United States Air Tour Ass'n v. Federal Aviation Administration

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued May 9, 2002 Decided August 16, 2002

No. 00-1201

United States Air Tour Association, et al.,
Petitioners

v.

Federal Aviation Administration, et al.,
Respondents

Grand Canyon Trust, et al.,
Intervenors

---------

Consolidated with
00-1212

On Petitions for Review of an Order of the
Federal Aviation Administration

William Davis Thode and Joseph F. Becker argued the
cause for petitioners United States Air Tour Association, et

al. With them on the briefs was William Perry Pendley.
Lorraine B. Halloway and Timothy M. Biddle entered ap-
pearances.

Alexander E. Dreier argued the cause for petitioners
Grand Canyon Trust, et al. With him on the briefs were
Michael L. Kidney, Catherine S. Stetson, Jeffrey C. Nelson,
and Robert Wiygul.

Ronald M. Spritzer, Attorney, U.S. Department of Justice,
argued the cause for respondents. With him on the brief was
Ellen J. Durkee, Attorney.

Michael L. Kidney, Catherine E. Stetson, Alexander E.
Dreier, Jeffrey C. Nelson, and Robert Wiygul were on the
brief of intervenors Grand Canyon Trust, et al.

Glenn M. Feldman argued the cause and filed the brief for
intervenor Hualapai Indian Tribe.

Before: Edwards, Henderson, and Garland, Circuit
Judges.

Opinion for the Court filed by Circuit Judge Garland.

Garland, Circuit Judge: As part of an ongoing effort to
reduce aircraft noise in Grand Canyon National Park, the
Federal Aviation Administration (FAA) promulgated a rule
limiting the number of air tours permitted to fly over the
Park. Two groups of petitioners, one led by the United
States Air Tour Association and the other by the Grand
Canyon Trust, challenge that rule. We reject the challenges
brought by the Air Tour Association, but conclude that the
challenges brought by the Trust raise issues that require
further consideration by the FAA.

I

The history of regulation of aircraft overflights at Grand
Canyon National Park is set out in Grand Canyon Air Tour
Coalition v. FAA, 154 F.3d 455, 460-64 (D.C. Cir. 1998)
[hereinafter Grand Canyon I]. We recount some of that
story here and explain subsequent developments to the extent
necessary to give context to the present controversy.

A

In 1987, Congress enacted the National Parks Overflights
Act, Pub. L. No. 100-91, 101 Stat. 674 (set out at 16 U.S.C.A.
s 1a-1 note). Section 3 of the Act declared that "[n]oise
associated with aircraft overflights at the Grand Canyon
National Park is causing a significant adverse effect on the
natural quiet and experience of the park." Overflights Act
s 3(a). To address this problem, Congress required the
Secretary of the Interior to submit to the Administrator of
the FAA:

recommendations regarding actions necessary for the
protection of resources in the Grand Canyon from ad-
verse impacts associated with aircraft overflights. The
recommendations shall provide for substantial restora-
tion of the natural quiet and experience of the park and
protection of public health and safety from adverse ef-
fects associated with aircraft overflights.

Id. s 3(b)(1) (emphasis added). Congress also required the
FAA to "prepare and issue a final plan for the management of
air traffic in the air space above the Grand Canyon." Id.
s 3(b)(2). That plan, the Act declared, "shall ... implement
the recommendations of the Secretary without change unless
the [FAA] determines that implementing the recommenda-
tions would adversely affect aviation safety." Id. Finally,
Congress directed the Secretary to submit, within two years
of the effective date of the plan, "a report discussing (A)
whether the plan has succeeded in substantially restoring the
natural quiet in the park; and (B) such other matters,
including possible revisions in the plan, as may be of inter-
est." Id. s 3(b)(3).

In response to the Overflights Act, the Secretary of the
Interior submitted recommendations to the FAA in Decem-
ber 1987. In May 1988, the FAA implemented those recom-
mendations in the form of Special Federal Aviation Regula-
tion (SFAR) 50-2. See Special Flight Rules in the Vicinity of
the Grand Canyon National Park, 53 Fed. Reg. 20,264 (June
2, 1988). The regulation applied to aircraft flying below
14,500 feet and established, inter alia, flight free zones (areas

into which aircraft may not fly), minimum altitudes, and other
rules constraining flight paths within the Park. It remained
in effect through 1997.

On September 12, 1994, the National Park Service (the
Park Service or NPS), on behalf of the Secretary of the
Interior, submitted the report to Congress required by sec-
tion 3 of the Overflights Act. See NPS, U.S. Dep't of the
Interior, Report on the Effects of Aircraft Overflights on the
National Park System (published in 1995) [hereinafter 1994
NPS Report]. In that report, the Park Service made several
foundational determinations. First, it decided that the appro-
priate measure for quantifying aircraft noise was the percent-
age of time that aircraft are audible. See id. at 60. Second,
the Park Service concluded that the key statutory phrase,
"substantial restoration of the natural quiet," required that
"50% or more of the park achieve 'natural quiet' (i.e., no
aircraft audible) for 75-100 percent of the day." Id. at 182.
Subsequently, the agencies determined that an aircraft was
audible at three decibels above the average natural ambient
sound level. See FAA, U.S. Dep't of Transp., Environmental
Assessment: Special Flight Rules in the Vicinity of Grand
Canyon National Park 4-4 to 4-5 (1996) [hereinafter 1996
Environmental Assessment].

Applying these principles, the agencies concluded that,
under SFAR 50-2, only 31% of the Park enjoyed "a substan-
tial restoration of natural quiet"--by which they meant that
only 31% of the Park experienced natural quiet for at least
75% of the day. Special Flight Rules in the Vicinity of Grand
Canyon National Park, 61 Fed. Reg. 69,302, 69,317 (Dec. 31,
1996) [hereinafter 1996 Final Rule].1 Moreover, the agencies

__________
1 When the agencies discuss their current progress toward "a
substantial restoration of natural quiet," they refer to the percent-
age of the Park experiencing natural quiet for at least 75% of the
day. When they discuss the overall statutory goal of "substantial
restoration of the natural quiet," however, they refer to a situation
in which at least 50% of the Park achieves natural quiet for at least
75% of the day. Depending upon the context, we will use the
phrase in the same two ways in this opinion.

predicted that without revisions to the existing regulations,
projected growth in the number of air tours would cause the
percentage of the Park enjoying substantial restoration of
natural quiet to drop to less than 10% by the year 2010. Id.

On December 31, 1996, the FAA issued a final rule that
adopted the definitions contained in the 1994 NPS Report,
including the definition of substantial restoration of the natu-
ral quiet. See 1996 Final Rule, 61 Fed. Reg. at 69,305-10.
Among other things, the 1996 Final Rule also established new
flight free zones, instituted flight curfews, and set a cap on
the number of aircraft that could fly over the park--although
not on the number of flights. See id. at 69,317, 69,332. In
addition to the 1996 Final Rule, the FAA proposed two
further rules: one to modify flight paths in the Park; the
other to require operators to use quieter aircraft. See Pro-
posed Air Tour Routes for the Grand Canyon National Park,
61 Fed. Reg. 69,356 (Dec. 31, 1996); Noise Limitations for
Aircraft Operations in the Vicinity of Grand Canyon National
Park, 61 Fed. Reg. 69,334 (proposed Dec. 31, 1996). The
FAA predicted that the 1996 Final Rule, in conjunction with
the two proposed rules, would meet the statutory goal of
substantial restoration of the natural quiet by the year 2008.
See Noise Limitations for Aircraft Operations, 61 Fed. Reg.
at 69,329.

In October 1997, the FAA discovered that it had signifi-
cantly underestimated the number of tour aircraft operating
in the Park, and that as a consequence the 1996 Final Rule
would be less effective than it had thought. See Special
Flight Rules in the Vicinity of Grand Canyon National Park,
62 Fed. Reg. 58,898, 58,899 (Oct. 31, 1997). After oral
argument in Grand Canyon I, the FAA informed the court
that it was considering placing a cap on the number of flights,
in addition to the 1996 Final Rule's cap on the number of
aircraft. See Grand Canyon I, 154 F.3d at 464.

B

In Grand Canyon I, several groups of petitioners chal-
lenged provisions of the 1996 Final Rule. The principal

challenges came, as they do here, from a group of air tour
operators (the Air Tour Coalition) that included members of
petitioner Air Tour Association, and from a group of environ-
mental organizations led by petitioner Grand Canyon Trust.
The air tour operators argued that the rule did "too much, too
soon," while the Trust argued that it did "too little, too late."
Grand Canyon I, 154 F.3d at 459-60. We upheld the rule
against both challenges.

In the course of our decision, we affirmed--against chal-
lenges from both the Coalition and the Trust--the Park
Service's definition of "substantial restoration of the natural
quiet" as "50% or more of the park achiev[ing] 'natural quiet'
(i.e., no aircraft audible) for 75-100 percent of the day." That
definition, we said, was a reasonable construction of an am-
biguous statutory phrase. Id. at 466-67 (citing Chevron
U.S.A. Inc. v. Natural Res. Def. Council, 467 U.S. 837, 841-43
(1984)). We also upheld as reasonable the agencies' three-
decibels-above-ambient measure of audibility, and we rejected
the Air Tour Coalition's contention that the agencies had
ignored their statutory obligation to consider the actual expe-
rience of Park visitors. Id. at 465-67, 469. We did not need
to decide whether there was such a statutory obligation,
because it was clear that the agencies' definition and audibili-
ty measure were specifically developed to address and to
enhance the experience of Park visitors. Id. Grand Canyon
I also observed that the Overflights Act "clearly divides the
institutional responsibilities between" the FAA and the Park
Service. Id. at 468. Because the Act directs the FAA to
"implement the recommendations of the Secretary without
change" unless they would adversely affect aviation safety, we
held that the FAA "had no choice but to adopt the Park
Service's recommendations" regarding substantial restoration
of the natural quiet. Id.

Finally, we noted the Trust's complaint that, using the
Park Service's definition and the FAA's recent reevaluation of
its data, neither the 1996 Final Rule nor the two additional
proposed rules would achieve Congress' goal of substantially
restoring natural quiet in the Park. We agreed that it would
be arbitrary and capricious for the government not to intend

to achieve the congressional goal on any timetable at all. Id.
at 477. We accepted, however, the FAA's assurance that it
still "anticipates meeting the goal of substantial restoration
by 2008" through implementation of the two proposed rules
and consideration of a cap on the total number of overflights.
Id. at 478. In so doing, we emphasized that the Trust could
raise its claim again if "the FAA does not issue additional
regulations reasonably promptly, or if those regulations do
not appear likely to achieve the statutory goal on a reason-
able timetable." Id.

C

On April 4, 2000, the FAA published two additional rules
governing flights over the Grand Canyon. One of those rules
(the Airspace Rule), not at issue here, modifies air flight
paths in the Park.2 The second rule, the Limitations Rule, is
the subject of the petitions for review filed in this case. That
rule imposes a cap on the total number of commercial air
tours that operators may run in the Park. See Commercial
Air Tour Limitation in the Grand Canyon National Park
Special Flight Rules Area, 65 Fed. Reg. 17,708 (April 4, 2000)
(codified at 14 C.F.R. ss 93.303-.325) [hereinafter Limitations
Rule]. Under the Limitations Rule, an air tour operator may
not conduct more flights in the Park than it conducted during
the base year of May 1, 1997 through April 30, 1998. 14
C.F.R. s 93.319(a), (b).

In developing the Limitations Rule, the FAA and the Park
Service issued three associated documents that detailed the

__________
2 See Modification of the Dimensions of the Grand Canyon Na-
tional Park Special Flight Rules Area and Flight Free Zones, 65
Fed. Reg. 17,736 (April 4, 2000). On October 17, 2001, this court
severed challenges to the Airspace Rule from those to the Limita-
tions Rule at issue here, and held the former challenges in abeyance
pending ongoing FAA administrative proceedings.

methodology they used to quantify noise levels in the Park
and to measure progress toward the goal of substantial
restoration of the natural quiet. First, the Park Service
announced in July 1999 that it was changing the threshold at
which it would regard aircraft noise as audible in part of the
Park. See Change in Noise Evaluation Methodology for Air
Tour Operations Over Grand Canyon National Park, 64 Fed.
Reg. 38,006 (July 14, 1999) [hereinafter Change in Noise
Evaluation Methodology]. As discussed above, the agencies
had previously adopted a threshold of three decibels above
the average natural ambient sound level, a measure of audi-
bility we affirmed in Grand Canyon I. In the Change in
Noise Evaluation Methodology, the Park Service divided the
Park into two zones: Zone One, encompassing about one-
third of the Park, includes the more developed areas; Zone
Two, encompassing two-thirds, contains the backcountry.
The Park Service announced that in Zone One it will continue
to consider aircraft audible at three decibels above the aver-
age natural ambient level. 64 Fed. Reg. at 38,006-08. For
Zone Two, however, the Park Service determined that air-
craft noise is audible if it is eight decibels below the average
natural ambient level. Id.

Second, in January 2000, the Park Service issued a review
of its Change in Noise Evaluation Methodology. NPS, Dep't
of the Interior, Review of Scientific Basis for Change in Noise
Impact Assessment Method Used at Grand Canyon National
Park (2000) [hereinafter 2000 NPS Review]. The review
explained in detail the acoustic model used in assessing noise
impacts in the Park. It also reaffirmed the Park Service's
1994 definition of "substantial restoration of the natural
quiet" as: "Fifty percent or more of the Park achieving
'natural quiet' (i.e., no aircraft audible) for 75-100 percent of
the day." Id. at 16. And it stated that "[t]his definition is a
threshold not to be exceeded on any given day ... and refers
to ... the 12 hour daylight period ... during which air tours
occur." Id.; accord id. at 4-5.

Third, in February 2000, the FAA issued a Final Supple-
mental Environmental Assessment in which it analyzed the
effects that it expected the Limitations Rule to have on noise

in the Park. See FAA, U.S. Dep't of Transp., Final Supple-
mental Environmental Assessment: Special Flight Rules in
the Vicinity of Grand Canyon National Park (2000) [hereinaf-
ter FSEA]. Of particular importance here, the FSEA stated
that the FAA intended to use an "average annual day"
standard when determining the percentage of "the day" that
is restored to natural quiet at a given location. See id. at 4-
12, 4-18, F-4. The assessment also made clear that the
FAA's noise model only accounts for noise from tour aircraft,
and does not consider noise from other aircraft that fly over
the Grand Canyon, including commercial jets, general avia-
tion, and military flights. Id. app. G, at 40.

Applying these standards, the FAA concluded that the
Limitations Rule would make "significant steps towards sub-
stantially restoring natural quiet," Limitations Rule, 65 Fed.
Reg. at 17,713, although the combined effect of all of the
agency's rulemakings would still not achieve the goal of
having 50% of the Park experience natural quiet for at least
75% of the day, id. at 17,711. The FAA estimated that only
32% of the Park currently achieved that mark, and that if no
further action were taken, future air tour growth would
reduce that to 25% of the Park in nine to ten years. Id. at
17,724. Adding the new 2000 rules, the FAA predicted,
would increase the percentage of the Park experiencing the
substantial restoration of natural quiet to above 41% and
maintain that level in the future. Id.; see FSEA at 4-18
(predicting substantial restoration of 43.6% of the Park
through 2003 and 43.5% in 2008). But the agency recognized
that additional steps, including implementation of the still-
pending quiet technology rule, would be necessary to achieve
Congress' goal in the Overflights Act. Limitations Rule, 65
Fed. Reg. at 17,714.

D

As noted above, two groups of petitioners have filed chal-
lenges to the Limitations Rule. The first is led by the United
States Air Tour Association, a trade organization whose mem-

bers fly air tours over the Park. The second, led by the
Grand Canyon Trust, is a group of six environmental organi-
zations. Petitioners level a number of challenges at the rule,
raising questions of statutory construction, regulatory inter-
pretation, and the rationality of the agencies' methodologies
and policy choices.

As to questions of statutory construction, where legislation
is "silent or ambiguous with respect to [a] specific issue," we
are obligated to defer to an agency's interpretation as long as
it is "based on a permissible construction of the statute."
Chevron, 467 U.S. at 843. Similarly, we defer to an agency's
reading of its own regulation, unless that reading is "plainly
erroneous or inconsistent with the regulation." Auer v. Rob-
bins, 519 U.S. 452, 461 (1997) (internal quotation marks
omitted). We also examine agency regulations to determine
whether they are "arbitrary, capricious, an abuse of discre-
tion, or otherwise not in accordance with law." Administra-
tive Procedure Act, 5 U.S.C. s 706(2)(A). In that regard, the
question for the court is whether the agency has considered
the relevant factors and articulated a " 'rational connection
between the facts found and the choice made.' " Motor
Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463
U.S. 29, 43 (1983) (quoting Burlington Truck Lines, Inc. v.
United States, 371 U.S. 156, 168 (1962)). Finally, by statute
the FAA's findings of fact "are conclusive" if they are "sup-
ported by substantial evidence." 49 U.S.C. s 46110(c).

We apply these principles to our analysis of the Air Tour
Association's challenges to the Limitations Rule in Part II
below, and to those of the Grand Canyon Trust in Part III.

II

The Air Tour Association asks this court to hold the
Limitations Rule unlawful for five principal reasons: (i) it was
prompted by an improper change in the definition of "natural
quiet"; (ii) the acoustic methodology that justifies the rule is
scientifically flawed; (iii) the FAA arbitrarily and capriciously

issued the rule without first promulgating a quiet technology
rule; (iv) in promulgating the rule, the FAA violated the
Regulatory Flexibility Act, 5 U.S.C. s 601 et seq.; and (v) the
rule arbitrarily and capriciously ignores the needs of the
elderly and disabled. These five contentions are considered
in the following sections.

A

In determining the need for the Limitations Rule, as well
as its predicted impact on noise in the Park, the FAA
employed the standards announced in the Park Service's 1999
Change in Noise Evaluation Methodology. Although the new
methodology continues to use the three-decibels-above-
ambient threshold for measuring audibility in the more-
developed areas of the Park, it substitutes an eight-decibels-
below-ambient threshold for use in backcountry areas. The
Air Tour Association contends that this reflects a new inter-
pretation of the statutory term "natural quiet." It argues
that while the previous interpretation was based on "noticea-
bility," measuring sounds that a person not engaged in active
contemplation of the Park likely would notice, the new inter-
pretation substitutes a "detectability" standard, measuring
any sound that is detectable to a vigilant observer.

The Air Tour Association further asserts that this change is
unlawful because it is inconsistent with our opinion in Grand
Canyon I, which upheld the three-decibels-above-ambient
threshold. The new threshold for Zone Two is contrary to
Grand Canyon I, the tour operators argue, because it is
unrelated to how visitors actually experience aircraft noise in
the Park. According to the Air Tour Association, visitors to
the Grand Canyon do not attentively listen for such sounds,
and would not notice below-ambient-level noise. Because the
new standard is inconsistent with Grand Canyon I, and
because it represents a change from the prior standard, the
Air Tour Association urges us to accord it less deference than
we would ordinarily extend to an agency determination.

We must begin with the last point, because it misperceives
the scope of our review. The Supreme Court "has rejected

the argument that an agency's interpretation 'is not entitled
to deference because it represents a sharp break with prior
interpretations' of the statute in question." Rust v. Sullivan,
500 U.S. 173, 186 (1991) (quoting Chevron, 467 U.S. at 862).
An agency is not required to establish "rules of conduct to
last forever," but rather "must be given ample latitude to
adapt [its] rules and policies to the demands of changing
circumstances." State Farm, 463 U.S. at 42 (internal quota-
tion marks omitted). Nor does the fact that we previously
affirmed an agency interpretation fix that interpretation in
stone. In Grand Canyon I we held that the standards
employed in the 1996 Final Rule reflected a reasonable
construction of the Overflights Act. 154 F.3d at 469. That
does not preclude a new standard, promulgated pursuant to
notice and comment as this one was, from being reasonable as
well. What the Park Service must do to sustain its decision
is justify the change in course with a "reasoned analysis."
State Farm, 463 U.S. at 57 (quoting Greater Boston Televi-
sion Corp. v. FCC, 444 F.2d 841, 852 (D.C. Cir. 1970)); see
Rust, 500 U.S. at 186-87.

The Park Service has provided that reasoned analysis.
First, it contends that it has not changed the underlying
definition of natural quiet. See Change in Noise Evaluation
Methodology, 64 Fed. Reg. at 38,006. The original 1994 NPS
definition, applied in the 1996 rule, was "no aircraft audible,"
1994 NPS Report at 182; 1996 Environmental Assessment at
4-2, which is the same definition we upheld as reasonable in
Grand Canyon I, see 154 F.3d at 461-62. According to the
agency, it continued to employ that definition during the
Limitations Rule rulemaking. See Change in Noise Evalua-
tion Methodology, 64 Fed. Reg. at 38,007 ("[A]udibility is the
basis for assessing progress toward the legislatively mandat-
ed goal of substantially restoring natural quiet."); id. at
38,011 ("Natural quiet remains the same as 'no aircraft
audible.' ").

What has changed, the Park Service says, is the threshold
it uses to measure audibility in Zone Two. The agency
agrees with the tour operators that the 1996 rule used a
noticeability threshold for determining when sounds become

audible, based on "the level at which visitors engaged in
activities other than contemplation of the national park are
likely to hear aircraft noise." Change in Noise Evaluation
Methodology for Air Tour Operations Over Grand Canyon
National Park, 64 Fed. Reg. 3969, 3971 (proposed Jan. 26,
1999) [hereinafter Proposed Change in Noise Evaluation
Methodology]; see also Change in Noise Evaluation Method-
ology, 64 Fed. Reg. at 38,007. That, it says, is what led to the
original three-decibels-above-ambient standard. But the
Park Service contends that since 1996 it has gathered more
data and performed additional research. Id. That data, col-
lected in Grand Canyon National Park, shows that "an active
listener could hear aircraft when their sound levels were
between 8 and 11 [decibels] below the A-weighted ambient."
FSEA at 4-5; see Change in Noise Evaluation Methodology,
64 Fed. Reg. at 38,007; Proposed Change in Noise Evaluation
Methodology, 64 Fed. Reg. at 3971-72.3

The new studies resulted in the Park Service's decision to
divide the Park into two zones, and to use different noise
thresholds for each zone. The agency continued to apply the
three-decibels-above-ambient threshold in the more-developed
areas of the Park, where visitors were often engaged in
activities other than contemplation. But it employed the new,
eight-decibels-below-ambient standard in the remaining,
mostly backcountry areas. In those areas, the Park Service
was concerned about the experience of visitors "sitting quietly
but actively seeking to experience the natural quiet and
solitude of the park." Proposed Change in Noise Evaluation

__________
3 The A-weighted level of a sound is a single number determined
by combining the sound levels in all frequencies. This combin-
ing de-emphasizes the low and high frequencies in a manner
similar to the sensitivities of human hearing. The A-weighted
level is widely accepted as one of the best over-all sound level
metrics for analysis of transportation noise. It has been shown
to correlate well with human assessment of the loudness or
noisiness of a sound.

Change in Noise Evaluation Methodology, 64 Fed. Reg. at 38,001.

Methodology, 64 Fed. Reg. at 3971. The Park Service's
evidence is that such visitors are able to perceive aircraft
noise at the lower decibel level. Id. As the agency explains,
"the threshold for Zone Two is set at 8 decibels below the
average ambient sound levels" because it is "a threshold
which reflects the point at which aviation noise can be heard
(i.e., audible) by ground visitors seeking to experience the
natural and cultural soundscapes of national parks." Id. at
3972. Thus, far from representing a lack of concern for
visitors' experiences, the agency adopted the distinction be-
tween the two zones precisely because visitors experience
sound differently in different areas.

Nor does the use of a threshold below the ambient indicate,
as the Air Tour Association contends, that aircraft noise is
banned even though it cannot be heard above the natural
ambient sound. Rather, it reflects the Park Service's new
understanding that audibility depends not just on volume
(loudness), but also on frequency (pitch). Change in Noise
Evaluation Methodology, 64 Fed. Reg. at 38,011. As the
agencies explain, "studies conducted in the [Park] for the
NPS ... have shown that individuals who are actively listen-
ing can hear aircraft at lower levels than the ambient A-
weighted sound levels ... because aircraft sound often con-
tains tones that are not present in the natural ambient
sound." FSEA at 4-5 (emphasis in original); see Change in
Noise Evaluation Methodology, 64 Fed. Reg. at 38,012.4
Hence, in order to preserve the definition of natural quiet as

__________
4 The Park Service offers this example:

[D]uring a concert if we listen for a high note on the piccolo,
only the portion of the background sound ... that is of nearly
the same pitch can interfere with our ability to hear the piccolo.
The base violins can play as loudly as they like without the
piccolo becoming inaudible.... Hence, the relative overall
"loudness" of the background and the target sounds is not the
key factor in the detection process. For a target sound to be
audible, it must ... be louder than ... background sounds in
the same frequency band.

Change in Noise Evaluation Methodology, 64 Fed. Reg. at 38,011.

"no aircraft audible" in the backcountry, the Park Service
concluded that it was necessary to change the threshold of
audibility to below the ambient level.

We conclude that the Park Service's explanation for its
change in methodology--as well as its explanation of why that
change remains true to the agency's original principles--are
reasonable, and that the change is neither arbitrary nor
capricious.

B

We next consider the Air Tour Association's "scientific"
challenges to the "noise model and related sound data" used
by the FAA and the Park Service in crafting the Limitations
Rule. Air Tour Ass'n Br. at 15. Again, our standard of
review is important. As a general matter, we "defer to the
agency's reasonable exercise of its judgment and technical
expertise" in the area of "aircraft noise." Grand Canyon I,
154 F.3d at 460; see also Sierra Club v. Dep't of Transp., 753
F.2d 120, 129 (D.C. Cir. 1985). When an agency uses a
computer model, it must "explain the assumptions and meth-
odology used in preparing the model and, if the methodology
is challenged, must provide a complete analytic defense."
Small Refiner Lead Phase-Down Task Force v. EPA, 705
F.2d 506, 535 (D.C. Cir. 1983) (internal quotation marks
omitted). Ultimately, however, the "scientific nature" of a
model "does not easily lend itself to judicial review," and our
review "proceeds with considerable deference to the agency's
expertise." Appalachian Power Co. v. EPA, 135 F.3d 791,
802, 814 (D.C. Cir. 1998). The principal question for us is
whether the agencies' explanation of the model's assumptions
and methodology is reasonable. See generally Small Refiner,
705 F.2d at 535.5

__________
5 Any agency model that satisfied this standard would also satisfy
the directive in the National Parks Air Tour Management Act of
2000, passed the day after the publication of the Limitations Rule,
that the agencies employ "reasonable scientific methods" when
assessing aircraft sound levels within the national parks. Pub. L.

 The agencies used a noise model developed by the FAA,
known as the Integrated Noise Model (INM), to assess noise
impacts in the Park. The INM is "the most widely used
civilian software program for analyzing noise effects relating
to changes in airspace use." 2000 NPS Review at 9. The Air
Tour Association does not challenge the use of the model
itself, but rather the way in which it was applied to the noise
problem in the Grand Canyon.

First, the Air Tour Association cites excerpts from the
report of a Park Service consultant, to the effect that "a more
complex metric is required to achieve satisfactory accuracy."
Air Tour Ass'n Br. at 17 (quoting 2000 NPS Review app. F, at
23). But as the government points out, the "more complex
metric" recommended by the consultant was the one ultimate-
ly used by the agencies, see 2000 NPS Review app. F, at 23,
and the consultant's overall conclusion was that "the science
being used by NPS for noise modeling is grounded on exten-
sive and valid scientific data," id. at Executive Summary 4;
see id. at 1, 8. The Air Tour Association also contends that
the INM overestimates aircraft noise because the agencies
disabled an algorithm that adjusts for "lateral attenuation"--
sound absorption by the ground and the air near the ground.
But the FAA reasonably responds that the algorithm was
removed because it is used to account for the effect of "flat,
acoustically soft terrain, such as grass, as would be found in
the vicinity of most major airports." FSEA at 4-7. As the
FAA explains, for visitors near the rim of the Canyon "there
is effectively no ground surface between the source and
receiver," and in the "vast majority of other locations ... the
ground surface is made up of acoustically hard rock and
packed dirt." Id.6

__________
No. 106-181, tit. VIII, s 808, 114 Stat. 185, 194 (set out at 49
U.S.C.A. s 40128 note).

6 The Air Tour Association further criticizes the Park Service for
excluding from its analysis data recorded at a number of collection
sites. The Service's experts explain, however, that this data was
collected before the plan for measuring sound levels was finalized,

 Finally, the Air Tour Association cites its own expert, John
R. Alberti, for the proposition that the audibility threshold
used by the Park Service in Zone Two, eight decibels below
the ambient sound level, is scientifically unsupportable be-
cause aircraft noise at that level cannot be heard by a listener
with normal hearing. Air Tour Ass'n Br. at 18 (citing Com-
ments of JR Engineering to NPS at 3 (Mar. 20, 1999)). The
2000 NPS Review contains a detailed critique, based on
analysis by the agency's own experts, of the Alberti study.
2000 NPS Review app. B, at 5-11. The most serious problem
with that study, the Park Service explains, is that it used data
"that cannot be characterized ... as representative of the
park or its many levels of ambient sound." Id. at 5. In
essence, Alberti began with a significantly lower ambient
level than that actually measured in the Park's acoustic
environments, making his calculation of eight decibels below
that level considerably quieter than the actual audibility
thresholds the agencies will employ in Zone Two. Id. at 5, 8.
Moreover, whatever the theoretical argument, the Park Ser-
vice reports as a fact that "technicians monitoring the sound
environment" in the Park "identified aircraft noise at A-
weighted levels of 8-12 decibels below the average A-
weighted natural ambient sound levels." Proposed Change in
Noise Evaluation Methodology, 64 Fed. Reg. at 3972.

We conclude that the agencies and their experts have
presented a satisfactory analytic defense of their model, and
therefore reject this challenge from the Air Tour Association.
See Marsh v. Oregon Natural Res. Council, 490 U.S. 360, 378
(1989) ("When specialists express conflicting views, an agency
must have discretion to rely on the reasonable opinions of its
own qualified experts.").

C

The Air Tour Association also contends that the FAA acted
arbitrarily and capriciously by issuing the Limitations Rule

__________
and that the excluded sites were not representative of the predomi-
nant acoustic environments in the Park. 2000 NPS Review app. C,
at 8.

without first promulgating a quiet technology rule. As de-
scribed by the Association, such a rule would establish stan-
dards for identifying "quiet aircraft technology" and would
implement "incentive" flight paths for aircraft employing such
technology. Air Tour Ass'n Br. at 20-21. According to the
Association, Congress first ordered the FAA to establish
quiet technology standards in the 1987 Overflights Act, and
then did so again in the National Parks Air Tour Manage-
ment Act of 2000, Pub. L. No. 106-181, tit. VIII, 114 Stat. 185
(set out at 49 U.S.C.A. s 40128 note). Nonetheless, the
FAA's 1996 proposed rule on quiet technology is still not
final.

The FAA notes that there is some irony in the air tour
operators' new-found interest in a quiet technology rule, as
they have previously opposed such a rule. Limitations Rule,
65 Fed. Reg. at 17,714. We also note that the operators have
misread both the Overflights Act and the Air Tour Manage-
ment Act. The Overflights Act does not contain a provision
regarding quiet technology. The Air Tour Management Act
does direct the FAA, by April 5, 2001, to "designate reason-
ably achievable requirements" for aircraft to be "considered
as employing quiet aircraft technology" for purposes of relief
from caps on flight operations. s 804(a), (c). However, the
Act also provides that "if the Administrator [of the FAA] ...
will not be able to make such designation" by that time, the
FAA shall report to Congress "the reasons for not meeting
such time period." s 804(a). The FAA submitted the re-
quired report. See FAA, U.S. Dep't of Transp., Quiet Air-
craft Technology for Grand Canyon (2001).7

__________
7 Another statute, the Airport and Airway Safety, Capacity, Noise
Improvement, and Intermodal Transportation Act of 1992, also
mentioned "quiet aircraft technology." Pub. L. No. 102-581, s 134,
106 Stat. 4872, 4888 (set out at 16 U.S.C.A. s 1-1a note). Like the
others, it did not direct the FAA to implement a quiet technology
rule; it merely instructed the agency to submit a plan of action to
Congress "to manage increased air traffic over Grand Canyon
National Park ... to meet the requirements established by [the
Overflights Act], including any measures to encourage or require
the use of quiet aircraft technology by air tour operators." Id.

 Moreover, the FAA has offered a reasonable explanation
for why it has not yet issued a quiet technology rule. Accord-
ing to the agency, there continue to be unresolved technical
issues regarding available technology and its noise effects
within the Grand Canyon, including the most basic issue:
"how to define quiet technology/noise efficiency." Limitations
Rule, 65 Fed. Reg. at 17,714. The FAA's brief represents
that the agency plans to issue a supplemental notice of
proposed rulemaking to invite further comment on the pro-
posed rule. Resp't's Br. at 32. In light of the technical
difficulties the FAA is currently having in finalizing a quiet
technology rule, its decision to go ahead with a cap on flights
to ensure continuing progress toward restoring natural quiet
is reasonable. As we explained in Grand Canyon I, "ordi-
narily, agencies have wide latitude to attack a regulatory
problem in phases." 154 F.3d at 471.

D

The Regulatory Flexibility Act requires an agency promul-
gating a final rule to prepare a regulatory flexibility analysis
that, inter alia, describes:

the steps the agency has taken to minimize the signifi-
cant economic impact on small entities consistent with
the stated objectives of applicable statutes, including a
statement of the factual, policy, and legal reasons for
selecting the alternative adopted in the final rule and
why each one of the other significant alternatives to the
rule considered by the agency which affect the impact on
small entities was rejected.

5 U.S.C. s 604(a). Although the FAA did prepare a regulato-
ry flexibility analysis for the Limitations Rule, see 65 Fed.
Reg. at 17,729-31, the Air Tour Association attacks that
analysis as defective in two principal respects: (1) it underes-
timates the burden the Limitations Rule imposes on tour
operators by using data regarding the number of tours from
May 1997 to April 1998, a period they contend was nonrepre-
sentative; and (2) it fails to consider significant alternatives

that would minimize the Rule's economic impact on small
entities.

Neither attack is justified. As the FAA explains, it select-
ed the 1997-98 base year because it was "the most accurate
and current data available during the period that this rule
was being drafted." Limitations Rule, 65 Fed. Reg. 17,718.
Indeed, selection of that year actually favors the air tour
operators, since subsequent data shows that in the following
year the number of air tours declined. Id. Moreover, the
agency plainly did consider alternatives to the rule, expressly
listing nine and explaining its reasons for rejecting them. Id.
at 17,729-30. The Air Tour Association does not point to any
alternative that it believes was unreasonably rejected other
than the quiet technology rule, which, as we have discussed
above, the FAA has reasonably put off for the present.
Because the FAA's analysis was reasonable, the Limitations
Rule survives the Tour Association's challenge under the
Regulatory Flexibility Act. See Allied Local & Reg'l Mfrs.
Caucus v. EPA, 215 F.3d 61, 79-80 (D.C. Cir. 2000); Grand
Canyon I, 154 F.3d at 470-71.

E

Finally, the Air Tour Association contends that the Limita-
tions Rule is arbitrary and capricious because it ignores the
needs of the elderly and disabled, who find it harder to visit
the Park on the ground. The Association asserts that "nearly
half of all Grand Canyon visitors never 'visit the Park on the
ground,' " and that "[a]s many as half of these 'air only'
visitors are elderly, disabled or mobility impaired and have no
meaningful way to experience Grand Canyon except by recre-
ational air tour." Air Tour Ass'n Br. at 22-23 (quoting
Limitations Rule, 65 Fed. Reg. at 17,716). By "reducing the
annual number of recreational air tours" that can be taken by
the elderly and disabled, the Association contends that the
Limitations Rule fails to accommodate their needs. Id.

Nearly every contention recounted in the preceding para-
graph contains a factual error. It is not true, for example,
that nearly half of Grand Canyon visitors never visit on the

ground. Rather, the authority cited by the Association actu-
ally states that over half of air tour visitors also visit the
Park on the ground. Limitations Rule, 65 Fed. Reg. at
17,716. Indeed, in another part of its brief, the Air Tour
Association states that of the more than 5 million people who
visit the Grand Canyon annually, only 750,000 visit by air
tour. Air Tour Ass'n Br. at 9. And the Association further
concedes that there is no record support for its claim that as
many as half of the "air only" visitors are elderly, disabled, or
mobility impaired. As the Association notes, "the Adminis-
trative Record lacks any reference to the demographics of 'air
only' Grand Canyon visitors." Id. at 23 n.5.

Finally, there is also no truth to the contention that the
agencies have failed to consider the needs of the elderly and
disabled. As the government explains, air tours are not the
only means by which these groups may view the Canyon, as
the Park has available such accommodations as handicapped-
accessible trails, mule rides, and raft trips. See Resp't's Br.
at 34. Nor does the Limitations Rule eliminate or even
"reduc[e] the annual number of recreational air tours"; it
merely caps the number of flights at current levels. Limiting
the number of visitors at a given time in a national park is a
standard measure used to protect park resources, see Limita-
tions Rule, 65 Fed. Reg. at 17,716, and its application to the
resource of "natural quiet" is not unreasonable. It is there-
fore not surprising that the Air Tour Association failed to
return to this argument in its reply brief.8

__________
8 The Air Tour Association also argues that the Limitations Rule
violates the equal protection component of the Fifth Amendment
because it exempts flights to and from the Hualapai Indian Reser-
vation from each tour operator's annual allocation. See 14 C.F.R.
s 93.319(f). The Association's argument is founded upon Adarand
Constructors, Inc. v. Pena, 515 U.S. 200, 227 (1995), in which the
Supreme Court held that "all racial classifications ... must be
analyzed by a reviewing court under strict scrutiny." In Morton v.
Mancari, 417 U.S. 535 (1974), however, the Court upheld an em-
ployment preference for Indians in the face of an argument similar
to that of the Air Tour Association, declaring that, "[a]s long as the
special treatment can be tied rationally to the fulfillment of Con-

 III

The Grand Canyon Trust raises two challenges of its own
to the decisions of the FAA. First, the Trust contends that
the agency unlawfully altered the Park Service's definition of
substantial restoration of the natural quiet: from 50% of the
Park experiencing natural quiet for 75% of "any given day,"
to 50% of the Park experiencing natural quiet for 75% of "the
average annual day." Trust Br. at 12. Second, the Trust
argues that the FAA's noise methodology is flawed because it
only accounts for noise from commercial air tours, while
ignoring noise from other types of aircraft. As a consequence
of both errors, the Trust contends that the FAA has greatly
overestimated the progress that the Limitations Rule makes
toward restoring natural quiet.

We consider these two challenges in sections B and C
below. Before doing so, we address the government's objec-
tions to our considering the Trust's arguments at all.

A

The Trust filed its petition for review pursuant to 49 U.S.C.
s 46110(a), which authorizes review in this circuit of certain
final "order[s] issued by the ... Administrator of the Federal

__________
gress' unique obligation toward the Indians," such treatment must
be upheld. Id. at 555. And there is no dispute that the Hualapai
exception is at least rationally related to "the government's interest
in fulfilling its trust obligation" to the Tribe. Narragansett Indian
Tribe v. National Indian Gaming Comm'n, 158 F.3d 1335, 1340
(D.C. Cir. 1998). Although the Air Tour Association contends that
Adarand effectively overruled Mancari, the Supreme Court has
made clear that the lower courts do not have the power to make
that determination. See Agostini v. Felton, 521 U.S. 203, 237 (1997)
("[I]f a precedent of this Court has direct application in a case, yet
appears to rest on reasons rejected in some other line of decisions,
the Court of Appeals should follow the case which directly controls,
leaving to this Court the prerogative of overruling its own deci-
sions." (internal quotation marks omitted)). And this circuit has
continued to apply Mancari post-Adarand. See Narragansett
Tribe, 158 F.3d at 1340.

Aviation Administration."9 The FAA does not dispute that
the Limitations Rule is a final order reviewable under that
section. See Limitations Rule, 65 Fed. Reg. at 17,708 ("This
final rule constitutes final agency action under 49 U.S.C.
46110."). The agency contends, however, that the Trust's
challenges are not attacks on the FAA's order, that the
decisions the Trust does attack are not final, and that the
Trust's challenges are not ripe for judicial review.

The FAA's first argument is that the Trust's challenges are
not to the Limitations Rule, but rather to its definition of "the
substantial restoration of the natural quiet" and to its meth-
odology for measuring aircraft noise. These, the agency
contends, are neither reviewable orders themselves nor parts
of the Limitations Rule. But whether the definition and
methodology are independently "orders" is a question we
need not decide, as there is no doubt that they are both part
of the rulemaking record and integral to the agency's ratio-
nale for promulgating the Limitations Rule. The fact that
they are contained in a separate document, the Final Supple-
mental Environmental Assessment, is of no moment; we
similarly reviewed the Park Service's original definition of
"substantial restoration of the natural quiet" in Grand Can-
yon I, even though that definition was contained in the 1994
NPS Report rather than in the 1996 Final Rule. The FSEA
was issued concurrently with the Limitations Rule, and the
rule's Federal Register notice repeatedly relies on it and
incorporates it by reference. See, e.g., Limitations Rule, 65
Fed. Reg. at 17,714 (noting that the noise methodologies
"used in support of this rule are explained further" in the
FSEA); id. at 17,709-10 (noting that the FAA placed the
draft Supplemental Environmental Assessment in the rule-
making docket for public comment, and that those comments
are "addressed in the [FSEA] issued concurrently with this
final rule"). Just as we may examine other record material
that provides the underpinnings for the Limitations Rule, so

__________
9 See also Puget Sound Traffic Ass'n v. Civil Aeronautics Bd.,
536 F.2d 437, 438-39 (D.C. Cir. 1976) (holding that review under the
predecessor to s 46110(a) extended only to final orders).

also may we review analytic documents issued by the agency
that elaborate upon the rule's rationale and assess its im-
pact.10

The FAA's second argument is that the definition and
methodology are not final because "the course the FAA may
follow in future rulemakings" is "uncertain," and the Park
Service "could revise" its recommendations. But as we have
only recently reiterated, "if the possibility ... of future
revision in fact could make agency action non-final as a
matter of law, then it would be hard to imagine when any
agency rule ... would ever be final as a matter of law."
General Electric Co. v. EPA, 290 F.3d 377, 380 (D.C. Cir.
2002) (citing Appalachian Power Co. v. EPA, 208 F.3d 1015,
1022 (D.C. Cir. 2000)). Both decisions at issue here represent
fully-crystallized determinations that the FAA used in pro-
mulgating the Limitations Rule, and that it has given every
indication it will employ in future rulemakings. Because they
represent the "consummation of the agency's decisionmaking
process" and are determinations from which "legal conse-
quences will flow," they are final for purposes of judicial
review. Bennett v. Spear, 520 U.S. 154, 177-78 (1997).

Third, the FAA contends that the Trust's challenges are
not ripe for review. To determine whether a challenge to
final agency action is ripe, we consider "the fitness of the
issues for judicial decision and the hardship to the parties of
withholding court consideration." General Electric, 290 F.3d
at 380 (quoting Abbott Labs. v. Gardner, 387 U.S. 136, 149

__________
10 Two circuit cases cited by the FAA to refute this proposition
are inapposite. In Molycorp, Inc. v. EPA, 197 F.3d 543, 545 (D.C.
Cir. 1999), we concluded that we were without jurisdiction to review
an EPA Technical Background Document. Unlike the FSEA, that
document was not issued as part of or in support of a rulemaking,
and was "intended solely to provide information to the public and
the regulated community." Id. at 545-46. In Florida Power &
Light Co. v. EPA, 145 F.3d 1414, 1418-19 (D.C. Cir. 1998), we held
that a statement in the preamble of a proposed rule was unreview-
able, noting that--unlike the FAA in this case--the EPA had yet to
take final action on the proposed rule to which the statement was
relevant.

(1967)). "In determining the fitness of an issue for judicial
review we look to see whether the issue is purely legal,
whether consideration of the issue would benefit from a more
concrete setting, and whether the agency's action is sufficient-
ly final." Id. We have already determined that the FAA's
definition and methodology are sufficiently final. The chal-
lenges to those decisions are fit for review in other respects
as well.

Whether the FAA may use an average annual day as part
of its definition of substantial restoration, and whether it may
ignore the noise of non-tour aircraft, are purely legal issues.
See Fox Television Stations, Inc. v. FCC, 280 F.3d 1027, 1039
(D.C. Cir. 2002) (finding an issue fit for review because
whether an agency determination is arbitrary and capricious
or contrary to law is a "purely legal" question). And they
arise in the concrete setting of the Limitations Rule. More-
over, we see no reason to believe that our consideration of
these issues would benefit from postponing review. As we
have already noted, there is nothing "tentative or interlocu-
tory" about the FAA's definition and methodology, Appala-
chian Power, 208 F.3d at 1022, and there is no prospect that
they will change in the near future unless questioned by this
court. At oral argument, the government suggested that the
Trust should wait until the FAA claims natural quiet has been
restored to 50% of the Park, and then appeal if it disagrees.
But the Trust's argument is that "under FAA's interpretation
of the [Overflights] Act it will never achieve that mandate."
Trust Br. at 23-24 (citing our statement in Grand Canyon I,
154 F.3d at 477, that it would be arbitrary and capricious for
the FAA to have no intention of achieving the Act's goal). If
we assume as we must for purposes of this ripeness analysis
that the Trust's challenges are valid,11 then its argument is
well-founded. For the FAA to substantially restore natural
quiet to 50% of the Park on "any given day" (as the Trust
contends is required), the agency would have to restore quiet
to significantly more than 50% of the Park on the "average

__________
11 See, e.g., Better Gov't Ass'n v. Dep't of State, 780 F.2d 86, 94
(D.C. Cir. 1986).

annual day." See infra Part III.B. And the FAA has never
indicated that it has any plans to achieve such a result.

Finally, again assuming the validity of the Trust's argu-
ments, it will suffer hardship if we withhold consideration
now.12 The Trust's members visit the Park and wish to
experience its natural serenity. As the FAA stated in the
Federal Register notice for the Limitations Rule, it promul-
gated that rule on the assumption that it would make "signifi-
cant steps towards substantially restoring natural quiet." 65
Fed. Reg. at 17,713. Indeed, the FAA predicted that the
Limitations Rule, together with the 2000 Airspace Rule and
other existing regulations, would substantially restore natural
quiet to more than 41% of the Park. Id. at 17,724. But the
Trust contends that the flaws in the agency's definition and
methodology cause it to significantly overestimate its prog-
ress toward the statutory goal (correctly understood)--so
much so that the true percentage of the Park that will be
substantially restored to natural quiet is below 19%.13

Crediting the government with a good-faith intent to
achieve the goal of the Overflights Act, the Trust contends
that if the FAA knew it was making as little progress toward
achieving that goal as in fact it is, it would have "to impose
more stringent regulation on air tours to make the progress it
sought to make in the [Limitations] Rule." Trust Reply Br.
at 3. The FAA might, for example, reduce rather than
merely cap the number of authorized air tour flights. We
cannot say that the prospects of the Trust achieving such
benefits are speculative merely because they depend upon the

__________
12 See City of Houston v. HUD, 24 F.3d 1421, 1431 n.9 (D.C. Cir.
1994) ("[W]here there are no institutional interests favoring post-
ponement of review, a petitioner need not satisfy the hardship
prong.").

13 As we discuss in Part III.B below, the Trust points to evidence
in the record that, using an "any given day" rather than an
"average annual day" standard, only 19% of the Park will be
substantially restored to natural quiet. Moreover, as we discuss in
Part III.C, even that figure is based on a model that ignores noise
from non-tour aircraft, and record evidence indicates that including
that noise may well further reduce the percentage of the Park to
which natural quiet will be substantially restored.

government's good faith. Cf. FEC v. Akins, 524 U.S. 11, 25
(1998) (holding that the plaintiff had standing even though the
agency might have reached the same decision had it agreed
with the plaintiff's view of the law); id. ("If a reviewing court
agrees that the agency misinterpreted the law, it will set
aside the agency's action and remand the case--even though
the agency ... might later, in the exercise of its lawful
discretion, reach the same result for a different reason.").
We are therefore satisfied that the Trust's contentions are
ripe for review and proceed to address them.

B

As we have discussed, section 3(b)(1) of the Overflights Act
gives the Secretary of the Interior the responsibility to
develop a definition of "substantial restoration of the natural
quiet," a responsibility that the National Park Service has
performed on behalf of the Secretary. See Grand Canyon I,
154 F.3d at 468. In its 1994 report, the Park Service defined
the term as requiring that 50% of the Park experience
natural quiet at least 75% of "the day." See 1994 NPS
Report at 182. In its 2000 Final Supplemental Environmen-
tal Assessment, the FAA stated that it interpreted the phrase
"the day" in the Park Service's definition to mean the "aver-
age annual day." FSEA at 4-18 to 4-19; see id. at 4-12.

The Trust contends that the FAA's interpretation is unlaw-
ful because it substitutes a new FAA definition, "the average
annual day," for what the Trust believes to be the Park
Service's meaning, "any given day." The Trust further con-
tends that the FAA's interpretation results in a definition of
the statutory term, "substantial restoration of the natural
quiet," that is arbitrary and unreasonable. The FAA replies
that because the phrase "the day" is ambiguous, this court
should defer to its interpretation.

In Grand Canyon I we held that the statutory term was
ambiguous, and we therefore deferred to the Park Service's
definition. 154 F.3d at 466-67 (citing Chevron, 467 U.S. at
841-43). There is no question that the phrase, "the day," in
the Park Service's definition is also ambiguous, and that the

Park Service is entitled to deference for its interpretation of
its own definition. See Auer, 519 U.S. at 461. The problem
for the FAA, however, is that it is not the Park Service, and
"deference is inappropriate when [an agency] interprets regu-
lations promulgated by a different agency." Office of Pers.
Mgm't v. FLRA, 864 F.2d 165, 171 (D.C. Cir. 1988); see Dep't
of the Treasury v. FLRA, 837 F.2d 1163, 1167 (D.C. Cir.
1988). Nor is the FAA entitled to deference for its own
interpretation of "substantial restoration of the natural
quiet," as Congress expressly reserved for the Secretary of
the Interior the authority to interpret that statutory term.
Overflights Act s 3(b)(2); see Grand Canyon I, 154 F.3d at
468; see also Dep't of the Treasury, 837 F.2d at 1167 ("Under
the law of this circuit, when an agency interprets a statute
other than that which it has been entrusted to administer, its
interpretation is not entitled to deference.").

As far as we can determine, the National Park Service has
addressed the meaning of "the day" in its definition of
substantial restoration on only one occasion: the January
2000 review of its Change in Noise Evaluation Methodology,
issued just three months before the FAA issued the Limita-
tions Rule.14 In that review, the Park Service stated that it
had defined the statutory term, "substantial restoration of the
natural quiet," as:

__________
14 The FAA points us to the "inside cover page" of the FSEA as
demonstrating Park Service approval of the FAA's average annual
day interpretation. Resp't's Br. at 49 n.8. The FSEA, however, is
a lengthy FAA document, the inside cover of which says nothing
more than that the Department of the Interior is a "cooperating
agency." The FAA also argues that the Park Service adopted an
"average annual day" standard in the Change in Noise Evaluation
Methodology, in which the Service declared that "the time period of
interest is the day (i.e., the average 12 daylight hours)." 64 Fed.
Reg. at 38,007. But that declaration merely states that substantial
restoration of natural quiet must be achieved during daylight, which
on average lasts 12 hours, and that quiet at night (when air tours do
not fly) does not count in the calculation. See Grand Canyon I, 154
F.3d at 475 n.19. It says nothing about averaging noise over an
entire year.

 Fifty percent or more of the Park achieving 'natural
quiet' (i.e., no aircraft audible) for 75-100 percent of the
day. This definition is a threshold not to be exceeded on
any given day ... and refers to the 12 hour daylight
period.... The NPS believes that it is a reasonable and
rational definition.

2000 NPS Review at 16 (emphasis added); accord id. at 4-5.
This statement clearly supports the Trust's contention that
"the day" means "any given day," and contradicts the FAA's
view that it means "the average annual day."15

__________
15 Although the government contends that the 2000 NPS Review
is merely a "technical document" that cannot be taken to represent
the views of the agency, Resp't's Br. at 51, the review gives every
appearance of being authoritative. Indeed, the introduction states
that "the National Park Service (NPS) has prepared the following
report" in "response to language in [a] House Appropriations Bill."
2000 NPS Review at 4. In any event, whether or not the review is
decisive affirmative evidence of the Secretary's view, it creates
more than enough doubt to justify a remand.

At oral argument, government counsel also suggested that, be-
cause the Park Service is a co-respondent in this case, we should
assume that the views expressed in the government's brief regard-
ing the meaning of "the day" are those of the Park Service.
Although deference to an agency's interpretation of its own regula-
tion "is not to be withheld merely because the agency's reading of
the regulation comes in form of a legal brief," that rule applies only
where there is " 'no reason to suspect that the interpretation does
not reflect the agency's fair and considered judgment on the matter
in question.' " Drake v. FAA, 291 F.3d 59, 68 (D.C. Cir. 2002)
(quoting Auer, 519 U.S. at 462). "In conducting this inquiry, we
consider whether the agency has 'ever adopted a different interpre-
tation of the regulation or contradicted its position on appeal.' " Id.
at 69 (quoting National Wildlife Fed'n v. Browner, 127 F.3d 1126,
1129 (D.C. Cir. 1997)). Because the 2000 NPS Review did adopt a
different interpretation that contradicts the government's position
on appeal, deference to the interpretation contained in the govern-
ment's brief is inappropriate.

 The FAA's "average annual day" interpretation appears to
be inconsistent not only with the 2000 NPS Review, but also
with the premise of the Park Service's definition of substan-
tial restoration, a premise that the FAA itself has repeatedly
endorsed. That premise is that aircraft noise should be
regulated to enhance the experience of Park visitors. As we
said in Grand Canyon I, both the 1994 NPS Report and the
notice announcing the 1996 Final Rule were replete with
concern for the manner in which aircraft noise affected
visitors' experiences. 154 F.3d at 465-66; see id. at 466
(noting that "in the Park Service's view, natural quiet is a
resource because it is relevant to visitor enjoyment"). And as
we made clear in Part II.A above, the same is true of the
Park Service's explanation of the methodology underlying the
Limitations Rule. Indeed, this premise derives in part from
the Overflights Act itself, which directed the Secretary of the
Interior to issue recommendations "that shall provide for
substantial restoration of the natural quiet and experience of
the park." Overflights Act s 3(b)(1) (emphasis added).

As the Trust points out, the use of an annual average does
not correspond to the experience of the Park's actual visitors.
People do not visit the Park on "average" days, nor do they
stay long enough to benefit from averaging noise over an
entire year. For the typical visitor, who visits the Grand
Canyon for just a few days during the peak summer season,
the fact that the Park is quiet "on average" is cold comfort.
Indeed, the FAA acknowledges that, "[b]ecause many park
visitors typically spend limited time in particular sound envi-
ronments during specific park visits, the amount of aircraft
noise present ... can have great implications for the visitor's
opportunity to experience natural quiet in those particular
times and spaces." Limitations Rule, 65 Fed. Reg. at 17,712.
The problem with an annual average is that it gives equal
weight to summer and winter days, notwithstanding that
there are many more visitors during the former than the
latter. Thus, theoretically the use of an annual average could
permit the statutory standard to be met despite an incessant-
ly noisy summer, assuming that the other seasons were

relatively quiet. On the FAA's view, it could then declare
that it had achieved substantial restoration of natural quiet
and cease any further efforts to restrict aircraft noise.

Nor are these consequences of using an annual average
merely theoretical. The FAA's Final Supplemental Environ-
mental Assessment projects that the 2000 Final Rules will
restore natural quiet in 43.6% of the Park on the "average
annual day." See FSEA at F-4. The same document, how-
ever, reveals that on the average summer day, natural quiet
will be substantially restored in only 31.3% of the Park, and
that on the day with the most air tour overflights the natural
quiet will be substantially restored in only 19.0% of the Park.
Id. This result appears inconsistent with the rationale that
the Park Service offered for its 1994 definition of substantial
restoration, and that we found reasonable in Grand Canyon
I: "Protecting 50% of the Park for 75% of the day gives
[backcountry visitors] at least a reasonable chance of seeing
the less-traveled areas in peace." 154 F.3d at 467; see also
1994 NPS Report at 184.

The FAA's only response to the seeming unreasonableness
of its definition is that its standard guidelines for evaluating
the impact of aircraft noise employ an annual average, and
that the courts have generally deferred to those guidelines.
Resp't's Br. at 54-56 (citing, e.g., City of Bridgeton v. FAA,
212 F.3d 448, 459 (8th Cir. 2000); Morongo Band of Mission
Indians v. FAA, 161 F.3d 569, 578-79 (9th Cir. 1998)). But
those guidelines were developed to evaluate aircraft noise at
airports, see 14 C.F.R. pt. 150; id. s 150.9(b), and it is in that
context that the courts have deferred. See, e.g., Bridgeton,
212 F.3d at 459; Morongo Band, 161 F.3d at 578-79. None
of the cited cases involved a statute that required the FAA to
achieve substantial restoration of the natural quiet, or to
employ a Park Service definition that demands quiet during
75% of "the day."

In sum, we conclude that the FAA's use of an annual
average day for measuring "substantial restoration of the
natural quiet" appears inconsistent with both the Park Ser-
vice's definition of the term and with the premise upon which
that definition was based. As noted in our discussion of
ripeness above, this is not merely a technical dispute. It has
a significant impact both on the FAA's estimate of the

progress its current and planned rules make, and on its
determination of the end point at which it will have fulfilled
the statutory objective. We must therefore remand this issue
for further consideration.

We recognize that, on remand, the Park Service may
declare that it did in fact choose what seems to us to be the
FAA's less reasonable reading of the Service's definition of
"substantial restoration." If it does, however, the Park Ser-
vice must understand that "the Overflights Act ... did not
relieve the government ... of its obligation not to promulgate
a rule that is 'arbitrary, capricious, an abuse of discretion, or
otherwise not in accordance with law.' " Grand Canyon I,
154 F.3d at 469 (quoting 5 U.S.C. s 706(2)(A)). To survive
review under that standard, the Service will have to explain
why its definition is reasonable in light of the premise upon
which it was assertedly based. At this point, the government
has offered no explanation at all.

C

The Trust's second challenge is to an aspect of the FAA's
noise methodology. The Trust argues that, in projecting the
amount of noise experienced by different parts of the Park,
the FAA's noise model only considers noise emitted from air
tour flights. By failing to account for noise from other
aircraft that fly over the Grand Canyon--for example, from
commercial jets, general aviation, and military flights--the
model arbitrarily overstates how quiet the Park really is.
The Trust asserts that if non-tour aircraft noise were includ-
ed in the calculation, the percentage of the Park in which
natural quiet would be substantially restored--even on the
FAA's average annual day--would be significantly less than
the projected 43.6%.

The FAA does not dispute that it excludes non-tour aircraft
from its model, and the Trust's description is in accord with
that of the Final Supplemental Environmental Assessment.
FSEA app. G, at 40. Nor does the FAA dispute that, at least
theoretically, the exclusion of non-tour aircraft could cause it
to overestimate the periods when, and number of places
where, Park visitors substantially experience natural quiet.
The FAA's only response to the Trust's challenge is that

excluding non-tour aircraft is reasonable because they "con-
tribute minimal noise compared to commercial air tour air-
craft.... [The FAA] did not believe the minimal amount of
noise they contribute would affect the accuracy of its esti-
mates." Resp't's Br. at 59.

The agency cites no direct evidence that the noise from
non-tour aircraft is minimal, basing that conclusion solely on
the fact that general aviation flights account "for about 3
percent of all aircraft" in the Park. FSEA app. G, at 40.
What is at issue here, however, is not the number of aircraft
that fly over the Grand Canyon, but rather the percentage of
time that they are audible. Although we might ordinarily
defer to the FAA's expert assurance that the number of non-
tour aircraft is too small to affect the accuracy of the latter
calculation, we cannot defer when the government's own data
suggests the contrary. Here, the record evidence indicates
that the exclusion of non-tour aircraft from the calculation of
the percentage of time that aircraft are audible may well have
a significant impact on the results.

The 1994 NPS Report provides a chart, based on actual
observations, that breaks down the percentage of time that
different types of aircraft were audible at different points in
the Park. 1994 NPS Report at 187. For example, it shows
that at Lipan Point, total aircraft noise--including noise from
air tours, commercial jets, general aviation, and military
flights--was audible during 43% of the observation period.
Air tours, however, could only be heard 30% of the time.
Hence, at that location, excluding non-tour aircraft from the
calculation would overestimate by 13 percentage points the
amount of time during which natural quiet prevailed. Other
locations reveal similarly significant differentials between the
amount of time that some type of aircraft was audible and the
amount of time that air tours could be heard: a 14% differen-
tial at Bright Angel Point, 10% at Toroweap Overlook, and
12% at Phantom Ranch Overlook. Id. When asked at oral
argument, the government could offer no explanation for why
these differentials were immaterial in calculating the areas or
periods in which the Park substantially experiences natural
quiet.

 The FAA also seeks to defend the exclusion of non-tour
aircraft from its noise model on the ground that "FAA's
regulations were not required to address every conceivable
type of aircraft." Resp't's Br. at 60. That is true, and the
FAA can certainly choose to achieve the substantial restora-
tion of natural quiet by regulating air tours alone. But the
FAA cannot dispute that whether or not non-tour aircraft are
regulated, natural quiet does not exist when the sound they
make is audible.16 Nor does the Overflights Act provide any
basis for ignoring noise caused by such aircraft.17 Hence, the
fact that the FAA has chosen not to regulate certain catego-
ries of aircraft does not justify ignoring the sound those
aircraft make when deciding how extensively to regulate
other categories. And in the absence of any reasonable
justification for excluding non-tour aircraft from its noise
model, we must conclude that this aspect of the FAA's
methodology is arbitrary and capricious and requires recon-
sideration by the agency.

IV

For the reasons discussed in Part II, the Air Tour Associa-
tion's petition for review is denied. For the reasons dis-
cussed in Part III, we grant the Grand Canyon Trust's
petition and remand the case to the FAA for further proceed-
ings consistent with this opinion.

__________
16 See, e.g., 2000 NPS Review at 6 (defining "natural quiet" as
"the sound levels associated with a given acoustic environment,
absent any mechanical or manmade noise sources" (emphasis add-
ed)); 1994 NPS Report at 182 (defining "natural quiet" as "no
aircraft audible").

17 See Overflights Act s 3(a) (declaring that "[n]oise associated
with aircraft overflights at the Grand Canyon National Park is
causing a significant adverse effect on the natural quiet and experi-
ence of the park").