REAVLEY, Circuit Judge:
In these consolidated appeals, petitioners challenge various actions by the United States Air Force (Air Force) and the Federal Aviation Administration (FAA) in connection with the Realistic Bomber Training Initiative (RBTI).
Petitioners allege
that the Air Force and FAA failed to follow procedures mandated by the National Environmental Policy Act, 42 U.S.C. §§ 4321-4370Í (NEPA) and its implementing regulations, 40 C.F.R. §§ 150Ó.1-1508.28 (2003) (CEQ regulations), 32 C.F.R. §§ 989.1-989.38 (2004) (Air Force regulations), and ask this court to set aside those agency actions and remand to the agencies for NEPA-sufficient procedure.
We agree that the Environmental Impact Statement (EIS) prepared by the Air Force and adopted by the FAA does not satisfy NEPA and therefore remand to the agencies to prepare a supplemental EIS in accordance with this opinion.
I.
Background
The basis of petitioners’ complaints is the RBTI, a plan to provide airspace and ground-based assets for realistic and integrated B-52 and B-l Bomber flight training within 600 miles of Barksdale and Dyess Air Force Bases. The RBTI includes a Military Operations Area (MOA), linked to a Military Training Route (MTR) by an Electronic Scoring Site system. The MOA provides space, identified to civil and commercial aircraft, where military aircraft can practice air-to-ground and air-to-air training. The MTR is a flight corridor where pilots can practice low-altitude navigation and maneuvers.
Concluding that implementation of the RBTI would constitute a “major action” under NEPA, the Air Force prepared an EIS.
The FAA participated in the NEPA process as a cooperating agency.
The EIS analyzed three alternative locations for the RBTI and a no action alternative. Two months after issuing the final EIS, the Air Force issued a Rule of Decision (ROD) adopting its preferred alternative (Alternative B). Alternative B, located mostly in western Texas, would modify and enlarge existing MTR Instrument Route 178 (IR-178) and create Lancer MOA by consolidating and expanding three existing MOAs. The FAA adopted the final EIS and approved Lancer MOA and the IR-178 modifications.
Petitioners are Davis Mountains Trans-Pecos Heritage Association (DMTPHA), a nonprofit corporation whose members are farmers, ranchers, and business people living and working in the areas underlying the RBTI airspace, and similarly situated named individuals. Concerned with potential impacts of the RBTI on underlying land, petitioners challenged the NEPA compliance of the Air Force and several named federal defendants in the district court.
Davis Mountains Trans-Pecos Heritage Association v. U.S. Air Force,
249 F.Supp.2d 763 (N.D.Tex.2003);
Welch v. U.S. Air Force,
249 F.Supp.2d 797 (N.D.Tex.2003) (hereinafter “Air Force cases”). Petitioners seek review of that court’s summary judgments in favor of defendants as well as the FAA’s approval of Lancer MOA and modified IR-178.
II.
Jurisdiction
This court has jurisdiction to review the district court’s grants of summary judgment in the Air Force cases under 28 U.S.C. § 1291. We have jurisdiction to review the FAA’s approvals under 49
U.S.C. § 46110(a), providing for review of FAA orders in the Courts of Appeals. We lack jurisdiction, however, to hear any claims of the Welch intervenors in the FAA appeal not raised by petitioners in that case.
United Gas Pipe Line Co. v. FERC,
824 F.2d 417, 434-38 (5th Cir.1987). In
United Gas,
we held that intervenors in a suit challenging FERC action under the Natural Gas Act could not raise issues in addition to those raised by petitioners, in order to prevent intervenors from effectively appealing outside the sixty day statutory period for appeal.
Id.
The same reasoning applies in the present case, where intervenors did not appeal the FAA decisions and filed their motion to intervene well outside the sixty day period for appeal provided for in § 46110(a). Therefore, we will not address intervenors’ argument that the FAA failed to adequately consider the effects of the RBTI on Lubbock, Texas.
III.
Standard of Review
We review the district court’s grants of summary judgment in the Air Force cases
de novo
Our review of the FAA orders is also
de novo,
and we may “affirm, amend, modify, or set aside any part” of the orders approving Lancer MOA and modified IR-178.
As petitioners in both the Air Force cases and FAA appeal challenge those agencies’ NEPA compliance, we must determine whether the actions complained of were arbitrary or capricious under the Administrative Procedure Act.
Generally, agency action is arbitrary and capricious
if the agency has relied on factors which Congress has not intended it to consider, 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.
Preparation of an EIS under NEPA furthers two broad goals. First, it ensures that the agency will consider relevant factors when making its decision. Second, its disclosure requirements foster meaningful public participation in the decisionmaking process.
NEPA does not, however, mandate a particular result.
In determining the adequacy of an EIS, this court considers three factors:
(1) whether the agency in good faith objectively has taken a hard look at the environmental consequences of a proposed action and alternatives;
(2) whether the EIS provides detail sufficient to allow those who did not participate in its preparation to understand and consider the pertinent environmental influences involved; and
(3) whether the EIS explanation of alternatives is sufficient to permit a reasoned choice among different courses of action.
The EIS must provide information satisfying these criteria, and its conclusions must
be supported by evidence in the administrative record.
IV.
Environmental Effects of the RBTI
A.
Livestock
Petitioners raise several challenges to the EIS’s analysis of the RBTI’s environmental effects. First, petitioners claim that the Air Force, and the FAA in adopting the EIS, did not adequately consider the effects of the proposal on the livestock on ranches underlying the RBTI route. Presumably relying on the principle that agencies must follow their own rules
, petitioners argue that the Air Force failed to take the requisite “hard look”
at livestock impacts because it did not follow its 1993 handbook, “The Impact of Low Altitude Flights on Livestock and Poultry” (Handbook).
Petitioners argue that, because the Air Force did not undertake the county — and individual-level inquiry outlined in the Handbook, but instead relied on several studies of the effects of low-level overflights on livestock and a general overview of the underlying region, its analysis was inadequate under NEPA.
Petitioners rely on
Idaho Sporting Congress, Inc. v. Rittenhouse,
in which the Ninth Circuit invalidated a Forest Service EIS, because it analyzed impact on certain species on a “home range” scale, contrary to a Forest Service report stating, “the habitat needs of these species must be addressed at a landscape scale.”
Contrary to
Rittenhouse,
however, cases have generally required that an agency pronouncement have the force and effect of law in order to bind the agency.
To have
the force and effect of law, an agency pronouncement normally “must have been promulgated pursuant to a specific statutory grant of authority and in conformance with the procedural requirements imposed by Congress.”
Petitioners do not argue, nor does the record show, that the Air Force’s Handbook was promulgated according to the APA’s procedural requirements.
See
5 U.S.C. § 553. Thus the Air Force retained discretion to analyze impacts on livestock by methods other than those contained in the Handbook, and we must address the adequacy of the Air Force’s chosen method according to the arbitrary and capricious standard and the relevant criteria announced in
Westphal.
Because determining whether the RBTI overflights will have a significant adverse effect on livestock requires resolution of issues of fact, we defer substantially to the Air Force’s expert analysis of the relevant data.
The EIS and administrative record reveal that the Air Force considered several studies and comments regarding potential impacts on livestock, including those indicating adverse effects. “[I]n making the factual inquiry whether an agency decision was ‘arbitrary or capricious,’ the reviewing court ‘must consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment.’ ”
After reviewing the administrative record, we conclude that the Air Force’s determination that no conclusive evidence showed adverse effects, based on its consideration of relevant studies, was not a clear error of judgment. In addition, the Air Force included a discussion of these studies in the main body of the EIS and its appendices, providing “detail sufficient to allow those who did not participate in its preparation to understand and consider the pertinent environmental influences involved.”
We therefore find the EIS’s analysis of livestock impacts adequate. Because the Air Force’s analysis complied with NEPA, the FAA’s adoption of this portion of the EIS did not violate its obligations under that statute.
B.
Economic Effects
Petitioners’ second challenge to the EIS’s adequacy concerns its analysis of the RBTI’s economic impacts. Specifically, petitioners fault the Air Force and FAA for failing to analyze in depth the effect that the RBTI will have on the values of underlying land for ranching, eco-tourism, and hunting lease income.
As studies regarding the effects of low level overflights on rural land values were unavailable, 40 C.F.R. § 1502.22 governed the Air Force’s duty to obtain this information. That section provides: “[w]hen an agency is evaluating reasonably fore
seeable significant adverse effects on the human environment in an environmental impact statement and there is incomplete or unavailable information, the agency shall always make clear that such information is lacking.”
Id.
It also mandates certain procedures, but only where adverse effects are “reasonably foreseeable.”
Id.
In response to facts similar to the present case, two courts have held that impacts of overflights on land values are not reasonably foreseeable and thus do not require detailed analysis.
We find the reasoning of these courts persuasive. As in
Lee v. U.S. Air Force,
the flights in the present case will take place along a corridor miles wide, and primarily over areas that have been overflown for years, and potential noise increases experienced by owners of land underlying the RBTI are not significant.
In addition, the Air Force examined available studies indicating that aircraft overflights near air bases and airports did not cause significant economic impacts. We find the Air Force’s consideration of economic impacts adequate. Accordingly, neither the Air Force’s nor the FAA’s determination that economic impacts were unlikely was arbitrary or capricious.
C.
Wake Vortex Effects
Petitioners also allege that the Air Force and FAA failed to take a “hard look” at the effects of wake vortices (trails of disturbed air) that would be generated by aircraft training in the RBTI. Petitioners argue that wake vortices damage ground structures like the windmills used by ranchers to provide water to livestock and wildlife. The Air Force responds that the EIS’s discussion of wake vortex effects is adequate, because it “provides a narrative description of what causes vortices and points out that actual, not modeled, B-52 aircraft flying as low as 300 feet [above ground level] ... would generate a surface wind speed of less than 4 mph.” Although CEQ regulations require agencies to “make explicit reference by footnote to the scientific and other sources relied upon for conclusions in the statement,”
the EIS does not reveal the source of this data. Petitioners point out that the information came from an e-mail from the Boeing Company, stating that tests conducted between 1970 and 1986 “at flight level 300” resulted in “[n]o effect on the ground from the B-52 vortexes.”
The Air Force presumably contends that “flight level 300” refers to 300 feet above ground level. In fact, it refers to 30,000 feet above ground level.
It is not clear whether the Boeing e-mail was a miscommunication, because the Air Force did not include the actual Boeing study in the administrative record. Therefore, the e
mail alone cannot provide an adequate basis for the Air Force’s conclusion that flights at 300 feet above ground level would generate low surface winds. To uphold that conclusion, we must find a more satisfactory basis than the Boeing email.
The Air Force also relied on a graph providing a “rough estimate” of Bl-B wake vortex effects at low altitudes. The administrative record shows that the equation used to generate the chart came from a 1949 aerodynamics text by James Dwinnell, but the Air Force did not include the equation or its inputs in the EIS or administrative record.
Petitioners urge this court to consider two extra-record doeuments-excerpts from the Dwinnell text and its expert’s declaration-to determine whether the Air Force’s chart was reliable and thus constituted a hard look at wake vortex effects.
Generally, the “record rule” limits judicial review of agency action to the administrative record before the agency at the time of its decision.
This court has recognized an exception to the general rule, however, where examination of extra-record materials is necessary to determine whether an agency has adequately considered environmental impacts under NEPA.
In the present case we find it necessary to look at the Dwinnell text to determine whether the Air Force’s use of the equation therein was sound. Because we lack technical expertise in aerodynamics, we also consider extra-record materials to aid our understanding of the science involved.
Our review of the Dwinnell text and the declarations of petitioners’ and the Air Force’s experts reveal that the Air Force failed to take a hard look at the possible effects of wake turbulence on ground structures. Although an illustration in the EIS shows that the wake turbulence of an airplane at 300 feet above ground would generate wind speed around two mph at thirty-five feet (the height of a windmill as depicted on the illustration), the Air Force’s own expert, Dr. Ojars Skujins, admits that a Bl-B at this altitude could generate wind speeds as high as forty-seven mph just twenty-two feet above ground. Dr. Skujins also declares that the chart generated by the Air Force based on the Dwinnell equation is “oversimplified” and “does tend to underestimate the maximum vortex strength.” Dr. Skujins concludes, however, that the Air Force was correct in finding that vortices would not create a significant impact, because average wind speeds in the RBTI area are similar to wind speeds generated by wake vortices.
The Air Force is entitled to rely on its own qualified experts’ reasonable opinions in determining the significance of impacts.
The Air Force did not rely on Dr. Skujins’s opinion, however, in addressing the wake vortex issue in the EIS process, but rather relied on the Boeing e-mail and the chart generated from the Dwinnell equation. As discussed above, neither document presents a reliable picture of the impact of wake vortices on surface structures, misinforming both public participation and the Air Force’s conclusion.
The Air Force’s reliance on this data cannot satisfy the hard look requirement of NEPA and thus this portion of the EIS is inadequate.
This determination applies equally to the FAA, which, as an adopting agency, was required to satisfy itself that the wake vortex discussion in the EIS complied with NEPA.
D.
Effects on Civil and Commercial Aviation
Petitioners’ final challenge to the EIS’s analysis of environmental effects concerns potential conflicts between training flights in IR-178 and Lancer MOA and civil and commercial aviation in western Texas. Petitioners contend that the Air Force’s conclusion in the EIS that the RBTI would have little effect on airspace management is contradicted by an FAA study in the administrative record. In addition, petitioners claim that the Air Force violated its own regulations by failing to adequately address mitigation measures proposed by the FAA study in the EIS.
The Air Force argues that effects on aviation are “aeronautical” rather than “environmental,” and thus do not require discussion in an EIS. Counsel for the Air Force acknowledged in oral argument, however, the difficulty involved in drawing a bright line between effects that are purely “aeronautical” and those that are “environmental.” Because “ ‘[environment’ means something more than rocks, trees, and streams, or the amount of air pollution [-i]t encompasses all the factors that affect the quality of life,”
we are reluctant to draw such a line. Civil and commercial aviation are part of the modern human environment broadly defined, and because the RBTI would impact aviation, NEPA required the Air Force to address that impact in the EIS.
“It is a familiar rule of administrative law that an agency must abide by its own regulations.”
The Air Force regulations implementing NEPA provide that an EIS must include “responses to comments on the Draft EIS by modifying the text and referring in the appendix to where the
comment is addressed or providing a written explanation in the comments section, or both.”
In the present case the Air Force responded to the FAA solely by modifying the text. It did not refer in the appendix to where the FAA’s comments were addressed or provide any written explanation, neglecting much of its responsibilities under the regulation. We therefore conclude that this portion of the EIS is also inadequate.
V.
Mitigation
A.
Omission of Mitigation Discussion in Draft EIS
In addition to their complaints regarding the EIS’s environmental inadequacies, petitioners take issue with several aspects of the EIS’s discussion of mitigation measures. First, they argue that the Air Force and FAA violated NEPA by failing to discuss mitigation measures in the draft EIS. CEQ regulations require agencies to prepare a draft EIS prior to issuance of a final EIS.
The draft “must fulfill and satisfy to the fullest extent possible the requirements established for final statements.”
A final EIS must contain a discussion of possible mitigation measures.
Whether the draft EIS must also contain a discussion of mitigation measures is a question of first impression in this circuit.
The Supreme Court has stated that, absent a discussion of possible mitigation measures, “neither the agency nor other interested individuals can properly evaluate the severity of the adverse effects.”
Although the Court there referred to inclusion of a mitigation discussion in a final EIS, the same reasoning can apply to the draft. Under the structure created by the CEQ regulations, the lead agency must request comments from other agencies and the public on the draft EIS before preparing the final EIS.
Following that structure in the present case, the Air Force provided a public comment period on the draft which closed before the Air Force issued the final EIS. Thus, by excluding mitigation measures from the draft, the Air Force prevented the public from commenting on those measures during the comment period.
On the other hand, even if the agency omits the mitigation discussion from the draft, nothing prevents the public from commenting on the mitigation measures once the agency issues the final EIS, and petitioners do not argue that they were prevented from commenting during the two months between the issuance of the final EIS and the Air Force’s ROD.
Giv
en these considerations, we find it unnecessary in the present case to adopt a rigid rule that a draft EIS
must
contain a mitigation discussion, although we note that inclusion of such a discussion is ideal.
B.
Adequacy of Mitigation Discussion in Final EIS
Petitioners also attack the discussion of mitigation measures in the final EIS and those adopted by the Air Force in its ROD.
First, petitioners argue that the final EIS does not adequately discuss measures to mitigate potential adverse effects on underlying livestock operations. Contrary to petitioners’ assertions, however, the final EIS does recognize that overflights may injure livestock and provides mitigation in the form of a claims process for ranchers whose livestock suffer injury. In light of the Air Force’s non-arbitrary conclusion that adverse effects on livestock were unlikely, we find the Air Force’s limited discussion of measures to mitigate those effects reasonable.
Petitioners also argue that reducing the annual number of sorties from the proposed 2,600 to 1,560 and utilizing existing military airspace to the maximum extent possible in creating Lancer MOA did not provide any mitigation because the RBTI would still impose more overflights on certain areas than they had experienced before implementation of the RBTI. This argument is premised on a misunderstanding of the term “mitigation.” The CEQ regulations define “mitigation” as “[a]voiding the impact altogether by not taking a certain action or parts of an action” or “[m]inimizing impacts by limiting the degree or magnitude of the action and its implementation.”
By reducing the number of sorties proposed for Alternative B by over 1,000 and avoiding creation of new airspace, the Air Force limited the magnitude of the RBTI. Thus, petitioners’ argument that these measures did not truly “mitigate” is without merit, and the EIS is not invalid for failure to adequately address mitigation measures.
VI.
Extra-Record Materials
In addition to the evidence pertaining to wake vortex effects, petitioners sought in the Air Force cases to introduce extra-record evidence regarding livestock, socioeconomic, and noise effects. The district court excluded all extra-record submissions. Petitioners argue that, by not considering the extra-record evidence, the district court could not adequately review the Air Force’s NEPA compliance.
Because district courts have discretion to consider extra-record evidence, we review the district court’s decision not to consider such evidence for abuse of discretion.
“A district court abuses its discre
tion if it: (1) relies on clearly erroneous factual findings; (2) relies on erroneous conclusions of law; or (3) misapplies the law to the facts.”
In the present case, the district court correctly stated the law regarding extra-record evidence in NEPA cases.
Without discussing its rationale, however, it excluded all of petitioners’ proffered extra-record evidence.
As discussed in section IV.C., consideration of the Dwinnell text and expert declarations is necessary to determine whether the Air Force took a hard look at wake vortex effects. Thus, by excluding that evidence, the district court “misapplie[d] the law to the facts.” Because this court has reviewed the extra-record submissions in its
de novo
review, however, we need not remand to the district court, but instead dispose of this issue by remanding to the Air Force to prepare an adequate supplemental EIS.
The remaining items of evidence consist of declarations of DMTPHA members and experts on livestock, economic, and noise effects of the RBTI. We conclude that the district court did not abuse its discretion in excluding this evidence. The DMTPHA members’ declarations are largely cumulative of evidence already in the administrative record. In addition, the Air Force was entitled to rely on the reasonable opinions of its own experts regarding livestock, economic, and noise effects.
None of petitioners’ proffered evidence on these issues shows that those experts’ opinions were unreasonable, but instead presents opposing expert opinions. Because the Air Force’s reliance on its own experts does not render its decisions arbitrary and capricious, admission of petitioners’ opposing expert opinions would not show that the Air Force failed to take a hard look at these effects. Thus, admission of petitioners’ extra-record evidence on all issues other than wake vortex was unnecessary to determine whether the Air Force adequately considered environmental impacts of the RBTI
, and the district court’s exclusion of that evidence was not an abuse of discretion.
VII.
NEPA Documentation for Existing IR-178
Petitioners also claim that the Air Force failed to prepare necessary supplemental EIS’s for IR-178 due to changes in the route and underlying land since the route’s creation in 1985. CEQ regulations require agencies to supplement an EIS if the agency makes substantial changes to the proposed action or significant new circumstances or information arise bearing on the proposed action or its impacts.
A claim asserting that NEPA documentation must be supplemented has three elements: (1) ongoing or remaining federal action and (2) new circumstances or information relevant to the environmental impact of the proposed action that are (3) significant enough to warrant supplementation of existing NEPA documents.
The district court held this claim time-barred, finding that the Air Force’s alleged
NEPA failures occurred more than six years before petitioners filed suit.
Although NEPA and the APA do not contain limitations periods, this court has held that claims under the APA are subject to the general six-year statute of limitations for claims against the government.
The limitations period begins to run when the right of action first accrues.
Because petitioners allege agency inaction or delay under 5 U.S.C. § 706(1), we must determine whether this cause of action accrued more than six years before petitioners brought suit.
Petitioners argue that the limitations period does not apply to its IR-178 claim, because the Air Force’s actions regarding IR-178 are ongoing. At least one court has concluded that the six-year limitations period does not apply to claims of unlawful delay under § 706(1), reasoning that unlawful delay of a statutory duty is a continuing violation of the statute.
Applying this line of reasoning in the present case would effectively remove the limitations period from claims that an agency has unlawfully delayed supplementation of NEPA documents, because a necessary element of such a claim is ongoing agency action.
We find the better view to be that a claim for agency delay in supplementing NEPA documents accrues when circumstances requiring supplementation first arise. Such a view prevents plaintiffs from circumventing the limitations period by phrasing their complaints against agencies as continuous delay (from the moment they failed to do something required by NEPA) rather than a failure to act at a discrete point in time. Petitioners argue that certain modifications to IR-178 required supplemental NEPA documentation and that the Air Force did not prepare it. That cause of action accrued when the modifications were implemented without the required documentation. Because all modifications that may have warranted supplementation occurred more than six years before petitioners filed suit, petitioners’ supplementation claim is barred.
VIII.
FAA’s Procedure on Limited Remand
As published in the National Flight Data Digest, modified IR-178 in-
eluded eleven segments with floor altitudes lower than those evaluated in the EIS. The FAA claimed this was an inadvertent error and this court granted a limited remand to correct it. Petitioners now argue that the FAA failed to follow its own regulations in making the correction.
The FAA’s Order on Special Military Operations, FAA Order 7610.4J, provides certain procedures for establishing or modifying a MTR. Order 7610.4J requires,
inter alia,
a certain form, coordination with the Regional Air Traffic Control Center and others, and consideration of minimization of disturbance to persons and property on the ground. The FAA did not follow these procedures on remand, and argues that Order 7610.4J does not apply to corrections like those at issue, which originate within the FAA. We find the FAA’s argument persuasive. Order 7610.4J speaks of route revisions sought by “military unit[s],” not ministerial revisions to correct internal error. Moreover, the FAA sought the remand to correct the altitudes to conform to those in the EIS, which had already considered minimization of ground disturbance. Because the result would be the same- — modification of the altitudes to conform to the EIS — whether the FAA followed the procedure of Order 7610.4J or not, petitioners have not been prejudiced by the FAA’s chosen procedure on remand, and we see no reason to invalidate the correction.
IX.
ROD for IR-178 Modifications
Lastly, petitioners argue that the FAA faded to issue a ROD for the IR-178 modifications.
The FAA responds that, because IR-178 and Lancer MOA are “environmentally and aeronautically linked,” its Non-Rulemaking Decision Document (NRDD) of December 11, 2001 for Lancer MOA serves as the ROD for both Lancer MOA and modified IR-178. Because we find the EIS inadequate and therefore must set aside both the Air Force’s and FAA’s RODs approving the RBTI, we need not address this issue.
X.
Conclusion
For the foregoing reasons we vacate the decisions of the district court, the Air Force ROD and the FAA orders approving the RBTI. We remand to the Air Force
and FAA to prepare a supplemental EIS which adequately addresses wake vortex impacts and FAA comments as required by CEQ and Air Force regulations.
Appendix
1. APA-Administrative Procedure Act
2. CEQ-Council on Environmental Quality
3. DMTPHA-Davis Mountains Trans-Pecos Heritage Association
4. EIS-Environmental Impact Statement
5. FAA-Federal Aviation Administration
6. IR-Instrument Route
7. MOA-Military Operations Area
8. MTR-Military Training Route
9. NEPA-National Environmental Policy Act
10. NRDD-Non-Rulemaking Decision Document
11. RBTI-Realistie Bomber Training Initiative
11. ROD-Record of Decision