American Farm Bureau Federation v. Environmental Protection Agency

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued September 15, 2008 Decided February 24, 2009

No. 06-1410

AMERICAN FARM BUREAU FEDERATION AND NATIONAL PORK
PRODUCERS COUNCIL,
PETITIONERS

v.

ENVIRONMENTAL PROTECTION AGENCY,
RESPONDENT

AMERICAN CHEMISTRY COUNCIL, ET AL.,
INTERVENORS

Consolidated with 06-1411, 06-1415, 06-1416, 06-1417

On Petitions for Review of an Order
of the Environmental Protection Agency

Michael J. Myers, Assistant Attorney General, Attorney
General’s Office of State of New York, argued the cause for
State Petitioners and State Amici. With him on the briefs were
Andrew M. Cuomo, Attorney General, Barbara D. Underwood,
Solicitor General, and Katherine Kennedy, Assistant Attorney
General, Terry Goddard, Attorney General, Attorney General’s
Office of the State of Arizona, Joseph P. Mikitish, Assistant
 2

Attorney General, Edmund G. Brown, Jr., Attorney General,
Attorney General’s Office of the State of California, Theodora
Berger, Senior Assistant Attorney General, Susan L. Durbin,
Deputy Attorney General, Richard Blumenthal, Attorney
General, Attorney General’s Office of the State of Connecticut,
Kimberly Massicotte, Assistant Attorney General, Joseph R.
Biden, III, Attorney General, Attorney General’s Office of the
State of Delaware, Valerie S. Csizmadia, Deputy Attorney
General, Lisa Madigan, Attorney General, Attorney General’s
Office of the State of Illinois, Thomas Davis, Assistant Attorney
General, G. Steven Rowe, Attorney General, Attorney General’s
Office of the State of Maine, Gerald D. Reid, Assistant Attorney
General, Douglas F. Gansler, Attorney General, Attorney
General’s Office of the State of Maryland, Kathy M. Kinsey,
Roberta R. James, and Mary Raivel, Assistant Attorneys
General, Martha Coakley, Attorney General, Attorney General’s
Office for the Commonwealth of Massachusetts, William L.
Pardee, Assistant Attorney General, Kelly A. Ayotte, Attorney
General, Attorney General’s Office of the State of New
Hampshire, Maureen D. Smith, Senior Assistant Attorney
General, Ann Milgram, Attorney General, Attorney General’s
Office of the State of New Jersey, Kevin P. Auerbacher and
Jung Kim, Deputy Attorneys General, Gary King, Attorney
General, Attorney General’s Office of the State of New Mexico,
Stephen R. Farris, Assistant Attorney General, Tracy M.
Hughes, William G. Grantham, and Karen L. Reed, Special
Assistant Attorneys General, Hardy Myers, Attorney General,
Attorney General’s Office of the State of Oregon, Philip
Schradle, Special Counsel, Kristen M. Campfield, Associate
Counsel, The Commonwealth of Pennsylvania, Department of
Environmental Protection, Patrick C. Lynch, Attorney General,
Attorney General’s Office of the State of Rhode Island, Tricia
K. Jedele, Special Assistant Attorney General, William H.
Sorrell, Attorney General, Attorney General’s Office of State of
Vermont, Kevin O. Leske, Assistant Attorney General, Peter J.
 3

Nickles, Interim Attorney General, Attorney General’s Office of
the District of Columbia, Todd S. Kim, Solicitor General, Donna
M. Murasky, Senior Assistant Attorney General, Kurt R. Wiese,
District Counsel, South Coast Air Quality Management District,
and Barbara Baird, Principal Deputy District Counsel. J. Jared
Snyder, Assistant Attorney General, Attorney General's Office
of State of New York, and Kristen C. Furlan, Attorney, The
Commonwealth of Pennsylvania, Department of Environmental
Protection, entered appearances.

Paul R. Cort argued the cause for Environmental
Petitioners. With him on the briefs were Erin M. Tobin,
Deborah S. Reames, and David S. Baron.

Denise W. Kennedy and Richard E. Schwartz argued the
cause for Industry Petitioners. With them on the briefs were
Kirsten L. Nathanson, Lawrence E. Volmert, Robert T. Connery,
John F. Shepherd, Gary H. Baise, and Julie Anna Potts. John
J. Zimmerman entered an appearance.

Hope M. Babcock was on the joint brief for Health Amici.

Norman L. Rave, Jr., and Brian H. Lynk, Attorneys, U.S.
Department of Justice, argued the cause for respondent. With
them on the brief were John C. Cruden, Deputy Assistant
Attorney General, and Steven E. Silverman and John T. Hannon,
Attorneys, U.S. Environmental Protection Agency.

F. William Brownell, Allison D. Wood, Lucinda Minton
Langworthy, Leslie A. Hulse, Stacy R. Linden, Richard S.
Wasserstrom, Robin S. Conrad, and Amar D. Sarwal were on
the joint brief of Fine PM Industry Intervenors in support of
respondent.

Richard E. Schwartz, Kirsten L. Nathanson, Peter S.
 4

Glaser, Robert R. Gasaway, Ashley C. Parrish, Julie Anna Potts,
and Harold P. Quinn, Jr., were on the joint Coarse PM NAAQS
brief for industry intervenors in support of respondent.

Erin Tobin, Paul Cort, Deborah S. Reames, and David S.
Baron were on the brief for intervenors American Lung
Association and Environmental Defense.

Thomas J. Ward, Robert R. Gasaway, and Ashley C.
Parrish were on the brief of amicus curiae National Association
of Home Builders in support of respondent. Duane J. Desiderio
entered an appearance.

Before: GINSBURG, GARLAND, and GRIFFITH, Circuit
Judges.

Opinion for the Court filed PER CURIAM.

PER CURIAM: In these consolidated cases, we consider
several challenges to the Environmental Protection Agency’s
most recent revision of the National Ambient Air Quality
Standards for particulate matter. Because the agency
promulgated standards for fine particulate matter that were, in
several respects, contrary to law and unsupported by adequately
reasoned decisionmaking, we grant the petitions for review in
part and remand those standards to the agency for further
proceedings. We deny the petitions for review of the agency’s
standards for coarse particulate matter because those standards
are not arbitrary, capricious, or otherwise contrary to law.
 5

I.

A. Background to the 2006 Rulemaking

Particulate Matter Pollution

This case is about the Environmental Protection Agency’s
(EPA) regulation of particulate matter (PM), an air pollutant.
PM includes “a broad class of chemically and physically diverse
substances that exist as discrete particles (liquid droplets or
solids) over a wide range of sizes.” Final Rule: National
Ambient Air Quality Standards for Particulate Matter, 71 Fed.
Reg. 61,143, 61,146 (2006). Within this general definition, PM
is classified based on factors such as particle size, origin, and
chemical composition. The EPA primarily uses particle size to
classify PM, distinguishing between “fine PM” and “coarse
PM.”

Fine and coarse PM differ in ways other than size. Fine PM
is produced chiefly by combustion processes and atmospheric
reactions of gaseous pollutants; sources include motor vehicles,
power generation, and residential fuel burning. Coarse PM
tends to result from mechanical processes or the resuspension of
dusts in the air; sources include construction and demolition
activities as well as agricultural and mining operations. The two
types of PM also exhibit different atmospheric behavior: while
fine PM can remain suspended for long periods of time and
travel great distances, coarse PM tends to deposit rapidly and
does not travel as far.

The EPA further notes that “it is appropriate to draw a
distinction between two general types of ambient mixes of
coarse particles: ‘urban’ and ‘non-urban.’” Id. at 61,185 n.66.
“Urban” coarse “characterizes the mix in more heavily
populated urban areas, where sources such as motor vehicles and
 6

industry contribute heavily to ambient coarse particle
concentrations and composition.” Id. In contrast, “non-urban”
coarse “encompasses mixes in a variety of other locations
outside of urbanized areas, including mixes in rural areas which
are likely to be dominated by natural crustal materials.” Id. The
EPA cautions, however, that this is not a sharp distinction
because “some types of sources are present in both urban and
non-urban areas.” Id.

Studies have demonstrated that both fine and coarse PM can
have negative effects on public health and welfare. For
example, each is associated with increased mortality (premature
death) rates and morbidity (illness) effects such as
cardiovascular disease and decreased lung function.
Specifically, scientific evidence supports these associations for
long- and short-term exposure to fine PM as well as short-term
exposure to urban coarse PM. With regard to public welfare,
high levels of fine PM in the air can impair visibility, while both
fine and coarse PM can damage vegetation, disrupt ecosystems,
corrode metals, and erode paints and other building materials.

The Clean Air Act

The Clean Air Act (CAA), 42 U.S.C. §§ 7401–7671p
(2000), a comprehensive statutory scheme designed to reduce air
pollution, requires the EPA to set national ambient air quality
standards (NAAQS) for air pollutants such as PM. The statute
directs the agency to identify air pollutants that “may reasonably
be anticipated to endanger public health or welfare.” Id.
§ 7408(a)(1). Once a pollutant is identified, the agency staff
develop and issue air quality criteria, collected in a Criteria
Document, that “accurately reflect the latest scientific
knowledge useful in indicating the kind and extent of all
identifiable effects on public health or welfare which may be
expected from the presence of such pollutant in the ambient air.”
 7

Id. § 7408(a)(2). Although not required by the statute, in
practice EPA staff also develop a Staff Paper, which discusses
the information in the Criteria Document that is most relevant to
the policy judgments the EPA makes when it sets the NAAQS.

For each pollutant identified, the EPA must propose and
promulgate two sets of NAAQS: (1) primary NAAQS, “the
attainment and maintenance of which in the judgment of the
Administrator, based on [the air quality criteria] and allowing an
adequate margin of safety, are requisite to protect the public
health,” id. § 7409(b)(1) (emphasis added); and (2) secondary
NAAQS specifying a level of air quality “the attainment and
maintenance of which in the judgment of the Administrator,
based on such criteria, is requisite to protect the public welfare
from any known or anticipated adverse effects associated with
the presence of such air pollutant in the ambient air,” id.
§ 7409(b)(2) (emphasis added). In setting both standards, the
EPA takes into account the Criteria Document, the Staff Paper,
and the recommendations of the Clean Air Scientific Advisory
Committee (CASAC), a seven-member, independent scientific
review committee appointed by the Administrator pursuant to 42
U.S.C. § 7409(d)(2). The EPA must review the air quality
criteria and the NAAQS and revise them as necessary at least
once every five years. Id. § 7409(d)(1).

Each NAAQS has four components: the indicator, the
level, the averaging time, and the form. The “indicator” defines
the parameters of the substance that the EPA will measure — for
example, the size or composition of the particles to which a PM
standard will apply. The “level” specifies the acceptable
concentration of that indicator in the air. The “averaging time”
specifies the span of time across which the amount of a pollutant
in the air will be averaged. For example, some NAAQS require
a certain average annual level, while others require a certain
average daily level. The “form” of a NAAQS describes how
 8

compliance with the level will be determined within this
averaging time. A NAAQS with a daily averaging time, for
example, might require that the level not be exceeded on more
than one day each year.

Previous PM Rulemakings

In 1971, the EPA identified PM as an air pollutant under
the CAA and promulgated the first set of PM NAAQS. It
selected as the indicator “total suspended particulate” (TSP),
which is measured using a device that collects from an air
sample all PM up to a diameter of 25 to 45 µm (micrometers).
The EPA first revised the PM NAAQS in 1987. It changed the
indicator from TSP to PM10, which includes all particles with a
diameter less than or equal to 10 µm. The EPA also changed
both the level and the form of the primary annual standard,
setting the new standard at 50 µg/m3 (micrograms per cubic
meter) measured as an expected annual arithmetic (rather than
geometric) mean. It also changed the level of the 24-hour
standard from 260 to 150 µg/m3, but retained the form of one
exceedance per year. Finally, the EPA revised the secondary
standards to be identical to the new primary standards.

The next revision took place in 1997 and allowed separate
regulation of fine and coarse PM for the first time. The EPA
added a second PM indicator, PM2.5, which includes all particles
with a diameter less than or equal to 2.5 µm. It set the primary
and secondary NAAQS for fine PM at a level of 15 µg/m3 based
on a three-year average of annual arithmetic mean
concentrations, and a level of 65 µg/m3 based on a three-year
average of the 98th percentile of daily concentrations. For
coarse PM, the EPA revised the form of the primary and
secondary 24-hour PM10 standards to be based on the 99th
percentile of concentrations at each monitoring site within a
geographic area.
 9

Several parties filed petitions for review in this court
challenging the 1997 revision to the NAAQS for PM and ozone.
In May 1999, we upheld the EPA’s decision to regulate fine PM
separately from coarse PM. See Am. Trucking Ass’ns, Inc. v.
EPA (ATA I), 175 F.3d 1027, 1055–56 (D.C. Cir. 1999). We
held, however, that the selection of PM10 as the indicator for
coarse PM was arbitrary and capricious. Because a PM10
indicator includes particles that are also part of a PM2.5 indicator,
the PM10 indicator was “inherently confounded,” and the EPA
had not explained how the two standards would work together
rather than lead to overregulation of the fine fraction of PM10
and underregulation of the coarse fraction. Id. at 1054. We also
held that the EPA’s interpretation of the CAA’s requirement to
set NAAQS at the “requisite” protective level lacked an
intelligible principle to guide the agency’s decisionmaking and
therefore “effect[ed] an unconstitutional delegation of legislative
power.” Id. at 1033–34. We subsequently granted rehearing in
part to revise a section of the original opinion dealing with the
NAAQS for ozone, but made no change to our conclusions
about PM. See Am. Trucking Ass’ns, Inc. v. EPA (ATA II), 195
F.3d 4, 10 (D.C. Cir. 1999).

In February 2001, the Supreme Court reversed our holding
as to unconstitutional delegation and resolved certain other
issues not relevant to this case. See Whitman v. Am. Trucking
Ass’ns, Inc., 531 U.S. 457 (2001). On remand, we considered
the remaining challenges to the NAAQS not resolved by ATA I
and II or Whitman. We rejected each of the challenges to the
new PM2.5 NAAQS. See Am. Trucking Ass’ns, Inc. v. EPA (ATA
III), 283 F.3d 355, 358 (D.C. Cir. 2002). At the conclusion of
the litigation, therefore, the EPA’s new PM2.5 NAAQS remained
intact and its revised PM10 NAAQS were rescinded; the
preexisting 1987 PM10 NAAQS remained in place.
 10

B. 2006 Rulemaking

The EPA began the review process that resulted in the
revisions challenged here in October 1997, shortly after the 1997
NAAQS were promulgated. In 2003, several environmental
groups brought a lawsuit challenging the EPA’s failure to
complete its review within five years as required by the CAA.
Pursuant to a consent decree entered in that lawsuit, the EPA
published its proposed revisions to the PM NAAQS on January
17, 2006. The EPA proposed several changes to the existing
suite of PM standards: (1) reducing the level of the primary
daily fine PM standard from 65 to 35 µg/m3; (2) changing the
indicator for coarse PM to PM10–2.5, in order to measure only
particles with a diameter between 2.5 and 10 µm; (3) qualifying
the coarse PM indicator to include ambient mixes of PM
“dominated by” the type of particles found in urban areas while
excluding ambient mixes “dominated by” particles typical of
rural areas; (4) revoking the annual coarse PM standards; (5)
reducing the level of the primary daily coarse PM standard from
150 to 70 µg/m3 and changing the form to measure compliance
based on the 98th percentile of annual daily concentrations
averaged over three years; and (6) adjusting all secondary
standards to be identical to the revised primary standards.

The CASAC, along with medical and public health groups
who submitted comments, challenged the EPA’s proposal to
retain the existing level of the primary annual fine PM standard
at 15 µg/m3. They urged the EPA to lower the level to
somewhere between 12 and 14 µg/m3. The CASAC and several
public commenters also objected to setting the secondary
standards for fine PM at the same level and averaging time as
the primary standards, arguing that they were insufficient to
protect against adverse visibility effects. The CASAC supported
 11

the EPA’s revocation of the annual standard for coarse PM;
some commenters, however, objected.

While commentary on most of the proposals was split, the
EPA’s proposed qualification of the coarse PM indicator drew
an almost entirely negative response. Commenters challenged
the agency’s interpretation of the scientific evidence and its
conclusion that nonurban coarse PM is not associated with
adverse effects on public health. The EPA’s proposed approach
for distinguishing between urban and nonurban coarse PM came
in for criticism as well. Commenters argued that the EPA had
not adequately defined which substances were subject to
regulation, but had instead adopted an indicator defined
arbitrarily by the requirements for placing coarse PM monitors
in particular geographic areas.

The EPA issued its final rule on October 17, 2006.
Although it made no changes to its proposed fine PM standards,
the EPA reversed its position on the qualified coarse PM
indicator and retained an unqualified indicator. The EPA
selected PM10 rather than PM10–2.5 in an attempt to differentiate
indirectly between urban and rural coarse PM. As discussed
above, PM10 includes all particles with a diameter less than or
equal to 10 µm — including fine PM particles with a diameter
up to 2.5 µm. Because urban areas tend to have higher levels of
fine PM than rural areas, a coarse PM limit based on PM10
would effectively require urban areas to have lower levels of
true coarse PM — that is, particles between 2.5 and 10 µm in
diameter — than rural areas. The EPA thus believes the
standard will target protection toward urban areas, where it is
most needed. Finally, based on the recommendation by the EPA
staff and the CASAC that the existing level of protection from
coarse PM was appropriate, and on its decision to retain the
existing PM10 indicator, the EPA concluded that no change to
the level of the 24-hour coarse PM standard was necessary.
 12

C. Petitions for Review

Three sets of petitioners, joined by several intervenors and
amici, filed petitions for review of the EPA’s final rule. The
American Lung Association, Environmental Defense, and the
National Parks Conservation Association (environmental
petitioners) challenge the primary annual and secondary
standards for fine PM as well as the elimination of the annual
standard for coarse PM. Several states and state agencies (state
petitioners) challenge the primary annual fine PM standard. The
American Farm Bureau Federation, the National Pork Producers
Council, the National Cattlemen’s Beef Association, and the
Agricultural Retailers Association (industry petitioners)
challenge the EPA’s retention of the PM10 indicator for coarse
PM and the 150 µg/m3 level for the daily coarse PM standard.

We review the actions of the EPA to determine whether
they are “(A) arbitrary, capricious, an abuse of discretion, or
otherwise not in accordance with law; (B) contrary to
constitutional right, power, privilege, or immunity; [or] (C) in
excess of statutory jurisdiction, authority, or limitations.” 42
U.S.C. § 7607(d)(9). Although we perform a “searching and
careful” inquiry into the facts underlying the agency’s decisions,
we will “presume the validity of agency action as long as ‘a
rational basis for it is presented.’” ATA III, 283 F.3d at 362
(quoting Lead Indus. Ass’n v. EPA, 647 F.2d 1130, 1145 (D.C.
Cir. 1980)). We give an “extreme degree of deference to the
agency when it is evaluating scientific data within its technical
expertise,” reviewing the agency’s action to “ensure that the
EPA has examined the relevant data and has articulated an
adequate explanation for its action.” City of Waukesha v. EPA,
320 F.3d 228, 248 (D.C. Cir. 2003) (internal quotation marks
omitted). And with regard to the EPA’s interpretation of the
CAA — a statute that the agency administers — we follow the
 13

rule announced in Chevron U.S.A., Inc. v. Natural Resources
Defense Council, 467 U.S. 837 (1984). We first consider
whether the Congress has directly addressed the question at
issue. If it has, “that is the end of the matter; for the court, as
well as the agency, must give effect to the unambiguously
expressed intent of Congress.” Id. at 842–43. If the statute is
silent or ambiguous on the issue, we ask “whether the agency’s
answer is based on a permissible construction of the statute.” Id.

In Part II we grant in part the petitions for review of the
primary annual fine PM standard, brought by the states and
environmental groups, and remand the standard to the EPA for
reconsideration. The EPA failed to explain adequately why an
annual level of 15 µg/m3 is “requisite to protect the public
health,” including the health of vulnerable subpopulations, while
providing “an adequate margin of safety.” 42 U.S.C.
§ 7409(b)(1).

In Part III we grant in full the petition for review of the
secondary NAAQS for fine PM brought by the environmental
groups and remand them to the EPA for reconsideration. The
EPA unreasonably concluded that the NAAQS are adequate to
protect the public welfare from adverse effects on visibility.

In Part IV we deny the petitions for review of the primary
daily standards for coarse PM brought by the industry groups.
We do not reach the question raised by the amicus National
Association of Home Builders whether the EPA lawfully could
have distinguished between urban and nonurban coarse PM in
selecting the coarse PM indicator.

Finally, in Part V, we deny the petition for review of the
EPA’s revocation of the primary annual standard for coarse PM
brought by the environmental groups.
 14

II.

The state and environmental petitioners claim the EPA’s
decision to set the primary annual NAAQS for PM2.5 at 15 µg/m3
pursuant to its authority under § 109(b)(1) of the CAA is
“arbitrary, capricious, an abuse of discretion, or otherwise not in
accordance with law.” 42 U.S.C. § 7607(d)(9). In assessing
their arguments, we defer to the EPA’s scientific judgment while
examining the record to ensure the agency has considered the
relevant factors and reasonably explained how it reached its
conclusions. See Carus Chem. Co. v. EPA, 395 F.3d 434, 441
(D.C. Cir. 2005) (under Administrative Procedure Act, court
defers to agency’s technical expertise but asks whether it “has
‘examined relevant data and has articulated a rational
explanation for its actions’”); Chem. Mfrs. Ass’n v. EPA, 28
F.3d 1259, 1263 (D.C. Cir. 1994) (substantive review under
CAA mirrors that under APA).

We conclude the EPA failed adequately to explain why, in
view of the risks posed by short-term exposures and the
evidence of morbidity resulting from long-term exposures, its
annual standard is sufficient “to protect the public health [with]
an adequate margin of safety,” 42 U.S.C. § 7409(b)(1).
Accordingly, we grant the petitions for review in part and
remand the annual NAAQS to the EPA for reconsideration.

A. Risk of Short-Term Exposure

An agency’s failure adequately to consider a relevant and
significant aspect of a problem may render its rulemaking
arbitrary and capricious. See Chamber of Commerce of U.S. v.
SEC, 412 F.3d 133, 140 (D.C. Cir. 2005); see also Motor
Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S.
29, 43 (1983) (“agency must examine the relevant data and
articulate a satisfactory explanation for its action”). In setting
 15

the primary annual NAAQS for PM2.5, the EPA relied
exclusively upon studies of long-term exposure, placing the
greatest weight upon the Harvard Six Cities and the American
Cancer Society (ACS) studies. See 2006 Final Rule, 71 Fed.
Reg. at 61,176/2. The state and environmental petitioners argue
the EPA did not adequately explain why, even if the studies of
long-term exposure were “most directly relevant” to setting the
annual standard, id. at 61,174/3, the studies of short-term
exposure were not at all relevant.

In the EPA’s view, studies of long-term exposure are most
directly relevant to the annual standard because the human body
responds differently to long-term than to short-term exposure
and because a long-term study evaluates “periods of exposure”
closer to a year than the periods evaluated in a short-term study.
See id.; Proposed Rule: National Ambient Air Quality Standards
for Particulate Matter, 71 Fed. Reg. 2620, 2627/2–3 (2006); U.S.
ENVTL. PROT. AGENCY, AIR QUALITY CRITERIA FOR
PARTICULATE MATTER (CRITERIA DOCUMENT) § 8.2.1, at 8-18
(2004). Be that as it may, the EPA did not assert that short-term
studies provide no relevant information. The EPA concluded
merely that it would be “more appropriate to consider the short-
term exposure studies as a basis for . . . the 24-hour standard and
to consider the long-term exposure studies as a basis for the . . .
annual standard.” 2006 Final Rule, 71 Fed. Reg. at 61,174/2.

Two premises underlie this conclusion: the agency can (1)
“appropriately . . . evaluate[]” where to set the level of the
annual standard based solely upon the long-term studies and (2)
“provide an appropriate degree of protection” from short-term
exposure through a daily standard alone. Id. at 61,174/3. The
petitioners challenge both.

As to the first, the petitioners start by pointing out that the
staff and the CASAC, both of which relied upon three short-
 16

term studies, proposed an annual standard lower than 15 µg/m3.
See OFFICE OF AIR QUALITY PLANNING AND STANDARDS, U.S.
ENVTL. PROT. AGENCY, REVIEW OF THE NATIONAL AMBIENT
AIR QUALITY STANDARDS FOR PARTICULATE MATTER: POLICY
ASSESSMENT OF SCIENTIFIC AND TECHNICAL INFORMATION
(STAFF PAPER) § 5.3.1.1, at 5-7 (2005); Letter from Dr. Rogene
Henderson, CASAC, to Administrator Stephen L. Johnson, EPA
3–4 (Mar. 21, 2006) (“Studies described in the PM Staff Paper
indicate that short term effects of PM2.5 persist in cities with
annual PM2.5 concentrations below [15 µg/m3]”). By statute the
EPA must explain its rejection of the CASAC’s
recommendation, 42 U.S.C. § 7607(d)(3), and the staff’s
analysis is something we consider when determining whether
the EPA has adequately addressed the relevant considerations
and reasonably reached its conclusions, see Natural Res. Def.
Council v. EPA, 902 F.2d 962, 967–68, 970 (D.C. Cir. 1990)
(noting staff paper’s relevance as “bridge” over “gap” between
criteria document and EPA’s policy judgment).*

We reject the EPA’s initial response, to wit, that its
approach is consistent with those recommendations. See 2006
Final Rule, 71 Fed. Reg. at 61,174/3 n.45. The CASAC directly
challenged the EPA’s proposal to retain the annual standard at
15 µg/m3. See Letter from Dr. Rogene Henderson, CASAC, to
Administrator Stephen L. Johnson, EPA 7 (June 6, 2005). The
staff recommended that if the agency kept the annual standard
at 15 µg/m3 while setting the daily standard with a 98th
percentile form, then it should set the daily standard at the
“middle to lower end” of the range from 25 to 35 µg/m3. See
*
We will not consider the letter the CASAC sent to the EPA on September
29, 2006, shortly after the Administrator had signed the final rule, as it was
not part of the administrative record. See Citizens to Preserve Overton
Park, Inc. v. Volpe, 401 U.S. 402, 420 (1971) (generally judicial review is
limited to “record that was before the [agency] at the time [it] made [its]
decision”).
 17

STAFF PAPER § 5.3.7, at 5-46; see also id. § 5.3.5.1, at 5-32
(“[S]taff continues to believe that an annual standard cannot be
expected to offer an adequate margin of safety against the
effects of all short-term exposures”). The EPA instead set the
daily standard at 35 µg/m3 with a 98th percentile form, see 2006
Final Rule, 71 Fed. Reg. at 61,165/2; id. at 61,171/3, while
acknowledging it did not consider the short-term studies when
setting the annual standard. As both the CASAC and the staff
reasoned, the studies of Eight Canadian Cities, of Santa Clara
County (CA), and of Phoenix are relevant to setting an annual
standard because each reports adverse health effects associated
with short-term exposures in places where the annual PM2.5
concentration is below 15 µg/m3. See CASAC Mar. 21 Letter at
3–4. The EPA failed adequately to explain its reason for not
accepting the CASAC’s recommendations, instead stating only
that it did not “disagree with CASAC’s factual statements
regarding the findings of [the short-term studies],” but “believes
. . . it . . . more appropriate to consider the short-term exposure
studies as a basis for the level of the 24-hour standard.” 2006
Final Rule, 71 Fed. Reg. at 61,174/2.

The petitioners also point to the EPA’s unexplained change
of position since it last dealt with this issue, which was in 1997.
Then the EPA believed short-term studies were indeed relevant
to the setting of an annual standard, stating, “the strongest
evidence for short-term PM2.5 effects occurs at concentrations
near the long-term (e.g., annual) average.” See Final Rule:
National Ambient Air Quality Standards for Particulate Matter,
62 Fed. Reg. 38,652, 38,676/1 (1997). The agency did not
thereby commit itself irrevocably, however if the relevant facts
have changed or the EPA has reasonably made a different policy
judgment, then it need only explain itself and we will defer. See
State Farm, 463 U.S. at 57 (“An agency’s view of what is in the
public interest may change, either with or without a change in
circumstances” (internal quotation marks omitted)).
 18

The EPA’s only explanation for its change of position is
that it relied upon short-term studies in 1997 because the studies
of long-term exposure then available were less reliable than the
short-term studies. See 2006 Final Rule, 71 Fed. Reg. at
61,174/3. At the time, however, the agency also said, “an
annual standard that controls an area’s attainment status is likely
to reduce aggregate risks associated with both short- and long-
term exposures with more certainty than a 24-hour standard.”
1997 Final Rule, 62 Fed. Reg. at 38,670/3. Even if the long-
term studies available today are useful for setting an annual
standard, therefore, it is not clear why the EPA no longer
believes it useful to look as well to short-term studies in order to
design the suite of standards that will most effectively reduce the
risks associated with short-term exposure.

We conclude the EPA failed adequately to explain its first
premise, viz. that an annual standard could be “appropriately . . .
evaluated based” solely upon long-term studies. See 2006 Final
Rule, 71 Fed. Reg. at 61,174/3. Therefore, we need not reach
the state petitioners’ argument that the EPA should have relied
in part upon the short-term studies because, unlike the ACS
Study, those studies do not underestimate the magnitude of risk
from exposure to PM2.5. That argument does not rest upon the
ACS Study having examined long-term rather than short-term
exposure, but instead questions whether the ACS Study involved
a representative cohort. If, however, the EPA can adequately
explain why studies of short-term effects are not relevant to
setting an annual standard, then it may disregard those studies
regardless whether the short-term studies are based upon a more
representative cohort than is the ACS Study.

Nor do we consider the EPA’s argument that the short-term
studies showed an annual standard would not provide more
protection against short-term exposure than would a daily
 19

standard. That argument was advanced for the first time in its
brief and therefore is not properly before us. See SEC v.
Chenery Corp., 332 U.S. 194, 196 (1947) (“[A] reviewing court
. . . must judge . . . [administrative] action solely by the grounds
invoked by the agency” when it acted); Ashland Oil, Inc. v. FTC,
548 F.2d 977, 981 & n.6 (D.C. Cir. 1976) (court cannot rely
upon post hoc position unless “no special agency expertise is
involved”).

The petitioners also challenge the EPA’s second premise,
that a primary daily standard for PM2.5 at 35 µg/m3 with a 98th
percentile form could “provide an appropriate degree of
protection” from short-term exposure, thus eliminating the need
to calibrate the annual standard to address short-term exposure.
2006 Final Rule, 71 Fed. Reg. at 61,174/3; see id. at 61,165/2
(selecting percentile form of standard); id. at 61,171/3 (selecting
level of standard). The EPA’s daily standard will require each
state to make sure that the average of the seventh highest
concentration measured each year for three years does not
exceed 35 µg/m3, thus permitting approximately six days a year
to have concentrations above that level. See id. at 61,164–65
(describing form of standard). The petitioners argue the EPA’s
reliance upon the daily standard to address all the health risks
from short-term exposure was unreasonable for several reasons.

First, the environmental petitioners contend the record
shows that low- to mid-range concentrations of PM2.5 cause most
health effects from short-term exposure, noting that both the
staff and the CASAC warned of risks at ambient concentrations
below 35 µg/m3. See STAFF PAPER § 5.3.5.1, at 5-31–32;
CASAC June 6 Letter at 7. In particular, the CASAC observed
that, because some cities have “relatively high annual PM
concentrations” but would “rarely” exceed ambient
concentrations of 35 µg/m3, it would be necessary to lower the
annual standard below 15 µg/m3 to provide sufficient protection
 20

from the risks associated with short-term exposure. CASAC
June 6 Letter at 7. The EPA maintains lowering the daily
standard to 35 µg/m3 will provide additional protection in all
areas where it is required, but as we discuss below, it is not clear
why the EPA believes areas that already meet the daily standard
either do not require additional protection or will enjoy
improved protection against short-term exposure.

Second, the environmental petitioners question the EPA’s
methodology. The EPA focused upon the 98th percentile air
quality values in the short-term studies, concluding those values
would shed light upon which overall distributions of air quality
likely cause harm from short-term exposure. Because some
studies did not find a statistically significant association between
short-term exposure and health effects when the 98th percentile
values were below 35 µg/m3, the EPA selected 35 µg/m3 as the
level, and the 98th percentile as the form, of the primary daily
standard. 2006 Final Rule, 71 Fed. Reg. at 61,169. The EPA
does not contend, however, that lowering the peak
concentrations to 35 µg/m3 would provide sufficient protection.
Indeed, the EPA acknowledges that to provide adequate
protection against short-term exposure, it must lower “a broad
distribution of PM2.5 air quality values in an area,” but explains
that lowering the daily standard will lower concentrations not
only at the peak of the distribution curve, but also those at the
trough because state implementation plans, “[a]lthough . . . not
yet defined, . . . are likely to” require the necessary
improvements in air quality. Id. at 61,168/3.

In support of this prediction the EPA points to a study
finding that as peak daily levels of PM2.5 decreased from “one
year to the next” in Philadelphia and Los Angeles, all off-peak
air quality values, or “low daily PM2.5 levels[,] decreased
proportionally.” See ABT ASSOCS., INC., PARTICULATE MATTER
HEALTH RISK ASSESSMENT FOR SELECTED URBAN AREAS (RISK
 21

ASSESSMENT) 17 (2005); see also STAFF PAPER § 4.3.1.2, at 4-
18 (study showed “concentrations at different points in the
distribution of 24-hour PM2.5 values . . . decreased by
approximately the same percentage”). Although the EPA is
entitled to rely upon its experience, it must have a reasonable
explanation of how its experience supports its conclusion. See
DSE, Inc. v. United States, 169 F.3d 21, 30 (D.C. Cir. 1999)
(court “will defer to [agency’s] experience provided that the
agency has offered a reasoned explanation . . . [and] the result is
in accord with material facts . . . in the administrative record”).
The environmental petitioners rightly question the EPA’s
conclusion based upon past trends in Philadelphia and Los
Angeles; according to the same study, they point out, the
proportional decline in peak and off-peak daily levels of PM2.5
in those cities was “not . . . the result of a PM2.5 control strategy,
but likely result[ed] from control programs for PM10 and control
programs for other pollutants.” RISK ASSESSMENT at 17. The
EPA offers no basis in the experience of those two cities — thus
understood — for expecting controls on the peak levels of fine
particles will lead to a proportional decline in levels of PM2.5 at
off-peak times simply because controls for coarse particles and
for other pollutants did so in the past. The EPA may not and
perhaps need not fully understand why off-peak PM2.5 levels
decreased as peak PM2.5 levels decreased in the past, but it
nevertheless needs plausibly to explain why it believes future
controls for PM2.5 will, like other controls for other pollutants in
the past, trigger the phenomenon.

Even assuming off-peak levels of PM2.5 will decrease in
proportion to decreases at peak levels, the environmental
petitioners argue there will be no such decline in areas that
already comply with the daily standard and therefore will not
have to reduce ambient daily concentrations. It is not enough
for the EPA to respond that, as it implied during the rulemaking,
those areas do not need further improvement; the agency itself
 22

observed that short-term health effects occur across a “broad
distribution of PM2.5 air quality values.” See 2006 Final Rule, 71
Fed. Reg. at 61,168/3.

We conclude the EPA has failed reasonably to explain why
it believes its daily standard will “provide an appropriate degree
of protection from health effects associated with short-term
exposures to PM2.5.” Id. at 61,174/3. We therefore remand the
annual standard to the EPA for further consideration of whether
it is set at a level requisite to protect the public health while
providing an adequate margin of safety from the risk of short-
term exposure to PM2.5.

B. Morbidity Among Vulnerable Subpopulations

The state petitioners believe the EPA unreasonably
concluded morbidity studies did not provide a basis for setting
a primary annual standard for PM2.5 below 15 µg/m3. First, they
argue the EPA arbitrarily refused to rely upon a study finding an
association between irreversible lung damage in children and
long-term exposure to PM2.5 at levels below 15 µg/m3. See W.
James Gauderman et al., Association Between Air Pollution and
Lung Function Growth in Southern California Children, 162
AM. J. RESPIRATORY & CRITICAL CARE MED. 1383 (2000). The
staff thought the Gauderman Study, if given “appreciable
weight,” indicated an annual standard of 13 µg/m3. See STAFF
PAPER § 5.3.4.1, at 5-22–23. Although the CASAC did not
advert to the Gauderman Study, it recommended a level of 13 to
14 µg/m3 based upon the risk assessment and the short-term
studies. See CASAC Mar. 21 Letter at 1. The EPA
acknowledged the Gauderman Study presented “important new
findings” but decided not to rely upon it because it was the only
study measuring “decreased growth in lung function” over time
and it focused solely upon children in one location. 2006 Final
Rule, 71 Fed. Reg. at 61,172/2-3.
 23

The state petitioners complain that the EPA’s rejection of
the Gauderman Study was based upon the faulty premise that
“further study . . . would be needed to increase confidence in the
reported associations” because no other study in the record was
consistent with the results of the Gauderman Study. Id. at
61,172/3. According to the state petitioners, the findings of the
Gauderman Study are supported by Mark Raizenne et al., Health
Effects of Acid Aerosols on North American Children:
Pulmonary Function, 104 ENVTL. HEALTH PERSP. 506 (1996)
(24-Cities Study). The EPA disagreed: the 24-Cities Study
measured decreased lung function “at a single point in time”
while the Gauderman Study found “decreased growth in lung
function” over time. 2006 Final Rule, 71 Fed. Reg. at 61,172/3.

We believe the EPA’s approach to the Gauderman and 24-
Cities studies was unreasonable in light of the agency’s
obligation to explain how the annual standard it set would
protect “not only average healthy individuals, but also ‘sensitive
citizens.’” Am. Lung Ass’n v. EPA, 134 F.3d 388, 389 (D.C. Cir.
1998). As the American Lung Association argued during the
rulemaking — and the EPA agreed — with each breath a child
aspirates more pollution relative to its body weight than does an
adult. See American Lung Association (ALA) et al., Comments,
2006 Rulemaking, Pub. Dkt. No. OAR-2001-0017-1890.1, at 17
(filed Apr. 17, 2006); 2006 Proposed Rule, 71 Fed. Reg. at
2637/1 (“Several factors may make children susceptible to PM-
related effects, including the greater ventilation rate per
kilogram body weight in children”). We therefore doubt
whether the studies of adult mortality upon which the EPA
relied provide the necessary confidence concerning the risk of
morbidity in children.

We doubt the reasonableness of the EPA’s analysis for a
second reason: the agency imputed undue significance to one
 24

difference, despite many similarities, between the Gauderman
Study, which focused upon the growth over time of lung
function in children, and the 24-Cities Study, which focused
upon the level of lung function of children at a particular time.
Both studies found an association between exposure to PM2.5 at
levels below 15 µg/m3 and impairment of lung function in
children, although the association in the 24-Cities Study was
“not statistically significant” below 15 µg/m3, 2006 Final Rule,
71 Fed. Reg. at 61,172/3. Indeed, the 24-Cities Study predicted
the children it examined might “continue on [a] track [of
reduced] . . . growth of their lung function, as suggested by
previous studies,” Raizenne et al. at 513, which was consistent
with the findings of the later Gauderman Study, see Gauderman
et al. at 1388. The EPA also relied upon its 1997 conclusion
that the 24-Cities Study does not show a statistically significant
association between health effects and annual average
concentrations of PM2.5 below 15 µg/m3. See 2006 Final Rule,
71 Fed. Reg. at 61,172/3. Viewed in isolation, of course, the
studies are far from conclusive. Viewed together in the context
of the studies the EPA considered when deciding whether to
revise the standard for PM2.5, however, the conclusion reached
after comprehensive scientific review by the EPA’s National
Center for Environmental Assessment in the Criteria Document
seems the only reasonable one: the findings of the Gauderman
and 24-Cities studies are related and together indicate a
significant public health risk. See CRITERIA DOCUMENT
§ 8.4.6.4, at 8-314; cf. 2006 Final Rule, 71 Fed. Reg. at 61,154/1
(“[T]he Criteria Document found that new studies of a cohort of
children in Southern California have built upon earlier limited
evidence to provide fairly strong evidence that long-term
exposure to fine particles is associated with development of
chronic respiratory disease and reduced lung function growth”).
On this record, therefore, it appears the EPA too hastily
discounted the Gauderman and 24-Cities studies as lacking in
significance. See Am. Radio Relay League, Inc. v. FCC, 524
 25

F.3d 227, 241 (D.C. Cir. 2008) (agency’s inadequate explanation
for dismissing empirical studies rendered decision arbitrary and
capricious); cf. ATA I, 175 F.3d at 1052–53 (EPA arbitrarily and
capriciously placed upon some studies “higher information
threshold” than it placed upon others).

Second, the state petitioners argue the EPA should have
explained how an annual standard of 15 µg/m3 would provide an
adequate margin of safety for vulnerable subpopulations, such
as children, the elderly, or those with conditions that expose
them to greater risk from fine particles. The EPA said there was
insufficient evidence that vulnerable subpopulations will be
exposed to adverse health effects at annual levels of PM lower
than 15 µg/m3. See 2006 Final Rule, 71 Fed. Reg. at 61,173/1.
Having concluded the EPA failed adequately to consider the
Gauderman and 24-Cities studies, however, we are constrained
to agree with the state petitioners that the EPA was unreasonably
confident that, even though it relied solely upon long-term
mortality studies, the revised standard would provide an
adequate margin of safety with respect to morbidity among
children. Notably absent from the final rule, moreover, is any
indication of how the standard will adequately reduce risks to
the elderly or to those with certain heart or lung diseases despite
(a) the EPA’s determination in its proposed rule that those
subpopulations are at greater risk from exposure to fine particles
and (b) the evidence in the record supporting that determination.
See 2006 Proposed Rule, 71 Fed. Reg. at 2637/1; STAFF PAPER
§ 3.3.2, at 3-19–22 & fig.3-2 (charting risk of morbidity for
persons with various diseases, but noting associations between
exposure to PM2.5 and certain heart diseases and stroke are not
“statistically significant”); CRITERIA DOCUMENT § 8.4.9, at 8-
327–28 (citing studies discussing risks to “older adults and
children”); ALA Comments at 52 (discussing risk to elderly at
level of 12 µg/m3). We therefore grant the petition for review
and remand this matter for the EPA to explain why it believes
 26

the NAAQS will provide, as required by the CAA, an adequate
margin of safety against morbidity in children and other
vulnerable subpopulations. See Am. Lung Ass’n, 134 F.3d at
393 (remanding because EPA had “failed adequately to explain
[why there is] . . . no public health threat” to asthmatics).

C. Other Challenges

Both groups of petitioners object also to the EPA’s analysis
of the long-term studies; the state petitioners, furthermore, claim
it was unreasonable of the agency not to rely upon the risk
assessment. We find neither objection persuasive.

First, the petitioners argue the EPA unreasonably focused
upon the long-term mean ambient concentrations of PM2.5 in the
ACS Study (17.7 µg/m3) and the Six Cities Study (18 µg/m3)
and then set a level below those concentrations to address long-
term exposure. See 2006 Final Rule, 71 Fed. Reg. at
61,176/2–3. The state and environmental petitioners claim this
approach violates the requirement of the CAA that the NAAQS
provide an “adequate margin of safety,” 42 U.S.C. § 7409(b)(1),
which requires that the EPA “err on the side of caution.” Lead
Indus., 647 F.2d at 1154–55. Because the most recent data from
the ACS and the Six Cities studies showed adverse health effects
in years when the mean ambient concentration of PM2.5 was
below 15 µg/m3, see DANIEL KREWSKI ET AL., REANALYSIS OF
THE HARVARD SIX CITIES STUDY AND THE AMERICAN CANCER
SOCIETY STUDY OF PARTICULATE AIR POLLUTION AND
MORTALITY: A SPECIAL REPORT OF THE INSTITUTE’S PARTICLE
EPIDEMIOLOGY REANALYSIS PROJECT 175 (2000), the
petitioners argue the EPA had to lower the level below 15 µg/m3
in order to address the problem created by long-term exposure.
They point to the staff’s conclusion that the ACS and the Six
Cities studies indicated an annual standard “somewhat below 15
µg/m3.” STAFF PAPER § 5.3.4.1, at 5-23. As the staff and the
 27

authors of the two studies explained, however, “it is not easy to
differentiate the role of historic exposures from more recent
exposures.” See id. § 3.6.5.4, at 3-53. In other words, a person
exposed to relatively high concentrations of PM2.5 in the past
may suffer long-term effects in the present even if he is exposed
to relatively low concentrations of PM2.5 in the present. The
EPA, mindful of its obligation to set a standard “not lower or
higher than is necessary . . . to protect the public health,”
Whitman, 531 U.S. at 475–76, reasonably decided to address
long-term exposure with an annual standard somewhat below
the long-term mean concentrations in the ACS and Six Cities
studies. See 2006 Final Rule, 71 Fed. Reg. at 61,172–73. We
approved a similar approach to assuring an adequate margin of
safety in ATA III and we do so again here. See 283 F.3d at 372
(denying petition for review of standard set “just below the
range of mean annual PM2.5 concentrations observed in studies
showing a statistically significant association between fine [PM]
and health effects”). We therefore deny the petitions for review
as to the EPA’s analysis of the key long-term studies.

Second, the state petitioners object on four grounds to the
EPA’s decision not to rely upon a quantitative risk assessment,
prepared for this rulemaking by Abt Associates, Inc. and
analyzed and summarized in the EPA’s staff paper, which
estimated how variations in the ambient concentration of PM
affect the incidence of adverse health effects. See 2006 Final
Rule, 71 Fed. Reg. at 61,171/2–3. First, if a threshold of 7.5
µg/m3 is assumed as the level below which exposure to PM is
not harmful, then the risk assessment predicts the EPA’s
standards will result in 3,700 premature deaths in the nine cities
studied, whereas lowering the primary annual standard to 12
µg/m3, as the petitioners suggest the EPA should do, may reduce
that number by 80%. See RISK ASSESSMENT at 117, E-17–31;
see also ALA Comments at 40 fig.4 (synthesizing data). The
EPA reasonably responds, however, that the comparison
 28

between the petitioners’ preferred NAAQS and the one the EPA
selected is based upon unreliable assumptions; in particular, the
risk assessment is “based on studies that do not resolve the issue
[whether there is] a threshold” below which PM2.5 is harmless.
See 2006 Final Rule, 71 Fed. Reg. at 61,168/2. The risk
assessment therefore “necessarily predicts that ever lower
standards result in ever lower risks,” which is not supported by
statistically significant epidemiological evidence. Id.

The state petitioners also claim the EPA is inconsistent
because it relied upon a risk assessment when setting the annual
standard for PM in 1997. Although the EPA stated the risk
assessment then before it contained “reasonable estimates . . .
given the available information,” it did not in fact use those
estimates to set the annual standard. See 1997 Final Rule, 62
Fed. Reg. at 38,656/2–3. Indeed, the EPA expressed the same
concern about the earlier risk assessment that it has about the
new one. We deferred to its decision then not to rely upon the
risk assessment in setting the level of the NAAQS. See ATA III,
283 F.3d at 374 (“EPA persuasively explains that it discounted
the quantitative predictions of the risk assessment in light of
‘inherent scientific uncertainties,’ including the ‘possibility of
. . . thresholds’ below which PM2.5 has little or no effect”).

The state petitioners nonetheless argue we should not
similarly defer now because the current risk assessment is more
reliable than the one available in 1997; this time the staff used
“only health endpoints for which . . . the overall weight of the
evidence supports the conclusion that PM2.5 is likely causally
related.” STAFF PAPER § 4.3.4, at 4-34. They note both the
CASAC and the staff thought the risk assessment was reliable.
See CASAC Mar. 21 Letter at 3; STAFF PAPER § 5.3.7, at 5-46.
Even so we must defer to the EPA’s assessment of “scientific
data within its technical expertise” as long as the agency has
examined the data and adequately explained itself. City of
 29

Waukesha, 320 F.3d at 247 (internal quotation marks omitted).
Here, unlike in its analysis of vulnerable subpopulations, the
EPA considered all aspects of the problem, catalogued its
concerns, and took the same decision — not to rely upon the risk
assessment — it took before; we deferred to that decision before
and we do so again now.

Finally, the state petitioners argue the EPA contradicted
itself when it relied upon the ACS and the Six Cities studies but
not upon the risk assessment, which was based in part upon
those studies. The record, however, indicates the risk
assessment magnified the uncertainties inherent in the
underlying studies. See STAFF PAPER § 4.3.4, at 4-34 (reliance
of risk assessment upon single-city studies in addition to multi-
city studies introduces additional uncertainty). Moreover, the
EPA’s decision not to rely upon the lower mean concentrations
in the multi-city studies is consistent with its decision that the
risk assessment overestimated the risk of harm from lower
concentrations. We cannot say the EPA’s distinction between
the studies and the risk assessment was unreasonable.

Because we believe the EPA reasonably analyzed the risk
assessment, we deny the petition for review in that respect. We
do not reach the industry intervenors’ defense of the EPA’s
action — except to note it was repudiated by the agency. See
State Farm, 463 U.S. at 50 (court can look only to “basis
articulated by the agency”); EPA, Responses to Significant
Comments, 2006 Rulemaking, Pub. Dkt. No. OAR-2001-0017-
3203, at 166–67 (filed Sept. 25, 2006).

D. Conclusion

In sum, the EPA did not adequately explain why an annual
level of 15 µg/m3 is sufficient to protect the public health while
providing an adequate margin of safety from short-term
 30

exposures and from morbidity affecting vulnerable
subpopulations. We therefore grant in part the petitions for
review filed by the States and by the environmental petitioners
(Nos. 06-1416 and 06-1411) and remand for reconsideration the
primary annual NAAQS for PM2.5. We deny the same petitions
insofar as they seek review of the EPA’s analysis of the long-
term mortality studies and the agency’s decision not to rely upon
the risk assessment. We do not, however, vacate the annual
standard. First, the EPA’s failure adequately to explain itself is
in principle a curable defect. Second, vacating a standard
because it may be insufficiently protective would sacrifice such
protection as it now provides, making the best an enemy of the
good. See Allied-Signal, Inc. v. U.S. Nuclear Regulatory
Comm’n, 988 F.2d 146, 150–51 (D.C. Cir. 1993) (when
selecting remedy court should consider: “the seriousness of the
[rule’s] deficiencies . . . and the disruptive consequences of an
interim change that may itself be changed” (internal quotation
marks omitted)).

III.

The environmental petitioners challenge as arbitrary and
capricious the EPA’s decision to set the secondary NAAQS for
fine PM, which protect the public welfare from adverse visibility
effects, at the same level as the primary NAAQS, which protect
public health.

The EPA staff and the CASAC recommended a secondary
standard level of 20 to 30 µg/m3 of PM2.5, averaged across a 4-
or 8-hour midday period each day, and using the 92nd to 98th
percentile of these daily measurements. Both groups explained
that this standard would be shielded from the confounding
effects of humidity. Because both humidity and fine PM can
impair visibility, measuring PM concentration only at midday,
when humidity is lowest, would ensure that the standard is
 31

targeting visibility effects caused by fine PM rather than by
humidity. The EPA staff and the CASAC based the
recommended ranges of levels and percentile forms on studies
in which participants rated the acceptability of visibility levels
as depicted in actual photographs or computer-generated
simulations. In each study, a majority of participants rated
visibility ranges of 40 to 60 kilometers as “acceptable.” The
EPA staff explained that the recommended standard was
“consistent with” this target range.

The EPA rejected these recommendations. Although it
agreed with its staff and the CASAC that a sub-daily averaging
time had “strong technical merit,” the EPA characterized the
evidence supporting the recommended level and form as
“limited and uncertain,” a conclusion based on “the generally
subjective nature of the public welfare effect involved.” 2006
Final Rule, 71 Fed. Reg. at 61,206–08. Having rejected the
recommended target level of visibility protection, the EPA did
not identify its own. Instead, it first considered the extent to
which the revised primary NAAQS would protect visibility and
compared this with the protection that would be provided by the
recommended secondary standard.

To make this comparison, the EPA looked to two tables in
the Staff Paper predicting the percentage of counties where the
existing air quality was unlikely to meet several potential
standards. One table showed that 27% of counties contained
PM monitors with a recent history of recorded fine PM levels
that would not meet the new primary PM NAAQS. The second
table showed that 24% of counties contained PM monitors with
a record of fine PM levels that would not meet one possible
standard within the range recommended by the EPA staff and
the CASAC. From this data, the EPA concluded that the
primary NAAQS would be more protective of visibility than the
standard recommended by its staff and the CASAC. Because a
 32

greater percentage of counties would fail to meet those
standards, more counties would be required under the CAA to
reduce fine PM levels and, incidentally, improve visibility.
Accordingly, the EPA adopted secondary fine PM NAAQS
identical to the primary fine PM NAAQS.

The petitioners argue that the EPA’s decision lacks a
reasoned basis. First, they assert that the EPA never determined
what level of visibility is “requisite to protect the public
welfare.” 42 U.S.C. § 7409(b)(2). The petitioners argue that
the EPA unreasonably rejected the starting point provided by its
staff — visibility of 40 to 60 kilometers — based on the inherent
“uncertainty” in subjective value judgments about visibility.
Second, the petitioners challenge the EPA’s method of
comparing the protection expected from potential standards.
They contend that the EPA relied on a meaningless numerical
comparison, ignored the effect of humidity on the usefulness of
a standard using a daily averaging time, and unreasonably
concluded that the primary standards will achieve a level of
visibility roughly equivalent to the level the EPA staff and the
CASAC deemed “requisite to protect the public welfare.” Id.

The EPA responds that it did not need to identify a specific
level of visibility to achieve because the evidence did not
support a need for secondary standards to be more stringent than
the primary standards. Moreover, the EPA argues that it was
reasonable to reject the target level of visibility recommended
by the Staff Paper because it relied on the subjective
perspectives of study participants. The EPA characterizes this
subjective evidence as “uncertain” because the studies could not
identify the precise level or percentage of days of visibility
impairment at which there is an adverse effect on public welfare.
Such uncertain evidence, the EPA argues, cannot substitute for
the sound policy judgment of the EPA Administrator. Finally,
the EPA defends its reliance on the comparative analysis of the
 33

protection afforded by different standards: it claims it did not
rely solely on the finding of equivalent protection, but also
found that the primary standards were in fact within the range
recommended by the EPA staff and the CASAC.

The EPA’s assertion that it need not determine what level
of visibility protection is requisite to protect the public welfare
fails under the plain language of the statute. The CAA provides:
“Any national secondary ambient air quality standard shall
specify a level of air quality the attainment and maintenance of
which . . . is requisite to protect the public welfare from any
known or anticipated adverse effects . . . .” 42 U.S.C.
§ 7409(b)(2) (emphases added). The EPA’s failure to identify
such a level when deciding where to set the level of air quality
required by the revised secondary fine PM NAAQS is contrary
to the statute and therefore unlawful. Furthermore, the failure
to set any target level of visibility protection deprived the EPA’s
decisionmaking of a reasoned basis. Because the EPA failed to
identify any target level, we need not decide whether it was
reasonable for the agency to reject the target recommended by
the Staff Paper and the CASAC because it was based on
uncertain subjective evidence.

The EPA’s substitute for identifying a target level — its
analysis of the relative protection expected from the
recommended standards and the revised primary NAAQS —
cannot save its decision. The equivalence analysis fails on its
own terms. The EPA compared the predicted 27% of counties
that would not meet the revised primary fine PM standards with
the predicted 24% that would not meet a standard of 30 µg/m3
using a 4-hour averaging time and a 95th percentile form — one
standard within the range recommended by the CASAC. The
EPA concluded that because fewer counties would meet the
primary standards, and would therefore have to take steps to
reduce fine PM levels, that standard was more protective than
 34

the one recommended by the CASAC. But the same table
shows that six other standards within the recommended range
would be more “protective” under the EPA’s definition than the
primary standards: from 47% to 85% of counties would not
meet these alternative standards. The alternative cited by the
EPA is one of only three standards that would offer equal or
greater protection than the primary standards. In other words,
two-thirds of the potential standards within the CASAC’s
recommended range would be substantially more protective than
the primary standards. The EPA failed to explain why it looked
only at one of the few potential standards that would be less
protective — and only slightly so — than the primary standards.

More fundamentally, however, the EPA’s equivalence
analysis demonstrates nothing about the relative protection
offered by the different standards. The CASAC put the EPA on
notice of this fact in a letter submitted in response to the
proposed revisions to the NAAQS: “The cited comparability
between percentages of counties not likely to meet a lenient sub-
daily secondary standard and the proposed 24-hour primary
standard is a numerical coincidence.” CASAC Mar. 21 Letter
at 5. Indeed, the tables both state that the “estimates are not
based on the same air quality data that would be used to
determine whether an area would attain a given standard or set
of standards” and “should be interpreted with caution.” STAFF
PAPER app. 5B at 2, app. 7A at 2. The tables simply offer no
valid information about the relative visibility protection
provided by the standards, and yet the EPA relied on them
almost exclusively in selecting the secondary NAAQS for fine
PM.

Finally, as the Staff Paper makes clear, a visibility standard
using a daily averaging time will be confounded by regional
differences in humidity. The EPA acknowledged this problem,
recognizing that a sub-daily averaging time, using the daylight
 35

hours when humidity tends to be lowest, has “strong technical
merit.” 2006 Final Rule, 71 Fed. Reg. at 61,208. Yet the EPA’s
equivalence analysis — its basis for concluding that the primary
standards would be sufficiently protective of visibility — did not
address this issue at all.

The EPA’s decision to set secondary fine PM NAAQS
identical to the primary NAAQS was unreasonable and contrary
to the requirements of 42 U.S.C. § 7409(b)(2). Accordingly, we
grant the petition for review (No. 06-1411) and remand for
reconsideration the secondary NAAQS for fine PM.

IV.

The industry petitioners lodge three challenges to the
EPA’s regulation of coarse PM. First, they challenge the EPA’s
decision to retain a daily standard for all coarse PM, including
that in nonurban areas. Second, they claim that the 150 μg/m3
daily NAAQS that the EPA set is not requisite to protect public
health under Whitman v. American Trucking, 531 U.S. 457.
Finally, they argue that the PM10 indicator that the EPA chose to
retain for coarse PM is confounded and impermissible under this
court’s decision in ATA I, 175 F.3d 1027. We find the EPA’s
decisions on all three issues to be reasonable, and we therefore
deny these petitions (Nos. 06-1410, 06-1415, and 06-1417).

A. Retention of a Daily Standard for All Coarse PM

In their challenge to the EPA’s decision to maintain a daily
NAAQS for all — not just urban — coarse PM, the industry
petitioners make two arguments, one contesting the EPA’s
authority to regulate nonurban coarse PM and the other
contesting its evidence of coarse PM’s dangerousness.

First, the industry petitioners contend that, “to promulgate
a NAAQS for nonurban PM coarse, the EPA must show that
 36

such particles present a significant risk to public health.”
Industry Reply Br. 4. To support this contention, the petitioners
rely on 42 U.S.C. § 7408(a)(1), which states:

For the purpose of establishing national primary and
secondary ambient air quality standards, the Administrator
shall . . . publish, and shall from time to time thereafter
revise, a list which includes each air pollutant — (A)
emissions of which, in his judgment, cause or contribute to
air pollution which may reasonably be anticipated to
endanger public health or welfare.

On the basis of this section, the petitioners argue that the EPA
lacks authority to establish a NAAQS for coarse PM that
includes nonurban coarse PM unless the agency first makes a
finding that such nonurban coarse PM “may reasonably be
anticipated to endanger public health or welfare.” Id.

The petitioners misunderstand § 7408(a)(1). Under that
section, the EPA must make a dangerousness finding in order to
place a pollutant on the list of pollutants regulated by NAAQS.
But that requirement to find dangerousness does not apply when
the EPA sets the NAAQS for a pollutant already on the list. At
issue in this review is not the listing of PM, but rather the review
and revision of the PM NAAQS that was mandated by
§ 7409(d)(1). See id. § 7409(d)(1) (“Not later than December
31, 1980, and at five-year intervals thereafter, the Administrator
shall complete a thorough review of the criteria published under
section 7408 of this title and the national ambient air quality
standards promulgated under this section and shall make such
revisions in such criteria and standards and promulgate such
new standards as may be appropriate in accordance with section
7408 of this title and subsection (b) of this section”).

In ATA I, this court confirmed that the establishment of an
 37

indicator for PM2.5 was not a new listing under § 7408(a)(1)
because PM2.5 had previously been regulated as part of the PM10
indicator. ATA I, 175 F.3d at 1055. Similarly, nonurban PM10
has been regulated under the general PM10 standard, and
therefore the EPA’s continued regulation of all PM10 is not a new
listing that would require an endangerment finding under
§ 7408(a)(1).

The industry petitioners also seek support for their lack-of-
authority claim in the Supreme Court’s opinion in Whitman. In
particular, they argue that Whitman’s citation of Industrial Union
Department, AFL-CIO v. American Petroleum Institute, 448 U.S.
607 (1980), a case that addressed a provision of the Occupational
Safety and Health (OSH) Act of 1970, means that “a NAAQS is
‘requisite’ to protect public health only if EPA has made a
threshold showing of real and significant risk.” Industry Br. 16.
But Whitman compared the specificity of the CAA to that of the
OSH Act — citing Industrial Union in the course of that
comparison — solely for the purpose of deciding whether the
CAA violated the nondelegation doctrine, Whitman, 531 U.S. at
473-74, a purpose not relevant here. It did not suggest that
Industrial Union has any implication for the EPA’s exercise of
its authority under section 109 of the CAA. The industry
petitioners’ contention that the EPA lacks authority to regulate
nonurban PM10 therefore lacks merit.

The industry petitioners’ second argument against the
EPA’s regulation of nonurban coarse is that scientific evidence
“shows that nonurban coarse PM is not associated with adverse
health effects.” Industry Br. 17. Drawing on statements in the
final rule, the petitioners argue that the scientific evidence shows
danger only from urban coarse, and not from nonurban coarse.
See 2006 Final Rule, 71 Fed. Reg. at 61,185-86. The petitioners
argue both that scientific evidence demonstrates that nonurban
coarse PM is not dangerous and that the EPA’s “cautious
 38

approach” is unlawful absent evidence of nonurban coarse PM’s
dangerousness. Industry Br. 17-22.

In assessing the scientific evidence, the petitioners have
mistakenly equated an absence of certainty about dangerousness
with the existence of certainty about safety. The petitioners
selectively quote from the final rule to bolster their argument that
nonurban coarse is not dangerous, see id. at 19, but they ignore
passages that weaken the force of their contention. For example,
the petitioners note that studies of exposure to coarse PM from
dust storms do not show significant health effects. 2006 Final
Rule, 71 Fed. Reg. at 61,186; Industry Br. 18. But the petitioners
ignore the EPA’s qualifier on the dust storm studies: people in
such situations may practice avoidance behavior and limit their
exposure to the dust. 2006 Final Rule, 71 Fed. Reg. at 61,191.
The petitioners similarly quote the portion of the final rule that
discusses the lack of health impacts observed in studies focusing
on volcanic ash from Mt. St. Helens, id.; Industry Br. 19, but
omit the subsequent discussion of possibilities for toxic
contamination in more typical nonurban coarse PM. 2006 Final
Rule, 71 Fed. Reg. at 61,191.

By contrast, the EPA has provided evidence that suggests
nonurban coarse PM likely is not safe. The EPA points to
dosimetric, toxicological, and occupational exposure studies that
all indicate danger from nonurban coarse. EPA Br. 95-97.
Dosimetric evidence shows that all types of coarse PM can
deposit in the “sensitive regions of the lung of most concern, the
tracheobronchial and alveolar regions.” 2006 Final Rule, 71 Fed.
Reg. at 61,197. Toxicological evidence demonstrates that all
types of coarse can be contaminated by motor vehicles and
industrial emissions, molds, fungi, endotoxins, polycyclic
aromatic hydrocarbons (PAHs), and resuspended fine PM. Id. at
61,189-92. Some rural coarse, such as that found in dry
lakebeds, may also “be highly contaminated with metals, salts,
 39

and other toxic constituents.” Id. at 61,191. The EPA notes that
contamination means there is not a clear division between urban
and nonurban coarse; rather “there is a continuum of sources and
contamination, so that the difference between them is a question
of degree.” EPA Br. 96. In addition, occupational studies
showing health effects from nonurban coarse at occupational
exposure levels “lend[] further support to a cautious approach in
considering revisions to the standards affording protection from
thoracic coarse particles.” 2006 Final Rule, 71 Fed. Reg. at
61,191.

Although the evidence of danger from coarse PM is, as the
EPA recognizes, “inconclusive,” id. at 61,193, the agency need
not wait for conclusive findings before regulating a pollutant it
reasonably believes may pose a significant risk to public health.
The evidence in the record supports the EPA’s cautious decision
that “some protection from exposure to thoracic coarse particles
is warranted in all areas.” Id. As this court has consistently
reaffirmed, the CAA permits the Administrator to “err on the
side of caution” in setting NAAQS. Lead Indus., 647 F.3d at
1155; see also ATA III, 283 F.3d at 369 (“The [Clean Air] Act
requires EPA to promulgate protective primary NAAQS even
where, as here, the pollutant’s risks cannot be quantified or
‘precisely identified as to nature or degree’” (quoting 1997 Final
Rule, 62 Fed. Reg. at 38,653)); Am. Petroleum Inst. v. Costle,
665 F.2d 1176, 1186 (D.C. Cir. 1981) (“In setting margins of
safety the Administrator need not regulate only the known
dangers to health, but may ‘err’ on the side of overprotection by
setting a fully adequate margin of safety”).

This court’s role is “limited to determining if the
Administrator made a rational judgment” and is “not to weigh
the evidence anew and make technical judgments.” Am.
Petroleum Inst., 665 F.2d at 1185. On the basis of the evidence
in the record, we find the Administrator’s decision to regulate all
 40

coarse PM to be rational and therefore deny the industry
petitioners’ claims on this issue.

B. Level of the Daily Coarse PM NAAQS

The industry petitioners next challenge the level of the daily
coarse PM standard. The EPA used PM10 as the indicator for
coarse PM and set the 24-hour coarse PM standard at 150 μg/m3,
the same level it retained in the 1997 NAAQS revision. See
1997 Final Rule, 62 Fed. Reg. at 38,679. It did not differentiate
between urban and nonurban coarse particles.

The industry petitioners first contend that the agency should
have set “different limits for urban and nonurban PM coarse.”
Industry Br. 27. In response, the EPA maintains that it is not
possible, given the “current state of science,” to set different
standards for urban and nonurban areas because there is no
reliable way to determine which ambient mixes are urban and
which are nonurban. 2006 Final Rule, 71 Fed. Reg. at 61,195.
The agency explains that, “given the variety of sources
contributing to PM10-2.5 concentrations in different locations, a
wide variety of ‘ambient mixes’ are likely to exist, greatly
complicating the determination of the appropriate standard level
for each location.” Id. The agency further notes that “there is
insufficient evidence regarding coarse particle composition in
different areas to allow for the proper assignment of different
standard levels in different locations, and the technical
capabilities necessary to make such determinations are currently
lacking.” Id. This is a reasonable explanation for declining to
set different levels for “urban” and “nonurban” coarse. Because
we affirm the EPA’s decision not to set different urban and
nonurban coarse PM standards, we do not reach the argument of
the amicus National Association of Home Builders that the EPA
does not have the authority to set different NAAQS for urban
areas or for different emissions sources. See NAHB Amicus Br.
 41

2-6.

The industry petitioners also contend that the EPA’s single
standard of 150 μg/m3 is not requisite to protect the public health.
Again, the EPA has a reasonable response. The EPA explains
that the Staff Paper undertook an examination of epidemiologic
studies, and that the studies showed mortality and morbidity
effects in areas that exceed the 150 μg/m3 standard. 2006 Final
Rule, 71 Fed. Reg. at 61,200 (“[T]he Staff Paper found little
basis for concluding that the degree of protection afforded by the
current PM10 standards in urban areas is greater than warranted,
since potential mortality effects have been associated with air
quality levels not allowed by the current 24-hour standard, but
have not been associated with air quality levels that would
generally meet that standard, and morbidity effects have been
associated with air quality levels that exceeded the current 24-
hour standard only a few times”); see also STAFF PAPER
§ 5.4.4.1, at 5-67. The Administrator agreed with the Staff
Paper’s findings, and on that basis concluded that the 150 μg/m3
standard did not provide more protection than necessary. 2006
Final Rule, 71 Fed. Reg. at 61,202. The Administrator further
concluded that the “very high degree of uncertainty in the
relevant population exposures implied by the morbidity studies
suggests there is little basis for concluding at this time that a
greater degree of protection is warranted.” Id.

The industry petitioners next argue that the 150 μg/m3
standard for PM10 will result in arbitrarily varying levels of
coarse PM, and that the agency should instead have used a PM10-
2.5 indicator. The EPA does not dispute that using the PM10
indicator will result in coarse PM levels that vary within the limit
of 150 μg/m3. As the EPA explains: “Because the PM10
indicator includes both coarse PM (PM10-2.5) and fine PM (PM2.5),
the concentration of PM10-2.5 allowed by a PM10 standard set at a
single level declines as the concentration of PM2.5 increases.
 42

Thus, the level of coarse particles allowed varies depending on
the level of fine particles present.” Id. at 61,195.

Although the EPA acknowledges that a PM10 indicator will
result in varying coarse PM levels, it does not agree that the
variance will be arbitrary. The EPA agrees with the industry
petitioners that protection from coarse particles should be
targeted at urban areas, where coarse particles have been shown
to pose the greatest danger. Id. at 61,194. But the agency argues
that targeting of urban areas is effectively accomplished by using
an indicator that permits the varying levels that the industry
petitioners challenge. As the EPA explains:

PM2.5 levels tend to be lower in rural areas and higher in
urban areas. Thus, to the extent that higher PM2.5 levels
lead to a lower allowable level of coarse particles in some
areas compared to others, this will occur in precisely those
locations — i.e. urban or industrial areas — where the
science has shown the strongest evidence of adverse health
effects associated with exposure to coarse particles.

Id. at 61,195-96 (citations omitted). In other words: “The
varying levels of coarse particles allowed by a PM10 indicator
will therefore target protection in urban and industrial areas
where the evidence of adverse health effects associated with
exposure to coarse particles is strongest.” Id.

The EPA also offers a further rationale for tying the
stringency of coarse PM regulation to increases in the level of
PM2.5. The EPA explains that the contamination of coarse
particles discussed above, which makes such particles more
dangerous to health, is linked to the presence of fine PM.
Specifically, “[m]any of these contaminants in PM10-2.5 come
originally from fine particles, which may become attached in the
atmosphere or be deposited and mixed into coarse materials on
 43

the ground.” Id. Because contamination increases the health
risks posed by coarse particles, the EPA argues that it is “logical
to allow lower levels of coarse particles when fine particle
concentrations are high. . . . [I]nclusion of PM2.5 in the PM10
indicator for purposes of coarse particle protection would
appropriately reflect the contribution that contaminants emitted
in fine particle form can make to the overall health risk posed by
coarse particles.” Id. Thus, “inclusion of the PM2.5 fraction in
the PM10 indicator . . . ensures that this risk of contamination of
coarse particles by PM2.5 is addressed in the suite of fine and
coarse PM standards.” Id.

It is true that the EPA relies on a qualitative analysis to
describe the protection the coarse PM NAAQS will provide. But
the fact that the EPA’s analysis is qualitative rather than
quantitative does not undermine its validity as an acceptable
rationale for the EPA’s decision. As this court held in ATA III,
“[t]he [Clean Air] Act requires EPA to promulgate protective
primary NAAQS even where, as here, the pollutant’s risks
cannot be quantified or ‘precisely identified as to nature or
degree.’” 283 F.3d at 369 (quoting 1997 Final Rule, 62 Fed.
Reg. at 38,653). The EPA’s qualitative explanation for the
targeting capacity of the 150 μg/m3 PM10 standard is reasonable
and warrants rejection of the petitioners’ challenge on this point.

In sum, we find that the EPA has provided a reasonable
explanation for its decisions not to set separate urban and
nonurban coarse PM standards, to set the single coarse PM
standard at 150 μg/m3, and to utilize a standard that allows
targeted variance in coarse PM levels in an inverse relationship
to the amount of fine PM in the air. Consequently, we reject the
industry petitioners’ challenge to the level of the coarse PM
standard.
 44

C. Choice of the PM10 Indicator

The industry petitioners also level a direct challenge to the
EPA’s choice of the PM10 indicator — a challenge that
significantly overlaps with that addressed in Part IV.B above.
The EPA initially proposed using a PM10-2.5 indicator instead of
a PM10 indicator. See 2006 Proposed Rule, 71 Fed. Reg. at 2665-
68. The PM10-2.5 indicator would have measured only the coarse
fraction (PM10-2.5) of PM smaller than 10 μg, see id., and would
have disposed of concerns about the PM10 indicator that this
court raised in ATA I. See 175 F.3d at 1054-55. In response to
comments expressing concern with the PM10-2.5 indicator, see
2006 Final Rule, 71 Fed. Reg. at 61,188-94, the EPA’s final rule
reverted to the PM10 indicator that the agency had adopted in the
1997 NAAQS review. Id. at 61,194-99.

The petitioners challenge the PM10 indicator on three
grounds. First, they argue that ATA I holds that the PM10
indicator is “inherently confounded,” and that no explanation by
the EPA for the indicator’s utility could make it acceptable.
Industry Reply Br. 13-14. For support, the petitioners rely on the
court’s statement that “it is the very presence of a separate PM2.5
standard that makes retention of the PM10 indicator arbitrary and
capricious.” ATA I, 175 F.3d at 1054. But ATA I did not
definitively preclude the use of a PM10 indicator. It held only
that the EPA had not offered an adequate explanation to “aid us
in understanding its decision” to select PM10 as the indicator for
coarse PM. Id. Without that explanation, we were constrained
to conclude the PM10 and PM2.5 indicators could lead to arbitrary
“‘double regulation’ of the PM2.5 component of PM10 and
potential underregulation of the PM10-2.5 component.” Id. As
recounted in Part IV.B, the EPA has now cured that failure of
explanation and provided a reasonable rationale for its choice of
PM10. See 2006 Final Rule, 71 Fed. Reg. at 61,193-97; EPA Br.
107.
 45

Second, the industry petitioners argue that the EPA’s choice
of a PM10 indicator also runs afoul of ATA I because it was based
solely on reasons of administrative convenience, which ATA I
found impermissible. ATA I, 175 F.3d at 1055 (“The
administrative convenience of using PM10 cannot justify
choosing an indicator poorly matched to the relevant pollution
agent”). This time, however, the EPA expressly disavowed
reliance upon administrative convenience, 2006 Final Rule, 71
Fed. Reg. at 61,195, and chose the PM10 indicator only after
considering and rejecting alternatives on the basis of flaws
discerned during the comment period. Id. at 61,193-97. For
example, the final rule explains that an unqualified PM10-2.5
indicator would not have been “requisite” because it would have
been too stringent in nonurban areas and insufficiently stringent
in urban areas. As discussed above, the agency rejected a
qualified indicator — one with different levels for urban and
nonurban areas — because “determining appropriate levels for
different kinds of ambient mixes is not feasible at this time.” Id.
at 61,195; see also id. at 61,193. Finally, as also discussed
above, the EPA explained its choice of PM10 on scientific rather
than administrative grounds: PM10 allows for targeting
regulation of PM10-2.5 concentrations in those areas that
experience high concentrations of PM2.5, which can contaminate
— and thus render more dangerous — coarse PM. Id. at 61,196-
97.

Third, the petitioners argue that the PM10 standard will
overregulate — or rather, double-regulate — fine particles
because such particles are also subject to the PM2.5 standard. The
discussion in Part IV.B, however, rebuts this argument as well.
As we concluded there, the EPA has reasonably explained that
the variance allowed by the PM10 indicator will target protection
by allowing less coarse PM in areas that experience high
concentrations of potentially contaminating fine PM. Id. at
 46

61,196. This will not double-regulate PM2.5. Fine PM will
primarily be regulated via the newly tightened daily PM2.5
standard. Id. at 61,196 n.72. Any residual regulation of PM2.5 by
the PM10 standard will serve a different, “non-duplicative
purpose[] in providing requisite protection from thoracic coarse
particles” that can be contaminated by PM2.5. Id. at 61,196.
Moreover, the petitioners’ over-regulation argument wrongly
assumes that an area in violation of the PM10 coarse standard
must decrease the fine fraction of PM to achieve compliance
with the standard. Such an area could, however, simply choose
to decrease the amount of coarse PM (PM10-2.5), a choice that
would yield no additional “regulation” of fine PM.

For the foregoing reasons, we reject these and all of the
industry petitioners’ other objections to the EPA’s revised
NAAQS for coarse PM and deny their petitions for review.

V.

Finally, the environmental petitioners challenge the EPA’s
revocation of the primary annual standard for coarse PM. The
petitioners argue that the CAA prohibits the EPA from revoking
this standard. Alternatively, they argue that the EPA’s
justification for the revocation was arbitrary and capricious. We
reject both arguments.

The statutory provision at issue, 42 U.S.C. § 7513(d)(2),
was enacted as part of the 1990 amendments to the CAA. In
those amendments, the Congress created a detailed enforcement
scheme to require that areas not in compliance with the NAAQS
(nonattainment areas) make faster progress toward meeting that
goal. The amendments established a procedure for classifying
nonattainment areas and setting the date by which an area must
attain compliance with the NAAQS. An area that is classified as
“moderate nonattainment” for coarse PM may receive a one-year
extension of its attainment date if certain criteria are met:
 47

Upon application by any State, the Administrator may
extend for 1 additional year . . . the date [for attainment] if:

(1) the State has complied with all requirements and
commitments pertaining to the area in the applicable
implementation plan; and
(2) no more than one exceedance of the 24-hour
national ambient air quality standard level for PM-10 has
occurred in the area in the year preceding the Extension
Year, and the annual mean concentration of PM-10 in the
area for such year is less than or equal to the standard
level.

Id. § 7513 (emphasis added).

Despite the explicit statutory reference to an annual
standard level of PM10, the EPA decided to revoke the PM10
annual standards while leaving in place the 24-hour standards.
The EPA interpreted the CAA to permit this revocation. We
review the EPA’s interpretation under the familiar standard set
forth by the Supreme Court in Chevron. See 467 U.S. at 842–43.

The EPA argues that the CAA does not unambiguously
require an annual standard for coarse PM and that its reading of
the statute is reasonable under Chevron step two. The EPA relies
primarily on this court’s decision in ATA I, in which we held that
the 1990 amendments, which referenced the existing primary
ozone standards, did not eliminate the agency’s authority to
revise those standards. See 175 F.3d at 1047–48.

The petitioners argue that the EPA’s action fails at step one
of the Chevron analysis because the 1990 amendments to the
CAA codified the use of an annual averaging time for the coarse
PM standard. The petitioners argue that the reference in
§ 7513(d)(2) to a standard “annual mean concentration of
 48

[PM10]” means that the EPA must have in place an annual PM10
standard.

The EPA is correct that this court’s precedent forecloses the
petitioners’ statutory argument; the decisive case, however, is not
ATA I but South Coast Air Quality Management District v. EPA,
472 F.3d 882 (D.C. Cir. 2006). Whereas ATA I dealt with the
EPA’s authority to revise the ozone standards in light of the 1990
amendments, South Coast considered the agency’s ability to
revoke a standard explicitly referenced by the text of those
amendments. The EPA had revoked the primary one-hour
standard for ozone and replaced it with a standard using an eight-
hour averaging time. The petitioners in South Coast pointed to
42 U.S.C. § 7511(a)(1), tbl.1, which sets out a scheme for
classifying areas into varying levels of nonattainment based on
the extent of their noncompliance with the one-hour ozone
standard in existence at the time. The petitioners argued that the
EPA could not revoke the one-hour standard because the
Congress had codified it as part of the 1990 amendments. See
South Coast, 472 F.3d at 899. We rejected this argument,
explaining that the “Congress contemplated . . . the possibility
that scientific advances would require amending the NAAQS.”
Id. Indeed, the 1990 amendments left intact § 7409(d)(1), which
directs the EPA to “complete a thorough review” of the NAAQS
every five years and to “make such revisions . . . as may be
appropriate.” Furthermore, another provision of the CAA
“regulates what EPA must do with revoked restrictions.” South
Coast, 472 F.3d at 899; see 42 U.S.C. § 7502(e) (requiring EPA
to promulgate so-called “anti-backsliding” regulations in the
event that the agency relaxes a NAAQS).

The petitioners here have failed to distinguish this case
from South Coast. As we explained there, it would frustrate the
purpose of the CAA to read the 1990 amendments as limiting the
EPA’s ability to revise the NAAQS based on advances in
 49

scientific understanding. The EPA still must review the NAAQS
every five years and make appropriate revisions, including
revoking a standard no longer warranted by the current scientific
understanding. See 42 U.S.C. § 7409(d)(1). In South Coast, we
held that the amendments’ incorporation of the existing one-hour
ozone standard did not prevent the EPA from revoking that
standard and replacing it with one based on an eight-hour
averaging time. Likewise, in this case, the reference in the 1990
amendments to an annual mean standard for PM10 does not
require the EPA to maintain an annual standard in the face of
scientific evidence counseling revocation. We reject the
petitioners’ argument to the contrary.

The petitioners alternatively argue that, even if the EPA had
the authority to revoke the annual coarse PM standard, its
decision to do so was arbitrary and capricious because it was an
unreasonable departure from the agency’s past practice and was
not based on the record. The petitioners point to the EPA’s 1997
revision to the PM NAAQS, in which the agency explained that
the annual coarse PM standard would “provide substantial
protection against short-term as well as long-term exposures to
particles.” 1997 Final Rule, 62 Fed. Reg. at 38,676. The
petitioners argue that, in light of the EPA’s previous decision to
control both short- and long-term exposure through the annual
standard, it was unreasonable for the EPA to revoke the annual
standard based only on the lack of adverse effects from long-
term exposure. The petitioners further assert that scientific
evidence still demonstrates an association between short-term
exposure to coarse PM and negative health effects.

But as the EPA points out, the petitioners have forfeited
their argument that an annual standard reduces the risk from
short-term exposure. The CAA provides that “[o]nly an
objection to a rule or procedure which was raised with
reasonable specificity during the period for public
 50

comment . . . may be raised during judicial review.” 42 U.S.C.
§ 7607(d)(7)(B); see also Appalachian Power Co. v. EPA, 135
F.3d 791, 818 (D.C. Cir. 1998) (“The purpose of the exhaustion
requirement is to ensure that the agency is given the first
opportunity to bring its expertise to bear on the resolution of a
challenge to a rule”). In the comments submitted in response to
the EPA’s proposed revocation of the annual standard, the
environmental petitioners argued only that the record evidence
demonstrated adverse effects from long-term coarse PM
exposure; they did not raise their current argument that an annual
standard is necessary to prevent adverse effects from short-term
exposure.

The EPA reasonably decided that an annual coarse PM
standard is not necessary because, as the Criteria Document and
the Staff Paper make clear, the latest scientific data do not
indicate that long-term exposure to coarse particles poses a
health risk. The CASAC also agreed that an annual coarse PM
standard is unnecessary. We deny the petition for review (No.
06-1411) with regard to the revocation of the annual coarse PM
standard.

VI.

For the foregoing reasons, we grant in part the petitions for
review of the EPA’s primary annual fine PM standard. We grant
in full the environmental petitioners’ petition for review of the
EPA’s secondary fine PM NAAQS. In all other respects, the
petitions for review are denied.

So ordered.