United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued January 24, 2012 Decided February 24, 2012
No. 10-1004
ATK LAUNCH SYSTEMS, INC.,
PETITIONER
v.
ENVIRONMENTAL PROTECTION AGENCY,
RESPONDENT
BRIAN MOENCH AND UTAH PHYSICIANS
FOR A HEALTHY ENVIRONMENT,
INTERVENORS
Consolidated with 10-1005, 10-1006,
11-1252, 11-1253, 11-1254
On Petitions for Review of a Final Action
of the Environmental Protection Agency
Michael A. Zody argued the cause for petitioners. With him
on the briefs were Michael L. Larsen, M. Lindsay Ford,
Elizabeth A. Schulte, and David W. Tundermann. Dylan M.
Fuge and Claudia M. O'Brien entered appearances.
Jessica O'Donnell, Attorney, U.S. Department of Justice,
argued the cause for respondent. With her on the brief was
2
Geoffrey L. Wilcox, Attorney, U.S. Environmental Protection
Agency.
Joro Walker was on the brief for intervenors Utah
Physicians for a Healthy Environment, et al. in support of
respondent.
Before: ROGERS and KAVANAUGH, Circuit Judges, and
GINSBURG, Senior Circuit Judge.
Opinion for the Court by Circuit Judge ROGERS.
ROGERS, Circuit Judge: In these consolidated petitions,
ATK Launch Systems, Inc., two Utah counties, and three Utah
cities seek partial vacation of a final rule designating certain
areas as nonattainment for the 2006 24-hour fine particulate
matter (PM2.5) standard. Air Quality Designations for the 2006
24-Hour Fine Particle (PM2.5) National Ambient Air Quality
Standards, 74 Fed. Reg. 58,688 (Nov. 13, 2009) (“Final Rule”).
In particular, petitioners challenge the inclusion of parts of
Tooele and Box Elder Counties within the Salt Lake City
nonattainment area. The Environmental Protection Agency
concluded, upon applying its nine-factor test for designations,
that emissions from eastern portions of both Box Elder County,
including Brigham City and ATK’s operations, and Tooele
County, including Tooele City and Grantsville City, contributed
to nearby violations of the 24-hour PM2.5 standard in and around
Salt Lake City.
Petitioners’ principal argument is that EPA was arbitrary
and capricious in applying the nine-factor designation analysis,
arguing dissimilar treatment as compared to EPA’s analysis of
the data for two east coast counties, Warren County, New Jersey
and Hartford County, Connecticut, which EPA designated
attainment. Petitioners also object to EPA’s use of a pollutant
3
transport model generally and its analysis of wind data for Box
Elder County specifically. Finally, they question EPA’s
decision to include ATK’s operations in the nonattainment
portion of Box Elder County. Because EPA’s nine-factor test is
intended to be applied on a case-by-case basis to account for
diverse considerations, including the varying effects of local
topography and meteorology on PM2.5 dispersion, and EPA
reasonably explained its designations, we deny the petitions for
review.
I.
Title 1 of the Clean Air Act (“CAA”) requires EPA to set
national ambient air quality standards for air pollutants that may
reasonably be anticipated to endanger public health and welfare.
42 U.S.C. §§ 7408–09. One such pollutant, PM2.5, consists of
airborne particles 2.5 micrometers or smaller in diameter. EPA
has promulgated both an annual and a 24-hour standard for
PM2.5. Effective December 18, 2006, EPA revised the 24-hour
PM2.5 standard downward from 65 micrograms/cubic meter to
35 micrograms/cubic meter. See National Ambient Air Quality
Standards for Particulate Matter, 71 Fed. Reg. 61,144 (Oct. 17,
2006) (codified at 40 C.F.R. pt. 50). Under § 107(d) of the
CAA, 42 U.S.C. § 7407(d), after new or revised standards are
promulgated, States are to submit proposed area designations to
EPA, classifying areas as attainment, nonattainment, or
unclassifiable. Areas are to be designated nonattainment if they
either violate the standard or contribute to a nearby area’s
violation. Id. § 7407(d)(1)(A)(i). The EPA Administrator may
modify the designations as deemed necessary, and States then
have an opportunity to respond to the modifications. Id.
§ 7407(d)(1)(B)(ii).
On June 8, 2007, EPA provided States with a guidance
document suggesting that they consider nine factors in making
4
designations: (1) emission data, (2) air quality data, (3)
population density and degree of urbanization, (4) traffic and
commuting patterns, (5) growth rates and patterns, (6)
meteorology, (7) geography/topography (e.g., mountain ranges
and other air basin boundaries), (8) jurisdictional boundaries,
and (9) level of control of emission sources. The list is neither
“mandatory nor an exclusive list of types of relevant
information.” Final Rule, 74 Fed. Reg. at 58,694-95. In
Catawba County v. EPA, 571 F.3d 20, 38–40 (D.C. Cir. 2009),
the court approved EPA’s interpretation of section 107 of the
CAA to permit it to apply this nine-factor test in determining
which areas contribute to violations in a nearby nonattainment
area.
The PM2.5 problem in the Salt Lake City area is driven by
topography and meteorology: the area is a valley almost
completely bounded by mountain ranges. Under normal
meteorological conditions, air temperature decreases as altitude
increases. In the Salt Lake City area, wintertime high pressure
systems cause temperature inversions; high altitude warm air
traps cold air below, with an inversion layer at about 1,500 feet.
The surrounding mountains, which extend above the inversion
layer, trap the ground level cold air and prevent dispersion.
Pollution then accumulates in the stagnant air mass, sometimes
for weeks at a time. The air quality worsens gradually until the
high pressure system lifts, at which point the polluted air can
disperse over the mountains. Without these seasonal inversions,
Salt Lake City would not likely be in violation of the air quality
standards – the State of Utah noted in its 2007 submission to
EPA that it attains the annual PM2.5 standard.
Utah submitted its proposed designations on December 18,
2007, including recommendations that Box Elder County be
designated attainment and Tooele County be designated
unclassifiable; Box Elder County is north, and Tooele County
is west, of the Great Salt Lake. Utah’s recommendations did not
5
include portions of these two “contributing” counties and also
made Utah County, which borders Salt Lake County to the
south, a separate nonattainment area. EPA disagreed and
classified parts of both Box Elder County (including Brigham
City and ATK’s operations) and Tooele County (including
Tooele City and Grantsville City) nonattainment and as part of
a single unified nonattainment area for the Salt Lake City area,
because, under EPA’s analysis, they contribute to the Salt Lake
City area’s violation of the 24-hour PM2.5 standard. EPA’s
classifications were consistent with Utah’s recommended
eastern boundary but differed relative to the western boundary.
Considering the nine factors together, EPA determined that
eastern portions of both Box Elder and Tooele Counties produce
emissions that contribute to nearby violations of the standard.
Final Utah Technical Support Document (Dec. 2008), at 53; see
Final Rule, 74 Fed. Reg. at 58,769–70 (table). EPA found that
the portions of Box Elder and Tooele Counties designated
nonattainment were in the same topographic airshed as the
greater Salt Lake City area, and that there was no physical
impediment to prevent their emissions from traveling into the
violating region. Using wind data collected from Salt Lake
International Airport, EPA analyzed PM2.5 measurements from
air quality monitors; when the Salt Lake County monitors
recorded violations, the prevailing winds were from the
northwest and southeast. EPA concluded this indicated that
“some portion of PM2.5 that influence[s] [violating] monitor[s]
[] originates from eastern Box Elder County to the north” and
“some portion . . . originates from the north and west of Salt
Lake County from sources in Tooele County.” Final Utah
Technical Support Document, at 39. EPA further explained that
modeling of emissions demonstrated that some contributions
came from Box Elder and Tooele Counties, the populated areas
of the counties had relatively high density, the percentage of
6
people commuting into Salt Lake County from the two counties
was high, and both had higher than average predicted growth.
EPA also examined the wind patterns, which showed that,
during the wintertime inversions, the emissions “sloshed”
around the valley, because winds move from higher elevations
to the low points during the night, and then in the opposite
direction during the day. In this manner, emissions from the
nonattainment portions of Box Elder and Tooele Counties would
be carried into violating areas in and around Salt Lake City at
night. EPA additionally considered the Contributing Emissions
Score (“CES”), which incorporates emissions, meteorological,
and air quality data to provide a relative ranking score of the
potential contribution of an area’s counties to the local air
quality problem. The scores are normalized to show relative
contribution within an area, with values ranging from 0 to 100;
the eastern portion of Box Elder County had a CES of 7 and the
eastern portion of Tooele County had a CES of 2. In conjunction
with the nine-factor test, the CES provided an additional method
of evaluating the potential contribution of nearby counties.
Based on its analysis, EPA defined the boundaries of the airshed
as the Wasatch Mountains to the east, the Promontory and North
Promontory Mountains to the west (for eastern Box Elder
County), and the Stansbury to the west (for eastern Tooele
County).
Petitioners sought review of the Final Rule in the U.S. Court
of Appeals for the Tenth Circuit, and subsequently filed
protective petitions for review in this court; the Tenth Circuit
transferred the case to this court, ATK Launch Sys., Inc. v. EPA,
651 F.3d 1194 (10th Cir. 2011). Our review of EPA’s
rulemaking pursuant to the CAA is under the same standard as
the Administrative Procedure Act, 5 U.S.C. § 706(2)(A).
Catawba, 571 F.3d at 41; Allied Local & Reg’l Mfrs. Caucus v.
EPA, 215 F.3d 61, 68 (D.C. Cir. 2000). The Final Rule may be
7
vacated only if “arbitrary, capricious, an abuse of discretion, or
otherwise not in accordance with law.” 42 U.S.C.
§ 7607(d)(9)(A). The court “must affirm the Rule if the record
shows EPA considered all relevant factors and articulated a
‘rational connection between the facts found and the choice
made.’” Catawba, 571 F.3d at 41 (quoting Burlington Truck
Lines v. United States, 371 U.S. 156, 168 (1962)). The court
“give[s] an ‘extreme degree of deference to [EPA] when it is
evaluating scientific data within its technical expertise.’” Id.
(quoting City of Waukesha v. EPA, 320 F.3d 228, 247 (D.C. Cir.
2003) (second alteration in original)). This level of deference is
“especially appropriate in [] review of EPA’s administration of
the complicated provisions of the Clean Air Act.” Id. (citing
Nat’l Ass’n of Clean Air Agencies v. EPA, 489 F.3d 1221, 1229
(D.C. Cir. 2007)); see also Bluewater Network v. EPA, 372 F.3d
404, 410 (D.C. Cir. 2004).
II.
In upholding EPA’s nine-factor test, this court concluded
that “nothing in the [CAA] compels EPA to quantify a uniform
amount of contribution below which counties will automatically
escape nonattainment designations or to quantify similar
thresholds for the nine factors EPA evaluated in making those
determinations.” Catawba, 571 F.3d at 39. Furthermore, the
court observed, “discrete data points” are not determinative,
because elevating them “ignore[s] the very nature of the nine-
factor test, which is designed to analyze a wide variety of data
on a case-by-case basis.” Id. at 46 (internal quotation marks and
citation omitted). Thus, “EPA’s holistic assessment of
numerous factors [] drives the process – no single factor
determines a particular designation.” Id. Nonetheless, the nine-
factor test may not be applied “inconsistently, resulting in
similar counties being treated dissimilarly.” Id. at 40 (emphasis
added).
8
A.
The bulk of petitioners’ challenge rests on a comparison
between the data underlying EPA’s decision to classify Box
Elder and Tooele Counties nonattainment and that underlying
EPA’s decision to classify two east coast counties, Warren
County, New Jersey and Hartford County, Connecticut,
attainment. Petitioners’ premise is flawed. The significant
topographical and meteorological differences between the Salt
Lake City area and the two east coast counties make a direct
one-to-one comparison of the data underlying the analyses
inappropriate.
As the State of Utah explained in recommending
designations, in the Salt Lake City area “it is not enough to
simply have an urban area with an urban mix of emissions [to
cause a violation]; there must also be a barrier to dispersion . .
. which allows PM2.5 concentrations to build up over a period of
several days.” Utah Area Designation Recommendation for the
2006 PM2.5 NAAQS (“Utah Recommendations”), at 3. In other
words, fewer emissions on any given day are necessary to cause
a problem in the Salt Lake City area because the inversion layer
forms a lid and the mountains a wall, trapping accumulating
emissions for days or weeks. EPA pointed out that the 24-hour
PM2.5 nonattainment areas on the east coast are distinct; no such
temperature inversion exists trapping pollutants at the ground
level, and no mountain ranges limit dispersion. See Final
Connecticut Technical Support Document (Dec. 2008), at Factor
7; Final New Jersey Technical Support Document (Dec. 2008),
at 18. There, the sheer volume of emissions on a given day,
rather than topographical and meteorological influences that
cause accumulation of emissions over time, suffices to cause
violations of the 24-hour standard. Because the Salt Lake City
area’s problem is cumulative, a comparison of data on a per-day
basis is not informative. It is no surprise that the data for the
four counties led EPA to reach opposite conclusions, and
9
petitioners thus fail to demonstrate that “similar counties [have
been] treated dissimilarly” by EPA. Catawba, 571 F.3d at 40
(emphasis added).
Petitioners also raise several objections concerning
dissimilar treatment that go beyond a simple comparison of the
data among the four counties. The first two, involving how EPA
characterized the population density of the Utah Counties and its
consideration of projected, in addition to historical, growth rates
for Utah but not for Hartford and Warren Counties, are readily
disposed of. Both Box Elder and Tooele Counties are physically
large, with vast expanses of rural, unpopulated desert regions
separated from the Salt Lake City airshed by mountain ranges.
It was reasonable for EPA to examine the population of only the
portions of the counties considered for nonattainment separately
from the counties as a whole – indeed, it would be illogical to do
otherwise. See Final Rule, 74 Fed. Reg. at 58,696. Although
EPA could have taken the next step of determining the actual
population density for the nonattainment-designated portions of
the counties, its conclusion of relatively high density is
supported by the data in the record. Likewise, the State of Utah
provided EPA with projected growth rate data, see Utah
Recommendations, at 27–28, and it was appropriate for EPA to
consider “the best available information,” Catawba, 571 F.3d at
44.
Petitioners also object that in considering the traffic and
commuting patterns for Box Elder and Tooele Counties, EPA
relied on percentages of commuters traveling to other counties,
rather than raw numbers of commuters, to support its findings of
contribution. In Warren and Hartford Counties, EPA focused on
the raw number of commuters. The court noted in Catawba that
EPA stated that “the relevant factor on which it relied was not
the percentage of commuters but rather the raw number of
commuters.” Id. at 50. Catawba, however, dealt with the
10
annual PM2.5 standard. Because the 24-hour standard is at issue
here, and the Salt Lake City area’s violations of that standard
occur after trapped pollution accumulates over days or weeks,
reliance on daily raw numbers of commuters would understate
the potential contribution of Box Elder and Tooele Counties.
EPA, in fact, provided the raw numbers (both commuters and
vehicle miles traveled) while noting this data “does not
adequately take into account . . . [a] large volume of diesel truck
traffic[] on the major highways running through th[e] area.”
Final Utah Technical Support Document, at 36. EPA explained
that the relatively high percentage of commuters traveling from
Box Elder and Tooele Counties to Salt Lake County
demonstrates a linkage between the areas, suggesting that the
two counties contribute to the nearby nonattainment status. Id.
The high projected growth rates in population and vehicle miles
traveled for both counties further supports this conclusion. In
any event, “no single factor determines a particular designation.”
Catawba, 571 F.3d at 46; see Final Rule, 74 Fed. Reg. at 58,695.
Additionally, petitioners question EPA’s treatment of the
CES on two grounds: the use of partial-county scores for Box
Elder and Tooele Counties and EPA’s treatment of those
counties’ low scores, compared to the higher scores for Warren
and Hartford Counties that EPA described as demonstrating a
low potential for significant contribution. EPA reasonably relied
on scores calculated for only the portions of Box Elder and
Tooele Counties designated nonattainment, however, because a
county-wide “CES cannot adequately account for the effects of
mountainous terrain which would essentially split a county into
different parts, each having [its] own potentially unique effect
on the violating county,” EPA Derivation of Contributing
Emission Score (Dec. 15, 2008) (“CES Derivation”), at 9. And
although the CES is normalized to show relative contribution,
EPA cautioned that it “only provides relative contribution within
the area under evaluation and does not provide a reliable means
11
for comparison between counties in different areas.” Final Rule,
74 Fed. Reg. at 58,695 n.16. In explaining the derivation of the
CES, EPA emphasized that the scores require careful
interpretation, “particularly in the western United States.” CES
Derivation at 9. EPA addressed that concern by calculating
partial-county scores to account for the size and geographic
separation of the counties, and further noted that mountain
ranges can act as barriers to transported emissions, implying
further caution was warranted because of the effect of the
inversion layer, which causes emissions to accumulate over
time. In any event, the CES was “merely an additional method
for evaluating the potential contribution of nearby counties,”
Final Rule, 74 Fed. Reg. at 58,695, and other of the nine factors,
such as topography, may weigh more heavily than the CES. Cf.
Catawba, 571 F.3d at 49. EPA’s application of the CES data is
the type of “‘evaluati[on] [of] scientific data within its technical
expertise,’” for which the court gives EPA an “‘extreme degree
of deference,’” Id. at 41 (quoting City of Waukesha, 320 F.3d at
247).
Further, petitioners maintain that in applying the first factor
(emissions), EPA applied a “significant contribution” standard
to Hartford County but an “any contribution” standard to Box
Elder and Tooele Counties. In the Final Rule, EPA rejected
adoption of a “significant contribution” standard, 74 Fed. Reg.
at 58,691–92, and the court upheld its similar rejection in
Catawba, 571 F.3d at 38–39. Although EPA’s technical
analysis used the word “significant,” it more often referred to
the potential contribution of Hartford County without using the
word, suggesting that it was used as an adjective to describe the
data, rather than as a dissimilar standard. Compare Final
Connecticut Technical Support Document, at Factors 1, 6, and
Conclusion, with id., at Factors 3, 4, 5, 8, and Conclusion. It is
far from the dissimilar treatment the court rejected in Catawba,
where one Region’s bright-line test would have resulted in a
12
county in a different Region being designated attainment, had
the same test been used in the second Region. See id. at 51–52.
Petitioners fail to demonstrate that EPA employed a different
standard or acted arbitrarily or capriciously in its analysis of Box
Elder and Tooele Counties as compared to its analysis of
Hartford County.
B.
In addition to asserting dissimilar treatment compared to
Warren and Hartford Counties, petitioners challenge EPA’s use
of modeling data in analyzing the potential transport of
emissions from Box Elder and Tooele Counties as technically
infirm, and its analysis and characterization of the wind data for
Box Elder County as both unsupportive of inclusion in the
nonattainment area, and nonresponsive to new information in
the record. Neither objection has merit.
As part of its analysis of factor six, meteorology, EPA used
a modeling program, “HYSPLIT,” which analyzes wind
conditions to determine dispersion of pollutants; this analysis
yields “back-trajectories” that help determine the source of
pollutants. See Final Utah Technical Support Document, at 41.
The modeling demonstrated a degree of transport of emissions
from Box Elder and Tooele Counties into the violating Salt Lake
City monitors during inversion episodes. Petitioners point to
EPA’s statements acknowledging the “HYSPLIT” model’s
limitations, particularly in areas with topographic features, such
as mountains. See EPA State and Tribal Comment Summary
and Response Document (Dec. 22, 2008) at 13, 188. EPA,
however, independently analyzed surface meteorological data
from the State of Utah’s monitoring stations, which confirmed
the results of the “HYSPLIT” model. Id. By confirming the
modeling results with on-the-ground data, EPA “took reasonable
steps to ensure that” the “HYSPLIT” model’s limitations were
considered, “and its choices are not arbitrary or capricious,” Am.
13
Coke & Coal Chems. Inst. v. EPA, 452 F.3d 930, 943 (D.C. Cir.
2006).
Likewise, petitioners’ objection to EPA’s analysis of wind
direction data is without merit. In its initial analysis modifying
Utah’s recommended designations, EPA used wind direction
and speed data from the Salt Lake City International Airport and
the Pocatello Regional Airport in Idaho to analyze the pollution
data for Box Elder County. In response to comments from the
State of Utah that the two sites were too far from Box Elder
County to adequately account for local topographic influences
on wind direction and speed, EPA replaced the data with that
from Hill Air Force Base, which is closer. The original data
showed that when PM2.5 levels in Box Elder County were
highest, wind was coming from the northwest. The replacement
data, using the more representative Hill Air Force Base wind
measurements, indicated that when PM2.5 levels in Box Elder
County were highest, wind was coming from the southeast.
Petitioners point out that EPA’s conclusion did not change
despite the substitution of new data for Box Elder County. In
both the preliminary and final technical analyses, EPA
concluded that “the data presented . . . for the Salt Lake City-
Ogden-Clearfield [area] . . . appear to show that some
component of measured elevated PM2.5 values may originate
from the” northwest, supporting the conclusion that Box Elder
County contributes to nearby violations. Compare EPA
Modification to Designations (Aug. 18, 2008) at 36 with Final
Utah Technical Support Document, at 39.
Nonetheless, petitioners’ claim that EPA was arbitrary and
capricious in failing to address the new data in its analysis is
incorrect. EPA’s analysis and conclusions were of the wind data
for the Salt Lake City air basin as a whole, not just for Box
Elder County. As EPA reasonably explained in responding to
comments about the changed data, “[t]he difference between
14
[the original and substituted data] is not unexpected given the
influence of local topography. The Salt Lake International
[data] used by EPA is likely more representative of large scale
wind patterns in the basin, given the relative distance of the
airport from topographic features.” EPA Public Comment
Summary and Response Document (Dec. 22, 2008), at 166–67;
see also EPA State and Tribal Comment Summary and
Response Document, at 189. In other words, to determine what
areas are contributing to nearby violations, the most relevant
wind data is that which corresponds with the violating monitors.
The record supports the conclusion that, when PM2.5 levels are
most severe in Salt Lake City, wind direction is sometimes from
the northwest, indicating contribution from Box Elder County.
EPA’s analysis of the wind data and air basin conclusion about
pollution transport was reasonably based upon “the best
available information,” Catawba, 571 F.3d at 44, and petitioners
thus fail to demonstrate that EPA ignored new information or
otherwise was arbitrary or capricious.
C.
Finally, petitioners challenge EPA’s designation of the
portion of Box Elder County containing ATK’s operations as
arbitrary and capricious. ATK is engaged in the aerospace and
defense industries. EPA may designate partial counties, see
Catawba, 571 F.3d at 42, and it reasonably relied on the
Promontory Mountains, and the coinciding intra-county
jurisdictional boundaries, as the western topographic boundary
for the airshed in Box Elder County. Petitioners do not dispute
that ATK’s operations occur below the inversion layer, which is
at about 1,500 feet, and, as discussed, EPA reasonably
concluded that meteorological data indicated that emissions
from eastern Box Elder County, where ATK’s operations occur,
contribute to nearby violations of the PM2.5 standards.
Petitioners fail to demonstrate that EPA was arbitrary or
15
capricious by including ATK’s operations within the
nonattainment area.
Accordingly, we deny the petitions for review.