The City of Aurora, Colorado and its mayor (hereinafter referred to collectively as the City) petition for review of a final rule of the Federal Aviation Administration implementing a new approach procedure at Stapleton International Airport, Denver, Colorado. The FAA planned to implement the procedure on January 19, 1984, but we stayed implementation pending this decision. The City claims that the FAA has deviated from its own safety rules in violation of the Administrative Procedure Act, and has failed to comply with the National Environmental Policy Act.1 *1459We are not persuaded, and accordingly we deny the City’s petition.
I.
BACKGROUND
Stapleton is the nation’s fifth busiest airport. It provides service for Denver and is a “hub” airport for connecting flights to the Rocky Mountain area and points further west. Stapleton is also the nation’s second worst airport for weather-related delays that affect air service both to and from Denver and to other cities nationwide.
The delays occur primarily because of the alignment and location of the airport’s four major runways. Runways 35R and 35L are aligned in a north/south direction, are parallel to each other, and are separated by 1,600 feet. Runways 26R and 26L are aligned in an east/west direction, are parallel to each other, and are 900 feet apart. When good weather prevails, simultaneous landings can be made on parallel runways. When the cloud cover drops below 2,200 feet, however, pilots are no longer able to land visually but must rely on signals from ground-based electronic instruments that guide their approach until their aircrafts emerge from the clouds. Because pilots are effectively blind while in the clouds, the sole means of determining the proper direction, slope, and speed for the approach are the signals received from the ground instruments and the radio communications from air traffic control. During such instrument approaches, the FAA requires a separation between aircraft of at least 4,300 feet. Because neither pair of runways at Stapleton is separated by 4,300 feet, instrument landings can be conducted on only one parallel runway at a time. During weather that requires instrument landing approaches, the arrival rate at Sta-pleton is thus cut in half, with concomitant delays.
In response to this problem, in 1980 the FAA established a committee consisting in part of representative^ from the FAA, the Air Transport Asspeiation, the Air Line Pilots Association, and Denver’s Director of Aviation. From among various alternatives, the committee opted to devise a new approach procedure centered on runways 35R and 35L, the parallel runways with the greatest separation between them. Under the proposed procedure, aircraft would continue to land on runway 35L using a straight-in approach, and aircraft would also be permitted to land simultaneously on the parallel runway 35R using an offset or angled approach.
The offset approach for runway 35R is known in the industry as an LDA/DME approach2 procedure, and is conducted as follows. Arriving aircraft are given directions (or vectors) by air traffic control personnel to a point about fifteen miles southeast of the airport. At that point, the aircraft is turned toward the airport to receive and follow radio signals from a navigational approach aid known as a localizer. The localizer directs the aircraft on a compass heading of 341 degrees toward a point southeast of the extended centerline of runway 35R called the “missed approach point.” The missed approach point is about three nautical miles south of runway 35R. Following the localizer and using its DME, the aircraft must have the runway or runway lights in sight by the missed approach point. Furthermore, weather conditions must be such that the pilot can conduct the remaining three miles of the approach visu*1460ally, without further reliance on the instrument guidance devices. Unless both of these conditions occur, the landing is discontinued. If visual conditions do exist, the pilot continues his approach and lands on runway 35R.
The advantage of the proposed procedure is that it maintains a separation of at least 4.300 feet between simultaneously approaching aircraft during the entire time the aircraft are guided by the ground-based instrument guidance devices. At the point when both aircraft are turned to their final approach, they are separated by at least 10,000 feet. At the missed approach point, when visual meteorological conditions must exist, the two aircraft are still more than 4.300 feet apart. Moreover, because runway 35L commences a mile south of 35R, aircraft approaching 35L are lower than those simultaneously bound for 35R. Thus, when an aircraft landing on runway 35L passes over that runway’s threshold, it is separated laterally from the aircraft landing on 35R by 2,500 feet. The aircraft landing on 35R continues to descend for another mile on the 341 degree heading, makes a ten degree turn to a 351 degree heading, and proceeds to land on runway 35R.
The procedure will be implemented only on those occasions when a north wind requires using runway 35R for landing,3 and when the weather conditions permit visual contact with the runway or runway lights at the missed approach point (three miles visibility and clouds no lower than one thousand feet above the ground). If weather conditions are worse, the approach cannot be used. If conditions are better but north approaches are still required because of wind, aircraft can land on 35R under normal visual approach procedures, without the need for localizer and DME guidance.
The offset approach was first published as a final rule in May 1983. See FAA Standard Instrument Approach Procedures, 48 Fed.Reg. 24,036 (1983). Because the FAA had failed to provide notice and opportunity for public comment before publishing the rule, it cancelled the rule and republished it as a proposed rule in July 1983. See 48 Fed.Reg. at 33,838. After considering substantial public comment and other newly-obtained information about the proposed rule, the FAA again published the procedure as a final rule in December 1983. See 48 Fed.Reg. at 56,344. This lawsuit followed.
II.
THE ADMINISTRATIVE PROCEDURE ACT
The City claims that the FAA’s promulgation of the new procedure violates the Administrative Procedure Act. 5 U.S.C. §§ 551 et seq. (1982) (APA). When reviewing agency action under the APA, a court must “hold unlawful and set aside agency action, findings, and conclusions found to be ... arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law....” Id. § 706(2)(A).
The City contends that the FAA’s promulgation of the proposed procedure is not in accordance with law because the FAA procedure violates its own safety rules, specifically the United States Standard for Terminal Instrument Procedures (Pet. Ex. 64, hereinafter cited as TERPS).4 The City argues that the TERPS are rules that bind agency discretion, see, e.g., Nader v. NRC, 513 F.2d 1045, 1051 (D.C.Cir.1975); Bonita, Inc. v. Wirtz, 369 F.2d 208, 212 n. 4 (D.C. *1461Cir.1966), and that the promulgation of the proposed procedure violates these rules. The new offset approach varies from at least two substantive TERPS paragraphs.
The handbook permits the FAA to waive or amend TERPS requirements under certain circumstances. The City argues that the FAA has failed to follow the TERPS procedures for effecting a waiver. Paragraph 141 of the TERPS provides:
“NONSTANDARD PROCEDURES. The standards contained in this manual are based on reasonable assessment of the factors which contribute to errors in aircraft navigation and maneuvering. They are designed primarily to assure that safe flight operations for all users result from their application.... Every effort shall be made to formulate procedures in accordance with these standards; however, peculiarities of terrain, navigation, information, obstacles, or traffic congestion may require special consideration where justified by operational requirements. In such cases nonstandard procedures which deviate from these criteria may be approved provided they are fully documented and an equivalent level of safety exists. A nonstandard procedure is not a substandard procedure but is one which has been approved after special study of the local problems has demonstrated that no derogation of safety is involved.”
(Emphasis added.) The City interprets this paragraph to require a special study consisting of
“comprehensive, scientific testing and evaluation both of the equipment to be used for the procedure and of the operation of the procedure itself by pilots and controllers in actual flight conditions which will be experienced utilizing the procedure. That study should be sufficiently extensive to provide ‘full documentation’ that ‘no derogation of safety is involved’ and that an ‘equivalent level of safety exists.’ ”
Petitioners’ Revised Brief at 35. Because the FAA did not undertake this kind of testing and evaluation, the City claims that the promulgation of the new procedure violates section 706(2)(A) of the APA because it is not in accordance with law.
The FAA concedes that it has waived or amended certain substantive TERPS paragraphs in promulgating the proposed procedures, but argues that the TERPS are not rules or regulations binding on the agency. See, e.g., Massachusetts Department of Correction v. Law Enforcement Assistance Administration, 605 F.2d 21, 25 (1st Cir.1979); Doe v. Hampton, 566 F.2d 265, 280-82 (D.C.Cir.1977). The FAA also argues that even if the TERPS are binding, it has fully complied with paragraph 141 as it interprets that paragraph. We need not decide in this case whether the TERPS are rules which bind FAA discretion for we conclude that even if they are, the agency in this case has complied with the procedural requirements set out in paragraph 141. Accordingly, the waivers of certain substantive paragraphs of the TERPS are in accordance with law.
The FAA interprets paragraph 141 to require substantially less than does the City. The agency points out that paragraph 141 does not say that “a” special study is required, but rather it states that “a nonstandard procedure ... is one which has been approved after special study of the local problems has demonstrated that no derogation of safety is involved.” (Emphasis added.) Paragraph 141 does not mention what kind of special study is required, nor does it dictate comprehensive, scientific testing and evaluation. Indeed, the FAA asserts that none of the agencies subject to the TERPS have ever interpreted paragraph 141 that way. Under the FAA’s interpretation, when it seeks to implement procedures that deviate from the TERPS, paragraph 141 requires only that it give particularized attention to the procedure before issuing a waiver of the TERPS requirements. According to the FAA the degree of particularized attention or special study depends on the circumstances involved.
An agency’s administrative interpretation of its own regulations is “of controlling weight unless it is plainly erroneous or *1462inconsistent with the regulation.” Bowles v. Seminole Rock Co., 325 U.S. 410, 414, 65 S.Ct. 1215, 1217, 89 L.Ed. 1700 (1945); see also United States v. Larionoff, 431 U.S. 864, 872, 97 S.Ct. 2150, 2155, 53 L.Ed.2d 48 (1977); United States v. City of Chicago, 400 U.S. 8, 10, 91 S.Ct. 18, 20, 27 L.Ed.2d 9 (1970); INS v. Stanisic, 395 U.S. 62, 72 (1969); 2 K. Davis, Administrative Law Treatise § 7:22 (2d ed. 1979). Even if an agency’s interpretation is not the only one permitted by the language of the rule, courts must respect it if it is at least a reasonable interpretation. Cf. Udall v. Tollman, 380 U.S. 1, 4, 85 S.Ct. 792, 795, 13 L.Ed.2d 616 (1965). We conclude that the FAA’s interpretation of paragraph 141 is reasonable and not plainly erroneous or inconsistent with that paragraph’s language. All that a plain reading appears to require is a careful examination of any safety concerns that a nonstandard procedure may present and full documentation of what the procedure entails.
Adopting the FAA’s interpretation of paragraph 141, the record supports the view that the agency fully complied with the procedural requirements outlined in that paragraph. The FAA ran simulated tests of an approach procedure that required a five degree offset, an earlier version of the proposed procedure. It also conducted a limited flight test of the ten degree offset procedure. The FAA provided full opportunity for public comment on the proposed rule in writing and at a hearing at which representatives of the City, the Air Line Pilots Association (ALPA), and the National Transportation Safety Board (NTSB) were present. The FAA fully reviewed these comments in the final rule, which devotes five pages to discussion of safety aspects alone and refers to the hundreds of documents compiled in this case. See 48 Fed.Reg. at 56,344-53. We believe that this constitutes sufficient special study of the proposed procedure within the meaning of paragraph 141. Since the FAA has complied with the procedural requirements of paragraph 141, we conclude that its action was “in accordance with law.”
Under the APA we must also determine whether the FAA abused its discretion by concluding that “an equivalent level of safety exists” which justifies the waivers under paragraph 141. “Courts owe great deference to FAA determinations in an area — such as flight patterns — where the agency has expertise and is the vehicle chosen by Congress to accomplish the regulations.” DiPerri v. FAA, 671 F.2d 54, 58 (1st Cir.1982). For these reasons, courts are “exceedingly slow” to find an FAA action relating to air safety to be “arbitrary” or “irrational.” Id.; see, e.g., Virginians for Dulles v. Volpe, 541 F.2d 442, 447 (4th Cir.1976); Air Line Pilots Association v. FAA, 454 F.2d 1052, 1054 (D.C.Cir.1971). In this case, the FAA has determined that it can waive or amend certain TERPS requirements while retaining a level of safety equivalent to that which would be achieved were these requirements fully followed.
We note initially that the part of the proposed procedure that governs approaches to runway 35R — which involves a ten degree turn for final alignment and requires a sixty degree turn in the event of a missed approach — is substantially the same as routine LDA/DME approaches used at other airports throughout the country. Thus, nothing appears inherently unsafe about this aspect of the proposed procedure. The proposed procedure is unique because the offset approach to runway 35R will be performed simultaneously with straight-in approaches to the parallel runway 35L. This difference is significant and raises the question whether the routine LDA/DME approach to runway 35R can be performed simultaneously with straight-in approaches to runway 35L while preserving the level of safety achievable through strict compliance with the waived TERPS requirements.
The City contends that the proposed procedure varies from three TERPS paragraphs. Paragraph 990 provides that “[s]imultaneous ILS [Instrument Landing System] approach procedures, using ILS installations parallel to each other, may be authorized when the minimum standards in this Section and Section 1 are met.” (Emphasis added.) The ILS instal*1463lations employed in the proposed procedure use two LDA’s that are offset by ten degrees rather than parallel. The City maintains that the ten degree offset violates paragraph 990. The FAA’s response to this argument in the final rule appears entirely correct.
“With respect to paragraph 990, the FAA agrees that the current terminology of that paragraph refers to TLS installations parallel to each other.’ While offset ILS installations are not referred to in paragraph 990, it should be noted that the basic definition of ‘simultaneous ILS’ in paragraph 901 includes any TLS installations which serve parallel runways ’ (italics added). The key to the safety of simultaneous approaches is the use of appropriate procedures in authorizing simultaneous service to the parallel runways, not the technical questions concerning whether the ILS facilities themselves are precisely parallel. The offsetting of one or both ILS facilities (full or partial) in no way affects the safety objectives of paragraph 990.”
48 Fed.Reg. 56,348-49. While we are unable to assess how offsetting may affect safety, we need only determine that paragraph 990 does not compel the ILS installations to be parallel. We cannot say that the agency’s interpretation of paragraph 990 is arbitrary or capricious.
Paragraphs 992 and 996 require more careful analysis, for the FAA clearly has waived the criteria contained in these TERPS. Both of these paragraphs contemplate a 4,300 foot separation between parallel runway centerlines when simultaneous, instrument-aided approaches are used.5 The proposed procedure allows simultaneous landings on runways 35R and 35L which are separated by only 1,600 feet.
Under visual weather conditions, aircraft may land simultaneously on runways 35R and 35L, even though they are separated by only 1600 feet. Indeed, most landings at Stapleton occur simultaneously on runways 26R and 26L, which are separated by only 900 feet. Only when weather conditions require instrument approaches must parallel runways be 4,300 feet apart under the TERPS. If the proposed procedure called for simultaneous landings by instruments for the entire approach until touchdown, there would indeed be a problem. Under the proposed procedure, however, the simultaneously approaching aircraft are instrument-guided only until the missed approach point some three nautical miles southeast of runway 35R. If by then the pilot is unable to conduct the remainder of the approach visually, the approach is abandoned. Until the missed approach point, the proposed procedure ensures at least a 4,300 foot separation between approaching aircraft. Thus, during the portion of the proposed procedure that is instrument-guided, the separation between simultaneously landing aircraft is always in excess of that commanded by paragraphs 992 and 996.
The TERPS, however, mandate a 4,300 foot separation between the runways and not merely the aircraft. The question is whether by adopting a procedure that appears to fulfill the underlying goals of paragraphs 992 and 996 while varying from their explicit requirements, the FAA has acted arbitrarily or capriciously. We conclude it has not. We are persuaded in large part by the fact that in the final rule the FAA carefully addressed all of the concerns expressed by the Air Line Pilots Association (ALPA) and the National Transportation Safety Board (NTSB), two *1464organizations with the technical expertise necessary to discern any safety problems with the proposed procedure.
Both the ALPA and the NSTB responded to the FAA’s request for public comments with written statements. Neither the ALPA nor the NSTB expressed the view that the FAA could not make a determination of an equivalent level of safety without conducting further simulated tests. In particular, the ALPA stated that it would participate in an evaluation if the FAA met several stated requirements. Apparently the FAA did so. In the final rule the FAA expressly adopted the ALPA suggestions to restrict “heavy” category aircraft to runway 35L and to make a Visual Approach Slope Indicator available for the procedure. See 48 Fed.Reg. at 56,345 and 56,349. The FAA also modified a localizer restriction to allow pilots to maneuver earlier onto the final course alignment when authorized by air traffic control. See id. at 56,345. In response to the ALPA’s recommendation that the procedure should not be utilized when there is a cross-wind component of greater than 15 knots, the FAA noted that a pilot “always has the final authority of accepting or rejecting an approach clearance based on his own safety judgment” and that “[mjaximum crosswind component values will continue to be evaluated.” Id. at 56,349. Finally, the ALPA requested that the proposed procedure for runway 35R remain unauthorized except when simultaneous approaches are in progress. The FAA was unwilling to preclude the possibility that, under certain circumstances, the proposed procedure might be used even for single-stream approaches. It noted, however, that “the normal use of the procedure will be for simultaneous approaches.” 48 Fed.Reg. at 56,347. The implication is that the proposed procedure will be used only when warranted by safety considerations. The FAA also carefully addressed in the final rule each of the concerns expressed by the NTSB. See 48 Fed. Reg. at 56,347. Nothing in the record on appeal indicates that the ALPA and the NSTB are not fully satisfied with the manner in which the FAA addressed their concerns.
We lack the expertise necessary to assess whether the proposed procedure is in fact safe, and in any event we are not empowered on review to substitute our judgment for that of the FAA. See Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 416, 91 S.Ct. 814, 823, 28 L.Ed.2d 136 (1971). The sole purpose of our review under the APA is to ensure that the agency has not abused its discretion or acted arbitrarily or capriciously. The FAA here has followed proper procedure in promulgating this final rule, and to the extent that the proposed procedure deviates from the TERPS requirements, it has provided reasoned explanations for its decisions. See 2 K. Davis, at § 7:13; see also Volpe, 401 U.S. at 416, 91 S.Ct. at 823. Absent a clear error of judgment by the FAA, see id., we are unwilling to conclude that the FAA’s actions in this instance were “so indefensible as to be ‘arbitrary.’ ” See DiPerri, 671 F.2d at 59. Accordingly, no grounds exist under the APA to enjoin the FAA in this case.
III.
NATIONAL ENVIRONMENTAL POLICY ACT
The City also attacks the proposed procedure on the ground that the FAA has violated the National Environmental Policy Act, 42 U.S.C. §§ 4321 et seq. (1982) (NEPA). Under NEPA, if an agency determines that its proposed action may “significantly affect” the environment, the agency must prepare a detailed statement on the environmental impact of the proposed action known as an Environmental Impact Statement (EIS). See 42 U.S.C. § 4332(2)(C). If the agency determines that its proposed action will not significantly affect the environment, it may issue a Finding of No Significant Impact (FONSI). The agency’s decision to issue either an EIS or a FONSI is based on an “environmental assessment” of the proposed action. In the present case, the FAA’s environmental assessment indicated that the new procedure would not have a significant impact on the environment and it issued a FONSI.
*1465The City makes four contentions under NEPA: (1) the FAA failed to consult adequately with other agencies regarding environmental issues before it issued its FONSI; (2) the environmental assessment contains an inadequate discussion of alternatives to the proposed procedure; (3) the environmental assessment fails to discuss the safety risks of the new procedure; and (4) the FAA should have issued an EIS in this case rather than a FONSI. We consider each of these contentions in order.
A. Consultation With Other Agencies
The City first complains that the FAA issued a FONSI without consulting the Environmental Protection Agency (EPA). Council on Environmental Quality regulations require an agency to “involve environmental agencies, applicants, and the public, to the extent practicable, in preparing [environmental] assessments.”6 40 C.F.R. § 1501.4(b) (1983). An FAA order similarly requires that:
“[t]he affected local units of Government, and pertinent Federal and State agencies ... should be consulted early in the process of preparing a [Departmental] EIS, FONSI, or environmental assessment. Comments on the environmental impacts of the proposed action shall be considered, as appropriate, in determining whether the proposed action requires an EIS or FONSI and in preparing the DEIS or FONSI.”
FAA Order 1050.1C, 45 Fed.Reg. 2244 (1980). The City alleges that the FAA issued its original FONSI only three days after it sent a copy of its environmental assessment to the EPA and well before the EPA had an opportunity to respond. Thus, allegedly the FAA failed to consult properly with the EPA before issuing the FONSI.
While it may be true that the FAA failed to consult with the EPA before issuing its original FONSI in May 1983, the FAA clearly considered the EPA’s comments in its final rule of December 1983. See 48 Fed.Reg. at 56,351. The EPA raised three concerns by letter dated September 7, 1983. First, the EPA complained that the environmental assessment was vague concerning the conditions under which the proposed procedure will be used at Stapleton. The final rule unequivocally states that the proposed procedure “will be used only when the narrow ceiling and visibility window exists.” 48 Fed.Reg. at 56,351 (emphasis in original). Second, the EPA challenged the use of the Ldn Noise Metric as the sole impact assessment methodology for noise. The FAA responded in the final rule with an extensive defense of the use of the Ldn Noise Metric. See id. at 56,351-52. Finally, the EPA complained that the FAA had failed to discuss adequately alternatives to the proposed procedure. As we discuss more fully below, the discussion of alternatives in the final rule was adequate. See id. at 56,352.
The Supreme Court recently restated the role of courts under NEPA in reviewing agency actions.
“NEPA has twin aims. First, it ‘places upon an agency the obligation to consider ¿very significant aspect of the environmental impact of a proposed action.’ Second, it ensures that the agency will inform the public that it has indeed considered environmental concerns in its decisionmaking process. Congress in enacting NEPA, however, did not require agencies to elevate environmental concerns over other appropriate considerations. Rather, it required only that the agency take a ‘hard look’ at the environmental consequences before taking a major action. The role of the courts is simply to ensure that the agency has adequately considered and disclosed the environmental impact of its actions and *1466that its decision is not arbitrary or capricious.”
Baltimore Gas & Electric Co. v. Natural Resources Defense Council, Inc., 462 U.S. 87, 103 S.Ct. 2246, 2252-53, 76 L.Ed.2d 437 (1983) (citations omitted). The FAA is required only to consider other agencies’ comments, and to involve them in preparing the environmental assessment. That was done here. The FAA’s responses in the final rule show that the agency took a “hard look” at the EPA comments and at the environmental consequences generally. Nothing more is required. The FAA adequately fulfilled its obligation under NEPA, the CEQ regulations, and the FAA Order to consult with the EPA.
B. Discussion of Alternatives in the Environmental Assessment
The City’s second argument under NEPA is that the FAA’s environmental assessment inadequately discussed alternatives to the proposed procedure. When an agency’s actions require an EIS, the agency must include within the EIS a thorough discussion of “alternatives to the proposed action.” 42 U.S.C. § 4332(2)(C)(iii). Because an EIS was deemed unnecessary in the present case, section 4332(2)(C)(iii) is inapplicable. However, even when an agency’s actions do not require an EIS, NEPA imposes a separate mandate on agencies to “study, develop, and describe appropriate alternatives to recommended courses of action in any proposal which involves unresolved conflicts concerning alternative uses of available resources.” Id. § 4332(2)(E);7 see Aertsen v. Landrieu, 637 F.2d 12, 20 (1st Cir.1980); Nucleus of Chicago Homeowners Association v. Lynn, 524 F.2d 225, 232 (7th Cir.1975) cert. denied, 424 U.S. 967, 96 S.Ct. 1462, 47 L.Ed.2d 734 (1976).
In reviewing the FAA’s actions under section 4332(2)(E), we are guided by the Supreme Court’s decision in Strycker’s Bay Neighborhood Council v. Karlen, 444 U.S. 223, 100 S.Ct. 497, 62 L.Ed.2d 433 (1980), which clarified the scope of judicial review of a federal agency’s consideration of alternatives in an environmental assessment. The Court stated:
“[0]nce an agency has made a decision subject to NEPA’s procedural requirements, the only role for a court is to insure that the agency has considered the environmental-consequences; it cannot ‘interject itself within the area of discretion as to the final choice of the action to be taken.’ ”
Id. at 427-28 (citations omitted).
The FAA’s environmental assessment in this case contains a ten-page discussion of alternatives. Initially, the FAA outlined five possibilities: (1) implement the proposed procedure; (2) take no action; (3) develop a new approach procedure to either runway 26R or 26L; (4) develop a new approach procedure to runway 35L; or (5) reschedule flight operations in accordance with airport acceptance rates. See Pet.Ex. 3, Environmental Assessment at 6-7. The FAA rejected the last three alternatives “due to national implications, passenger inconvenience, schedule disruptions, minimal benefits, and/or environmental impacts.” Id. at 7. The FAA devoted the rest of its discussion to a comparison between implementing the proposed procedure and taking no action.
The City does not complain about the general discussion comparing the two considered alternatives, but contends the FAA failed to fulfill its NEPA duties with respect to the three alternatives it rejected. Specifically, the City argues:
“[Ajll three ‘alternatives’ are summarily rejected, without analysis, for superficial and facially inadequate reasons. The assessment contains absolutely no analysis of the environmental impacts of any of the ‘alternatives,’ but merely includes unsubstantiated conclusory statements such as that air traffic delays cause ‘an unsatisfactory environmental condition,’ *1467or that a converging offset approach procedure for Runway 35L is impractical, due to ‘obvious environmental impacts.’ ”
Petitioners’ Brief at 47.
On the contrary, the FAA provided reasonable explanations for rejecting these three alternatives. The environmental assessment notes that the development of a new approach procedure for either runway 26L or 26R “would provide minimal benefit and provide no relief to the existing delay situation when utilization of Runways 35L and 35R is required.” Pet.Ex. 3, Environmental Assessment at 7. The FAA also explained why a new approach procedure to 35L was impractical.
“[Development of this procedure] would provide the same capabilities as the proposed action. Development of this procedure is complicated by precipitous terrain to the west, major high-rise complexes, existing and proposed, and dense development along the approach course. Preliminary evaluation disclosed that while feasible, the practicality of this alternative is questionable, due to lack of maneuvering area, higher mínimums and obvious environmental considerations.”
Id. Finally, the environmental assessment explains that the third alternative of restructuring flight operations would “impact air travel throughout the entire United States .... The ramifications regarding passenger inconvenience and economic impacts would be substantial.” Id.
These are entirely reasonable explanations. An agency need not analyze the environmental consequences of alternatives it has in good faith rejected as too remote, speculative, or as in this case, impractical or ineffective. See National Indian Youth Council v. Watt, 664 F.2d 220, 226 (10th Cir.1981); Environmental Defense Fund, Inc. v. Andrus, 619 F.2d 1368, 1375 (10th Cir.1980). The City does not contend that alternatives exist that were brought to the FAA’s attention and ignored by it. The FAA took into account all the alternatives that would fulfill the goal of reducing weather-created delays at Staple-ton. It considered and compared the environmental consequences of these alternatives. See Strycker’s Bay, 444 U.S. at 227-28, 100 S.Ct. at 499-500. The agency has substantial discretion in framing its discussion of an alternative and in deciding which alternative to adopt. NRDC v. NRC, 606 F.2d 1261, 1272 (D.C.Cir.1979); see also National Indian Youth Council, 664 F.2d at 226. We discern no basis for overturning the FAA final rule under section 4332(2)(E).
C. Discussion of Safety Risks in the Environmental Assessment
The environmental assessment prepared by the FAA consisted primarily of a detailed analysis of the noise levels anticipated from the new landing procedure. The City argues that the FAA should also have addressed potential environmental impacts due to any safety risks associated with the proposed procedure. The City contends that such assessment is required because of “the risk of ... disastrous [environmental] impacts which would result from a possible aircraft collision or crash.” Petitioners’ Revised Brief at 50.
The purpose of an environmental assessment is to determine whether the proposed change constitutes a major federal action significantly affecting the quality of the human environment. See 42 U.S.C. § 4332(C); 40 C.F.R. § 1508.9; FAA Order 1050.1C, 45 Fed.Reg. at 2251. The assessment must contain brief discussions of the need for the proposal, alternatives as required by 42 U.S.C. § 4332(E), the environmental impacts of the proposed action and its alternatives, and a listing of agencies and persons consulted. See 40 C.F.R. § 1508.9; 45 Fed.Reg. 2251. An assessment of safety risks is not expressly required by the governing regulations.
In this case, the proposal was the new offset approach procedure to runway 35R for use in particular weather conditions. Under this procedure, aircraft would continue to approach Stapleton from the south, and any missed approach would still be diverted to the northeast, which is the current missed approach route. Thus, the flight paths of a landing and missed ap*1468proach for the proposed procedure differ little from those currently used during restricted or unrestricted visibility. Moreover, it does not appear that the environmental impact of a crash or collision along the newly proposed flight path would differ from that along the currently used flight path to runway 35R. Both flight paths pass over urban areas, and nothing in the record suggests that one urban area is more dense or congested than the other.
Finally, throughout all stages of the rule-making procedure including the environmental assessment, the FAA was guided by the requirement that there be no significant change in the level of flight safety, and that an equivalent level of safety be preserved. See TERPS, 11141. In order to assess the environmental impact caused by a decline in the level of safety due to implementation of this procedure, the FAA would have had to assume a set of conditions that would bar use of the procedure at all.
The FAA’s prime concern is unquestionably aviation safety and the controversy between the City and the FAA is concerned almost exclusively with this issue. The FAA considered and responded to a vast number of safety questions and concerns, throughout the rulemaking process and in the final rule itself. Because of the FAA’s finding of an equivalent level of flight safety and, a fortiori, no significant environmental impact due to aircraft collisions or crashes, the aims and policies of NEPA were fulfilled.8 See Baltimore Gas & Electric Co., 103 S.Ct. at 2252-53.
D. The Issuance of the FONSI
The City’s fourth claim under NEPA is the general assertion that “this radically new flight path” should have been accompanied by an EIS, not merely a FONSI. The initial determination concerning the need for an EIS lies with the agency. Image of Greater San Antonio, Texas v. Brown, 570 F.2d 517, 522 (5th Cir.1978). If an agency decides not to file an EIS, the reviewing court must consider whether the agency reasonably has concluded that its actions will have no significant environmental consequences. Preservation Coalition, Inc. v. Pierce, 667 F.2d 851, 855 (9th Cir.1982); see also Greater San Antonio, 570 F.2d at 522; Wyoming Outdoor Coordinating Council v. Butz, 484 F.2d 1244, 1248-49 (10th Cir.1973).
We are convinced that the FAA reasonably concluded an EIS was unnecessary in this case. To begin with, the proposed approach procedure is projected to be used very rarely. A study of landing data covering the period between October 1979 through September 1982 indicates the proposed procedure would have been used during that period for only 1.64% of the total yearly landings. Moreover, by alleviating landing delays during poor weather conditions, the proposed procedure may actually reduce noise exposure. Most of Staple-ton’s air traffic occurs between 7:00 A.M. and 8:00 P.M. daily. When backlogged by weather delays, however, landing operations can continue until after midnight. By compressing the landings into shorter time periods, arrivals will continue to occur more often during the more acceptable day and early evening hours. Finally, even when the procedure is used, the change in cumulative noise level will be less than one decibel, normally unperceivable by most people. See Affidavit of John E. Wesler, Director of Environment and Energy for the FAA, at 4.
The FAA has responded to all comments received on the procedure and the City has identified no environmental concerns that have not been adequately addressed. We find nothing in the record to suggest the FAA’s decision was unreasonable. Accordingly, we are without grounds under NEPA to overturn the FAA’s proposed procedure.
*1469IV.
CONCLUSION
We have carefully reviewed all of the City’s claims and can discern no basis for preventing the FAA’s final rule from becoming effective. Accordingly, we lift the stay we imposed on instituting the new approach procedure. The FAA may now use the procedure in the manner contemplated by the final rule promulgated at 48 Fed.Reg. 56,344.
The City’s motion to set aside the final rule is DENIED.
. The City also seeks a declaratory judgment under 28 U.S.C. § 2201 (1982), that an Ad Hoc Committee formed to study Stapleton's delay problems was constituted and conducted in violation of the Federal Advisory Committee Act, 5 U.S.C.App. § 1 (1982). This claim was raised for the first time in this court and we do not address it because the City has failed to comply with the applicable procedural requirements, see Fed.R.Civ.P. 57, and because there is no independent jurisdictional basis necessary to *1459support a declaratory judgment action in this court. See, e.g., Amalgamated Sugar Co. v. Bergland, 664 F.2d 818, 822-23 (10th Cir.1981); Fry Brothers Corp. v. H.U.D., 614 F.2d 732, 733 (10th Cir.1980).
. LDA stands for Localizer Directional Aid. A localizer is an aircraft navigation aid that operates by emitting a radio beacon from a ground location that is captured by aircraft instruments tuned to the proper frequency. The localizer produces a signal on cockpit flight instruments that indicates to the pilot whether the aircraft is headed in the direction of the radio signal. It provides horizontal (i.e., left-right), not vertical (i.e., up-down) guidance, and is only one component of a full instrument landing system. DME or Distance Measuring Equipment is another radio signal device also transmitted from a ground facility. When translated by cockpit instruments, the DME’s signal will provide a numerical depiction of the distance, in tenths of a mile, between the aircraft and the radio facility.
. Most of the time aircraft land at the airport toward the west, using runways 26R and 26L. However, when a strong north wind is present, aircraft must land toward the north, requiring the use of runways 35L and 35R.
. The TERPS is a handbook containing criteria "which shall be used to formulate, review, approve, and publish procedures for instrument approach and departure of aircraft to and from civil and military airports." TERPS ¶ 1. The first chapter of the TERPS is devoted to administrative or procedural aspects such as the scope of the TERPS, the jurisdiction of the various civil and military agencies, and how proposed instrument-aided procedures should be prepared. Id., ¶¶ 1-149. The rest of the handbook consists of substantive TERPS paragraphs that give criteria for various kinds of instrument-aided approaches.
. Paragraph 992 provides:
"RUNWAY SEPARATION. Simultaneous approaches require a minimum of 4300 feet of separation between parallel runway center-lines where simultaneous operations [using instruments] are authorized. When less than 4300 feet is available, prescribed separation by ATC [Air Traffic Control] shall be provided.”
Similarly, paragraph 996 provides:
“FINAL APPROACH SEGMENT. Except as stated in this paragraph, criteria for the final approach segment are contained in Section 3 of this chapter. Air Traffic Control provides a ‘Zone of No Transgression’ at least 2000 feet wide, 1150 feet from each runway ceterline whenever simultaneous operations are authorized. Aircraft observed to enter this zone are instructed to alter course as appropriate to return to the desired course.”
. By Presidential. Order, the Council on Environmental Quality developed uniform regulations to guide federal agencies in complying with NEPA. See 40 C.F.R. Parts 1500-08. Pursuant to these CEQ regulations, see § 1507.3, the FAA issued its regulatory order 1050.1C which established the policies and procedures it would use in considering the environmental impacts of its proposed actions. See 45 Fed.Reg. 2244 (1980). The FAA order both incorporates and supplements the CEQ regulations, and together the CEQ regulations and the FAA order govern the FAA’s compliance with NEPA. Id.
. We note that initially this section was numbered 102(2)(D) of the NEPA and was codified at 42 U.S.C. § 4332(2)(D) (1970). Congress later redesignated it as 102(2)(E), and it was recodi-fied in its present form, 42 U.S.C. § 4332(2)(E). See Act of Aug. 9, 1975, Pub.L. 94-83, 89 Stat. 424.
. We do not believe that our conclusion is contrary to the Supreme Court's decision in Metropolitan Edison Co. v. People Against Nuclear Energy, 460 U.S. 766, 103 S.Ct. 1556, 75 L.Ed.2d 534 (1983). While an agency may be required to consider the effects that will occur if a risk is realized, where no increase in risk is permitted, as here, no significant safety impact exists to trigger further assessment or inquiry under NEPA.