Journal article reporting that the National Institutes of Health's Office of Scientific Integrity "is reviewing 'a claim that there could be serious errors in the data base' of the so-called Six Cities study." Wall Street Journal, Nov. 30, 1989, at B-5.

This court faced a similar request in Lead Industries Association v. EPA, 647 F.2d 1184 (D.C.Cir.), cert. denied, 449 U.S. 1042, 101 S.Ct. 621, 66 L.Ed.2d 503 (1980). In that case, EPA denied the petition for reconsideration before the court handed down its decision in the lead standards case. See id. at 1186. But the petitioner sought the court's deferral of action on the appeal of the lead standards until it could obtain judicial review of EPA's decision denying its petition for reconsideration. See id. The court ruled that in order to conclude that such a delay is justified, "the court must be convinced that the 'new information' which provides the basis for the reconsideration petition raises substantial questions about the validity of the Agency's analysis." Id. at 1187.

We do not believe that the alleged "new information" cited in the petition for reconsideration warrants delay of our review of the standards. First, even if we assume that there are serious errors in the Six Cities studies, these studies were only part of the evidence on which EPA relied in selecting the twenty-four hour and annual particulate matter standards. Other studies support the selection of the PM10 standards at the levels promulgated in the final rule. Second, there is considerable doubt about whether the Six Cities data are actually infected with error. All that the "new information" indicates is that the National Institutes of Health ("NIH") is investigating charges that one of the authors of the studies suppressed evidence of errors. Both the outcome and the duration of NIH's investigation, and any investigation EPA might undertake, are completely beyond our ken. In short, the significance of the opening of the investigation is too speculative to provide grounds for delaying our decision on AISI's petition for review of the PM10 standards. Thus we deny AISI's motion to hold this appeal in abeyance.

F. Compliance with Control Techniques Requirements

Finally, AISI claims that EPA has unlawfully failed to issue up-to-date PM10 control techniques information to the states, in violation of 42 U.S.C. Sec. 7408(b)(1) and (c).32 EPA argues that since the rules under review here involve the revision of particulate matter criteria and ambient air quality standards, EPA's duty to revise and reissue control techniques information is controlled by 42 U.S.C. Sec. 7408(c)--not by Sec. 7408(b)(1)--and that it has met the statutory requirements.

Under either provision, it is to "the States and appropriate air pollution control agencies" that EPA has a duty to provide cited information. See 42 U.S.C. Sec. 7408(b)(1). On this record, we conclude that AISI lacks standing to challenge EPA's alleged failure to furnish adequate control techniques information to the states. Article III of the Constitution limits the exercise of the "Judicial Power" of the United States to the resolution of "Cases" and "Controversies." U.S. CONST. art. III, Sec. 2. The Supreme Court has defined the essential elements of a case or controvery as follows:

[A]t an irreducible minimum, Art. III requires the party who invokes the court's authority to "show that he personally has suffered some actual or threatened injury as a result of the putatively illegal conduct of the defendant," and that the injury "fairly can be traced to the challenged action" and "is likely to be redressed by a favorable decision."

Valley Forge Christian College v. Americans United for Separation of Church and State, 454 U.S. 464, 472, 102 S.Ct. 752, 758, 70 L.Ed.2d 700 (1982) (citations omitted).

We find here that the "injury" asserted by the petitioner is too speculative and the causation and redressability too attenuated and uncertain for AISI to meet the constitutional requirements for standing. AISI's petition does not identify "concrete organizational interests detrimentally affected by the particular" agency action it challenges. See Action Alliance of Senior Citizens v. Heckler, 789 F.2d 931, 937 (D.C.Cir.1986). The injury AISI appears to allege is that--in giving effect to the new PM10 primary standards--states might adopt implementation plans that are needlessly stringent or EPA might approve needlessly stringent state implementation plans, by relying on "out-dated" control techniques information.33

But, as counsel for AISI conceded at oral argument, it is clear that the control techniques information does not have a regulatory effect on AISI. Indeed, it is not even clear that the states are bound to use the information. Although EPA must provide control techniques information to the states, the choice of control techniques is within the state's discretion. See, e.g., Train v. Natural Resources Defense Council, 421 U.S. 60, 79, 80, 95 S.Ct. 1470, 1481, 1482, 43 L.Ed.2d 731 (1975); Union Elec. Co. v. EPA, 427 U.S. 246, 250, 96 S.Ct. 2518, 2522, 49 L.Ed.2d 474 (1976); Florida Power & Light Co. v. Costle, 650 F.2d 579, 586-87 (5th Cir.1981). EPA has only a "secondary role in the process of determining and enforcing the specific, source-by-source emission limitations." Train, 421 U.S. at 79, 95 S.Ct. at 1481; see also 42 U.S.C. Secs. 7401(a)(3), 7407(a) (1982). So long as the technique employed by the state is "adequate to achieve compliance with national air quality standards, the Administrator is not authorized to reject it in favor of another technique." Mission Indus. v. EPA, 547 F.2d 123, 129 (1st Cir.1976). Thus, it appears that the causal relationship between the Administrator's issuance of control techniques information and the control techniques implemented by the states is very attenuated.

Moreover, we assume that affected parties will contribute to the states' consideration of suitable control techniques in the development of state implementation plans. See, e.g., United States v. Ford Motor Co., 814 F.2d 1099, 1104 (6th Cir.) (noting that emission sources are "offered the opportunity to have significant input on the setting of emission limits," for example, in the state's promulgation of the state implementation plan and EPA's approval of it), cert. denied, 484 U.S. 822, 108 S.Ct. 83, 98 L.Ed.2d 45 (1987); see also United States v. Alcan Foil Products, 889 F.2d 1513, 1516 (6th Cir.1989) (company's proposed revision to Kentucky state implementation plan approved by air pollution control region), petition for cert. filed, 58 U.S.L.W. 3480 (U.S. Jan. 17, 1990) (No. 89-1104). We further assume that a state implementation plan founded on poor control-techniques information would be open to challenge by affected parties.

The Senate Report accompanying the Clean Air Amendments of 1970 "recognize[d] that the States will continue to need this information to develop meaningful programs for implementation of ambient air quality standards on a regional basis." See S.REP. No. 1196, 91st Cong., 2d Sess. 9 (1970).34 The legislative history shows that Congress intended the control-techniques information to "serve as guidance to States, not as limitations on control technology innovation." Id. AISI has not persuaded us that the "guidance" EPA has made available to the states has or will wreak injury on it. We do not determine here that it would be impossible under all circumstances for AISI, or another organization, to establish standing to challenge EPA's failure to comply with its obligation to release control-techniques information. We conclude only that AISI lacks standing in the case before us to challenge EPA's compliance with the requirements of 42 U.S.C. Sec. 7408(b)(1) and (c).

We conclude that each of AISI's challenges to EPA's final rule must fail.

Congress amended the Clean Air Act in 1977 in response to the widespread failure of states to attain the NAAQS promulgated in 1971. Under a new Sec. 107(d) of the Act, every state must classify each of its air quality regions for each air quality standard as "attainment," "nonattainment," or, in the event that insufficient information exists, "unclassifiable." For regions identified as attainment or unclassifiable, Congress adopted a Prevention of Significant Deterioration ("PSD") program in a new Part C of the Act, 42 U.S.C. Secs. 7470-7479. The PSD provisions are designed to preserve air quality, in regions that have already satisfied the NAAQS, by imposing "maximum allowable increases" in the emission of pollutants from new and existing sources. See 42 U.S.C. Sec. 7473. The program requires each state to adopt preconstruction review of any "major emitting facility" to ensure that emissions in the region where the facility is located will not exceed allowable PSD increments. See 42 U.S.C. Sec. 7475. Congress delegated to EPA the authority to promulgate PSD increments for hydrocarbons, carbon monoxide, photochemical oxidants, nitrogen oxides, and other "pollutants for which national ambient air quality standards are promulgated after August 7, 1977." 42 U.S.C. Sec. 7476. But Congress specified in Sec. 163(b) the exact numerical increments for particulate matter and sulfur dioxide. See 42 U.S.C. Sec. 7473(b). Although Sec. 163(b) does not explicitly mention what size indicator (Total Suspended Particulates ("TSP") or PM10 ) would be the standard of measurement, EPA has consistently interpreted the statute as mandating the use of the TSP standard.35

Petitioner American Mining Congress ("AMC"), an industry association representing the coal mining industry, challenges EPA's decision to retain the TSP indicator for purposes of the PSD program once the agency adopted the PM10 indicator, the successor measuring device, for the particulate matter NAAQS, see 52 Fed.Reg. 24,685, 24,699-702 (1987). EPA did announce that the agency would promulgate new PSD increments for particulate matter using the PM10 reference through a rulemaking proceeding pursuant to Sec. 166. See 52 Fed.Reg. 24,672, 24,685 (1987). The AMC contends, however, that EPA should have kept the same numbers specified in Sec. 163(b) but "administratively redefined" them as referring to a PM10 indicator rather than TSP. The PSD program expressed in TSP terms, the AMC believes, severely disrupts coal mining because such operations release large-particle "fugitive dust" into the air.36 AMC maintains that as long as the TSP standard is retained for the PSD provisions, coal mines will likely exceed the maximum particulate matter increments.37 Petitioner argues that EPA's refusal to modify the PSD increment calculations contravenes congressional intent and our circuit's precedent. Furthermore, AMC believes that the retention of the admittedly flawed TSP standard constitutes arbitrary and capricious decisionmaking. We find none of the petitioner's arguments persuasive.

Although Sec. 163(b) does not define "particulate matter" by size, both EPA and the AMC agree that Congress intended to apply the TSP indicator when it enacted the PSD increments in 1977. The numbers Congress specified in Sec. 163(b) reflect percentages of the then-existing NAAQS for particulate matter--which was measured only in TSP terms--and indicate Congress' judgment on what constitutes a significant deterioration of air quality. See H.R.REP. No. 294, 95th Cong., 1st Sess. 153 (1977); S.REP. No. 127, 95th Cong., 1st Sess. 11 (1977) U.S.Code Cong. & Admin.News 1977, pp. 1077, 1232. Consequently, the only issue is whether EPA was authorized to alter the TSP measurement embedded in the statute when the agency adopted a new indicator for NAAQS purposes.

It hardly bears noting that EPA's discretion cannot include the power to rewrite a statute and reshape a policy judgment Congress itself has made. See, e.g., Louisiana Public Serv. Comm'n v. FCC, 476 U.S. 355, 376, 106 S.Ct. 1890, 1902, 90 L.Ed.2d 369 (1986); MCI Telecommunications Corp. v. FCC, 765 F.2d 1186, 1195 (D.C.Cir.1985). The AMC argues that the legislative history of the 1977 Amendments to the Clean Air Act demonstrates a congressional intent to authorize EPA to modify the TSP reference in the Sec. 163(b) PSD increments if and when the agency adopts a new size cutoff for measurement purposes. The petitioner directs us to a short passage in the 1977 Senate Committee report suggesting that EPA use "administrative good sense" in responding to imprecisions in the TSP measurement. See S. REP. No. 127, 95th Cong., 1st Sess. 98 (1977). And according to the AMC, the absence of any specific standard of measurement in Sec. 163(b) indicates that Congress meant to give EPA flexibility to change the standard even though the numbers were to be constant.

The AMC's position would require us to rely on this snippet of legislative history to set aside what appears to us to be a specific and unambiguous congressional policy judgment. By establishing PSD increments numbers expressed in what could only have been TSP terms, Congress determined the permissible amount of increases in emissions of particulate matter. Congress arrived at those Sec. 163(b) numbers based on its estimate of the particulate matter emissions that would normally accompany moderate and intensive economic growth. See H.R.REP. No. 294, 95th Cong., 1st Sess. 152-53 (1977). Simply changing the measurement reference from TSP to PM10 while maintaining the same numbers, as AMC urges EPA to do, would dramatically relax the PSD restrictions on particulate matter since the TSP indicator is a much more inclusive measure of pollutants. The absence of a specific indicator in Sec. 163(b) hardly suggests the flexibility that AMC reads into the statute; only the TSP indicator existed at the time of the 1977 Amendments and Congress could have thought it superfluous and excessively technical to mention it. If Congress had intended to delegate the power to alter the Sec. 163(b) increments, it could have done so explicitly--as it did for other pollutants in Sec. 166. It would be wholly irrational for Congress to set forth specific numbers for particulate matter increments and yet authorize a choice of indicators which radically alters the numbers' significance and therefore ultimately how much particulate matter may be emitted. Congress has plainly expressed an intent to allow a certain quantum of particulate matter emissions, and we must give that intent effect. See Chevron U.S.A. Inc. v. Natural Resources Defense Council, 467 U.S. 837, 842-43, 104 S.Ct. 2778, 2780-81, 81 L.Ed.2d 694 (1984).

The AMC also contends that in Alabama Power v. Costle, 636 F.2d 323 (D.C.Cir.1980), we previously authorized EPA to redefine particulate matter to exclude the larger particle fugitive dust from the PSD provisions. In a footnote in Alabama Power, we did suggest, in dicta, that EPA could define particulate matter to exclude those of a size or composition believed to be harmless. See id. at 370 n. 134. That footnote in Alabama Power, however, did not directly confront Congress' explicit direction in Sec. 163(b) detailing the permissible particulate matter increment and is thus hardly binding on either EPA or this court.

Finally, the AMC urges us to reverse as arbitrary and capricious EPA's decision to retain the TSP indicator. Petitioner presented a welter of scientific evidence demonstrating that the TSP measure is flawed, large particle fugitive dust is harmless, and therefore the TSP standard fails a cost benefit analysis. Unfortunately for the AMC, however, its real dispute is not with EPA but with Congress, which decided for itself what increases in particulate matter emissions were acceptable. EPA may not be faulted for following Congress' clear direction in Sec. 163(b) and deciding that the agency possessed no discretion to consider the costs and benefits of maintaining the TSP indicator.

The AMC also challenges, in a most cursory fashion, three highly technical and minor features of the particulate matter NAAQS: the shift from a geometric to arithmetic mean38 in expressing the annual particulate matter standard, the continued adjustment of particulate matter concentrations to standard temperature and pressure, and the refusal to modify the technique for adjusting particulate matter concentrations due to high wind speeds. These challenges are nothing more than an invitation to second-guess an expert agency's resolution of scientific and technical issues of its rulemaking--the paradigm instance of agency discretion--and we will defer to EPA as long as its explanation is rational. See American Petroleum Inst. v. Costle, 665 F.2d 1176, 1184 (D.C.Cir.1981), cert. denied, 455 U.S. 1034, 102 S.Ct. 1737, 72 L.Ed.2d 152 (1982); Lead Indus. Ass'n v. EPA, 647 F.2d 1130, 1162 (D.C.Cir.), cert. denied, 449 U.S. 1042, 101 S.Ct. 621, 66 L.Ed.2d 503 (1980).

EPA changed the form of its annual particulate matter standard from an annual geometric mean to an expected arithmetic mean. See 52 Fed.Reg. 24,634, 24,640 (1987). The AMC complains, with virtually no explanation, that the arithmetic mean is too sensitive to aberrationally high values and is therefore unjustified. According to EPA, because the arithmetic mean is proportional to the sum of the daily means, it may better reflect health data based on total exposure. Indeed, the public health studies underlying the annual standard applied the arithmetic mean. The geometric mean, moreover, may be too sensitive to aberrationally low data points since it would multiply the terms together and find their Nth root. Thus, EPA noted that a single day of zero concentration would result in a geometric mean calculation of zero for the entire year, regardless of the concentrations present in the 364 other days. The "choice of statistical methods is committed to the sound discretion of the Administrator," Chemical Mfrs. Ass'n v. EPA, 870 F.2d 177, 227 (5th Cir.1989), and we find no basis for questioning the rationality of EPA's decision to shift the method of calculation.

EPA also continued its longstanding requirement that samples of ambient air be corrected to standard temperature (298 degrees K) and pressure (101.3 kPa). See 40 C.F.R. Part 50, Appendix J, Sec. 11. The AMC objects, again with very little elaboration, that this adjustment exaggerates particulate matter concentration when sampling ambient air at higher altitudes. The agency explained that even at moderate altitudes exercise can stimulate significant differences in ventilation between residents at those altitudes and comparable conditions at sea level. The decreased oxygen pressure, moreover, could adversely affect sensitive individuals. The AMC made no attempt to rebut this explanation, and we conclude that EPA's retention of this adjustment is reasonable.

Finally, the AMC asserts that EPA's measurement techniques do not adequately account for high wind speeds that might skew the particulate matter concentration results. See 40 C.F.R. Parts 53 and 55. EPA denied the AMC's petition because the AMC did not raise this issue during the rulemaking. The AMC has not offered any explanation why it might have been impractical to raise the wind speed issue during rulemaking or special circumstances that might justify consideration of its petition. Therefore, we find that EPA properly denied the AMC's petition. See 42 U.S.C. Sec. 7607(d)(7)(B); see also Northside Sanitary Landfill, Inc. v. Thomas, 849 F.2d 1516, 1519 (D.C.Cir.1988), cert. denied, --- U.S. ----, 109 S.Ct. 1528, 103 L.Ed.2d 833 (1989).

IV. THE NATURAL RESOURCES DEFENSE COUNCIL'S CLAIMS

The following opinions by Chief Judge Wald and Circuit Judges Edwards and Silberman consider the claims of the Natural Resources Defense Council ("NRDC") concerning EPA's conduct vis-a-vis a secondary national ambient air quality standard for particulate matter to protect the public welfare from visibility impairment and acid deposition. Although Chief Judge Wald's opinion lays out the statutory and regulatory framework of the issue, the three opinions differ in their analyses and, to some degree, in their conclusions. In sum, the judgment of the court is as follows. We hold that EPA has not taken final action with regard to the appropriateness of a fine particle standard protecting the public welfare against visibility impairment; accordingly, we do not have jurisdiction under Sec. 307 of the CAA over NRDC's petition to order that such action be taken. The portion of NRDC's petition concerning a visibility impairment standard is, consequently, dismissed. We also conclude that EPA has taken final action, for purposes of this rulemaking, not to issue a fine particle standard protecting the public from acid deposition; accordingly, we do have jurisdiction over the section of NRDC's petition concerning an acid deposition standard. We remand this portion of the case to the agency and order it to provide, within sixty days of the issuance of the mandate of this opinion, a statement of reasons for its decision not to initiate a rulemaking on a secondary standard protecting against acid deposition.

This Part considers a suit by the Natural Resources Defense Council ("NRDC") and affiliated petitioners1 claiming that the Environmental Protection Agency ("EPA" or "agency") improperly failed to set a secondary national ambient air quality standard ("NAAQS") for particulate matter to protect the public welfare from visibility impairment and acid depositions. NRDC asks us to establish a one-year timetable for EPA to implement this aspect of its secondary standard. Our reading of the Clean Air Act leads us to conclude (1) that we do not have jurisdiction to consider NRDC's contention as to visibility impairment, and (2) that while we do have jurisdiction to consider NRDC's contention as to acid deposition, the thinness of the record requires us to remand for a more adequate explanation by EPA of its decision not to initiate rulemaking on a particulate matter standard dealing with acid deposition.

I begin by briefly setting out the statutory context in which NRDC's suit arises. Section 109 of the Clean Air Act ("CAA" or "Act"), 42 U.S.C. Sec. 7409 (1982), instructs the EPA Administrator to publish a primary and secondary NAAQS for each air pollutant for which air quality criteria have been issued. Primary standards are designed to protect the public health. Secondary standards, at issue here,

shall specify a level of air quality the attainment and maintenance of which in the judgment of the Administrator, based on such criteria, is requisite to protect the public welfare from any known or anticipated adverse effects associated with the presence of such air pollutant in the ambient air.

Id. Sec. 7409(b)(2). The Act elsewhere explains that

[a]ll language referring to effects on welfare includes, but is not limited to, effects on soils, water, crops, vegetation, manmade materials, animals, wildlife, weather, visibility, and climate, damage to and deterioration of property, and hazards to transportation, as well as effects on economic values and on personal comfort and well-being.

Id. Sec. 7602(h).

In 1977, after nearly ten years of review and evaluation of the ambient air quality standards, see H.R.REP. No. 294, 95th Cong., 1st Sess. 179-82 (1977), Congress enacted Sec. 109(d) of the Act, 42 U.S.C. Sec. 7409(d). It states in relevant part:

Not later than December 31, 1980, and at five-year intervals thereafter, the Administrator shall complete a thorough review of the criteria published under section 7408 of this title and the national ambient air quality standards promulgated under this section and shall make such revisions in such criteria and standards and promulgate such new standards as may be appropriate in accordance with section 7408 of this title and subsection (b) of this section. The Administrator may review and revise criteria or promulgate new standards earlier or more frequently than required under this paragraph.

Id. Sec. 7409(d)(1).

Thus, EPA bears continuing regulatory responsibilities for updating secondary standards. The CAA's complex scheme for judicial review of EPA's performance, however, has given rise to conflicting interpretations as to which court has jurisdiction over claims that EPA has not fulfilled its obligations in this regard. Section 307(b)(1), id. Sec. 7607(b)(1), provides in relevant part that

[a] petition for review of action of the Administrator in promulgating any national primary or secondary ambient air quality standard, ... any other nationally applicable regulations promulgated, or final action taken, by the Administrator under this chapter may be filed only in the United States Court of Appeals for the District of Columbia.

At the same time, Sec. 304 of the Act, id. Sec. 7604, the "citizen suit" provision, provides that district courts have jurisdiction over suits alleging the EPA Administrator's failure to perform a non-discretionary act or duty.2 In appropriate cases, the district court may order the Administrator to perform such an act or duty. Id. Sec. 7604(a)(2).

In March 1984, the Administrator proposed secondary standards protecting the public welfare against soiling and nuisance caused by airborne particulate matter. Proposed Revisions to the National Ambient Air Quality Standards for Particulate Matter, 49 Fed.Reg. 10,408, 10,418-19 (1984). In the same document, the Administrator explained that three other principal welfare effects--visibility impairment, climatic effects, and acid deposition--"are most strongly related to regional scale fine particle levels and any standards and controls would likely involve regional sulfur oxide emissions." Id. at 10,419. Thus, the Administrator made a preliminary determination that EPA can best protect the public welfare from these effects through a separate standard involving only "fine" particulate matter.³ Id. To that end, the Administrator noted "the advantages of recognizing the interrelated aspects of the known and potential effects of fine particles on visibility and climate together with the acidic deposition phenomenon when considering a possible fine particle standard." Id. As a result, the Administrator "decided to defer a decision on a possible fine particle secondary standard until it is possible to link such a standard with a coherent, scientifically based strategy for these related regional air quality problems." Id. "In parallel with" its evaluations of various approaches to acidic deposition, EPA would continue to examine "the implications of acidic deposition control strategies on visibility and other air quality values." Id. The information gathered by EPA would be used "in preparing an advance notice of proposed rulemaking soliciting public comment regarding a possible fine particle secondary standard." Id.

In July 1987, EPA issued its Revisions to the National Ambient Air Quality Standards for Particulate Matter. 52 Fed.Reg. 24,634 (1987) [hereinafter cited as Revisions]. The Revisions included a final secondary standard protecting the public welfare against soiling and nuisance. Id. at 24,645-46. In addition, the agency explained that it was

reassessing its position with regard to consideration of a secondary fine particle standard for visibility. In particular, the Agency is considering whether, given the time that would be required to develop, propose, promulgate, and implement a visibility based standard, it would now be appropriate to proceed with consideration of a visibility based standard in parallel with work on acid deposition, so that compatible strategies for dealing with the two problems can be developed at the implementation stage.

Id. at 24,646-47. Accordingly, on the same day as it issued the Revisions, EPA published an advance notice of proposed rulemaking (ANPR) "soliciting public comment on the appropriateness of a separate secondary fine particle standard designed to protect visibility, and on a number of issues that would have to be resolved in proposing such a standard." Id. at 24,64 7. In the ANPR itself, EPA identified three major issues on which it solicited public comment relevant to a possible fine particle standard: the appropriate level of protection, the relationship between ambient fine particulate matter and visibility, and the desirability of developing a fine particle standard to protect visibility while scientific research continues to resolve uncertainties surrounding acid deposition. In light of that uncertainty, the ANPR did not propose a particulate matter standard to deal with acid deposition. Comments pursuant to the ANPR were to be submitted by September 29, 1987. Review of the National Secondary Ambient Air Quality Standards for Particulate Matter, 52 Fed.Reg. 24,670-71 (1987) [hereinafter cited as ANPR].

NRDC argues that EPA's actions heretofore described amount to an "abandon[ment of] its responsibility to protect the environment by refusing to set secondary standards to protect against certain well known adverse effects of particulate matter, namely visibility impairment and acid rain." Brief for Petitioners NRDC, et al. ("NRDC Br."), at 8. Consequently, NRDC asks us to establish a timetable according to which EPA would propose and promulgate these standards. Id. at 31. At issue initially is whether we have jurisdiction to consider the merits of NRDC's challenge to EPA's failure to promulgate secondary standards in these areas.

I begin by recasting the issue before us in the jurisdictional terms of art employed by the CAA. CAA Sec. 307(b)(1), 42 U.S.C. Sec. 7607(b)(1), asserted by petitioners as the basis of our jurisdiction, confers review authority on this court if EPA's action falls into one of the following categories: "action of the Administrator in promulgating any national ... secondary ambient air quality standard," "any other nationally applicable regulations promulgated," or "final action taken" by the agency. 42 U.S.C. Sec. 7607(b)(1). Neither of the first two categories is applicable here. The word "promulgate" in the CAA refers only to the original issuance of a standard, while the word "revision" refers to subsequent modifications of that standard. See, e.g., 42 U.S.C. Sec. 7607(d)(1)(A) ("This subsection applies to ... the promulgation or revision of any national ambient air quality standard ..."). Thus, the omission of the word "revision" from the first two jurisdictional heads of Sec. 7607(b)(1) renders them irrelevant to our consideration of the EPA's "Revisions" of particulate matter NAAQS. Revisions, 52 Fed.Reg. 24,634. EPA's action on visibility impairment and acid deposition must constitute "final action" for NRDC's petition to fall within our scope of review.

We have explained in the past that the determination of finality encompasses the question of ripeness, by focusing on "the appropriateness of the issues presented for judicial review." Sierra Club v. Gorsuch, 715 F.2d 653, 657 (D.C.Cir.1983); see also Gulf Oil Corp. v. Dep't of Energy, 663 F.2d 296, 310 (D.C.Cir.1981) (finality is "weighty" but not "indispensable" factor in determination of ripeness). In part, then, our task is to determine whether EPA's rulemaking on visibility impairment and acid deposition is in a state that is ripe for our review. This determination is diffi cult,, however, because as described supra, the precise status of action by EPA on both of these secondary effects of particulate matter is uncertain: a fine particle visibility standard is the subject of an ANPR whose comment period expired over two years ago, and an acid deposition standard appears to have been indefinitely deferred. The statutory scheme and legislative history provide guidance as to the finality of EPA's actions.

B. The Statutory Play Between Sections 307 and 304

As described supra, the CAA's jurisdictional scheme is laid out in Secs. 307 and 304. Section 307 was originally enacted as part of the CAA in 1970. Congress committed review of "action of the Administrator in promulgating" any NAAQS to this court. 42 U.S.C. Sec. 1857h-5(b) (1976). The Senate Committee on Public Works explained that in situations requiring "even and consistent national application" of administrative actions, this court should exercise jurisdiction. S.REP. No. 1196, 91st Cong., 2d Sess. 41 (1970). In 1977, Congress added to Sec. 307(b)(1) the "final action" language relevant to our case. In a brief reference, the House Committee on Interstate and Foreign Commerce stated that Sec. 307(b)(1), as amended, "makes it clear that any nationally applicable regulations promulgated by the Administrator under the Clean Air Act could be reviewed only in the U.S. Court of Appeals for the District of Columbia." H.R.REP. No. 294, at 323, U.S.Code Cong. & Admin.News 1977, p. 1402; see Harrison v. PPG Industries, 446 U.S. 578, 589-92, 100 S.Ct. 1889, 1896-97, 64 L.Ed.2d 525 (1980) (discussing legislative history). Congress' thrust for Sec. 307 review, then, appears to have been that only final EPA regulations or other final decisions of national application should be reviewed exclusively in this court.

CAA Sec. 109(d)(1) states that, not later than December 31, 1980, and on five-year reviews thereafter, the Administrator "shall make such revisions in such ... [primary and secondary national ambient air quality] standards ... as may be appropriate." In my view, the statute clearly envisioned rulemaking cycles ending in 1980 and every subsequent five years; in each cycle, the agency would, in theory, make those NAAQS revisions that it deemed appropriate. See Environmental Defense Fund v. Thomas, 870 F.2d 892, 897-98 n. 1 (2d Cir.1989) (Sec. 109(d)(1) requires EPA to decide whether to revise standards "within the stated deadlines"). In practice, however, EPA may not state clearly in each cycle whether it has decided to revise standards, not to revise standards, or even whether it has decided to decide whether to revise standards. As we apply Sec. 307(b) to Sec. 109(d)(1), our task is to determine the circumstances under which EPA's statements--or silence--about revising standards amount to "final" action. Brief consideration of several possible scenarios renders somewhat more pellucid the otherwise rather opaque concept of "finality" in this context.4

Two cases seem clear: If the Administrator issues revised standards, they fall squarely within our jurisdiction because they involve actions of national application over which Congress gave this court exclusive supervisory review. Similarly, if the Administrator explicitly determines after review that revision of the NAAQS is not "appropriate," that decision would, in most situations, also be final and fall within our review jurisdiction.

A third situation also seems likely to constitute final action for a given rulemaking cycle. The agency may stand mute, indicating neither that a revision is contemplated nor that a decision against revision has been reached. In that circumstance, a rulemaking cycle will pass with no change in the relevant standards. The result is no different from a statement by EPA that a revision of standards is not appropriate for this rulemaking. Thus, it seems probable that for the given rulemaking cycle, EPA's muteness is simply a proxy for its determination that no "revisions in such ... standards ... may be appropriate." 42 U.S.C. Sec. 7409(d)(1). Alternatively, there may be contextual or historical evidence to persuade a court that the agency is not presenting its views honestly and is, instead, hiding a final decision not to revise behind its silence. See Sierra Club v. Thomas, 828 F.2d 783, 793 (D.C.Cir.1987) (inaction in form of "[e]ffectively final action not acknowledged" may constitute final agency action for jurisdictional purposes); Sierra Club v. Gorsuch, 715 F.2d at 660-61 (remanding to EPA for clarification of agency silence in rulemaking). Such evidence, even if circumstantial, would bring agency silence inside our scope of review. Or, EPA may have failed to act with regard to one or a few interconnected parts of an overall plan, as, for example, when the agency defers review of certain aspects of state implementation plans that required its approval. In such cases, courts of appeal have considered the partial deferrals to be reviewable because of their interconnectedness with--and indispensability to--the plan as a whole. Abramowitz v. EPA, 832 F.2d 1071, 1076 (9th Cir.1987) (quoting Indiana & Michigan Electric Co. v. EPA, 733 F.2d 489, 490 (7th Cir.1984)).5

Aside from these three polar situations--an explicit decision to revise, an explicit decision not to revise, and pure silence--EPA may take some preliminary action that does not conclusively establish whether or how the agency will act vis-a-vis revision of standards during a given rulemaking cycle. For instance, the Administrator may solicit information and comment as to the type or level of standard that would be appropriate, or as to whether an NAAQS is appropriate at all. Such tentative action may well fall short of the agency's nondiscretionary statutory responsibilities. See NRDC Br. at 13 (discussing regulatory requirements imposed by Sec. 109). Nonetheless, since the Administrator's decisionmaking for the cycle would still be in progress, the rulemaking could not, logically, be called "final" and would fall outside our jurisdiction. Although not asserting jurisdiction in this situation may enable EPA to avoid appellate review by putting out even preliminary rulemaking feelers, I believe that restricting our jurisdiction in this manner affords a healthy and substantial measure of deference to agencies dealing, however slowly, with complex issues.

This view of the finality requirement for review under Sec. 307 would, nonetheless, be highly troublesome if it left EPA free of all review whenever it declined to decide whether revision of the standards was appropriate or inappropriate. A statutory scheme enabling EPA to evade review of even the most egregious footdragging would be inconsistent with Congress' goal of carefully monitoring polluting emissions into the ambient air. H.R.REP. No. 294, at 182. Consequently, I examine review procedures in the CAA that do not depend on final agency action in order to backstop my conception of the availability of appellate review.

The CAA specifically provides for a route to judicial intervention whenever EPA fails to perform a nondiscretionary statutory duty. As noted above, CAA Sec. 304 authorizes plaintiffs aggrieved by EPA inaction to bring suits in federal district court for the purpose of requiring EPA to carry out its nondiscretionary duties. Although Congress originally envisioned this section as providing citizen oversight of inadequate EPA enforcement of air quality standards and regulations, S. REP. No. 1196, at 36-38, Sec. 304 suits have been successfully employed to enforce a broad range of EPA obligations to implement or issue air quality regulations. 1 W. Rodgers, Environmental Law: Air and Water Sec. 3.4, at 223-24 (1986). Petitioners in a citizen suit must show, of course, that the EPA Administrator has failed to perform some nondiscretionary duty. While the line between discretionary and nondiscretionary obligations is sometimes blurry, id., that determination is made in the first instance by the district court exercising jurisdiction over the citizen suit. The precise circumstances under which EPA would fail to meet its obligations under Sec. 109(d)(1) are, therefore, not at issue today.6

The Administrative Procedure Act's ("APA") mandate for courts to "compel agency action ... unreasonably delayed," 5 U.S.C. Sec. 706(1) (1988), is inapplicable to judicial review of a number of types of EPA rulemaking pursuant to the CAA, including rulemaking on NAAQS revisions. See CAA Sec. 307(d)(1), (9), 42 U.S.C. Sec. 7607(d)(1), (9). In Sierra Club v. Thomas, 828 F.2d 783 (D.C.Cir.1987), we noted that the Clean Air Act's citizen suit is available to remedy the Administrator's "failure ... to perform" a "nondiscretionary duty of timeliness." Id., 828 F.2d at 790. Thus, Sec. 304 may in appropriate circumstances provide a check against indefinite stalling by EPA to avoid or evade mandatory substantive decisions.7

Admittedly, the boundary lines between Secs. 304 and 307 jurisdiction have been blurred by judicial rulings responding to an infinite variety of factual scenarios so that now a "sizeable" overlap has developed between these provisions. 1 W. Rodgers, supra, at 225 (citing cases). Inevitably, then, prudential considerations often weigh into jurisdictional determinations. Claims alleging that EPA has not taken mandated action and requesting judicial intervention to order an agency decision are by common consensus better handled by the district court. Indiana & Michigan Electric, 733 F.2d at 491 (suggesting that suits "asking that the EPA be ordered to act" fall under rubric of Sec. 304); D. Currie, Air Pollution: Federal Law and Analysis Sec. 9.11 (1981) (Sec. 304 "properly provides a remedy for the failure of the Administrator to act at all, and its existence is a persuasive reason for declining to hold mere inaction 'final action' under ... Sec. 307"). Requests for the affirmance of standards or the vacation of EPA decisions, based on a well-developed administrative record, are, on the other hand, more properly brought in the court of appeals. Indiana & Michigan Electric, 733 F.2d at 490; D. Currie, Air Pollution Secs. 9.10-11 (Sec. 307 provides for review of administrative action, rather than for prepromulgation review).

Against this statutory backdrop of the CAA's delineation of review authority between the courts of appeals and district courts, I will now consider our jurisdiction to review EPA's determinations vis-a-vis visibility impairment and acid deposition.

With regard to any decision by the Administrator as to the appropriateness of a visibility impairment standard, the record before us is decidedly inconclusive. To recapitulate: In 1987, EPA issued an ANPR designed to solicit public comment on a fine particle standard protecting visibility. Since the closure of the public comment period over two years ago, there has been no formal agency decision as to whether it will or will not issue the fine particle standard.

In sum, EPA has taken some preliminary action in the direction of revising a standard, but that action certainly is not sufficiently developed to admit of nationwide application. While EPA's dishearteningly slow pace has not yet produced even a proposed visibility standard, but only an inconclusive ANPR, the Administrator remains technically engaged in at least the foreplay of formal rulemaking. NRDC's complaint about EPA's response to the evidence on particulates as a source of visibility impairment concerns the agency's gait, not the outcome of its rulemaking. Without a more definite determination by the Administrator as to how or whether he will act further on the visibility issue, it is impossible for us to conclude that he has taken "final action."8 The agency's use of the term "Final Rule" to explain its rationale for publishing an ANPR and for deferring a fine particle standard is not sufficient in itself to demonstrate that EPA has taken final action on the subject when the gist of the ANPR indicates just the opposite; "[i]t is the effect of the action and not its label that must be considered." Abramowitz v. EPA, 832 F.2d at 1075 (finding EPA action to be "final" despite agency claims to the contrary) (citing Harrison v. PPG Industries, Inc., 446 U.S. 578, 100 S.Ct. 1889, 64 L.Ed.2d 525 (1980)).9

Furthermore, NRDC's complaint about EPA's laggard tread on a visibility standard "is [not] embedded in a challenge" to other decisions whose finality is acknowledged. Compare Abramowitz, 832 F.2d at 1076. The visibility impairment standard that is under discussion, measured in terms of 2.5-micron particles, would be entirely separate from the soiling and nuisance secondary standard issued in July 1987, which is measured in terms of ten-micron particles. In no sense, then, would the fine particle standard be "embedded" in an overall secondary NAAQS for particulate matter. Consequently, EPA's final action on a soiling and nuisance standard cannot anoint the open-ended status of visibility rulemaking as "final action."

Prudential considerations also militate against our assertion of jurisdiction on the visibility impairment issue. NRDC seeks only that we order EPA to propose and promulgate a fine particle secondary visibility standard, pursuant to a fixed timetable set by us. NRDC does not request us to set aside the secondary standards for nuisance and soiling, or any other aspects of EPA's July 1987 Final Rule. In light of our acknowledged lack of authority to order direct, TRAC -type relief,10 the limited type of relief that NRDC seeks--judicial ordering of further agency action--is best granted by the district court. See Indiana & Michigan Electric, 733 F.2d at 491; D. Currie, Air Pollution Secs. 9.10-11.11

Whatever the merits of NRDC's claim of EPA's unreasonable delay in issuing a secondary standard on visibility, I conclude that our court does not have jurisdiction to pass on the issue because no "final action" has yet been taken by the Administrator. To the extent that EPA has a mandatory duty under Sec. 109(d)(1) to take some action on the evidence of visibility impairment presently before it--a question that is beyond the scope of the instant case--the district court would have jurisdiction to hear NRDC's suit under Sec. 304 of the CAA.

In its 1984 proposed rulemaking, EPA announced that it was inclined to defer a fine particle standard until research permitted it to develop a regional air control strategy dealing with both visibility impairment and acid deposition. By 1987, however, EPA apparently was inclined to pursue an independent visibility impairment standard because it considered research on acid deposition still inadequate for a secondary standard dealing with both effects. While EPA said it was still "considering whether ... to proceed with consideration of a visibility based standard in parallel with work on acid deposition," Revisions, 52 Fed.Reg. 24,646, the ANPR--the relevant operative document--focused exclusively on the feasibility of a separate visibility standard. With regard to acid deposition, the ANPR explained that "[a] decision on the need for additional emission controls for acid deposition has been deferred because of a lack of adequate scientific understanding." ANPR, 52 Fed.Reg. 24,671. In light of that uncertainty, the ANPR stated, "it may be prudent" to consider an independent visibility standard; inconsistencies between the visibility standard and a future acid deposition standard would be resolved during implementation. Id.

Thus, it appears that an acid deposition secondary standard is on indefinite hold until EPA is satisfied that the state of research art justifies one.12 And while EPA has taken some halting steps to devise a rule protecting visibility, the Administrator has apparently not yet made even a preliminary determination whether a secondary standard protecting the public welfare from acid deposition is appropriate or inappropriate. In effect, EPA has remained mute vis-a-vis a particulate matter standard regulating acid deposition.

I believe, however, that in the applicable statutory context, EPA's indefinite deferral is analogous to pure agency muteness, which, as we discussed in Part II.B, supra, amounts to a final decision not to revise. CAA Sec. 109(d)(1) requires that "[n]ot later than December 31, 1980, and at five-year intervals thereafter, the Administrator shall complete a thorough review" of the NAAQS, 42 U.S.C. Sec. 7409(d)(1) (emphasis supplied), and then revise these standards or promulgate new ones as appropriate, id. The current phase of revisions to particulate matter secondary standards represents EPA's response to the 1980 deadline. See Part I of per curiam opinion, supra, at 966-967 (describing evolution of 1987 Revisions). Nearly a decade has passed, and EPA has taken no formal action indicating that it is moving in the direction of making a decision whether or not to establish a secondary standard for acid deposition, a welfare effect that the agency has explicitly identified. Revisions, 52 Fed.Reg. 24,646. While we have no evidence suggesting that EPA's refusal to decide whether to issue an acid deposition standard is disingenuous, I believe that EPA's inaction for ten years beyond the statutory deadline is effectively a final decision not to revise in this NAAQS revision cycle. The agency's protestations of its open mind and continuing study are, without more, insufficient evidence that in this rulemaking, EPA's decision is anything but one not to establish an acid deposition standard as a result of this review and revision cycle. See Sierra Club v. Gorsuch, 715 F.2d at 659 ("Judicial review of decisions not to regulate must not be frustrated by blind acceptance of an agency's claim that a decision is still under study."); Environmental Defense Fund v. Ruckelshaus, 439 F.2d 584, 593 (D.C.Cir.1971) (agency cannot delay determination "indefinitely" in order to escape proper judicial review).

EPA has relied on only the most general claims of scientific uncertainty to justify its inaction. See ANPR, 52 Fed.Reg. 24,671. There is simply no substantive record available for us to consider the merits of EPA's lack of movement on this issue. Consequently, we remand to EPA for the agency to provide us with clear and cogent reasons why it has taken no action vis-a-vis an acid deposition standard. To avoid unnecessary prolongation of this overextended dispute, we order EPA to submit that statement of reasons within sixty days of the issuance of the mandate of this opinion.

Our decision today implies no conclusion as to the merits of EPA's lack of movement on acid deposition rulemaking. EPA may respond with a range of possible arguments, perhaps including claims detailing the nature of the claimed scientific uncertainty and the benefits of pursuing a solution to the acid deposition problem in the legislative arena. EPA's statement of reasons may indeed provide a cogent explanation for its conduct, worthy of substantial deference; we seek only a more complete explanation as to why EPA did not engage in a rulemaking proceeding vis-a-vis acid deposition in this cycle.

I admit to a great deal of frustration in resolving the jurisdictional questions on this appeal. It would be disingenuous to pretend that all of the various cases in the various circuits involving the appropriate provinces of Secs. 307 and 304 are reconcilable and of one piece. They are not. Similarly, we cannot wholly ignore the possible counterincentives that a restrictive scope for Sec. 307 review provides to a recalcitrant agency to delay decisionmaking or to offer little or no explanation of its inaction in order to avoid a final decision on whether or not to revise a standard. On the other hand, I am reluctant to read the finality requirement of Sec. 307 as permitting us to wade in and order action by EPA whenever its timetable for decisionmaking appears intolerably slow. Congress has provided some sort of rough roadmap for jurisdiction between the courts of appeals and district courts, assigning to this court jurisdiction over actions of national consequence that we can conveniently review on a fully developed administrative record, and assigning to district courts jurisdiction over petitions to require the Administrator to do what the Act demands, allowing opportunity in such suits for the parties to create their own record in the trial court.

The vague contours of the jurisdictional lines in the CAA have been widely commented upon and may well deserve reconsideration in the context of pending amendments to the Act. For the present, however, as judges we can only do our best to discern and follow the congressional plan, whatever its imperfections.

HARRY T. EDWARDS, Circuit Judge, concurring in Parts I-III of the per

curiam opinion, concurring in Part IV of the per curiam only in the judgment on the acidic deposition issue and dissenting on the visibility impairment issue:

In my view, this court has jurisdiction over the claims of the Natural Resources Defense Council ("NRDC") pursuant to section 307 of the Clean Air Act, 42 U.S.C. Sec. 7607 (1982).1 I agree with Chief Judge Wald that, in considering NRDC's claims, we are bound by the jurisdictional framework of sections 304 and 307 of the Clean Air Act ("Act"), 42 U.S.C. Secs. 7604, 7607 (1982). I also accept the conclusion that, under section 307, we have jurisdiction over challenges to the particulate matter standards revisions only if there has been "final action" by the Agency. I agree that the Agency has taken final action in failing to establish standards addressing acidic deposition. But I disagree with the conclusion that the Agency has not taken "final action" on whether to set standards to protect against the adverse public welfare effects of visibility impairment.

The record in this case shows that, following an extensive notice and comment period, EPA determined in a final rule not to revise the secondary national ambient air quality standards ("NAAQS") to address the effects of visibility impairment and acidic deposition as part of the 1980 revision required by section 109(d), 42 U.S.C. Sec. 7409(d) (1982). See 52 Fed.Reg. 24,634 (1987). That determination was a "final action" under section 307 and is reviewable by this court. On the merits, I would grant NRDC's petition and order EPA to finalize regulations within 240 days from the decision of this court.

Under the Clean Air Act, the courts of appeals have jurisdiction over challenges to "final actions" of the Administrator, provided that the petition for review is filed within sixty days from the date notice of the action appears in the Federal Register. See 42 U.S.C. Sec. 7607(b)(1). While the Act does not define the terms "action" or "final action" these "terms have traditional meanings in the administrative context," Sierra Club v. Gorsuch, 715 F.2d 653, 656 (D.C.Cir.1983); and, in applying these terms, the courts have used practical and functional definitions both in the context of section 307 and in the administrative context generally.2 Moreover, it is axiomatic that determination of whether an action is final does not turn on how the Agency labels the action. See, e.g., Sierra Club v. Gorsuch, 715 F.2d at 658-59 ("EPA's position--that final action has not been taken--does not affect our jurisdiction"); Abramowitz v. EPA, 832 F.2d 1071, 1075 (9th Cir.1987) ("mere fact that the Agency did not label its action 'final' does not preclude review in this court under section 307. It is the effect of the action and not its label that must be considered.").

In a case like the instant one, the confluence of two factors is central to the finding of a final action: (1) the completion of a rulemaking and (2) some clear indication that the disputed issue was considered during the rulemaking procedure. Where a rulemaking procedure has been completed and an issue has been properly raised during the course of the rulemaking, the Agency's failure to address or resolve that issue does not render it unreviewable by the court of appeals. Rather, in such circumstances, the Agency decision not to act or to defer a decision on a particular issue is a final action, reviewable by the court of appeals. See, e.g., Maine v. Thomas, 874 F.2d 883, 891 (1st Cir.1989) ("regulations, though entirely promissory in respect to regional haze, nevertheless amounted to 'final action taken' within the ambit of 42 U.S.C. Sec. 7607(b)(1)"); see also Environmental Defense Fund v. Costle, 448 F.Supp. 89, 93 (D.D.C.1978) (finding deferral a "final-action" under section 307--"Administrator had taken a final action as to the issue of whether Sec. 165 [of the Clean Air Act] was to be immediately effective"); 1 W. RODGERS, ENVIRONMENTAL LAW: AIR AND WATER Sec. 3.3 at 209 (1986) ("[t]he scope of any deliberation producing action includes things not done that are fairly considered part of the original choice"); D. CURRIE, AIR POLLUTION: FEDERAL LAW AND ANALYSIS Sec. 9.10 at 9-31 (1981 & Supp.1990) ("[A]llegations that the Administrator has failed to take action required by statute should not be permitted to circumvent the plain statutory command that judicial review of decisions respecting implementation plans and other regulations is to be in the courts of appeals").

At issue in this case is whether EPA's actions regarding visibility impairment and acidic deposition are "final actions." On July 1, 1987, EPA promulgated a final rule, pursuant to section 109(d), 42 U.S.C. Sec. 7409(d) (1982). Section 109(d) requires the Administrator to review and revise the particulate matter criteria and standards in accordance with section 7408 and section 7409(b) "[n]ot later than December 31, 1980, and at five-year intervals thereafter." Pursuant to this mandate, albeit late for the December 31, 1980, deadline, the Agency reviewed and revised the "criteria" for particulate matter and published its findings and conclusions regarding visibility impairment and acidic deposition.³ In its proposal for revisions to the standards, published in 1984, the Agency proposed deferring a decision on a fine particle secondary standard until it could develop a strategy for addressing visibility impairment and acidic deposition problems together.4 After the notice and comment period, EPA "announce[d] [its] final decisions regarding ... changes" in the particulate matter standards based on its review and revision of the criteria.5 In this final rule, the Agency decided not to establish standards to address visibility impairment and acidic deposition.6 The Agency's decision regarding visibility impairment was not to revise the standards as part of the "1980 revision," but to consider further whether to revise.7 The Agency's decision regard ing acidic deposition was to defer revision indefinitely "because of a lack of adequate scientific understanding."8 The "1980 revision" is finished, and NRDC now challenges the Agency's determinations in that revision as violative of the statutory mandate of 42 U.S.C. Sec. 7409.

Our decision in Sierra Club v. Gorsuch, 715 F.2d 653 (D.C.Cir.1983), provides the clearest precedent for the case before this panel. In that case, EPA declined to place strip mines on the list of pollutant sources subject to fugitive emissions regulations in both the proposed and the final regulations. Id. at 654-56. On review, EPA contended that, because it had not made a final decision and would further study the issue and consider the need for further regulation, there could be no jurisdiction under section 307. The court in Sierra Club v. Gorsuch rejected this contention, holding that "jurisdiction over Sierra Club's claim exists because its petition challenges EPA's list of sources as promulgated. The inclusion of strip mines was clearly an issue in that rulemaking ... and Sierra Club had submitted comments on the issue." Id. at 657 (emphasis in original).

In a similar case, the Second Circuit also has held that the Agency's failure to take any action on an issue that was a subject of the rulemaking did not deprive the court of appeals of jurisdiction to consider the merits of the challenge to EPA's action. See Vermont v. Thomas, 850 F.2d 99, 102 (2d Cir.1988) (reaching the merits of petitioner's claim for review of EPA's final rule that took "no action" on portions of Vermont's state implementation plan, even though the issue in question was not entwined with other issues in the plan).

In the instant case, there is no question that the issues in question were subjects of the rulemaking. EPA acted to revise the particulate matter secondary standards to protect the public welfare against adverse effects, as required by section 109(d). The criteria document set forth the latest scientific knowledge regarding visibility impairment and acidic deposition effects of particulate matter.9 And parties commented regarding those effects. Thus, under the controlling case law, I find no basis for denying jurisdiction over NRDC's claims pursuant to section 307.

I recognize that the line drawn between cases arising under section 307 and section 304 is sometimes difficult to discern. Whereas section 307 encompasses petitions to review final agency actions, section 304 is more narrowly confined to address citizen suits to compel agency observance of nondiscretionary duties. The district court alone has jurisdiction to hear suits under section 304, and "[f]inality-based jurisdiction under section and district court jurisdiction under section are mutually exclusive." See Maine v. Thomas, 874 F.2d 883, 886 n. 5 (1st Cir.1989). "Unsurprisingly, this jurisdictional dichotomy has created a confused class of circumforaneous litigants, wandering perplexedly from forum to forum in search of remediation." Id. at 884-85. Nonetheless, the courts have focused on the posture of a case in determining whether a challenge properly arises under section 307 or section 304. In so doing, the various courts of appeals that have considered the question have been consistent in saying that jurisdiction always lies under section 307 after the Agency has completed a rulemaking and the challenge is to that rulemaking.

For example, in Environmental Defense Fund v. Thomas, where EPA had failed even to commence rulemaking to revise the national ambient air quality standards, after revising the criteria document for sulfur oxides ("SOx") pursuant to the requirements of section 109(d)--the same provision under which EPA did rulemake to completion regarding particulate matter at issue in the instant case--the petitioner brought suit in the district court "to compel the Administrator to promulgate revised NAAQS for SOx." 870 F.2d 892, 895 (2d Cir.), cert. denied, --- U.S. ----, 110 S.Ct. 537, 107 L.Ed.2d 535 (1989). The Administrator failed to take any official public action either revising the existing standards or formally declining to revise them. Id. The Second Circuit concluded that the Administrator had a duty, under section 109(d), to reach some decision regarding the revision of the NAAQS and, because the Agency had subsequently begun the formal process of decisionmaking, the Second Circuit remanded to the district court for entry of an order directing EPA to continue the formal process of decisionmaking. Id. at 896. The Second Circuit noted that obtaining a formal decision is "all the relief" to which the appellants are entitled in a section 304 action. Id. at 900. The Second Circuit further noted that "[w]hether the decision when reached is wrong on the merits--even egregiously wrong--will be for the District of Columbia Circuit to resolve." Id.

Similarly, in Maine v. Thomas, 874 F.2d 883 (1st Cir.1989), the appellant filed suit in district court pursuant to section 304 seeking to compel EPA to promulgate regulations designed to deal with "regional haze." Id. at 885. The First Circuit concluded that the district court did not have jurisdiction to compel action because the deadline announced in the regulations had not passed, id. at 888, and hence EPA did not have a present nondiscretionary duty to act. However, the First Circuit explained in great detail that had the appellant challenged EPA's inaction following the final rule in which the Agency deferred taking action on regional haze, the court of appeals would have had jurisdiction to review EPA's deferral. Id. at 887.

In the final rule in Maine v. Thomas, EPA set standards only for plume blight and essentially promised to deal with the matter of regional haze in future rules and orders. Id. at 885. The First Circuit flatly rejected EPA's contention that its decision in the final rule regarding regional haze was not a final action. Id. at 886, 887. The First Circuit emphasized that the regulations "represented EPA's assessment of what might reasonably be done in 1980," id. at 886 (emphasis in original), and the Agency's announcement that " '[f]uture phases [of its regulations] will extend the visibility program by addressing more complex problems such as regional haze,' " id. at 885 (quoting 45 Fed.Reg. 80,084, 80,086 (1980)), was a final action when the regulations were promulgated in December 1980, id. at 887. The First Circuit could scarcely have been more explicit that had the rule been challenged within sixty days of promulgation, "the premises for postponed regulation could well have been reviewed by the District of Columbia Circuit." Id. at 888-89. "To mince no words, the decision to defer constituted a fully developed part of the final action taken on the statutory mandate." Id. at 887.

The First Circuit further noted that such a challenge would have been ripe for review, because it

would neither have forced a court to entangle itself "in abstract disagreement over administrative policies ...," nor have subjected EPA to "judicial interference [before] an administrative decision has been formalized and its effects felt in a concrete way by the challenging parties." Abbott Laboratories v. Gardner, 387 U.S. 136, 148-49, 87 S.Ct. 1507, 1515-16, 18 L.Ed.2d 681 (1967). While the plume blight regulations are paradigmatic examples of an order ripe for review, the temporizing "promise" anent uniform haze, given the congressional mandate and the state of the art, was a bit greener at the edges--but also ripe.

Id. "Surely a court could have assessed whether the record adequately supported EPA's decisions and decided whether the promise of future regulation adequately harmonized with the Agency's nondiscretionary duty to progress toward the national goal." Id. at 889. I cannot see how the deferrals in the instant case compel a different result or defy the reasoning set forth by the First Circuit in Maine v. Thomas.

In short, EPA's failure to take action in a final rule, either by postponing a decision or failing altogether to address an issue raised in comments, does not turn a petition for review into an action for failure to comply with a mandatory duty such that the district court alone has jurisdiction. Indeed, this view of "final action" would undo the basic design of notice-and-comment rulemaking, which contemplates notice from the agency of possible action, an agency response to comments, reasoned analysis by the agency to justify its decision and, finally, the possibility of judicial review to ensure compliance with these requirements. If the Agency can fend off review by simply deferring action at the conclusion of each rulemaking, with a vague promise to consider the matter in some later rulemaking, a critical component of notice-and-comment rulemaking would be lost. This is not to say that a decision to defer is necessarily inconsistent with the Agency's legal obligations; rather, it is to say that such a decision at the conclusion of a rulemaking is reviewable under section 307.

Contrary to the Agency's assertions in this case, EPA's promise of a new rulemaking does not render its failure to make certain revisions in the substance of the initial rule unreviewable. See Maine v. Thomas, 874 F.2d at 886-89. Nor does any alleged incompleteness of the record affect the court's jurisdiction under section 307. See Harrison v. PPG Indus., Inc., 446 U.S. 578, 592-93, 100 S.Ct. 1889, 1987-98, 64 L.Ed.2d 525 (1980) (rejecting argument that in section 307 Congress gave the courts of appeals jurisdiction only over final actions for which there was a developed record). If the record does not provide a sufficient basis for a determination by the court on the merits, the proper course is to remand to the Agency for further explanation. Id. at 592-94, 100 S.Ct. at 1897-98. The mischief latent in any other approach is all too clear.

Had EPA failed to initiate or complete the "1980 revision" of particulate matter standards, then jurisdiction might lie in the district court under section 304, 42 U.S.C. Sec. 7604. Alternatively, a party seeking to compel action to initiate revision might need first to file with EPA a petition for rulemaking. See Group Against Smog & Pollution, Inc. v. EPA, 665 F.2d 1284, 1289-91 (D.C.Cir.1981). There is no reason in the instant case to forge into the complex questions addressed in Sierra Club v. Thomas, 828 F.2d 783 (D.C.Cir.1987), regarding the proper forum for challenges to EPA's failure to initiate or complete rulemaking. In Sierra Club v. Thomas, rulemaking had not come to a conclusion; the petitioner in that case claimed that EPA had "unreasonably delayed in promulgating regulations," and it sought "declaratory and injunctive relief to compel EPA to conclude th[e] rulemaking begun on October 24, 1984." Id. at 787. But, here, EPA completed the review and revision of its particulate matter standards and, as part of its final rule, decided not to promulgate standards to address visibility impairment and acidic deposition. This is final Agency action that is subject to review under section 307.

The relief available under section 304 is to compel rulemaking; section 304 does not permit substantive review. See, e.g., Environmental Defense Fund v. Thomas, 870 F.2d at 896 ("substance of the Administrator's decision is beyond the power of the district court, ... its authority being limited to ordering the Administrator to make a formal decision"). The content of the Administrator's decision whether to revise is reviewable only in the District of Columbia Circuit. Id. at 899. We will create an enormous loophole in the congressional scheme of review under the Clean Air Act if we hold that EPA can avoid substantive review by saying, at the end of a rulemaking, that action is to be deferred. As the First Circuit has noted, if the rule in which the Agency defers action but promises to take further action

were not subject to review, then it would in turn simply spawn more section 304 actions, each resolvable, ultimately, by an unreviewable promise. Even Kafka would have found it difficult to devise a more twisted antilogy.

Maine v. Thomas, 874 F.2d at 889.

Such an interpretation of the scope of sectin 304 and 307 would leave parties circling in a Sargasso Sea of jurisdictional boundaries, out of reach of the substantive review they seek. It would indulge a potentially mischievous notion of final action--one which bears no relation to ripeness and which surrenders this court's authority to offer substantive review of EPA's decisions. The problem that we face in a case of this sort is that the merits appear to be so difficult that a court is justly reluctant to wade into the agency's domain of expertise. But, in my view, that is not a legitimate basis for a court to deny jurisdiction under section 307. Indeed, it is not our function to second-guess the Agency on substantive matters, so we need not languish under such concerns. In any event, as Justice Scalia recently noted in another context, "[o]ur cases have long supported the proposition that federal courts lack the authority to abstain from the exercise of jurisdiction that has been conferred." New Orleans Public Serv. v. Council of New Orleans, --- U.S. ----, 109 S.Ct. 2506, 2512, 105 L.Ed.2d 298 (1989). Here, too, we must act pursuant to the jurisdiction that has been conferred by Congress.

Because EPA promulgated a final rule to revise the particulate matter standards pursuant to section 109(d) and NRDC challenged the final rule, we have jurisdiction to reach the merits. The question for this court, then, is whether, having identified visibility impairment and acidic deposition as "known adverse effects," the Agency's failure to promulgate standards to protect the public welfare from these effects was within its authority and, if so, whether such action was reasonable.

The language of section 109 provides the Administrator discretion to determine the level of air quality standards that will protect the public welfare from any known or anticipated adverse effects associated with the presence of the pollutant in the air. See 42 U.S.C. Sec. 7409(b)(2). But, in revising the standards for particulate matter, the Administrator was required to select a level which in his judgment is requisite to protect the public welfare from any known or anticipated adverse effects associated with the presence of particulate matter in the air. Id.; see also id. Sec. 7409(d) ("shall ... promulgate such new standards as may be appropriate in accordance with ... subsection (b) of this section"); id. Sec. 7602(h) (1982) (expressly including effects on soils, water, vegetation, manmade materials, animals and visibility within the meaning of "welfare"). In specifically addressing the issue of acidic deposition in the Acid Precipitation Act of 1980, Congress expressly preserved regulatory authority under existing law. See 42 U.S.C. Sec. 8904(b) (1982).10

The language and legislative history are not ambiguous. Consequently, under Chevron step one, see Chevron U.S.A. Inc. v. Natural Resources Defense Council, 467 U.S. 837, 843 n. 9, 104 S.Ct. 2778, 2781-82 n. 9, 81 L.Ed.2d 694 (1984), the Administrator was required to select a level to protect the public from visibility impairment and acidic deposition. Moreover, as we noted in addressing American Iron and Steel Institute's challenges, Congress prohibited EPA from considering economic and technological considerations in setting the NAAQS and, as is all too clear from our discussion of the primary standards, EPA need not wait for scientific certainty to effect air quality standards.

In my view, the court should remand to the Agency with instructions for it to propose rules addressing visibility impairment and acidic deposition within ninety days and finalize the necessary regulations within 240 days from the decision of this court. See 42 U.S.C. Sec. 7409(a) (imposing even stricter deadlines for initial standards); see also Environmental Defense Fund v. EPA, 852 F.2d 1316, 1331 (D.C.Cir.1988) (imposing on EPA a schedule for fulfilling its statutory obligations under the Resource Conservation and Recovery Act), cert. denied, --- U.S. ----, 109 S.Ct. 1120, 103 L.Ed.2d 183 (1989); Sierra Club v. EPA, 719 F.2d 436, 470 (D.C.Cir.1983) (requiring EPA to fulfill its obligations under the Clean Air Act within six months, the period originally specified by Congress), cert. denied, 468 U.S. 1204, 104 S.Ct. 3571, 82 L.Ed.2d 870 (1984); Sierra Club v. Gorsuch, 715 F.2d at 661 n. 47 (absent cause shown, requiring action within ninety days). Obviously, if new legislation alters EPA's mandate, EPA may petition for appropriate action by the court with respect to any outstanding order. Until such time, however, the court must give effect to existing law.

SILBERMAN, Circuit Judge, joining Parts I-III of the per curiam opinion, joining only the judgment on the visibility issue, and dissenting on the acid deposition issue:

I agree that we do not have jurisdiction to hear the NRDC's petition concerning visibility impairment because of the absence of final agency action, but I disagree with the majority's conclusion that the EPA has made a final decision with respect to acid deposition. Therefore, I respectfully dissent.

By determining that we have jurisdiction to review the EPA's delay in proposing secondary standards for acid deposition, Chief Judge Wald's opinion (and Judge Edwards') expands our authority well beyond the boundaries Congress marked out for us. Section 307(b)(1) of the Clean Air Act, 42 U.S.C. Sec. 7607(b)(1), provides for jurisdiction exclusively in this court to review "final action" taken by the Administrator of the EPA in revising a national ambient air quality standard. See Harrison v. PPG Indus., 446 U.S. 578, 100 S.Ct. 1889, 64 L.Ed.2d 525 (1980). Although the language of the statute strongly suggests that agency inaction is not reviewable at all, we have, in other contexts, carved out a limited number of exceptions.

To be sure, the Administrative Procedure Act imposes a general duty on agencies to avoid "unreasonable delay" in reaching a final decision and thereby, in that context, expressly permits judicial review of agency inaction. See 5 U.S.C. Sec. 706(1). But Congress specifically blocked that avenue for this court to use in reviewing EPA inaction by broadly exempting the Clean Air Act from the requirements of the APA and then selecting those provisions it did want to be subject to the APA. See 42 U.S.C. Sec. 7607(d)(1). Section 307(d)(1)(A) explicitly shields the "promulgation or revision" of any NAAQS from the unreasonable delay provisions of the APA. See 42 U.S.C. Sec. 7607(d)(1)(A).

We have nevertheless authorized review of EPA behavior when we thought the record indicated that the agency has cloaked a final decision--even a decision not to act--from the public. Certainly the most extreme "stretch" in this regard is the case of Sierra Club v. Gorsuch, 715 F.2d 653 (D.C.Cir.1983), where we rested our jurisdiction on the determination that EPA had made a final, if covert, decision not to "list" fugitive emissions from coal mines among sources subject to fugitive dust regulations. We so concluded because we thought that the agency's previous policy of not regulating fugitive dust, its own self-announced deadline for responding to the issue, and its silence in the face of uncontroverted evidence of the impact of those emissions all combined to suggest a final agency decision not to act. Admittedly, the opinion seems really to be based on a determination that the agency should have acted--a "constructive" final action notion--and the covert decision theory may well be a fiction, but it is a fiction to which we have held fast until today. See, e.g., Sierra Club v. Thomas, 828 F.2d 783 (D.C.Cir.1987) (rejecting disguised agency action as a basis for jurisdiction because no such final decision was apparent from the record).

Both of my colleagues abandon the fiction and embrace a theory of constructive final action--although they do not call it that. Chief Judge Wald views pure agency silence--as opposed to silence which masks an unacknowledged final agency action--as "analogous" to a final decision and therefore it may form a basis for our jurisdiction. See Wald op. at 987. Judge Edwards, on the other hand, extends our jurisdiction to any matter EPA considered before promulgating a rule whether or not the rule addresses the matter.1 See Edwards op. at 989-990. Judge Edwards is, in my view, logically consistent with respect to both the visibility impairment and acid deposition issues, whereas Chief Judge Wald "splits the baby" based on what appears to me to be a tenuous distinction.

Part of the difficulty with these concepts of appellate jurisdiction is that they require the court to consider the merits of the claim explicitly or implicitly in order to determine whether it has jurisdiction. Creeping into the court's determination is the notion that the agency should have acted based on the record before it, and therefore we will force them to act. Of course the same can be said about review under the APA for unreasonable delay--but traditionally, and largely for that reason, we are rather circumspect in entertaining such petitions. See, e.g., United Steelworkers of Am. v. Rubber Mfrs. Ass'n, 783 F.2d 1117, 1120 (D.C.Cir.1986). In any event, Congress deliberately omitted that basis for our appellate jurisdiction. As I see it, both of my colleagues would simply add to the statute that which Congress omitted. However, "it is not our task to make an imperfect statute perfect." Central Vermont Ry. v. Brotherhood of Maintenance of Way Employees, 793 F.2d 1298, 1303 (D.C.Cir.1986).

Indeed, Congress did not totally ignore the generic problem of "unreasonable delay." Under section 304(a)(2), the district courts are given jurisdiction over claims that the Administrator failed "to perform any act or duty ... which is not discretionary with the Administrator." Of course, that formulation is narrower than the unreasonable delay cause of action under the APA, which can extend to discretionary matters. Therefore, quite clearly, judicial review of EPA's behavior is in this respect more restricted than that which applies to other agencies. This has bothered some courts. The Second Circuit, for instance, in Environmental Defense Fund v. Thomas, 870 F.2d 892 (2d Cir.), cert. denied sub nom. Alabama Power Co. v. Environmental Defense Fund, --- U.S. ----, 110 S.Ct. 537, 107 L.Ed.2d 535 (1989), concluded on appeal from a section 304 action that the EPA did not have authority to defer decisionmaking on a NAAQS revision because otherwise it would "leav[e] the matter in a bureaucratic limbo subject neither to review in the District of Columbia Circuit nor to challenge in the district court." Id. at 900. That sort of reasoning seems to me to reflect a judicial review tail wagging the legislative intent dog, and it is reasoning we have heretofore rejected: "It does not follow automatically that, if the district court lacks jurisdiction, then it must lie in this court, for both courts have just as much jurisdiction as Congress has provided by statute." Sierra Club v. Thomas, 828 F.2d 783, 792 (D.C.Cir.1987). The bureaucratic limbo to which the Second Circuit referred is merely an area of agency behavior (inaction) not subject to judicial review. I do not believe that Congress, by so deciding that the judiciary's role would be to this limited extent circumscribed, threatened the foundations of the Republic.

Chief Judge Wald's general discussion of our jurisdiction seems influenced by the Second Circuit's reasoning because, even with regard to the visibility impairment issue on which we agree that we do not have jurisdiction, she broadly implies that the district court would so as to avoid a "gap" in judicial review. Of course, as I have indicated above, I do not think that necessarily follows and, in any event, the scope of the district court's jurisdiction is not before us.

Apart from my disagreement with both of my colleagues' analysis of the jurisdiction provisions, Chief Judge Wald's explanation for her inconsistent treatment of the acid deposition and visibility impairment issues is, in my view, unpersuasive. Her untenable distinction between pure silence and "some preliminary action," see Wald op. at 984, only highlights the analytical difficulties this new concept of jurisdiction brings. Once we abandon a view of jurisdiction anchored in an actual (whether publicly announced or not) final decision in favor of "constructive" final action, we are cast adrift from logic and principle and are led inexorably into drawing arbitrary lines on the spectrum between reviewable agency silence and unreviewable agency behavior. In my view, when an agency has failed to decide, we do not have appellate jurisdiction--no matter how "close" the agency is to that decision.2

But even if we were to apply Chief Judge Wald's reasoning, the EPA's inaction on acid deposition should not be reviewable. She points out in justifying our declination of jurisdiction over the visibility issue that since the EPA has apparently taken "some preliminary action" on visibility impairment--in the form of a lapsed advanced notice of proposed rulemaking announced three years ago--the agency still remains in the "foreplay of formal rulemaking." See Wald op. at 986. Since we do not know how or whether the agency will act further on visibility, it is said that "it is impossible for us to conclude that [the Administrator] has taken 'final action.' " Id. at 986. Under this analysis, however, the procedural posture of the stalled acid deposition rulemaking, where Chief Judge Wald and thus the majority does find finality, is virtually indistinguishable from that of the visibility impairment standard.

Chief Judge Wald concedes that the agency has not made a final decision on acid deposition. At worst, according to her opinion, the agency has indefinitely deferred final action. But that categorization, under the opinion's own reasoning, should be sufficient to render EPA's inaction on acid deposition just as unreviewable as the agency's nondecision on visibility impairment. See id. at 986. The same "prudential" considerations that counselled against judicial interference on visibility impairment, see id. at 986, apply with equal force to the acid deposition issue. And like visibility impairment, the acid deposition standard is not "embedded" in a challenge to an admittedly final action, and so that alternative ground is similarly unavailable. See id. But the court nonetheless reaches a different conclusion on acid deposition, because, according to the opinion, the agency has remained "mute" on acid deposition in the context of section 109(d)'s statutory "deadline." See id. at 987. This "distinction," however, ignores the agency's ongoing efforts towards a final decision on acid deposition.³

The record simply does not support the opinion's suggestion that the agency is standing idle on the acid deposition matter, and I cannot see why whether the agency is publicly "mute" is at all relevant. The opinion concedes that the agency is free to delay making a decision as long as it acts in good faith. See id. at 983-984. The EPA announced that it had "deferred" its acid deposition decision because of "lack of adequate scientific understanding." 52 Fed.Reg. 24,671. Unlike visibility impairment, acid deposition, in the agency's view, presented "a topic of extreme scientific complexity," 53 Fed.Reg. 14,926, 14,935 (1988), and was not sufficiently well-understood to justify a secondary standard at this time. The EPA and the scientific community believe that acid depositions are caused by some complex interaction of particulate matter, sulfur oxides, and nitrogen oxides. See 53 Fed.Reg. 14,926, 14,936 (1988). (By contrast, ironically, EPA's independent Clean Air Science Advisory Committee (CASAC) determined that "[t]he criteria document [dealing with visibility] provides an excellent basis for agency deci sion making on these issues." 52 Fed.Reg. 24,634, 24,657.) In view of this scientific uncertainty, the agency has followed two tracks. First, pursuant to the Acid Precipitation Act of 1980, 42 U.S.C. Secs. 8901-05, the EPA is only now nearing the end of a ten-year study of the causes of and possible governmental responses to acid deposition. Second, the agency has been evaluating solutions for the acid deposition problem on non-particulate matter fronts, particularly the NAAQS for sulfur oxides. Recently, the EPA asked for public comments on its proposed decision to defer revisions to its sulfur oxide standard because of the scientific complexity of acid deposition. See 53 Fed.Reg. 14,926 (1988).

The picture that emerges from the record is not one of an agency sleeping on a serious public welfare problem but that of an agency moving, albeit slowly, towards a final decision. Until that time, even under the Chief Judge's reasoning, we are without jurisdiction to second-guess either the pace or outcome of that decision.