FOR PUBLICATION
 UNITED STATES COURT OF APPEALS
      FOR THE NINTH CIRCUIT

GREAT BASIN MINE WATCH,              
                       Petitioner,
               v.
UNITED STATES ENVIRONMENTAL
                                     
                                          No. 03-70231
PROTECTION AGENCY,
                                           OPINION
                     Respondent,
NEWMONT USA, LIMITED; STATE OF
NEVADA,
         Respondents-Intervenors.
                                     
        On Petition for Review of an Order of the
           Environmental Protection Agency

                Argued and Submitted
      November 1, 2004—San Francisco, California

                  Filed March 23, 2005

  Before: William C. Canby, Jr., Pamela Ann Rymer, and
          Michael Daly Hawkins, Circuit Judges.

                 Opinion by Judge Canby




                           3539
3542              GREAT BASIN MINE v. EPA


                         COUNSEL

Roger Flynn, Bradley A. Bartlett, Western Mining Action
Project, Boulder, Colorado, for the petitioner.

Andrew J. Doyle, Environmental Defense Section, United
States Department of Justice, Washington, D.C., for the
respondent.

William J. Frey, Deputy Attorney General, Carson City,
Nevada, for intervenor State of Nevada.


                         OPINION

CANBY, Circuit Judge:

   Great Basin Mine Watch petitions for review of a final rule
of the Environmental Protection Agency (“EPA”) allowing
Nevada to split one of its clean air areas (“area 61”) into two
(“lower 61” and “upper 61”). Great Basin contends that the
EPA, by approving the split, violated its statutory and regula-
tory duties under the Clean Air Act because it failed to con-
sider the effect of a major pollution emitter, Barrick
Goldstrike Mine, in area 61. We deny Great Basin’s petition
for review because we conclude that the EPA did not act arbi-
                   GREAT BASIN MINE v. EPA                    3543
trarily, capriciously, or contrary to law when it granted
Nevada’s request to divide area 61, and that the existence and
operation of the Mine did not preclude the division.

                          Background

   [1] The Clean Air Act includes a program for the preven-
tion of significant deterioration of air quality (“PSD” pro-
gram), which applies to areas that are actually or potentially
low in air pollution. Several of the program’s restrictions on
emissions are triggered when a major stationary source (i.e.,
a major pollution emitter) submits an application for a permit
for new construction or major modification within the area
under the appropriate regulations. 40 C.F.R. § 52.21(b)
(14)(ii). A central issue in the present appeal is whether the
PSD restrictions were triggered, or must be deemed triggered,
in area 61 by the actions of the Barrick Mine. Great Basin
contends that the PSD restrictions were triggered, and that as
a consequence EPA’s decision to allow division of area 61
was impermissible. We agree with the EPA, however, that the
restrictions were not triggered, and that division of area 61
was not arbitrary, capricious, or contrary to law. Before we
explain our reasons, we think it best to describe briefly the
relevant framework of the Act.

  1.   The Clean Air Act’s PSD Program

   [2] The Clean Air Act regulates air pollutants such as par-
ticulate matter, sulfur dioxide, and nitrogen oxide. See 42
U.S.C. §§ 7408-7409. The Act requires the division of states
into air quality planning areas (“baseline areas”). These base-
line areas are assigned one of three labels—attainment,
unclassifiable, or nonattainment—depending on the quality of
their air. If an area fails to meet national air quality standards,
it is classified as a nonattainment area. If an area meets
national standards, it is classified as an attainment area. If it
is unclear whether the area meets the standards, the area is
denominated unclassifiable. The PSD program applies to the
3544                  GREAT BASIN MINE v. EPA
latter two categories, and it is undisputed that area 61 falls
within them.1

   [3] In attainment and unclassifiable areas, the PSD program
attempts to maintain the relatively clean air by limiting the
total pollution “increment” per year. The PSD restrictions are
not automatic, however; they come into effect when an appli-
cation is filed under the appropriate regulations for new con-
struction of a major stationary source or for major
modification of an existing major source within the area. 40
C.F.R. § 52.21(b)(14)(ii). It is important for our purposes to
note that application for a permit is the trigger, not simply the
existence of substantial pollution emissions. The filing of
such an application establishes a “minor source baseline date”
for which the EPA determines the ambient “baseline concen-
tration” for the area. 40 C.F.R. § 52.21(b)(13), (14)(ii). There-
after, the PSD program places strict limits on aggregate
increases in pollution within the baseline area whether the
increases come from minor or major sources.

   [4] If no major source within a baseline area has applied for
a permit, however, the PSD restrictions are not triggered. In
that event, pollution emitters are free to operate under the less
restrictive national air quality standards, rather than the stric-
ter standards of the PSD program. See, e.g., Reno-Sparks
Indian Colony v. EPA, 

, 902-03 (9th Cir. 2003).

  2.    Discretion to Redesignate Baseline Areas

   Because the PSD program is administered by baseline
areas, the number and size of these areas has a very consider-
able effect on potential polluters. The greater the number and
the smaller the size of the baseline areas, “the less likely it is
that a major source has applied for a permit within any one
area, thereby establishing a baseline date.” 

Thus a
  1
    Area 61 was classified as attainment for sulfur dioxide and as unclassi-
fiable for particulate matter and nitrogen dioxide.
                  GREAT BASIN MINE v. EPA                 3545
minor source is more likely to “find an area in which to oper-
ate where it is not subject to the requirements of the PSD pro-
gram.” 

smaller scale, the division of one baseline
area into two might have a similar effect; if a major source
were to seek a permit in only one of the two newly-divided
areas, the other might escape PSD regulation that would have
applied had the original area not been split.

   [5] Despite this consideration, the EPA has broad discretion
to grant a state’s request to divide an area for which no base-
line date or baseline concentration has been established. The
EPA may grant redesignation requests “on the basis of air
quality data, planning and control considerations, or any other
air quality-related considerations the Administrator deems
appropriate.” 42 U.S.C. § 7407(d)(3)(A). This decision also
must rely on “sufficient data.” 40 C.F.R. § 81.300(a).

   [6] The situation is different when PSD restrictions have
already been triggered in an area. Division of such an area
into two new areas raises additional problems. One is the
question whether one of the new areas may or should be “un-
triggered” because the major source that triggered the PSD
restrictions lies in the other new area. Perhaps in recognition
of this and other problems, the discretion of the EPA is more
limited when dealing with redesignation of an area for which
PSD restrictions have been triggered. The EPA cannot
redesignate, for example, if the new area would “intersect or
be smaller than the area of impact of any major stationary
source or major modification which . . . [e]stablishes a minor
source baseline date.” 40 C.F.R. §§ 51.166(b)(15)(ii),
52.21(b)(15)(ii). Moreover, the EPA has indicated that it is
more likely to deny a request to split an area in which the
PSD caps apply because the redesignation may allow greater
deterioration of the air quality.

                  The Present Controversy

  This dispute arose when the State of Nevada submitted its
request to the EPA to divide baseline area 61 (550 square
3546               GREAT BASIN MINE v. EPA
miles) into two, lower 61 and upper 61. Nevada claimed that
the split would aid its air quality management because the two
new areas more accurately reflect the “local air transport pro-
cesses,” industrial development, and the region’s topography,
among other reasons. Much of the present dispute arises from
the fact that, in Nevada’s view, the Barrick Mine, although a
major source, has never applied or been required to apply for
a permit for new construction or a major modification. This
situation presumably results from the fact either that the Mine
was a major source before the Clean Air Act’s requirements
took effect, or that it became a major source by small incre-
ments not subject to permits. Nevada reflected its view in its
redesignation request, in which it asserted that “Area 61 does
not contain any PSD sources, and it has not been significantly
impacted by any major source or modification,” although it
acknowledged that Barrick Goldstrike Mine was “a major
source for PM10, NOx [nitrogen oxide], and SO2.”

   In response to Nevada’s request, the EPA issued a proposed
rule granting the request to split area 61 into lower 61 and
upper 61. See 67 Fed. Reg. 21194, 21197 (Apr. 30, 2002).
The EPA proposed to approve the request because of its pol-
icy to “provide States a fair degree of autonomy to balance air
quality management with economic planning” and because
the redesignation would not interfere with Nevada’s manage-
ment of air quality in the area. 

that the
change met the statutory and regulatory requirements of the
Act, the EPA relied on its conclusion that “no PSD source has
located in [area 61], . . . and the newly created baseline areas
. . . do not intersect the area of impact of any major PSD
source nor do they have boundaries that are smaller than such
impact area.” 

footnote, the proposed rule acknowledged that the Bar-
rick Mine was a major source, but disregarded it because “the
source has not been subject to PSD review.” 

n.3.
The proposed rule also stated that the redesignation did not
result in “an untriggering of the baseline area,” and thus there
                   GREAT BASIN MINE v. EPA                    3547
was “no elimination of already consumed [pollution] incre-
ment and no consumed increment would be added to the base-
line for the area.” 

Moreover, the division did not
“carve out small ‘postage stamp’ areas encompassing only the
significant impact area around a major PSD source.” 

The EPA also solicited public comment on its proposed
rule.

   Relying on the reasons listed in the proposed rule, the
EPA’s final rule granted Nevada’s request despite criticism
from several commentators, including Great Basin. See 67
Fed. Reg. 68769, 68771, 68776 (Nov. 13, 2002). Great Basin
petitioned for review under 42 U.S.C. § 7607(b)(1).

                      Standard of Review

   “In reviewing a final action by the EPA, we reverse only
if it is arbitrary, capricious, or contrary to law or if it exceeds
the statutory jurisdiction, authority, or limitations.” Exxon
Mobil Corp. v. EPA, 

, 1248 (9th Cir. 2000). We
will overturn a final action of the EPA, “if the agency has
relied on factors which Congress has not intended it to con-
sider, entirely failed to consider an important aspect of the
problem, offered an explanation for its decision that runs
counter to the evidence before the agency, or is so implausible
that it could not be ascribed to a difference in view or the
product of agency expertise.” Motor Vehicle Mfrs. Ass’n v.
State Farm Mut. Auto. Ins. Co., 

, 43 (1983).

                           Discussion

   The critical question is whether Barrick Goldstrike Mine
(hereafter “Barrick”) triggered the PSD pollution regulations
in area 61. If not, the choice whether to approve the rede-
signation is committed to the EPA’s discretion within broad
limits that we have already described. We conclude that Barr-
ick did not trigger the PSD restrictions, and that the EPA did
not act arbitrarily, capriciously, or contrary to law in approv-
3548               GREAT BASIN MINE v. EPA
ing division of area 61. We therefore deny the petition for
review.

  1. The Barrick Mine Did Not Trigger the PSD
  Restrictions

   [7] Great Basin attacks the final rule on four fronts, and we
address each in turn. Great Basin first argues that Barrick’s
status as a major source triggered the minor source baseline
date, thereby limiting the EPA’s discretion to redesignate area
61. The minor source baseline date is triggered by an applica-
tion for a permit under the PSD program. See 40 C.F.R.
§ 52.21(b)(14)(ii); 

. The PSD
program requires a permit for (1) any new “major emitting
facility” or (2) any modification of a major facility that
increases the net emissions of regulated pollutants. See, e.g.,
Alaska Dep’t of Envtl. Conservation v. EPA, 

,
472 (2004). Although the PSD program requires a permit for
the construction of a major emitting facility built after August
7, 1977, see 42 U.S.C. § 7475(a)(1), the record does not indi-
cate when Barrick was constructed, or whether it was initially
constructed as a major emitting facility. Therefore, Great
Basin’s first challenge fails.

   [8] Barrick did undertake modifications, however, and they
are the center of Great Basin’s final three challenges. There
is no dispute that Barrick modified its facilities three times in
2001. There also is no dispute that the PSD program requires
a permit for any “major modification” that causes a forty tons-
per-year (“tpy”) increase in nitrogen oxides. 40 C.F.R.
§ 51.21(b)(23)(I); Alaska Dep’t of Envtl. 

. Finally, it is undisputed that, in the aggregate,
Barrick’s three 2001 modifications increased nitrogen oxide
emissions over forty tpy. It is disputed, however, whether
these totals should be accumulated, and the EPA argues that,
when viewed separately, no single modification caused an
increase of forty tpy or more.
                       GREAT BASIN MINE v. EPA                           3549
   In April, August, and October, 2001, Barrick sought per-
mits from the State of Nevada’s Bureau of Air Quality for
each of the three modifications. For the first and largest emis-
sions modification, Barrick claimed that nitrogen oxide emis-
sions would increase by less than forty tpy.2 Barrick estimated
that its starting nitrogen oxide emissions were 383.3 tpy, and
its modified emissions would be 423.1 tpy. This figure (39.8)
falls narrowly below the forty tpy trigger.3

   Great Basin disputes the starting point, however. Great
Basin contends that Barrick reported its starting emissions as
374.6, not 383.3, resulting in an increase of 48.5 tpy of nitro-
gen oxides. The EPA responds that this is a “scrivener’s error
on Barrick’s part,” and the accurate starting point is 383.3 tpy.
The EPA and Nevada chose to rely on this higher starting fig-
ure, and their reasonable reliance is supported by the record.
We therefore view 383.3 as the correct starting point, and the
first modification causes less than a forty tpy increase in nitro-
gen oxides.

  The next modification increased the estimated nitrogen
oxide emissions from 423.1 to 426 tpy.4 Under either starting
point, then, if the nitrogen oxide emissions increases from the
two modifications are combined for purposes of the PSD trig-
ger, the modifications exceed forty tpy (i.e., 42.9).5 Thus,
Great Basin’s final two arguments contend that these totals
should be accumulated.
  2
     This modification put twelve of Barrick’s emergency electric genera-
tors into regular use.
   3
     Both the EPA and the State of Nevada noted the narrowness of this
margin, and Nevada conditioned the issuance of its permit on Barrick’s
agreement that the new generator use would not cause forty tpy or more
of additional nitrogen oxide emissions.
   4
     This modification replaced “existing backfill crushing, screening, and
shotcrete plants and [made] miscellaneous changes to related existing
facilities.” It is important to note that this modification is unrelated to the
previous modification.
   5
     The final modification did not change the nitrogen oxide emissions.
3550                  GREAT BASIN MINE v. EPA
   [9] Great Basin first argues that the modifications should be
accumulated because, in November 2001, Barrick submitted
an operating permit application to Nevada “incorporat[ing]”
all three “currently pending” modifications. Great Basin
claims that the EPA should treat this operating permit applica-
tion as a PSD permit application grouping the three modifica-
tions. Title V operating permits, however, are not part of the
Act’s PSD program. See 42 U.S.C. §§ 7661-7661f; United
States v. Marine Shale Processors, 

, 1356 (5th
Cir. 1996); New York v. Niagara Mohawk Power Corp., 

, 661 (W.D.N.Y. 2003). Therefore, the Act
does not require the EPA to treat PSD permits and operating
permits interchangeably.6

   Finally, Great Basin argues that it should prevail because,
under the EPA’s practice of accumulation, the April and
August modifications exceed the forty tpy threshold. The final
rules of the EPA—at the time of its decision—indicate that,
for major sources, minor emissions increases are accumulated
during a period “contemporaneous” with the modification,
with “contemporaneous” basically meaning within the preced-
ing five years. See Requirements for Preparation, Adoption,
and Submittal of Implementation Plans; Approval and Pro-
mulgation of Implementation Plans, 45 Fed. Reg. 52676,
52701-02 (Aug. 7, 1980). Obviously, the gap between April
and August of the same year is less than five years. Therefore,
according to Great Basin, the modifications should have trig-
gered the PSD caps, and the EPA failed to consider this
important factor in its decision.7
   6
     Our disposition of this issue makes it unnecessary for us to examine
Great Basin’s assumption that incorporating three separate modifications
in one permit application would require the effects of the modifications to
be accumulated.
   7
     The EPA counters that Great Basin failed to show prejudice because
the EPA could divide area 61 even if Barrick had triggered the PSD pro-
gram. This argument is unavailing for two reasons. First, the EPA stated
in its final rule that it relied on the conclusion that Barrick (or any other
entity) had not triggered the PSD caps.
                       GREAT BASIN MINE v. EPA                          3551
   [10] This argument fails for at least two reasons. First,
Great Basin waited far too long to raise it. Great Basin raised
the argument neither before the agency, nor in its petition for
review, nor in its opening brief in this court; it presented the
argument for the first time in its reply brief here. Second, and
more important, even if we were to agree with Great Basin’s
interpretation, vacating the EPA’s rule on that basis would
serve no purpose. In 2003, the EPA clarified the regulation,
and it now states that emission increases are not netted unless
the modification, in and of itself, is significant (i.e., results in
this case in a forty or more tpy increase of nitrogen oxides).
Therefore, even if the EPA incorrectly read the regulation then8
to mean exactly what it states now,9 remanding the decision
to the agency would not change the result because no individ-
ual modification met the threshold.

   Second, while the decision to redesignate is largely discretionary, see,
e.g., 42 U.S.C. § 7407(d)(3)(A), it does have its limits, particularly when
a major source has triggered the PSD program. The EPA is prohibited, for
example, from dividing area 61 if the new area would “intersect or be
smaller than the area of impact of any major stationary source or major
modification which . . . [e]stablishes a minor source baseline date.” 40
C.F.R. §§ 52.21(b)(15)(ii), 51.166(b)(15)(ii). Operating under the assump-
tion that Barrick did not trigger a minor source baseline date, the EPA did
not analyze whether lower 61 intersects with, or is smaller than, Barrick’s
area of impact. Counsel for the EPA analyzed it in his brief, and he
thoughtfully concluded that it did not. The EPA itself, however, did not
analyze the issue when it made its decision.
   8
     The regulation stated that a “ ‘major modification’ means any physical
change in . . . a major stationary source that would result in a significant
net emissions increase of [nitrogen oxides].” 40 C.F.R. § 52.21(b)(2)(I)
(2002).
   9
     The regulation states that a “ ‘major modification’ means any physical
change . . . that would result in: a significant emissions increase [of nitro-
gen oxides]; and a significant net emissions increase of that pollutant from
the major stationary source.” 40 C.F.R. § 52.21(b)(2)(I) (2004) (emphasis
added). A “significant emissions increase” in nitrogen oxides means an
increase that exceeds 40 tpy. 40 C.F.R. §§ 52.21(b)(23), (40). The EPA’s
comments on this change assert that the new regulation “clarif[ies] what
has always been [its] policy.” 67 Fed. Reg. 80186, 80190 (Dec. 31, 2002).
3552                  GREAT BASIN MINE v. EPA
  2. The EPA’s Rule Approving Redesignation Was Not
  Arbitrary, Capricious, or Contrary to Law

   [11] Having determined that Barrick did not trigger the
PSD program, we conclude that the EPA permissibly granted
Nevada’s request to redesignate. As we pointed out earlier,
the EPA has broad discretion to grant or deny requests. With
the exception of barriers irrelevant here, the EPA may grant
the request “on the basis of air quality data, planning and con-
trol considerations, or any other air quality-related consider-
ations the Administrator deems appropriate.” 42 U.S.C.
§ 7407(d)(3)(A); see also 40 C.F.R. § 81.300(a). The EPA’s
decision complies with these statutory and regulatory criteria.
The record adequately supports the EPA’s finding that the
two new areas more accurately reflect the region’s “local air
transport processes,” industrial development, and topography,
among other reasons. The record also does not indicate that
either Nevada or the EPA have a history of carving out “post-
age stamp” baseline areas, and upper and lower 61 are no excep-
tions.10 Therefore, EPA did not act arbitrarily or exceed its
authority in granting Nevada’s request.11

   PETITION FOR REVIEW DENIED.




  10
      The smaller of the two areas, lower 61, is over 200 square miles.
Although it was pointed out in oral argument that, in the West, 200 square
miles seems more like a postage stamp than it would elsewhere, we con-
clude that lower 61 still does not qualify.
   11
      In light of this disposition, we deny Great Basin’s request for attor-
neys’ fees. See Western States Petroleum Ass’n v. EPA, 

, 286
(9th Cir. 1996).