In the eastern United States, the average visual range in most national parks and wilderness areas designated as Class I Federal areas, see 42 U.S.C. § 7472(a), is less than 30 kilometers, about 20 percent of what it would be under natural conditions. See National Research Council, Protecting Visibility in National Parks and Wilderness Areas 1 (1993). In order to address this problem, the Environmental Protection Agency promulgated a Regional Haze Rule, 40 C.F.R. § 51.308, pursuant to Section 169A of the Clean Air Act ("CAA"), 42 U.S.C. § 7491. See Regional Haze Regulations and Guidelines for Best Available Retrofit Technology (BART) Determinations, 70 Fed.Reg. 39,104 (July 6, 2005) (the "Haze Rule"). The Haze Rule requires that under specified circumstances states impose best available retrofit technology ("BART") on any BART-eligible sources. The latter are a specific class of large stationary pollution sources that "were put in place between August 7, 1962 and August 7, 1977, and whose operations fall within one or more of 26 specifically listed source categories." 70 Fed. Reg. at 39,105/1; see also 40 C.F.R. § 51.301. The regulation calls for imposition of BART if the source "may reasonably be anticipated to cause or contribute to any impairment of visibility in any mandatory Class I Federal area." 40 C.F.R. § 51.308(e)(1)(ii). The Haze Rule also permits states to reduce haze by alternate means, including a regional approach, so long as the alternative would be "better-than-BART"—i.e., would improve visibility more rapidly than under BART. 40 C.F.R. § 51.308(e)(2). Aspects of the Haze Rule have been before this court twice before, Center for Energy and Economic Development v. E.P.A., 398 F.3d 653 (D.C.Cir. 2005) ("CEED"); American Corn Growers Ass'n v. E.P.A., 291 F.3d 1 (D.C.Cir.2002) ("Corn Growers"), and those opinions contain extensive discussions of the rule's statutory framework and regulatory history.

This case involves challenges from multiple groups, including the Center for Energy and Economic Development and the Utility Air Regulatory Group ("industry petitioners"), and the National Parks Conservation Association ("environmental petitioner"). In its brief, EPA succinctly summarizes the challenges: "Industry Petitioners generally challenge the rule as inappropriately requiring States to apply BART to too many sources, while the Environmental Petitioner argues that the rule improperly allows States to exempt too many sources from BART." Because we believe the Haze Rule is a reasonable interpretation of CAA § 169A, we affirm the rule against both sets of challenges.

As we explained in Corn Growers, § 169A(a)(1) of the Clean Air Act established a national goal of preventing and remedying existing visibility impairment at Class I areas, and CAA § 169A(b)(2) directs EPA to issue regulations requiring that states adopt measures—including BART—to make "reasonable progress" towards meeting this national goal. See Corn Growers, 291 F.3d at 5-6.

As outlined in § 169A(b)(2)(A) and implemented by the Haze Rule, the BART process consists of two steps. First, in the "Attribution Step" ("Step I"), states must review each "BART-eligible source" within the state to determine whether any such source emits "any air pollutant which may reasonably be anticipated to cause or contribute to any impairment of visibility in any mandatory Class I Federal area;" sources that do so are "subject to BART." See 40 C.F.R. § 51.308(e)(1)(ii). An earlier preamble to the Haze Rule required states to "find that a BART-eligible source is `reasonably anticipated to cause or contribute' to regional haze if it can be shown that the source emits pollutants within a geographic area from which pollutants can be emitted and transported downwind to a Class I area," an approach known as "collective contribution." Regional Haze Regulations, 64 Fed.Reg. 35,714, 35,740/1 (July 1, 1999). In Corn Growers we struck down such guidance as "inconsistent with the Act's provisions giving the states broad authority over BART determinations." 291 F.3d at 8 (emphasis added). In doing so, however, we did not foreclose the states themselves from deciding to take a collective approach in the Attribution Step, see id. at 18 (Garland, J., dissenting on other grounds), and the current rule identifies "collective contribution" as only one of at least three different approaches that a state may take in meeting its obligations under CAA § 169A(b)(2)(A). See 70 Fed. Reg. at 39,117/2. Under the current Haze Rule, a state can complete the Attribution Step by using collective attribution, by demonstrating that, cumulatively, none of its BART-eligible sources contributes to visibility impairment, or by analyzing each source's individual contribution. Id. States "may also use other reasonable approaches for analyzing the visibility impacts of an individual source or group of sources." 70 Fed.Reg. at 39,162/1.

The second step outlined in § 169A(b)(2)(A), the "Determination Step" ("Step II"), requires states to determine the particular technology that an individual source "subject to BART" must install. That determination requires consideration of five factors: "the cost of compliance, the energy and nonair quality environmental impacts of compliance, any existing pollution control technology in use at the source, the remaining useful life of the source, and the degree of improvement in visibility which may reasonably be anticipated to result from the use of such technology." 42 U.S.C. § 7491(g)(2); see also 40 C.F.R. § 51.308(e)(1)(ii); 70 Fed.Reg. at 39,163/3. In Corn Growers, we held that these five factors "were meant to be considered together by the states," 291 F.3d at 6, but that EPA could not require the states to evaluate the improvement factor collectively while mandating that the other four factors be evaluated separately for each individual source. Compare id. at 8 with id. at 8-9.

BART is not, however, the sole means by which states can meet their obligations under the Clean Air Act. The Haze Rule also permits states

to implement or require participation in an emissions trading program or other alternative measure rather than to require sources subject to BART to install, operate, and maintain BART. Such an emissions trading program or other alternative measure must achieve greater reasonable progress than would be achieved through the installation and operation of BART.

40 C.F.R. § 51.308(e)(2). We affirmed the use of such "better than BART" approaches in CEED, though we objected to the particular program under review there. See CEED, 398 F.3d at 660. We said nothing about how better-than-BART might be measured.

After our CEED decision, EPA introduced the following test to evaluate whether a BART-alternative achieves "greater reasonable progress" than BART:

If the distribution of emissions is not substantially different than under BART, and the alternative measure results in greater emission reductions, then the alternative measure may be deemed to achieve greater reasonable progress. If the distribution of emissions is significantly different, the State must conduct dispersion modeling. . . . The modeling would demonstrate "greater reasonable progress" if both of the following two criteria are met:

(i) Visibility does not decline in any Class I area, and

(ii) There is an overall improvement in visibility, determined by comparing the average differences between BART and the alternative over all affected Class I areas.

40 C.F.R. § 51.308(e)(3).

On March 10, 2005, EPA issued the Clean Air Interstate Rule ("CAIR"), requiring reductions in emissions of sulfur dioxide and nitrogen oxides in 28 eastern states and the District of Columbia. 70 Fed.Reg. 25,162 (May 12, 2005); see also 70 Fed.Reg. at 39,106/3. CAIR imposes specified emissions reduction requirements on each affected state, and enables states to meet the requirements by means of cap-and-trade programs. 70 Fed.Reg. at 39,106/3. In conjunction with the introduction of CAIR, EPA amended the Haze Rule to add a new regulation—contested here—providing that "[a] State that opts to participate in the Clean Air Interstate Rule cap-and trade . . . program . . . need not require affected BART-eligible EGUs [electric generating units] to install, operate, and maintain BART." 40 C.F.R. § 51.308(e)(4); [70 Fed.Reg. 39,156/3]; see also 70 Fed.Reg. 39,138-39 (noting that the CAIR-for-BART comparison is to be evaluated under the standards enumerated in § 51.308(e)(3)); 70 Fed.Reg. 39,142/3 (same).

In adopting the current version of § 51.308(e)(4), EPA provided analyses demonstrating that CAIR would achieve greater overall emission reductions than BART, and would make greater reasonable progress according to the two-pronged visibility test outlined in § 51.308(e)(3)—i.e., that CAIR would result in a greater aggregate visibility improvement (than BART) averaged over all Class I areas without reducing visibility at any individual area. See 70 Fed.Reg. at 39,136; see also Technical Support Document for the Final Clean Air Interstate Rule, March, 2005, at http://www.epa.gov/cair/pdfs/finaltech04.pdf ("CAIR TSD"). In doing so, however, EPA also noted that the "determination that CAIR makes greater reasonable progress than BART for EGUs is not a determination that CAIR satisfies all reasonable progress requirements in CAIR affected States . . . . [A state] cannot assume that CAIR will satisfy all of its visibility-related obligations." 70 Fed.Reg. at 39,143/3. In particular, despite the rule changes reflecting CAIR, the EPA retained a regulation specifying that states must establish reasonable progress goals "[f]or each mandatory Class I Federal area located within [a] State," and that such goals must "provide for an improvement in visibility for the most impaired days . . . and ensure no degradation in visibility for the least impaired days. . . ." 40 C.F.R. § 51.308(d)(1) (emphasis added).

On October 13, 2006, EPA once again promulgated revisions to the Haze Rule—revisions for some reason not called to our attention by any of the lawyers in this case. See Revisions to Provisions Governing Alternative to Source-Specific Best Available Retrofit Technology (BART) Determinations, 71 Fed.Reg. 60,612 (Oct. 13, 2006). Those revisions largely appear to respond to this court's decision in CEED. For example, the new rule both clarifies the process by which BART-alternatives are to be compared to BART and provides minimum elements for cap-and-trade programs adopted in lieu of BART. Id. at 60,612. But as the new rule does not become effective until December 12, 2006, and was not briefed or even mentioned by counsel, its specifics are not under consideration here. Our own perusal hasn't uncovered any changes undermining our conclusions; in fact, in at least one instance (discussed below), the new rule corresponds with concessions that EPA made at oral argument but not in its original briefs to this court.

* * * * * *

Industry petitioners argue that EPA acted contrary to statutory authority in two respects: first, by authorizing a state to infer, from evidence that its BART-eligible sources collectively contribute to visibility impairment in at least one Class I area, that all such sources may reasonably be anticipated to cause or contribute to visibility at such an area, without a source-by-source analysis—i.e., by authorizing the use of collective attribution—and, second, by issuing mandatory guidelines for the states' attribution determinations for power plants exceeding 750 megawatts ("MW"). We reject both arguments.

Industry petitioners claim that the collective attribution process allows states virtually to skip the Attribution Step; "once a State finds that a single BART-eligible source in the State affects visibility in a Class I area, other BART-eligible sources in the State may be swept into the BART Determination process without any analysis as to their effect on visibility." This is true, but because the substance of the impact issue remains open in Step II, it is of little consequence (with one exception, described below).

Industry petitioners' valid concern is that collective attribution will force sources to install BART even when such installations would serve no purpose whatsoever. But this fear is unwarranted. As EPA openly conceded at oral argument, if an individual source is found subject to BART in Step I because of collective attribution, that source can nonetheless challenge the necessity of installing BART in Step II—and have the impact issue resolved de novo. See Transcript of Oral Argument at 19-20. Recall that Step II involves the weighing of five factors, the last of which is the visibility impact of imposing BART. If that impact is zero because the source does not contribute to visibility impairment in the first place, then the source need not impose BART, regardless of the results dictated by the other four factors or the use of collective attribution in Step I. Counsel for EPA, commenting in oral argument on the passage in EPA's description of the BART determination process that industry found most alarming ("States, as a general matter, must require owners and operators of greater than 750 MW power plants to meet these BART emission limits," 70 Fed.Reg. at 39,131/3), repeatedly confirmed that a finding of zero impact at this stage would trump the four remaining factors and excuse the application of BART. See Transcript of Oral Argument at 17. (This interpretation is in part paralleled in the October 13 revisions to the Haze Rule: "Where a State takes this approach [i.e., collective attribution], the opportunity for assessing source-by-source visibility impact would still remain at the next step of setting the benchmark—the BART determination analysis." 71 Fed.Reg. at 60,615/2.) We adopt counsel's interpretation as our own understanding of the interplay between Steps I and II of the Haze Rule and between the impact criterion and the other factors.

That individual sources found subject-to-BART under collective attribution can nonetheless challenge the necessity of installing BART at the Determination Step does not render collective attribution a meaningless exercise. By setting a low threshold above which sources "may reasonably be anticipated to cause or contribute to any impairment," CAA § 169A(b)(2)(A), collective attribution essentially places on a source itself the burden of demonstrating that it doesn't contribute to visibility impairment. At oral argument, counsel for industry disclaimed any legal quarrel with EPA's assignment of the burden. See Transcript of Oral Argument at 12, 24. We find EPA's interpretation reasonable as against industry's challenges.

Industry petitioners' second argument is that EPA's guidelines for state attribution determinations for power plants exceeding 750 MW are mandatory for the states, contrary to industry's reading of the statute. But the industry briefs point to no such mandatory language. It is surely true that several elements of the Haze Rule purport to establish mandatory guidelines as to the Determination Step of the process. See, e.g., 70 Fed.Reg. at 39,131/3. Moreover, some passages identified in the briefs are ambiguous as to the force of EPA's provisions. See, e.g., 70 Fed.Reg. at 39,123 ("In the unlikely case that a State were to find that a 750 MW power plant's predicted contribution to visibility impairment is within a very narrow range between exemption from or being subject to BART, that the State can work with EPA. . . ."). 70 Fed.Reg. at 39,123/2. EPA's brief, though not focusing on this passage, argued that these attribution guidelines were "advisory, not mandatory, as to all sources, including 750 megawatt power plants." There appears to be no language in the rule which contradicts this claim; having defended the regulations on the basis that they are advisory here, EPA cannot later rely on a reading that they are mandatory. Of course states must still meet EPA's explicit—if rather general—requirement of a State Implementation Plan ("SIP") that can satisfy the statute's reasonable progress criterion, as construed by EPA. See 42 U.S.C. § 7491(b)(2) (requiring states to submit SIPs containing emission limits, schedules of compliance, and other measures necessary to make reasonable progress toward meeting the national visibility goal). But we do not understand industry to argue that such indirect compulsion violates the statute.

* * * * * *

The environmental petitioner argues that EPA's substitution of CAIR for BART contravenes the language and structure of the Clean Air Act because it cannot guarantee "reasonable progress" at all Class I areas. This argument is predicated on a belief that the Clean Air Act requires that BART-alternatives such as CAIR "do better" than BART at each individual Class I area (as opposed to simply in the aggregate), and, evidently, on every type of day (best days, worst days, etc.).

EPA's preliminary response is that environmental petitioner lacks standing because it has not been (and will not be) injured by the CAIR-for-BART rule. But EPA's own evidence shows the likelihood that states will adopt CAIR and that CAIR will be less effective at a few areas on some days. When CAIR was first introduced EPA noted that "[b]ased on our experience . . . we anticipate that States will choose to require EGUs to participate in the cap and trade programs administered by EPA." 69 Fed.Reg. 4,566, 4,586/3 (Jan. 30, 2004). EPA's technical analysis of CAIR then went on to predict that, at three (out of 156) Class I areas, such participation would result in visibility improvements on the best 20% of days that are less than the expected improvements at those areas under BART. See CAIR TSD 18. (The same analysis shows equal or superior visibility at all Class I areas on the 20% worst days. See id. at 23.)

Those findings, of course, do not in themselves show that petitioners' members will travel to the parks likely to be negatively impacted by the CAIR-for-BART Rule. Indeed, petitioner's affidavits do not clearly indicate that its members visit every park, let alone that they visit the three operative parks on the best 20% of days. But given the organization's large membership—over 320,000 members in all 50 states—we find it reasonable to infer that at least one member will suffer injury-in-fact. We do so with some hesitation, however. While some judicial opinions purport to reject reliance on mathematical likelihood, see, e.g., Sargent v. Mass. Accident Co., 307 Mass. 246, 29 N.E.2d 825, 827 (Mass.1940) ("It has been held not enough that mathematically the chances somewhat favor a proposition to be proved"), that viewpoint overlooks the reality that all empirical issues are matters of probability. But it is at least an imposition for a party to force courts to rely on statistical inference when the party presumably has better evidence within easy reach—here, a member's affidavit showing a high individualized probability of future visits to a particular park (presumably based on a pattern of past visits). Cf. United States v. Veysey, 334 F.3d 600, 605 (7th Cir.2003). Nevertheless, the size of petitioner's membership appears large enough here to indicate substantial probability of injury.

Having found that petitioner has standing, we nonetheless squarely reject its claim that the Clean Air Act requires EPA to ensure that any BART-alternative improves visibility at least as much as BART at every Class I area and in all categories of days. The plain language of the Act imposes no such mandate, and EPA's refusal to read one in is reasonable.

As we said in Corn Growers, "[t]he statutory goal enunciated in [CAA] § 169A(a)(1) is quite clear: `the prevention of any future, and the remedying of any existing, impairment of visibility.'" 291 F.3d at 10 (citing 42 U.S.C. § 7491(a)(1)). In order to meet this goal, the Clean Air Act specifically calls for regulations to assure that "reasonable progress" is made by the states. 42 U.S.C. § 7491(a)(4). Because "reasonable progress" is nowhere defined in the Act itself, we review EPA's interpretation of the term under the standard framework of Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984), and defer to the agency's interpretation so long as it is reasonable.

Recall that under the Haze Rule reasonable progress means that "[f]or each mandatory Class I Federal area . . . [states] must provide for an improvement in visibility for the most impaired days . . . and ensure no degradation in visibility for the least impaired days over the same period." 40 C.F.R. § 51.308(d)(1) (emphasis added). Moreover, unless there is some reasonable excuse, this progress must be sufficient to attain natural visibility conditions at every single Class I area by 2064. 40 C.F.R. § 51.308(d)(1)(ii). Indeed, EPA emphasized in its briefs that because "the regulatory scheme as a whole (and all the regulations promulgated pursuant to it) must be designed to achieve the goal [of reasonable progress] at every Class I area," EPA Br. at 66 (emphasis added), states must, if CAIR is substituted for BART and is not likely to achieve that goal, take "other measures as necessary to achieve reasonable progress goals including at each Class I area," id. at 67 (emphasis added). Thus, EPA not only agrees with petitioner that CAA § 169A(a)(1)'s declaration of a "national goal" that includes "the remedying of any existing [] impairment of visibility . . . [that] results from manmade air pollution" implies a need for ubiquitous improvement over time (emphasis added), but it has adopted regulations manifesting that goal.

Nonetheless, the Clean Air Act leaves wide discretion about how the goal is to be achieved. Notwithstanding the Act's discussion of BART in § 169A(b), we have already held in CEED that EPA may leave states free to implement BART-alternatives so long as those alternatives also ensure reasonable progress. 398 F.3d at 660. Moreover, nothing in § 169A(b)'s "reasonable progress" language requires at least as much improvement at each and every individual area as BART itself would achieve (much less improvement at each area at every instant); and EPA's requirement of some improvement at all areas on the worst days, coupled with no degradation at any area on the best days, 40 C.F.R. § 51.308(d)(1), appears a reasonable notion of reasonable progress. Finally, EPA allows use of a BART alternative only if it combines aggregate improvement (relative to BART) with universal, area-specific absence of degradation, 40 C.F.R. § 51.308(e)(3); on this metric CAIR-for-BART is far better than BART. 70 Fed. Reg. at 39,138-39,142.

Petitioner also appears to argue that the origin of CAIR in other clean air programs precludes EPA's decision to allow states the CAIR option in fulfillment of § 169A. But petitioner identifies no language requiring EPA to impose a separate technology mandate for sources whose emissions affect Class I areas, rather than piggy-backing on solutions devised under other statutory categories, where such solutions meet the statutory requirements.

Denied.