Full opinion text
PER CURIAM: The Congress enacted the Clean Air Act (the Act), 42 U.S.C. §§ 7401 et seq., “to protect and enhance the quality of the Nation’s air resources so as to promote the public health and welfare and the productive capacity of its population.” Id. § 7401(b)(1). At issue in this case is Title I of the Act, which requires the Environmental Protection Agency (EPA) to promulgate National Ambient Air Quality Standards (NAAQS), thus setting the maximum level of permissible pollutant concentration in the atmosphere. See id. §§ 7408(a)(1), 7409(a)-(b). After the EPA sets the NAAQS, it must determine whether each state is in compliance with these air-quality standards and, in the event of a NAAQS violation, how to establish the geographic boundaries around the non-compliant area. See id. § 7407(d)(1). In these consolidated petitions, several states, counties, industrial entities and environmental organizations challénge the EPA’s determination that certain geographic areas, are, or are not, in “attainment” with the EPA’s ground-level ozone NAAQS. Id. Some argue that the Act, as applied to them, violates various Constitutional provisions; others argue that the EPA misconstrued the terms of the Act. Virtually every petitioner argues that, for one reason or another, the EPA acted arbitrarily and capriciously in making its final NAAQS designations. But because the EPA complied with the Constitution, reasonably interpreted the Act’s critical terms and wholly satisfied — indeed, in most instances, surpassed — its obligation to engage in reasoned decision-making, we deny the consolidated petitions for review in their entirety. I. BACKGROUND The EPA began the odyssey resulting in these consolidated petitions nearly seven years ago. Along the way, it construed a variety of the Act’s provisions, promulgated regulations and issued informal guidance to assist in the collaborative area-designation effort between it and the states. Before discussing the substance of the issues, a brief overview of the Act and the underlying proceedings in this case is in order. A. The Clean Air Act Under the Act, the EPA must promulgate NAAQS, which set the maximum ambient, or outdoor, air concentrations for six pollutants that “may reasonably be anticipated to endanger public health or welfare.” 42 U.S.C. § 7408(a)(1). Once it establishes a NAAQS, the EPA must designate each “area” in the United States as “attainment” or “nonattainment.” See id. § 7407(d)(l)(A)(i)-(ii). Alternatively, the EPA may designate an area as “unclassifiable” if the area “permit[s] no determination given existing data.” Catawba Cnty., N.C. v. EPA, 571 F.3d 20, 26 (D.C.Cir.2009) (citing 42 U.S.C. § 7407(d)(l)(A)(i)-(iii)). The EPA treats an “unclassifiable” area as if it were in attainment. See 42 U.S.C. § 7471. Generally speaking, the EPA designates an area that meets the relevant NAAQS as in attainment, while areas that exceed the NAAQS receive a nonattainment designation. See Catawba Cnty., 571 F.3d at 26. But even if an area’s ambient air concentration complies with the relevant NAAQS, the EPA nonetheless designates it as non-attainment if it “contributes” to a NAAQS violation in a “nearby area.” See 42 U.S.C. § 7407(d)(1)(A)©. The Act does not define the terms “contributes,” “nearby” or “area.” The EPA works collaboratively with the states to determine the NÁAQS-attainment status for all areas within a respective state’s borders. No later than one year after the EPA promulgates a new or revised NAAQS, each state must submit recommended “initial designations” to the EPA. Id. § 7407(d)(1)(A). A state’s initial designations must suggest both the appropriate geographic boundaries for each “area” and whether the EPA should classify the suggested area as attainment, non-attainment or unclassifiable. See id. § 7407(d)(l)(A)-(B). Once it receives a state’s initial designations, the EPA may either promulgate them as submitted or modify them as it “deems necessary.” Id. § 7407(d)(l)(B)(ii). The Act gives the EPA discretion to change a state’s recommended designation, to alter a state’s proposed geographic area or both. See id. Although the EPA “has no obligation to give any quantum of deference to a designation that it ‘deems necessary’ to change,” Catawba Cnty., 571 F.3d at 40, it must nonetheless notify the state of any intended change and provide the state with at least 120 days “to demonstrate why any proposed modification is inappropriate,” 42 U.S.C. § 7407(d)(l)(B)(ii). These notifications are known as “120-day letters.” See Air Quality Designations for the 2008 Ozone National Ambient Air Quality Standards, 77 Fed.Reg. 30,088, 30,090 (May 21, 2012) [hereinafter 2008 Designations Rule]. While the EPA has ultimate authority to determine each area’s attainment status, each state has “primary responsibility” for ensuring that the geographic areas within its borders either maintain attainment or progress towards it. 42 U.S.C. § 7407(a). Accordingly, once the EPA finalizes its designations, each state must submit to the EPA a State Implementation Plan (SIP) specifying how the NAAQS “will be achieved and maintained.” Id. For areas in attainment, the SIP must simply “contain emission limitations and such other measures as may be necessary ... to prevent significant deterioration of air quality.” Id. § 7471. For a nonattainment area, however, the Act imposes more stringent requirements. A SIP from a state with a nonattainment area must demonstrate that the state intends to implement “all reasonably available control measures” and “reasonably available control technology” to bring the area into attainment. Id. § 7502(c)(1). The Act also imposes deadlines, or “attainment dates,” on an offending area. See id. § 7502(a)(2)(A). For a violation of a primary NAAQS, the offending state must reach attainment “as expeditiously as practicable, but no later than 5 years from the date such area was designated nonattainment.” Id. The EPA “may extend the attainment date to the extent [it] determines appropriate” but only “for a period no greater than 10 years from the date of designation as nonattainment.” Id. Taken together, these two requirements often mean that a state with a nonattainment area must implement potentially expensive technology or expensive process changes to reduce pollution levels over a relatively short period of time. If a state fails to reach attainment timely and the failure is due to inadequate implementation efforts, sanctions can be imposed, including loss of federal highway funds and increasingly severe restrictions on emissions sources within the state. See id. § 7509(a)-(b). B. The 2008 Ozone NAAQS and the EPA’S 2008 Guidance On March 12, 2008, the EPA promulgated new primary and secondary NAAQS for ambient ozone, a component of urban smog. See 2008 Designations Rule, 77 FecLReg. at 30,089. Even though ozone is an “essential presence in the atmosphere’s stratospheric layer,” it becomes harmful at ground level and “can cause lung dysfunction, coughing, wheezing, shortness of breath, nausea, respiratory infection, and in some cases, permanent scarring of the lung tissue.” S. Coast Air Quality Mgmt. Dist. v. EPA, 472 F.3d 882, 887 (D.C.Cir.2006) (quoting Henry A. Waxman, An Overview of the Clean Air Act Amendments of 1990, 21 Envtl. L. 1721, 1758 (1991)). It also “has a broad array of effects on trees, vegetation, and crops and can indirectly affect other ecosystem components such as soil, water, and wildlife.” Mississippi v. EPA, 744 F.3d 1334, 1340 (D.C.Cir.2013). Because ozone forms at ground level when “ozone precursors”— specifically, nitrous oxides (NOx) and volatile organic compounds (VOCs) — react with sunlight, NAAQS compliance largely depends on reducing emissions from ozone-precursor producers like power plants, industrial compounds, motor vehicles and combustion engines. See 2008 Designations Rule, 77 Fed.Reg. at 30,089. Complicating this task is-that ozone and ozone precursors travel easily through the atmosphere, which can result in NAAQS violations hundreds of miles away from the source of the ozone precursors. See id. Both the EPA’s 2008 primary and secondary ozone NAAQS reduced the maximum allowable daily average eight-hour level of ozone from 0.08 parts per million (ppm) to 0.075 ppm. See National Ambient Air Quality Standards for Ozone, 73 Fed.Reg. 16,436,16,436-37 (Mar. 27, 2008). By setting these new NAAQS, the EPA triggered the states’ responsibility to submit their initial designations. See 42 U.S.C. § 7407(d)(1)(A). To assist this process, the EPA issued a guidance titled “Area Designations for the 2008 Revised Ozone National Ambient Air Quality Standards” [hereinafter 2008 Guidance] on December 4, 2008, which included several matters relevant to the instant petitions. First, the 2008 Guidance instructed states on the quality of data it expected them to consider. Specifically, it recommended that the states “identify violating areas using the most recent three consecutive years of quality-assured, certified air quality data.” 2008 Guidance at 2. The 2008 Guidance also informed the states that “[i]n general, [NAAQS] violations [will be] identified using data from ... monitors that are sited and operated in accordance with [EPA regulations located at] 40 C.F.R. Part 58.” Id. Second, the 2008 Guidance provided instruction for establishing geographic boundaries around nonattainment areas, noting first that the “EPA believes it is important to examine ozone-contributing emissions - across a relatively broad geographic area.” 2008 Guidance at 3. Accordingly, the 2008 Guidance recommended that if an air-quality monitor reports a NAAQS violation, the state should consider using the Core Based Statistical Area (CBSA) or Combined Statistical Area (CSA) in which the monitor is located as the “presumptive” boundary. Id. If the violating monitor is not in a CSA or CBSA, the 2008 Guidance recommended using the county-in which the violating monitor is located as the presumptive boundary. Id. The 2008 Guidance made plain, however, that CSAs, CBSAs and county lines were merely presumptive boundaries, recognizing that “area-specific analyses ... may support nonattainment area boundaries that are larger or smaller than the presumptive area starting point.” Id. Stressing that “each potential nonattainment area should be evaluated on a case-by-case basis,” the 2008 Guidance instructed the states to consider nine factors when determining a nonattainment area’s borders. See id. at 2, Attach. 2. These include (1) air-quality data; (2) emissions data (such as location of emissions sources and contribution to ozone concentrations); (3) population density and degree of urbanization (including commercial development); (4) traffic and commuting patterns; (5) population growth rates and patterns; (6) meteorology (such as weather and air-transport patterns); (7) geography and topography (such as mountain ranges or other air-basin boundaries that could affect ozone dispersion); (8) jurisdictional boundaries (such as counties, air districts, existing nonattainment area boundaries and regional planning authority boundaries) and (9) the level of control of emissions sources. See id. Attach. 2. The 2008 Guidance stated that the EPA planned to consider these same factors, “along with any other relevant information,” in determining whether to modify the states’ initial designations. Id. C. The 2008 Ozone Designation Process By 2009, all states had submitted their initial designations to the EPA. Rather than immediately reviewing the initial designations, however, the EPA halted the designation process to consider whether to lower the ozone NAAQS even further. This delay prompted a lawsuit by Wild-Earth Guardians — an environmental-group petitioner in this case — that sought to compel the EPA to complete the stalled ozone NAAQS designation process. The EPA and WildEarth Guardians eventually entered into a consent decree that required the EPA to finalize its designations no later than May 31, 2012. See 2008 Designations Rule, 77 Fed.Reg. at 30,091. The EPA notified the states in September 2011 that it intended to finalize the ozone NAAQS designations by the May 31, 2012 deadline set forth in the consent decree. In accordance with the 2008 Guidance’s instruction to “identify violating areas using the most recent three consecutive years of quality-assured, certified air quality data,” 2008 Guidance at 2, virtually every state had already submitted air-quality data from 2008 to 2010 by the time the EPA resumed the designation process. Although the EPA assured the states that it still planned to consider the recommended designations and ozone data they had submitted initially, it recognized that some states may have collected more recent air-quality data for their regions. For this reason, the EPA allowed the states to provide updated recommendations and analyses — so long as any updated air-quality data was certified for quality — but assured them that they were under no obligation to do so. In response to this invitation, several states updated their initial designations and some submitted air-quality data from 2009 to 2011 to replace their older 2008 to 2010 data. The states seeking to use data from 2009 to 2011 agreed to certify their data for quality by February 29, 2012, so that the EPA had sufficient time to consider the more recent data in advance of its May 31, 2012 deadline to finalize the designations. The EPA then reviewed each state’s initial designations to determine whether to modify them. It first examined the air-quality submissions from the states to determine which monitors reported ozone NAAQS violations. If a state certified its air-quality data from 2011 by the February 29, 2012 deadline, the EPA generally considered its air-quality data from the years 2009 to 2011. For all other states, the EPA considered air-quality data from 2008 to 2010. After identifying NAAQS-violating monitors, the EPA decided whether to alter the states’ respective recommended nonat-tainment boundaries. To do so, the EPA used a multi-factor, weight-of-the-evidence test that tracked — but was not identical to — the nine-factor test in the 2008 Guidance. Specifically, the EPA collapsed the 2008 Guidance’s nine-factor test into a five-factor test, which examined (1) “Air Quality Data,” or whether an area’s monitor reported a NAAQS violation; (2) “Emissions Data,” including emissions levels and controls, population, population density, population growth, degree of urbanization and traffic and commuting patterns; (3) “Meteorology,” including wind speed and direction; (4) “Geography/Topography,” which examined the effect of physical land features on the distribution of ozone and (5) “Jurisdictional Boundaries,” which helped determine whether certain areas could effectively carry out air-quality planning and enforcement functions for nonat-tainment areas. Once attainment designations were made, the EPA notified the states of any proposed modifications it deemed necessary and invited them to submit any additional data or comments they wished to have the EPA consider. Although not required by statute, see 42 U.S.C. § 7407(d)(2)(B), the EPA also opened a 30-day public comment period on the proposed notifications. Several states, organizations and members of the public — including many of the petitioners in this case— submitted comments. The EPA considered the comments and then promulgated its final designations, which identified 48 nonattainment areas in 26 states, the District of Columbia and Indian country. The nonattainment areas included 192 counties in toto and 36 counties in part. The EPA published the majority of its final designations on May 21, 2012, see 2008 Designations Rule, 77 Fed.Reg. at 30,088, and in the case of certain Chicago-area designations, on June 11, 2012, see Air Quality Designations for the 2008 Ozone National Ambient Air Quality Standards for Several Counties in Illinois, Indiana, and Wisconsin; Corrections to Inadvertent Errors in Prior Designations, 77 Fed.Reg. 34,221, 34,221 (June 11, 2012). After the EPA received and denied 29 petitions for reconsideration, the parties in this consolidated case petitioned this Court for review. We have jurisdiction under 42 U.S.C. § 7607(b)(1). II. COMMON LEGAL PRINCIPLES Before addressing the petitioners’ individual challenges, we think it helpful to discuss several principles that bear on most, if not all, of the issues the petitioners have raised. First, we review the EPA’s NAAQS designations under the same standard we use in reviewing a challenge brought under the Administrative Procedure Act (APA). See Allied Local & Reg’l Mfrs. Caucus v. EPA 215 F.3d 61, 68 (D.C.Cir.2000). Accordingly, we will set aside a NAAQS designation by the EPA only if it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Catawba Cnty., 571 F.3d at 41 (quoting 5 U.S.C. § 706(2)(A)). We must, however, give an “extreme degree of deference” to the EPA’s evaluation of “scientific data within its technical expertise,” City of Waukesha v. EPA, 320 F.3d 228, 247 (D.C.Cir.2003), especially where, as here, we review the “EPA’s administration of the complicated provisions of the Clean Air Act.” Catawba Cnty., 571 F.3d at 41 (citing Nat’l Ass’n of Clean Air Agencies v. EPA 489 F.3d 1221, 1229 (D.C.Cir.2007)). Because the EPA’s “basic obligation” is to conduct “reasoned decisionmaking,” id. at 25, we will uphold its action if the record shows that the EPA “considered all relevant factors and articulated a ‘rational connection between the facts found and the choice made,’ ” id. at 41 (quoting Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962)). Second, we have long since rejected the argument that the EPA violates the Act if it uses a holistic, multi-factor, weight-of-the-evidence test for determining whether a given area contributes to a NAAQS violation. See ATK Launch Sys., Inc. v. EPA, 669 F.3d 330, 336-37 (D.C.Cir.2012) (challenge to 2006 fine particulate matter NAAQS designations); Catawba Cnty., 571 F.3d at 46 (challenge to 1997 fine particulate matter NAAQS designations). Indeed, in Catawba County, we made explicit that the EPA does not violate the Act even if it fails to adopt “a bright-line, ‘objective’ test” for determining contribution and we also held that the “EPA’s failure to quantify its analysis” does not render “its interpretation of ‘contribute’ arbitrary and capricious and therefore unreasonable.” 571 F.3d at 39. Rather, because “[a]n agency is free to adopt a totality-of-the-circumstances test to implement a statute that confers broad discretionary authority, even if that test lacks a definite ‘threshold’ or ‘clear line of demarcation to define an open-ended term,’” we have held that, “[t]o be reasonable, such an ‘all-things-considered standard’ must simply define and explain the criteria the agency is applying.” Id. With this background in mind, we now turn to the petitioners’ challenges. III. THE PETITIONERS’ CHALLENGES A. Delaware & Connecticut We begin with a challenge to the EPA’s construction of the key statutory provision in this case. Petitioners Delaware and Connecticut challenge the EPA’s refusal to designate broad, multi-state non-attainment areas to address the issue of long-range ozone transport. According to the States, the EPA’s final designations are inconsistent with its statutory mandate to designate areas as nonattainment if they “contributef ] to ambient air quality in a nearby area that does not meet [the NAAQS].” 42 U.S.C. § 7407(d) (emphasis added). We conclude, to the contrary, that the designations are consistent with the EPA’s reasonable interpretation of the ambiguous statutory term “nearby.” After the EPA reopened the designation process in 2011, Delaware proposed a non-attainment area that would stretch across 16 upwind states and the District of Columbia — to states as far west as Missouri. Connecticut similarly proposed an 18-state nonattainment area, also stretching west to Missouri. Both States argued for what Delaware described as a “more workable definition of ‘nearby ” — one that would ask “whether a source is ‘near enough to contribute’ to nonattainment or interfere with maintenance.” Letter from Del. Dep’t of Natural Res. & Envtl. Control to EPA 5 (Oct. 28, 2011) [hereinafter Delaware Response]. The EPA, however, had taken a different approach in the 2008 Guidance, instead interpreting “nearby” as presumptively including counties in the same metropolitan area as the violating county. 2008 Guidance at 3. In the Guidance, the EPA acknowledged that certain regions have ozone transport problems, but it concluded that the Act “does not require that all contributing areas be designated nonat-tainment, only the nearby areas.” Id. at 4. The agency explained that “[regional strategies, such as those employed in the Ozone Transport Region and EPA’s NOx SIP Call are needed to address the long-range transport component of ozone nonat-tainment.” Id. In keeping with this understanding of the statute, the EPA declined to designate “super-regional” nonattainment areas, see Responses to Significant Comments on the State and Tribal Designation Recommendations for the 2008 Ozone NAAQS at 8-9 (Apr. 30, 2012) [hereinafter Response to Comments], and instead made more limited nonattainment designations in both Delaware and Connecticut, see Delaware Area Designations for the 2008 Ozone NAAQS 2; Connecticut Area Designations for the 2008 Ozone NAAQS l. We evaluate the EPA’s interpretation of a Clean Air Act provision under the familiar two-step Chevron framework. See Util. Air Regulatory Grp. v. EPA, — U.S. —, 134 S.Ct. 2427, 2439, 189 L.Ed.2d 372 (2014) (citing Chevron, U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984)). The first question— “whether Congress has directly spoken to the precise question at issue,” Chevron, 467 U.S. at 842, 104 S.Ct. 2778 — has previously been resolved by this Court. In Pennsylvania Department of Environmental Protection v. EPA (PADEP), we held that the statutory term “nearby” in section 107(d) is ambiguous; indeed, we reached that conclusion in the course of addressing the precise argument that Delaware makes here. See 429 F.3d 1125, 1129-30 (D.C.Cir.2005). In Catawba County, we reached the same conclusion. See 571 F.3d at 35 (noting that section 107(d) does not define “nearby,” and that it is “the kind[ ] of word[ ] that suggests] a congressional intent to leave unanswered questions to an agency’s discretion and expertise”). Recognizing these precedents, Delaware and Connecticut conceded at oral argument that our analysis must be governed by Chevron’s second step, Oral Arg. Recording at 3:49-3:54, which requires us to ask only whether the EPA’s interpretation is reasonable, see, e.g., PADEP, 429 F.3d at 1130; But we have addressed that question once as well, also in PADEP, where we said that “Chevron requires that we defer to the agency’s reasonable interpretation of the term, and Delaware has given us no reason to think that EPA’s interpretation is unreasonable.” Id. We reach the same conclusion here. First, the agency’s interpretation of “nearby” — as presumptively including counties within the same metropolitan area as the violating county — falls readily within the dictionary definition of “nearby” as “close at hand; not far off; adjacent; neighboring.” RANDOM House College DICTIONARY 889 (rev. ed.1980). By contrast, neither the dictionary nor common parlance would regard Missouri as “nearby” to Connecticut or Delaware, as the petitioners’ proposals would require. Second, the EPA’s construction is consistent with the approach the agency has taken in prior designations proceedings— an approach that this Court has previously upheld as reasonable. See PADEP, 429 F.3d at 1127, 1129-30; 2008 Guidance at 3. Third, the EPA’s construction is consistent with the statutory scheme. The EPA selected the metropolitan area as the presumptive “nearby” area for its contribution analysis in part because the Congress itself chose the metropolitan area as the default boundary for ozone nonattainment areas classified as “serious,” “severe,” or “extreme.” ' See 42 U.S.C. § 7407(d)(4)(A)(iv); 2008 Guidance at 3 n. 5. The Congress’ choice is certainly evidence that the legislature envisioned broad but relatively local nonattainment areas. As in PADEP, the petitioners argue that the EPA’s interpretation.is unreasonable because it fails to appreciate the role of ozone transport, and consequently yields designations that fail to include the true contributors to their nonattainment status. See PADEP, 429 F.3d at 1129-30. Delaware notes, for example, that 84 to 94 per cent of its ozone results from the contributions of other states, including states as far west as Missouri. See Delaware Reply Br. 4. Without emissions reductions from those states, petitioners argue,- they cannot meet the 0.075 ppm standard. Thus, by failing to address the principal sources of their ozone pollution, the EPA’s interpretation eliminates any possibility that they will attain the NAAQS. Athough we are sympathetic to the petitioners’ concerns, our role is not to decide whether their proposed interpretation is reasonable. Instead, the sole question before us is whether the EPA interpreted the term reasonably and consistently with the statute. See PADEP, 429 F.3d at 1130 (noting that, although a broader “construction of ‘nearby’ may well be sensible, Chevron requires that we defer to the agency’s reasonable interpretation of the term”). Here, the EPA had already considered the problem thé petitioners raised. Part of the rationale for using the metropolitan area as the starting point for the contribution analysis was to account for ozone transported from outside the violating county. See -2008 Guidance at 3--A. A-though this approach does not fully account for longer-range, interstate transport, the EPA has addressed that problem in regulations promulgated under other provisions of the Act. See, e.g., Federal Implementation Plans: Interstate Transport of Fine Particulate Matter and Ozone and Correction of SIP Approvals, 76 Fed. Reg. 48,208 (Aug. 8, 2011) (promulgating the Cross State Air Pollution Rule, commonly referred to as the Transport Rule). Although the petitioners recognize the EPA’s reliance on those other regulatory options, they maintain that they “have been less than successful” up to this point. Delaware Br. 6; see also id. at 9. We, however, must defer to the EPA’s reasonable judgment that regional strategies adopted pursuant to other statutory provisions specific to long-range ozone transport remain the appropriate means for addressing this problem. See 2008 Guidance at 4. The petitioners note that our decision in PADEP rested in part upon the fact that there, Delaware had “offered no evidence that ‘in practice’ EPA will not enlarge a nonattainment area in response to [its then] eleven-factor analysis.” 429 F.3d at 1130. Indeed, in PADEP, Delaware had failed altogether “to produce an eleven-factor analysis.” Id. But we did not mean by this to suggest that, had Delaware produced the appropriate factor analysis, the EPA would have been required to adopt an interpretation of “nearby” that included states as far away as those within the-petitioners’ proposed nonattainment areas. The points discussed above — including the dictionary definition of “nearby” and the consistency of the EPA’s interpretation with the statute and its prior practice— strongly suggest that the EPA’s narrower interpretation would still be reasonable. Nonetheless, if the petitioners had submitted a persuasive five-factor analysis establishing contributions from farther-away states, that would be relevant to our assessment of the reasonableness of the EPA’s refusal to enlarge the nonat-tainment area beyond its presumptive scope. In this case, however, although the petitioning States did submit technical analyses, they failed to demonstrate the requisite linkages under the EPA’s 2008 Guidance. See, e.g., Delaware Response Attach. 2 at 5-7, 11-13 (disputing relevance of factors related to urbanization, traffic, and economic growth); id. at 14-15 (with respect to meteorology factor, describing long-range transport without describing weather patterns within the proposed 16-state nonattainment area). Hence, the petitioners did not show that the agency “will not enlarge a nonattainment area in response to” the (current) five-factor analysis, PADEP, 429 F.3d at 1130. Rather, the States’ analyses were simply insufficient to overcome the agency’s definitional presumption. In sum, we conclude that the EPA’s final designations of Delaware and Connecticut counties are consistent with a reasonable interpretation of the Clean Air Act. B. Uinta Basin Petitioner WildEarth Guardians (Wild-Earth) challenges the EPA’s designation of Uinta Basin, Utah, as “unclassifiable.” We find the EPA’s designation rational and in accordance with the Clean Air Act, and we therefore deny WildEarth’s petition. 1. Uinta Basin Background The EPA requires every state to establish a network of regulatory monitoring stations to collect ozone air-quality data. See 40 C.F.R. pt. 58. The number of regulatory monitors required in an area depends, in part, on the area’s population. See id. app. D. tbl.D-2. Areas with populations below 50,000 and many areas with fewer than 350,000 inhabitants require no regulatory monitors. Id. Many rural areas therefore lack monitors. Uinta Basin, Utah, had no regulatory monitoring until April 2011. The pre-2011 absence of regulatory-air-quality monitors in Uinta Basin meant that, when the EPA in 2013 conducted the designation process for the 2008 NAAQS, the agency had regulatory data for Uinta Basin for only two years — 2011 and 2012. The 2008 ozone NAAQS, however, reflect three-year averages of ozone levels. See 2008 Designations Rule, 77 Fed.Reg. at 30,089. Noting that “there are not yet three consecutive years of certified ozone monitoring data available [from Uinta Basin] that can be used to determine the area’s attainment status,” id., the EPA designated the area as “unclassifiable,” which the Clean Air Act defines as an area that “cannot be classified on the basis of available information as meeting or not meeting” the NAAQS, 42 U.S.C. § 7407(d)(l)(A)(iii). Although no regulatory data exist for Uinta Basin prior to 2011, private companies working under consent decrees have been required to operate ozone air-quality monitors in Uinta Basin since 2009. See Letter from Robin Cooley, Counsel, Wild-Earth Guardians to Lisa P. Jackson, Adm’r, EPA 3 (July 19, 2012). Under the terms of those consent decrees, the private monitors must comply with many of the same requirements as regulatory monitors. See Consent Decree ¶¶ 80-81, United States v. Kerr-McGee Corp., No. 1:07-cv-01034 (D. Colo. May 17, 2007). From 2009 to 2011, the private monitors provided raw data showing ozone levels significantly exceeding the 2008 ozone NAAQS. The EPA found the 2009 to 2011 private data insufficient to support a nonattainment designation. 2. The Private Monitoring Data Challenge WildEarth argues that, inflight of the private data, the EPA contravened the Act’s requirements when it designated Uinta Basin as unclassifiable rather than nonattainment. We disagree. The Act calls for the EPA to make designations “on the basis of available information.” 42 U.S.C. § 7407(d)(l)(iii). We have repeatedly found similar language to be ambiguous when assessing whether to defer to an agency’s construction. See Catawba Cnty., 571 F.3d at 35, 38 (finding the phrase “based on air quality monitoring data” to be ambiguous); Sierra Club v. EPA, 356 F.3d 296, 305-06 (D.C.Cir.2004) (finding the phrase “based on photochemical grid modeling” to be ambiguous). The EPA therefore may interpret the statutory language as it sees fit, as long as its interpretation is reasonable. Chevron, 467 U.S. at 845, 104 S.Ct. 2778. And even assuming the Act obligates the EPA to consider certain types of data, there would be no obligation for the agency to base its designations on data it reasonably considers to be unsound, at least if it “adequately explain[s] its reasons for rejecting ... data” on which it declines to rely. City of Waukesha, 320 F.3d at 248. We evaluate the EPA’s reasons cognizant of the “extreme degree of deference” we owe an agency “when it is evaluating scientific data within its technical expertise.” Catawba Cnty., 571 F.3d at 41. The EPA reasonably explained that the private monitoring data afforded an insufficient basis for a nonattainment designation because the agency was unable to perform post-collection quality assurance checks on the data. In particular, the EPA lacked quality assurance data needed to verify and audit the private data. As the agency explained: Quality assurance data consist, primarily, of biweekly single point quality control (QC) checks, used to assess the precision and bias a given instrument is displaying in its day-to-day measurements, and annual independent performance evaluations (audits) of equipment, which rely on independent staff and measuring systems to confirm that the monitors are operating as expected and required. Letter from Lisa P. Jackson, Adm’r, EPA to Robin Cooley, Counsel, WildEarth Guardians 5 (Dec. 14, 2012) (denying reconsideration of Uinta Basin designation). The agency determined that, without audits or quality control checks, it could not adequately verify the quality of the private data. That explanation comports with common sense and falls within the substantial deference accorded the EPA in evaluating the soundness of data available to it. WildEarth presses several counterarguments, none of which we find persuasive. First, WildEarth observes that the consent decrees required the private monitors to operate in “substantial compliance” with 40 C.F.R. Part 58, the quality assurance requirements under which regulatory monitors operate. But “substantial compliance” is not “full compliance,” and the EPA could reasonably draw a distinction between the two. Moreover, data from regulatory monitors — which must be collected in compliance with 40 C.F.R. Part 58 — undergo post-collection auditing and verification processes. See, e.g., .40 C.F.R. pt. 58, app. A, § 3. Those post-collection processes could not be conducted for the private monitor data. Accepting Wild-Earth’s argument would require us to conclude that the EPA must apply less stringent post-collection validation requirements to data collected from private monitors in “substantial compliance” with the agency’s data-collection regulations than the agency applies to data collected from régulatory monitors in actual compliance with those regulations. We see no reason to embrace that counterintuitive result. Second, WildEarth points out that the EPA has encouraged other federal entities to take notice of the private monitoring data. The EPA acknowledges that it argued, in a judicial proceeding supporting entry of the same consent decrees mandating the private monitoring, that the private monitors would provide data that would be “reliable and of good quality” and “useful in assisting regulators.” Resp’t’s Br. 57. And indeed the data have proven helpful to the EPA in other regulatory contexts. On the basis of the private data, for example, the EPA informed the Forest Service that Uinta Basin ozone concentrations “exceed the NAAQS” and are a “serious problem.” Supp. JA 387. We agree with WildEarth that an agency may be required to articulate why data are sufficiently reliable for one purpose but not for another. See Cnty. of L.A. v. Shalala, 192 F.3d 1005, 1022 (D.C.Cir.1999). But the EPA has done so here. That the data may be sufficiently reliable to warrant identifying ozone as a serious issue for a Forest Service analysis under one statutory provision does not necessarily mean that the data are reliable enough to compel a nonattainment designation under a different statutory regime. To hold otherwise would require the EPA wholly to blind itself to potentially useful private data for any purpose if it were to consider that data insufficiently reliable for one purpose. There is no basis for constraining the agency in that way. That the EPA partially relied on the private data in the course of this very designation process does not undercut that conclusion. While “unclassifiable” represents a single statutory designation, see 42 U.S.C. § 7407(d)(l)(A)(i)-(iii), the EPA further divided that classification into two sub-categories: “unclassifiable/attainment” and “unclassifiable.” See 2008 Designations Rule, 77 Fed.Reg. at 30,089. “Historically for ozone,” the EPA designates as “ ‘unclassifiable/attainment’ ” those areas for which “air quality information is not available because the areas are not monitored.” Id. at 30,090. But in Uinta Basin, the EPA instead designated the area “unclassifiable” after determining that the private monitoring “detected levels of ozone that exceed the NAAQS.” Id. at 30,089. There is no arbitrariness in the EPA’s choice partially — but not fully — to rely on the private data. At the outset, we note that the parties point us to no material differences between an “unclassifiable/attainment” and an “unclassifiable” designation, and we are aware of none. See 40 C.F.R. § 51.1100(g) (“Attainment area means, unless otherwise indicated, an area designated as either attainment, unclassifiable, or attainment/unclassifiable.”); cf. 42 U.S.C. § 7471 (instructing the EPA to give the same treatment to “unclassifiable” and “attainment” areas for SIP purposes). But give'n the EPA’s decision to create two different unclassifiable designations, we will assume arguendo that materially different regulatory burdens attend each designation. Even then, however, we agree with the EPA that it was reasonable to conclude that it would be inappropriate to label the Uinta Basin area “unclassifiable/attainment”: the private data, even if unverified, at least implied that a NAAQS violation was possible, even if not conclusively proven to the agency’s satisfaction. WildEarth, moreover, points to no other area for which private — but not regulatory — monitoring suggested a NAAQS violation. It thus appears that Uinta Basin differed from all other areas meriting an “unclassifiable/attainment” designation. We conclude that the EPA’s conclusion partially — but not fully — to credit the private data was reasonable and non-arbitrary, particularly in light of the “extreme deference” we owe the agency. See Catawba Cnty., 571 F.3d at 41. In sum, the EPA reasonably declined to rely on data that it considered of insufficient quality for designations purposes. With that conclusion, and having reviewed the remainder of WildEarth’s challenges and determined that they lack merit, we deny the group’s petition for review. See Catawba Cnty., 571 F.3d at 52. C. Sierea Club Petitioner Sierra Club challenges the EPA’s refusal to use uncertified 2011 air-quality data during the designation process, a decision that resulted in 15 counties avoiding nonattainment designations. Finding the EPA’s actions rational and in accordance with the Clean Air Act, we deny Sierra Club’s petition. 1. Sierra Club Background In furtherance of the Clean Air Act’s “ ‘core principle’ of cooperative federalism,” EPA v. EME Homer City Generation, L.P., — U.S. —, 134 S.Ct. 1584, 1602 n. 14, 188 L.Ed.2d 775 (2014), states take the lead in the collection of air-quality data. In doing so, states operate regulatory monitors under an array of “[exhaustive technical specifications” promulgated by the EPA. Catawba Cnty., 571 F.3d at 30; see 40 C.F.R. pt. 58. States “edit[]” and “validate[ ]” the collected data pursuant to the EPA-mandated procedures and report it to the EPA according to a prescribed schedule. See 40 C.F.R. .§ 58.16(b)-(c). Data collected in each quarter must be “edited, validated and entered” into the EPA’s system within ninety days of the end of the quarter. Id. “For example, the data for the reporting period January 1 — March 31 are due on or before June 30 of that year.” Id. § 58.16(b). Post-auditing, the data are still considered “uncertified” when submitted to the EPA. While uncertified data from the first quarter (ie., January 1 to March 31) become available to the EPA as of June 30, those data remain subject to continuing audits and edits by states. The data collection process reaches completion only when a state provides final certification that the necessary “ambient concentration and quality assurance data are completely submitted ... and ... are accurate.” Id. § 58.15(a). The EPA requires certification by May 1 of the following calendar year for all data collected in the previous year. Id. § 58.15(a)(2). States therefore had to certify their 2011 data by May 1, 2012. As explained, because the 2008 ozone NAAQS represent a three-year average, the EPA needs air-quality data from three sequential calendar years to classify an area as attainment or nonattainment (as opposed to unclassifiable). See 2008 Designations Rule, 77 Fed.Reg. at 30,089. In the designation process for the 2008 NAAQS, the EPA gave each state a choice between two options: (i) early-eertify 2011 data by February 29, 2012, in which event the EPA would consider 2009 to 2011 data for the designation process for that state (Option One); or (ii) decline to early-certify (and stick to the normal May 1 certification deadline), in which event the EPA would use 2008 to 2010 data for designations in that state (Option Two). See id. at 30,091. At least eight states selected Option Two. Sierra Club identifies over one dozen counties within those eight states for which the choice between Option One and Option Two ii.e., the choice between designations based on 2008 to 2010 data versus 2009 to 2011 data) allegedly meant that those counties avoided nonattainment designations. See Letter from Robert Ukei-ley, Counsel, Sierra Club to EPA, Re: Designations for the 2008 Ozone NAAQS Docket ID No. EPA-HQ-OAR-2008-0476 at 3 tbl.l (Feb. 3, 2012). Sierra Club contends that the EPA was compelled to use 2009 to 20Í1 data for those areas. We disagree and conclude that the EPA’s actions were non-arbitrary. 2. Uncertified Data Challenge Sierra Club first notes that, at the time of the designation process, the EPA possessed uncertified 2011 data for all areas. Because the agency’s regulations require the submission of uncertified data within ninety days of the end of the quarterly reporting period, see 40 C.F.R. § 58.16(b), the EPA had all 2011 uncertified data in its possession by the end of March. It should have used that data, Sierra Club argues, notwithstanding the lack of certification. We are unpersuaded. While the uncertified data must undergo preliminary auditing and quality checks before submission to the' EPA, see id. § 58.16(c), those preliminary quality control measures are just that — preliminary. As the EPA explains, the data remain subject to continuing checks and revisions by the states until final certification. Resp’t’s Br. 66. Accordingly, the EPA reasonably “does not presume that data [validation and auditing] processes are complete and accurate until” the final data certification. Id. at 46. Mindful of the significant deference we owe the EPA in matters concerning data quality or sufficiency, see Catawba Cnty., 571 F.3d at 41, we see no basis for second-guessing the EPA’s considered judgment on the issue. Sierra Club next argues that, even if the agency acted reasonably in refusing to rely on uncertified data, it acted arbitrarily in declining to delay the designation process until all states had certified their 2011 data by the standard May 1 deadline. After all, Sierra Club notes, the consent decree under which the EPA. conducted the designation process allowed the agency until May 31, 2012, to promulgate the final designations. 2008 Designations Rule, 77 Fed. Reg. at 30,091. Sierra Club, however, identifies no authority obligating the EPA to wait until the last possible minute to promulgate its designations. And in this case, doing so would have made little sense. The EPA entered into the consent decree precisely to settle allegations that it had already missed the Act’s statutory deadlines for promulgating the 2008 ozone NAAQS designations. See id. Accepting Sierra Club’s position would effectively call for the EPA to infringe the Act’s deadlines still further. In any event, as the EPA explained in denying Sierra Club’s petition for reconsideration of the designations after the May 1, 2012, certification deadline passed and 2009 to 2011 data were fully certified and available to the EPA, “[n]ew technical data become available on a regular basis.” Letter from Lisa P. Jackson, Adm’r, EPA to Robert Ukeiley, Counsel, Sierra Club enclosure p. 2 (Dec. 14, 2012). The EPA reasonably concluded that delay “to consider such new information would result in a never-ending process in which designations are never finalized.” Id. Indeed, Sierra Club itself has already filed a petition for reconsideration based on 2010 to 2012 data. See Sierra Club Reply Br. 8. The EPA could reasonably conclude that the process must end at some point. We conclude that the agency did not act arbitrarily in ending it here. Cf. Catawba Cnty., 571 F.3d at 51 (“New York’s underlying complaint is that the iterations should have continued, perhaps ad infini-tum. But such a process is inconsistent with the CAA: Congress imposed deadlines on EPA and thus clearly envisioned an end to the designations process.”). With that conclusion, and having reviewed the remainder of • Sierra Club’s challenges and determined that they lack merit, we deny the group’s petition for review. See Catawba Cnty., 571 F.3d at 52. D. Mississippi The State of Mississippi challenges the EPA’s use of 2008 to 2010 data to classify the counties within the Memphis, Tennessee area, an analysis that resulted in a nonattainment designation for part of De-Soto County, Mississippi. Because we conclude that the EPA’s actions were rational and in accordance with the Clean Air Act, we deny Mississippi’s petition for review. 1. Mississippi Background In Mississippi and elsewhere, the EPA conducted the designations for metropolitan areas through a two-step process. First, the EPA examined air-quality data from all regulatory monitors in a metropolitan area. If no monitors in the area showed a NAAQS violation, no county in the area would be designated nonattainment. In that event, there would be no second step. But if a single monitor from the area showed a NAAQS violation, the county housing the violating monitor would be designated nonattainment. See 2008 Guidance at 3-4. In that case, the EPA would proceed to the second step for that metropolitan area. The second step took account of the fact that the Act mandates nonattainment designations not only for areas themselves exceeding the relevant NAAQS, but also for all areas that “contribute[ ]” to a NAAQS violation in a “nearby area,” even if the “contributing” area’s air quality' — ■ considered alone — meets the NAAQS. See 42 U.S.C. § 7407(d)(l)(A)(i); 2008 Guidance at 3-4. In the second step, the EPA assessed each county in a metropolitan area with a violating monitor on a case-by-case basis to determine if the county contributed to the identified violation. If, on the basis of a multi-factor test, the EPA determined that a county “contributed” to the NAAQS exceedance at the violating monitor in another county, the EPA also designated the contributing county as non-attainment. We have repeatedly upheld multi-factor contribution analyses as consistent with the Act’s designation process under section 107 — a conclusion that Mississippi does not challenge here. See, e.g., ATK Launch Sys., 669 F.3d 330; Catawba Cnty., 671 F.3d 20. See generally supra § II. In 2011 and 2012, the EPA conducted that two-step designation process for the Memphis CBSA. The Memphis CBSA consists of several counties in Tennessee (Shelby, Tipton, and Fayette), Mississippi (DeSoto, Marshall, Tate, and Tunica), and Arkansas (Crittenden). See Office of Mgmt. & Budget, OMB Bulletin No. 10-02, Update of Statistical Area Definitions and Guidance on Their Uses 40 (Dec. 1, 2009). At the first step, the EPA evaluated 2008 to 2010 certified air-quality data and detected a NAAQS violation at the monitor in Shelby County, Tennessee. Proceeding to the second step, the EPA conducted the multi-factor analysis and determined that part of DeSoto County, Mississippi, contributed to the Shelby County violation. On December 9, 2011, the EPA notified Mississippi that it planned to designate part of DeSoto County as nonattainment when it promulgated the final designations in 2012. The EPA invited Mississippi (and all other states) to provide to the agency by February 29, 2012, any additional information for consideration in the final designation process — including any early-certified 2011 data. See Memphis, TN-MS-AR Area Designations for the 2008 Ozone NAAQS 3-4 [hereinafter Memphis Area Designations]. Mississippi responded to the EPA’s multi-factor analysis with its own multi-factor analysis, disputing the EPA’s conclusion that DeSoto County contributed to any violation in Shelby County. Additionally, Mississippi and Tennessee— two of the three states in the Memphis CBSA — early-certified their 2011 data before the February 29, 2012, deadline. Arkansas — the third state in the Memphis CBSA — declined to early-certify any 2011 data. On May 21, 2012, the EPA published its final designations for the Memphis CBSA. At the first step of the two-step designation pro'cess, the agency used 2008 to 2010 data and again identified a violation at the Shelby County monitor. The EPA then moved to the second step and, after considering Mississippi’s multi-factor analysis and updating its own analysis accordingly, reiterated its original conclusion that part of DeSoto County contributed to the Shelby County violation. The agency therefore designated part of DeSoto County as nonattainment. See Memphis Area Designations at 16. Mississippi claims that designation was arbitrary and capricious. We disagree. 2. Challenge to the First Step of the Designation Process First, Mississippi argues that the EPA acted arbitrarily in using 2008 to 2010 data for the first step of the two-step designation process (ie., identifying violating monitors within a CBSA) even though the EPA possessed early-certified 2011 data from Tennessee. The 2009 to 2011 data showed no NAAQS violation at the Shelby County monitor. Accordingly, Mississippi argues, no violation should have been identified at the first step of the two-step designation process. But the EPA declined to evaluate Shelby County using the early-certified 2009 to 2011 data, instead using the 2008 to 2010 data. True, the EPA must adequately explain why it declined to rely on the early-certified 2011 data. See City of Waukesha, 320 F.3d at 248. But the agency did so. At the time of the final designations, the EPA had in its possession early-certified data from Mississippi and Tennessee, but not from Arkansas. In the first step of its two-step designation process, the EPA evaluates all air-quality monitors in a metropolitan area. Without 2011 Arkansas data, the EPA did not have a full set of 2011 data for the Memphis CBSA. The EPA only had data from different time horizons — 2008 to 2010 data for the Arkansas portion of the Memphis CBSA, and 2009 to 2011 data for the Tennessee and Mississippi portions of that same CBSA. The agency declined to rely on this mismatched dataset. Instead, the EPA opted to rely on the most recent matched dataset in its posséssion: the complete set of 2008 to 2010 data. We see no reason — and Mississippi provides none — to declare irrational the EPA’s conclusion that comparing data from the same time period would be more appropriate than analyzing data from different time periods in the same evaluation process. Cognizant of the substantial deference we owe the EPA in that highly technical evaluation, see Catawba Cnty., 571 F.3d at 41, we find the EPA was entitled to rely on a matched dataset instead of a mismatched one. Even assuming the EPA’s choice to rely only on matched datasets for the Memphis CBSA was reasonable (as we conclude it to be), Mississippi argues that the EPA’s approach nonetheless was arbitrary because the agency required a matched dataset for Memphis-area designations but allegedly relied on a mismatched dataset for Chiea-go-area designations. “[I]nconsistent treatment,” we have found, is a “hallmark of arbitrary agency action.” Id. at 51. There was no inconsistent treatment here, however. In both Chicago and Memphis, the EPA relied only on matched datasets in the designation process. With regard to the Chicago metropolitan area, Illinois early-certified its 2011 data. Wisconsin and Indiana — portions of which also lie in the Chicago metropolitan area— did not early-certify. Illinois’s early-certified data showed a violating monitor in the Chicago area. At the first step of the Chicago-area designation process, the EPA relied on Illinois’s early-certified data, noted the violation, and thus proceeded to the second step’s multi-factor contribution analysis for all Chicago-area counties. Mississippi argues that, because the EPA only possessed early-certified data from Illinois, it used a mismatched data-set for Chicago’s designations. Consequently, Mississippi claims that the EPA took different approaches to dataset selection between Memphis and Chicago. Mississippi’s argument rests on a flawed understanding of the EPA’s designation process. At the first step of the process, a single violating monitor suffices to conclude the analysis and move to the second step. Though only Illinois had early-certified its data, that data showed a violating monitor. That was enough to terminate the first step of the process and move to the second step. It thus became irrelevant whether Wisconsin or Indiana data showed any violations: the EPA would proceed to the second step of the analysis regardless, based on the Illinois violation alone. The EPA therefore had a sufficient matched dataset of 2009 to 2011 data (albeit data from only one state, Illinois) to proceed to the second step of the designation process using 2009 to 2011 data alone. By contrast, the EPA had no matched dataset of 2009 to 2011 data in the Memphis area sufficient to complete the first step of the two-step process using that data alone. While data showing a single violating monitor are enough to end the first step and proceed to the second step, data showing all monitors in compliance would be needed to avoid proceeding to the second step’s multi-factor analysis — ie., to terminate the two-step process at the first step. As a result, when Arkansas declined to early-certify its 2011 data, the EPA could not determine if the. entire Memphis CBSA showed NAAQS compliance at all monitors for the 2009 to 2011 period; the agency lacked a sufficient 2009 to 2011 matched dataset with which to do so. The EPA then relied on the most recent matched dataset sufficient to complete the first-step analysis (the 2008 to 2010 data), just as the EPA selected the most recent matched dataset sufficient for the first-step analysis of the Chicago area. The EPA therefore acted in a consistent manner in both areas, each time using the most recent matched datasets sufficient to complete the first step of the two-step designation process. 3. Challenge to the Second Step of the Designation Process Mississippi also challenges the EPA’s application of the second step of the designation process. The EPA acted arbitrarily, the state argues, in applying the multi-factor test and concluding that De-Soto County contributed to the Shelby County violation. We find no reason to disturb the EPA’s analysis. First, Mississippi challenges the EPA’s differing articulations of the multi-factor test. As pronounced in the 2008 Guidance, the EPA originally conceived of that test as consisting of nine factors. In making the final designations, the EPA applied a five-factor test. See supra § I.B-C, The state argues that the EPA’s “consolidation]” of the test from nine to five factors was arbitrary and capricious. State & County Br. 15. We disagree.' At the outset, we do not necessarily agree that the EPA was required to adhere to the 2008 Guidance. The 2008 Guidance did not purport to be a legislative rule, and it explicitly provided that it was “not binding on states, tribes,' the public or the EPA.” 2008 Guidance at 4; cf. Catawba Cnty., 571 F.3d at 33-34 (materially similar guidance for PM2,5 NAAQS designations did not “create or modify legally binding rights”). But even if we assume that the 2008 Guidance was binding, the EPA did not deviate from it in the final designations. The “consolidation” of the factors was just that — a consolidation.. It effected no deletion. During the final designation process, the agency simply grouped several of the 2008 Guidance factors into a single factor, the consideration of which necessarily entailed consideration of the multiple 2008 Guidance factors now residing within it. We find no examples of a final designation that failed to consider a factor identified in the 2008 Guidance. With “no bright line for any of the factors,” and with each factor “weighted considering the unique circumstances of each nonattainment area,” Response to Comments at 61, the consolidation worked no substantive change and thus affords no basis for setting aside the EPA’s analysis. Second, Mississippi challenges the EPA’s specific application of the multi-factor test to DeSoto County. We accord the EPA “extreme deference” in applying that test, and will overturn the EPA’s designations only if the agency applied the test “so erroneously in a particular case that it could not have reasonably concluded that a county was contributing to nearby violations.” Catawba Cnty., 571 F.3d at 40-41. This is not such a case. The agency provided data showing that DeSoto County’s NOx and S02 (ozone precursors) emissions were the second-highest in the Memphis CBSA. Memphis Area Designations at 8. The county also had the second highest number of workers commuting to counties with violating monitors, the second highest number of vehicle miles traveled in the CBSA, and the highest percentage population growth over the last decade. Those factors led the EPA to conclude that DeSoto County was integrated with Shelby County in a way that indicated ozone contribution. Id. at 9-10. Additionally, meteorological analysis at the Shelby County monitor showed weather patterns characterized in part by winds blowing in from DeSoto County. Id. at 12. On those bases, the EPA reasonably concluded that DeSoto County contributed to the Shelby County violation. Mississippi principally argues that significant “commerce activity” occurring outside of DeSoto County (including interstate highway traffic, rail and barge transportation, diesel fuel sales, and air traffic) means that other' counties contribute to the Shelby County violation more than DeSoto County does — and that, because some of those counties avoided nonattainment designations, DeSoto County should, too. Miss. Dep’t of Envtl. Quality, Air' Div., 2008 Ozone Standard Designation Recommendation for DeSoto County, Mississippi 8-12 (Feb. 2012). But the EPA considered that argument and determined in a well-reasoned analysis that the data from Mississippi was only one consideration in the designation process. See Response to Comments at 97; see also Memphis Area Designations 1-31. The EPA concluded that DeSoto County did contribute to Shelby County’s vi