Commercial navigation on the Upper Mississippi River and the Illinois River is made possible through a series of 27 locks and dams known as the Upper Mississippi River Navigation System. Locks and Dam 26, which is located on the Upper Mississippi near Alton, Illinois, is a vital link in this system. Because it stands just south of the juncture of the Illinois and Mississippi Rivers, and just north of the juncture of the Missouri and Mississippi Rivers, it serves as a funnel through which all traffic along these waterways must pass.1 Almost 13 years ago the Army Corps of Engineers proposed that the existing structure be replaced with a new lock and dam. The Corps believed that a new facility was necessary because the existing structure was deteriorating and lacked sufficient capacity to accommodate increasing barge traffic. After years of planning and several vigorously fought court battles, the Corps obtained legislative approval for the project in 1978, when Congress enacted a bill specifically authorizing construction of the new facility. Act of Congress, October 21, 1978, Pub.L.No. 95-502, 92 Stat. 1693 (P.L. 95-502). The Corps began to make construction plans shortly thereafter.2

Before the Corps could proceed with construction of the facility, appellants, 18 midwestern railroads3 and three environmental groups,4 renewed an action against the Government5 that they had originally commenced in 1974. In their amended complaint they sought to halt further work on the project.6 The Association for the Improvement of the Mississippi River (AIMR) intervened in the District Court on the side of the Government. Although appellants raised a variety of claims, their objections were, in essence, that: (1) the cost-benefit analysis prepared by the Corps prior to receiving congressional approval of the project violated the Water Resources Planning Act of 1965, 42 U.S.C. § 1962 et seq. (1976), various other statutes, and the Corps' own regulations, 33 C.F.R. Parts 290-295 (1979)7; (2) the Corps failed to comply with the National Environmental Policy Act (NEPA), 42 U.S.C. § 4321 et seq. (1976)8; and (3) the planning conducted by the Corps after receiving congressional authorization did not comply with various statutes and Corps regulations.9 The District Court dismissed the first set of claims for lack of jurisdiction. Atchison, Topeka & Santa Fe R. Co. v. Callaway (Atchison IV), 480 F. Supp. 972 (D.D.C.1979). After conducting a short trial it concluded that the Corps had adequately fulfilled its obligations under NEPA. Atchison, Topeka & Santa Fe R. Co. v. Alexander (Atchison V), 480 F. Supp. 980, 994-1002 (D.D.C.1979). As for the arguments regarding post-authorization planning, the District Court found that most of the appellants' claims lacked merit. Id. at 987. It concluded, however, that the Corps violated its own regulations when it failed to hold a public meeting to discuss implementation of the project. Id. at 993-994. It decided not to grant an injunction requiring the Corps to hold such a meeting. Id. at 1002-1003.

In this appeal the appellant railroads and environmental groups seek review of the District Court's decision. In general we are satisfied with the District Court's disposition of the case. The District Court was clearly correct in holding that it was without jurisdiction to review claims that the cost-benefit analysis prepared by the Corps failed to comply with the Water Resources Planning Act and various other statutes and regulations. We also agree that the Corps' Final Environmental Impact Statement (FEIS) fulfilled its obligations under NEPA. Finally, we agree that the Corps should have held a public meeting after receiving congressional authorization so that it could solicit comments on implementation of the project. We disagree, however, with the District Court's decision not to require such a meeting. Thus we affirm in part and reverse in part, remanding so that the District Court may amend its judgment to require the Corps to hold a public meeting. This meeting should be held within 30 days of the time the judgment, as amended, becomes final. To ensure that the meeting is not an empty formality, the District Court should also enter an order requiring the Corps to respond in writing to the objections made at the meeting. This response should be completed no later than 30 days after the meeting is held.

The Corps of Engineers first considered replacing existing Locks and Dam 26 with a new structure in the mid-1960's, when it became concerned about two problems: lack of capacity and structural instability. FEIS Vol. 1 at 1-2. The existing facility consists of a main lock, which is 600 feet long, and an auxiliary lock, which is 360 feet long. It has a maximum annual capacity of 73 million tons. Because the waterways to the north have a capacity of 108 million tons per year, and the waterways to the south have a capacity of 148 million tons, the Corps feared that the existing structure would act as a bottleneck that would restrict full use of the waterways system and delay shipping. Id. at ii, 1-2. These fears were apparently justified. Barges must now wait an average of at least ten hours to pass through the locks. Id. at 2; Board of Engineers for Rivers and Harbors Report (BERH Rep.) at 16 (Plaintiffs' Exhibit (PX) 427). Locks and Dam 26, which is founded on wooden piles driven into sand, also has a history of structural problems, including settlement, underseepage, and loss of foundation material. Physical deterioration has led to increased maintenance and operation costs, and may ultimately result in complete breakdown of the facility. FEIS Vol. 1 at 2, 6-11; BERH Rep. at i, 16-18.

Identifying the problems that afflict the current structure has proven to be far easier than finding an acceptable solution. Below we recount the long and complicated history of the Corps' efforts to replace Locks and Dam 26.

The first proposal to replace existing Locks and Dam 26 was made in 1968, when the Corps' St. Louis District Engineer recommended that a new dam and two 1,200-foot locks be constructed two miles downstream from the existing structure. This recommendation was reviewed by the Board of Engineers for River and Harbor (BERH), which was established by Congress to provide an independent review of water-related projects.10 The BERH submitted a report to the Chief of the Corps of Engineers in which it approved the project and recommended immediate implementation. The Chief of Engineers then sent a report to the Secretary of the Army, in which he asked the Secretary to approve construction. The Secretary granted approval, acting pursuant to Section 6 of the Rivers and Harbors Act of 1909, 33 U.S.C. § 5 (1976), which authorizes him to order maintenance and repair of existing navigation facilities. Between 1968 and 1974 the Corps prepared a series of design memoranda, see Design Memoranda (DM) (Defendants' Exhibit (DX) 33-43), and an Environmental Impact Statement (EIS) for this two-lock project, see EIS (PX 507). Congress appropriated construction funds in 1974. See Atchison V, supra, 480 F.Supp. at 984-985.

In August 1974, before construction of the replacement facility could start, the same environmental organizations and midwestern railroads that are appellants here filed suits in the District Court seeking to enjoin the project.11 The District Court immediately granted a temporary restraining order that halted bidding on the construction. Shortly thereafter the court issued a preliminary injunction stopping all further activity. Atchison, Topeka & Santa Fe R. Co. v. Callaway (Atchison I), 382 F. Supp. 610 (D.D.C.1974). The court stated that a preliminary injunction was justified because the railroads and environmentalists were likely to prevail on their claim that the Secretary of the Army lacked authority to approve the project under Section 6 of the Rivers and Harbors Act of 1909, 33 U.S.C. § 5 (1976).12 According to the District Court, that Act applies only when the Corps plans to repair an existing structure. Here, however, the Corps had proposed an entirely new facility. Before such a project can be implemented, Section 9 of the Rivers and Harbors Act of 1899, 33 U.S.C. § 401 (1976), requires that congressional authorization be obtained.13 382 F.Supp. at 616-617. The District Court also stated that a preliminary injunction was justified because the plaintiffs were likely to prevail on their claim that the Corps' EIS was inadequate. The court noted that the EIS failed to discuss the systemwide effects of the proposal and to consider reasonable alternatives to the plan. Id. at 620-623.

After the District Court issued the preliminary injunction the Corps decided to seek congressional authorization under Section 9 of the Rivers and Harbors Act of 1899 and to revise its EIS. In 1975 the District Engineer for the St. Louis District prepared a Formulation Evaluation Report (FER). See FER (DX 43). The FER, which contained a cost-benefit analysis, evaluated a variety of alternatives, including rehabilitation of the existing structure. Id., Vol. 1 at 7-1 to 9-18. It concluded that replacement would be safer and less costly than rehabilitation, and recommended a project identical in scope to that originally proposed in 1968: construction of a new dam with two 1,200-foot locks. Id., Vol. 1 at 5-17, 10-1 to 10-8. In addition to the FER, the District Engineer prepared a Draft Supplemental Environmental Impact Statement (DSEIS), which discussed the systemwide impact of the two-lock proposal and carefully examined reasonable alternatives to the proposal. See DSEIS Vol. 1 at 4-1 to 4-101, 5-1 to 5-9, 6-1 to 6-139 (DX 46).

The BERH reviewed the proposal and issued a report recommending construction of a single 1,200-foot lock rather than a double-lock facility. The BERH agreed with the Corps' determination that, because of the existing facility's structural instability, rehabilitation or replacement was essential. It also agreed with the Corps' conclusion that replacement would be safer and less costly than rehabilitation. BERH Rep. at 89. It argued, however, that the environmental impact of the two-lock proposal, which would more than double the capacity of Locks and Dam 26, was uncertain. Id. at 88. A single-lock project, on the other hand, would have an insignificant environmental impact. Id. at 89. Although the economic benefits provided by such a project would be less than those provided by the two-lock project, there would be some gain: the capacity of Locks and Dam 26 would still be increased by 18 percent, so shipping delays would be reduced. See id. The BERH also noted that the single-lock project could be justified on structural grounds.14 Id. at iv-vi, 89.

The Corps decided to accept the BERH's recommendation. A new Revised Draft Supplemental Environmental Impact Statement (RDSEIS), see RDSEIS (PX 526), and a Final Environmental Impact Statement (FEIS), which incorporated by reference the FER and the RDSEIS, see FEIS at 1,15 were prepared. In March 1976 the Chief of Engineers issued a report describing the new project to the Secretary of the Army. The Secretary terminated his approval of the pending two-lock proposal. He then formally recommended to Congress that it authorize construction of a new dam and single 1,200-foot lock, and that it not authorize a second lock until new studies indicated that one should be constructed. Proposed legislation and the FEIS were transmitted to Congress along with the Secretary's recommendation. See Communication from the Assistant Secretary of the Army, Locks and Dam No. 26, Mississippi River, Alton, Illinois, H. Doc. 94-504, 94th Cong., 2d Sess. (August 26, 1976). The Corps also provided Congress with a Supplemental Economic Data Report (SED), updating and revising the cost-benefit analyses contained in the FER. See SED (PX 534).

While the Secretary's proposal was pending before Congress, the railroads and environmental groups continued to seek judicial relief. In May 1977 the District Court dissolved the outstanding preliminary injunction, since the Secretary's decision to withdraw approval of the two-lock project removed the possibility of harm to the plaintiffs. Atchison, Topeka & Santa Fe R. Co. v. Callaway (Atchison II), 431 F. Supp. 722, 725 (D.D.C.1977). At the same time the court reviewed plaintiffs' standing to seek relief. It recognized a private right of action to enforce Section 102(2)(C) of NEPA, 42 U.S.C. § 4332(2)(C) (1976), which provides that an environmental impact statement must accompany proposals for legislative action. It also held that appellants possessed standing to enforce this right. 431 F. Supp. at 725-730. Approximately one year later the court considered a motion for summary judgment filed by the railroads and environmental groups. In their motion they argued that the FEIS submitted to Congress by the Corps failed to comply with NEPA and that the cost-benefit analysis performed by the Corps was inadequate.16 The court denied this motion because it believed there were material issues of fact concerning the systemwide environmental impact of the project and the Corps' calculation of costs and benefits. Atchison, Topeka & Santa Fe R. Co. v. Callaway (Atchison III), 459 F. Supp. 188, 190-192 (D.D.C.1978). The court also held that an evidentiary hearing was appropriate to assess the sufficiency of an environmental impact statement under NEPA; review of the FEIS would not be limited to the contents of the administrative record. Id. at 192-193. Finally, the court granted a Government motion to dismiss appellants' claims seeking to enjoin the proposal of authorizing legislation.17 Id.

Throughout the period preceding authorization by Congress, the railroads and environmental groups not only sought relief in the courts; they also participated actively at the administrative level. Representatives of appellants attended each of the four public meetings held by the Corps during the pre-authorization stage.18 Appellants also had an opportunity to comment on the DSEIS, which the Corps circulated to interested parties. They provided lengthy responses, in which they argued that: rehabilitation of the existing structure or investment in railroads were environmentally and economically preferable alternatives to replacement; the traffic increase resulting from the increase in capacity would have an adverse environmental impact; upstream locks would not be able to absorb the traffic increase; and a programmatic environmental impact statement was required because the replacement project was part of a broader program to expand the Upper Mississippi Waterway System. These responses were included and discussed in the FEIS and RDSEIS. See FEIS Vol. 2 at 243-247, 265-267, 271-457; RDSEIS Vol. 1, Parts 2, 3; see generally FEIS Vol. 1; RDSEIS Vol. 1. Finally, appellants were given an opportunity to comment on the FER, which contained the Corps' economic evaluation of all project alternatives. Again, their responses were included and discussed in the FEIS and the RDSEIS. See FEIS Vol. 2 at 243-247, 271-457; RDSEIS Vol. 2, 3. The Corps revised its cost-benefit analysis in part because of appellants' objections. See SED Vol. 1 at 5-1 to 5-9.

Congress considered the Secretary's proposal to replace existing Locks and Dam 26 during 1976 and 1977. Extensive hearings were conducted before House and Senate committees. See Locks and Dam 26: Hearings Before the Subcommittee on Water Resources of the Senate Committee on Public Works, 94th Cong., 2d Sess. (1976) ("Senate Hearings 1976"); Proposed Waterway User Charges and Replacement of Locks and Dam 26: Hearings on S. 712, S. 790, and S. 923 Before the Subcommittee on Water Resources of the Senate Committee on Environment and Public Works, 95th Cong., 1st Sess. (1977) ("Senate Hearings 1977"); Replacement of Locks and Dam 26 at Alton, Illinois: Hearings Before the Subcommittee on Water Resources of the House Committee on Public Works, 95th Cong., 1st Sess. (1977) ("House Hearings 1977"). The merits of the project were also debated on the floor of Congress. See, e. g., 124 Cong.Rec. S18043-S18052 (daily ed. October 10, 1978) (Senate debate); 124 Cong.Rec. H12695-H12702 (daily ed. October 13, 1978) (House debate).

The appellants participated in the hearings before Congress, raising essentially the same objections they made during the pre-authorization stage. The environmental groups argued that the traffic increase expected from the single-lock facility would have a substantial adverse environmental impact. See, e. g., H.R.Rep.No.95-545, 95th Cong., 1st Sess. 12 (1977); Senate Hearings 1976 at 417-428; House Hearings 1977 at 341-362. Both the environmentalists and the railroads opposed construction of the replacement facility on the ground that it was part of a plan to expand capacity throughout the waterways system. See, e. g., S.Rep.No.95-215, 95th Cong., 1st Sess. 6 (1977). The railroads also emphasized the argument that the Corps had failed to give adequate attention to the rehabilitation alternative. See, e. g., Senate Hearings 1977 at 270-292, 344-350. They presented a rehabilitation scheme that they claimed would be safer and less expensive than replacement. See, e. g., H.R.Rep.No.95-545, supra, at 14-15. All of the appellants argued that the Corps had not properly computed the project's economic cost-benefit ratio. See, e. g., Senate Hearings 1976 at 849-909.

Despite the objections raised by appellants, both the House and the Senate committees issued reports recommending authorization of a single-lock replacement facility. S.Rep.No.95-215, supra ; H.R.Rep.No.95-545, supra. After the floor debate, Congress enacted Public Law 95-502, in which it authorized construction of a single-lock replacement for Locks and Dam 26 as part of a broad scheme for management of the Upper Mississippi Waterway System. Act of October 21, 1978, P.L. 95-502, 92 Stat. 1693 (1978). P.L. 95-502 also:

(1) requires development of a "master plan" for the Upper Mississippi River System following comprehensive studies of the environmental and economic effects of expanded navigation;

(2) prohibits further expansion of the capacity of any lock, dam, or channel in the system, apart from Locks and Dam 26, until the study has been completed and Congress has approved the master plan; and

(3) establishes a user charge on the inland waterways in order to recover a portion of the public investment in the system.

President Carter signed the bill in October 1978.

C. Post-Authorization Planning and Litigation

After P.L. 95-502 was enacted, the Corps of Engineers, acting pursuant to its own regulations, see Engineering Regulation (ER) 1110-2-1150 (describing post-authorization procedures), reprinted at Statutory Appendix (SA) 178-185, commenced its post-authorization planning. The Chief of Engineers issued a memorandum in which he stated that because the project as authorized did not differ from the project as proposed, and because there had been no major changes in external conditions since authorization, full reevaluation of the project was not necessary; neither the cost-benefit analysis nor the FEIS need be reviewed. Chief of Engineers Memorandum, January 10, 1979, reprinted at Joint Appendix (JA) 2406, 2408-2409. The memorandum also directed the St. Louis District Engineer to restate the cost-benefit calculations in current prices and to prepare final engineering plans and specifications. It noted that most of the engineering work had been accomplished in the design memorandum prepared before the District Court enjoined construction of the two-lock project in 1976. Id. Acting pursuant to the directive of the Chief of Engineers, the Corps prepared Design Memorandum Supplement No. 2 (GDM Supp. No. 2). See GDM Supp. No. 2 (PX 540). Congress appropriated funds for the project in October 1979 and work commenced the next month.

Appellants responded to the new legislation by renewing their claims for judicial relief. They filed an Amended Complaint in the District Court in which they claimed that the pre-authorization cost-benefit analysis performed by the Corps failed to comply with the Water Resources Policy Act of 1965, 42 U.S.C. § 1962 et seq. (1976), the Water Resources Council Principles and Standards for Planning Water and Related Land Resources (Principles and Standards), 38 Fed.Reg. 24778 (1973), reprinted at SA 191, the Department of Transportation Act, 49 U.S.C. § 1656(a) (1976), the River and Harbor and Flood Control Act of 1970, P.L. No. 91-611, 84 Stat. 1823, and the Corps' own planning regulations, 33 C.F.R. Parts 290-295 (1979).19 They also argued that the Corps had failed to comply with NEPA.20 Finally, they claimed that the Corps' post-authorization planning was inadequate in several respects.21

Before trial the Government moved to dismiss appellants' claim that the pre-authorization cost-benefit analysis had violated the Water Resources Act, the Corps' own regulations, and various other statutes. The District Court conceded that each of the statutes and regulations in question established standards and principles that the Corps must apply in evaluating federal water projects. But the court also noted that the statutes and regulations established these standards primarily for the benefit of Congress; they were intended to aid Congress in deciding whether a particular project should be undertaken. Once the legislature had exercised its judgment by deciding to authorize a project, judicial review of the analysis on which that judgment was based would interfere with the legislative process. Atchison IV, supra, 480 F.Supp. at 976-980. The District Court refused to grant the Government's motion to dismiss appellants' NEPA claims. It distinguished NEPA from the other statutes, reasoning that NEPA set forth guidelines for agency conduct, not just to benefit the legislature, but also to benefit the general public. Id. at 974-975.22

In September 1979 the District Court conducted a five-day trial, at which it confronted the two remaining issues: (1) whether the Corps had complied with NEPA, and (2) whether the Corps' post-authorization planning was adequate. Appellants made a variety of arguments with respect to the first question. They claimed that the record was inadequate for review because various calculations made by the Corps in evaluating the environmental and economic impact of the project were based on "secret data"; that the FEIS failed to give enough attention to alternatives to the replacement project; that the Corps should have prepared a programmatic environmental impact statement because the Locks and Dam 26 project was part of a larger plan to expand the entire Upper Mississippi River System; that the FEIS underestimated the adverse environmental impact of the Locks and Dam 26 project; and that the cost-benefit analysis conducted by the Corps, which was referred to in the FEIS; understated the environmental and economic costs attributable to the replacement project. Atchison V, supra, 480 F.Supp. at 994-1001. Appellants presented their case through the testimony of six expert witnesses two economists, two biologists, an engineer, and a railroad president. Deposition excerpts and hundreds of documentary exhibits were also produced. The Government responded by putting four of its own economists and engineers on the stand. Id. at 989-990. After carefully reviewing the evidence, the District Court rejected each of appellants' arguments. It held that the Corps' FEIS adequately fulfilled the Corps' obligations under Section 102 of NEPA, 42 U.S.C. § 4332 (1976). It also ruled that the Corps had fulfilled its substantive obligation under Section 101 of NEPA, 42 U.S.C. § 4331 (1976), to balance the economic benefits of the project against its environmental costs. 480 F. Supp. at 994-1001.

Appellants also made several arguments with respect to the question whether the Corps' post-authorization planning was adequate. First, they claimed that under the Water Resources Planning Act and the Corps' own planning regulations, 33 C.F.R. Parts 290-295 (1979), the Corps should have performed a de novo cost-benefit analysis before deciding to go ahead with construction. The District Court rejected this claim, suggesting that a new cost-benefit analysis was not required absent a material change in circumstances. Atchison V, supra, 480 F.Supp. at 987.23 Second, appellants argued that the Corps had violated its regulations when it failed to hold a public meeting before implementing the project. ER 1110-2-1150, reprinted at SA 148, 33 C.F.R. §§ 209.405(d), (f)(5), and 209.410(n)(ii) (1979). The District Court agreed, holding that the Corps must conduct at least one public meeting during post-authorization planning. Atchison V, supra, 480 F.Supp. at 993-994. It entered a declaratory judgment in favor of appellants on this issue. After balancing the various interests at stake, however, the court decided not to grant injunctive relief requiring the Corps to hold a meeting before going ahead with the project. The court reasoned that the passage of P.L. 95-502 reflected a strong public interest in proceeding with construction of the new lock and dam. Appellants had already participated vigorously at the administrative and legislative levels, and would be unlikely to benefit from an additional meeting. Moreover, the public had an opportunity to comment on the project through their representatives in Congress. Finally, any additional delay would undoubtedly harm the Corps. Id. at 1002-1003.

II. REVIEWABILITY OF PRE-AUTHORIZATION COST-BENEFIT ANALYSIS

Before submitting the Locks and Dam 26 proposal to Congress, the Corps, acting pursuant to the Water Resources Planning Act of 1965, 42 U.S.C. § 1962 et seq. (1976), the Water Resources Council Principles and Standards, supra, SA 191, the Department of Transportation Act, 49 U.S.C. § 1656(a) (1976), the River and Harbor and Flood Control Act of 1970, P.L. No. 91-611, 84 Stat. 1823, and its own planning regulations, 33 C.F.R. Parts 290-295 (1979), conducted a cost-benefit analysis in which it examined the economic and environmental impact of the single-lock project, as well as several alternatives. It concluded on the basis of this analysis that the single-lock project was justified. The analysis was submitted to Congress along with the request that construction of the new facility be authorized. Appellants sought review of the analysis in the trial court, arguing that the Corps failed to comply with the relevant statutes and regulations. The court dismissed this claim, holding that it did not have jurisdiction to review pre-authorization cost-benefit analyses conducted by the Corps. We agree with this conclusion.

The Water Resources Planning Act of 1965 establishes the inter-agency Water Resources Council and directs the Council to establish "principles, standards, and procedures" for the "formulation and evaluation" of federal water projects. 42 U.S.C. § 1962a-2 (1976). The primary goal of this legislation was to provide Congress with guidance in its planning efforts in the water resources field; Congress hoped that it would be better equipped to evaluate various development projects. See H.R.Rep.No.169, 89th Cong., 1st Sess. 3, 5 (1965). In 1973 the Water Resources Council issued its Principles and Standards, supra, 38 Fed.Reg. 24778-24868, SA 191-283. The Council stated that, in general, projects should be evaluated by measuring their costs and benefits in light of two objectives: national economic development and environmental quality. SA 194. More specifically, the Principles and Standards provide that a planning organization should consider several alternative plans for dealing with a problem. It should evaluate the economic and environmental costs and benefits of each of these alternatives. A plan should be recommended to Congress only if its net economic benefits are positive, unless economic benefits have been foregone or economic costs incurred in order to enhance environmental quality. SA 199-200.24

The Department of Transportation Act provides the Water Resources Council with additional authority to establish standards governing economic evaluation of water resource projects. 49 U.S.C. § 1656(a) (1976). It also sets forth a definition of "primary direct navigation benefits" and states that this definition should be employed whenever agencies evaluate water projects. Id. Like the Water Resources Planning Act, this statute was intended primarily to assist the legislature in deciding whether to approve particular projects.25 The River and Harbor and Flood Control Act of 1970 requires the Corps to develop guidelines to assure that "possible adverse economic, social, and environmental effects" relating to any project are fully considered. P.L. No. 91-611 § 122, 84 Stat. 1818. This statute also was clearly intended to assist Congress in evaluating water projects; the statute provides that the guidelines are to be applied in all proposals that the Corps submits to Congress.

The Corps' cost-benefit analysis regulations, 33 C.F.R. Parts 290-295 (1979), implement the commands embodied in these statutes and in the Principles and Standards. They are based primarily on the criteria described by the Water Resources Council in its Principles and Standards, supra.26 They provide that projects are to be evaluated in light of their economic and environmental costs and benefits. The planning regulations also contain the guidelines required by the River and Harbor and Flood Control Act of 1970. See 33 C.F.R. Part 294 (1979); see generally id. Parts 290-295 (describing multi-objective planning framework outlined in River and Harbor and Flood Control Act of 1970). Like the statutes, the regulations reflect the fact that the cost-benefit analysis is intended to assist Congress; the regulations state that they "provide( ) the basis for selecting (a plan) and, if appropriate, recommending it for authorization." Id. § 290.11.27

Because these statutes and regulations indicate that the cost-benefit analysis is prepared for the benefit of Congress, we do not believe that they give the federal courts jurisdiction to review that analysis, once the legislature has made its decision to approve or not to approve a project. In deciding whether authorization should be granted, Congress will itself review the analysis and determine whether it has been conducted properly. Once it has made this determination, the courts should not interfere. See Atchison V, supra, 480 F.Supp. at 990. See also Environmental Defense Fund, Inc. v. Alexander, 501 F. Supp. 742, 754-760 (N.D.Miss.1980). We note, in any event, that judicial review would be contrary to the express language of the Water Resources Planning Act, which states that it does not in any way "limit the authority of Congress to authorize and fund projects(.)" 42 U.S.C. § 1962-1 (1976).

The absence of any need for judicial interference is amply demonstrated by the facts of this case. The legislative record of P.L. No. 95-502 shows that Congress gave extensive consideration to claims the Corps had not properly computed the replacement project's costs and benefits. For example, appellants argued before Congress that a modest increase in traffic from Locks and Dam 26 would cause congestion on the Illinois River and lead to significant "delay costs." See, e. g., Senate Hearings 1976 at 849-909. Congress explicitly rejected this and all other arguments challenging the cost-benefit ratios; both the House and Senate Reports adopt, with minor adjustments, the cost-benefit analysis prepared by the Corps. H.R.Rep.No.95-545, supra, at 17-18; S.Rep.No.95-215, supra, at 5.28

Additional support for our conclusion is provided by decisions that have confronted the question whether the courts have jurisdiction to review cost-benefit analyses performed by the Corps pursuant to the Flood Control Act of 1936, 33 U.S.C. § 701a (1976). That statute, like the statutes and regulations at issue here, directs the Corps to submit cost-benefit analyses to Congress. It has consistently been held that determination of economic costs and benefits is a matter of legislative judgment and that, since the analyses are performed for Congress, the legislative branch rather than the judiciary is best suited to review their adequacy. See Oklahoma ex rel. Phillips v. Guy F. Atkinson Co., 313 U.S. 508, 527-528, 61 S. Ct. 1050, 1060, 85 L. Ed. 1487 (1941); United States v. West Virginia Power Co., 122 F.2d 733, 738 (4th Cir.), cert. denied, 314 U.S. 683, 62 S. Ct. 187, 86 L. Ed. 547 (1941) ("So far as the benefit exceeding the cost is concerned, this was a matter of legislative policy * * *."); Environmental Defense Fund, Inc. v. Froehlke, 368 F. Supp. 231, 240-241 (W.D.Mo.1973), aff'd, 497 F.2d 1340 (8th Cir. 1974) ("questions of fact involved in connection with the ultimate resolution of conflicting data in regard to the determination of benefit-cost ratios 'are solely a matter for congressional determination' "); Cape Henry Bird Club v. Laird, 359 F. Supp. 404, 413 (W.D.Va.), aff'd per curiam, 484 F.2d 453 (4th Cir. 1973); Environmental Defense Fund, Inc. v. Corps of Engineers, 348 F. Supp. 916, 925 (N.D.Miss.1972), aff'd, 492 F.2d 1123 (5th Cir. 1974) ("any question as to the adequacy or accuracy of defendants' economic and technical analysis must be addressed to the Congress"); Environmental Defense Fund, Inc. v. Corps of Engineers, 325 F. Supp. 728, 740 (E.D.Ark.1970), aff'd, 470 F.2d 289 (8th Cir. 1972), cert. denied, 412 U.S. 931, 93 S. Ct. 2749, 37 L. Ed. 2d 160 (1973) (it is for Congress to determine in authorizing a water resources project whether benefits are in excess of the estimated costs). These decisions clearly lend support to our conclusion that review of cost-benefit analyses should not be available under the statutes and regulations relied on by appellants.29

Our conclusion that the cost-benefit analysis may not be reviewed to determine compliance with the Water Resources Planning Act, the Department of Transportation Act, the River and Harbor and Flood Control Act of 1970, and the Corps' cost-benefit regulations does not completely insulate the analysis from judicial review. Later in this opinion we hold that the Corps' analysis is reviewable to determine whether the Corps complied with its substantive obligations under Section 101 of NEPA, 42 U.S.C. § 4331 (1976), which requires agencies to weigh the economic benefits of a project against its environmental costs. See Part IV-A, G infra. The Corps relied on its cost-benefit analysis in concluding that the Locks and Dam 26 project would not have an unjustifiable adverse environmental impact.

A. Post-Authorization Decision to Implement the Project

Appellants argue that after Congress has passed authorizing legislation the Corps makes a de novo evaluation of the project, repeating all pre-authorization economic and environmental studies. On the basis of these studies the Corps computes a new cost-benefit ratio, and makes a new determination whether to build or not to build. But appellants seriously misunderstand the nature of the post-authorization planning conducted by the Corps. The Corps does not reevaluate ab initio every project that has been approved by Congress. The Corps' Engineering Regulation, ER 1110-2-1150, SA 148, states that the central purpose of post-authorization planning is to execute the congressional decision, or "to establish the most suitable overall plan for accomplishment of the authorized improvement and to establish the basic design of the project features." Id. Post-authorization planning consists largely of advanced engineering and design work on the project that results in the issuance of one or more "General Design Memoranda." Id. 150.30

It is true that the Corps makes a post-authorization decision whether to implement the approved project. This decision is more limited in scope than appellants suggest, however. Changes in external conditions may require a new project design or otherwise fundamentally alter the assumptions on which economic feasibility was originally premised. Thus at the outset of post-authorization planning the Corps will determine whether the project should be reaffirmed as authorized, or whether changes in circumstances warrant reformulation of the project. ER 1110-2-1150 at SA 148, 150, 153-159.31 In those rare situations where the Corps concludes that such changes have occurred and that reformulation is necessary, a new cost-benefit analysis may be conducted pursuant to the cost-benefit guidelines set forth at 33 C.F.R. Parts 290-295 (1979). See 33 C.F.R. § 290.12(d) (1979).32 The Corps may also decide to revise the environmental impact statement. Or the Corps may decide that because of a change in conditions an authorized project should be shelved.33 See McKinney Deposition at 21-29, 48-49 (PX 294); Cohn Deposition at 68-69, 74, 142-143 (PX 301).34

The post-authorization planning in this case followed the general pattern just outlined. Planning commenced in January 1979, about three months after enactment of P.L. 95-502. In a memorandum addressed to the District Engineer for the St. Louis District, the Corps' Chief of Engineers stated that no significant change had occurred since authorization, and that no updating of the economic analysis or the environmental impact statement was required. See Chief of Engineers Memorandum, January 10, 1979, JA 24-26, 2408-2409. After reaffirming the project, the Chief ordered preparation of a Design Memorandum containing the advanced engineering and design work necessary before construction could begin. He also ordered that the Design Memorandum restate project costs and benefits in current prices. Id. The GDM was completed on May 17, 1979. See GDM Supp. No. 2.35

Appellants argue that even if the Corps did not conduct a de novo analysis of this project, and even if its internal regulations, ER 1110-2-1150, do not ordinarily require such analysis, the Water Resources Planning Act, the Water Resources Council's Principles and Standards, and the Corps' own cost-benefit regulations, 33 C.F.R. Parts 290-295 (1979), require a new analysis. But as we have already seen, the Water Resources Planning Act and the Water Resources Council's Principles and Standards develop criteria to be used in evaluating water projects before they are submitted to Congress for authorization. Nothing in the Act or the Principles and Standards requires a post-authorization analysis. See Part II supra. And the Corps' cost-benefit regulations, which implement the Act and the Principles and Standards, are also intended primarily to apply to pre-authorization studies. It is true that they may be used if the Corps decides that a substantial change in circumstances warrants a complete reevaluation. See 33 C.F.R. § 290.12(d) (1979); text and note at note 32 supra. But nothing in the regulations requires the Corps to reevaluate every project during the post-authorization stage.

In addition to their claims about the scope of the Corps' post-authorization decisions, appellants raise several claims regarding: (1) the procedures the Corps must follow in making its post-authorization decisions, and (2) judicial review of those decisions. First, they suggest that under the Due Process Clause of the Constitution the Corps must conduct a full adjudicatory hearing before deciding to go forward with a project. If it fails to conduct such a hearing, then any disputed issues of fact must be tried de novo in the courts. Second, appellants suggest that they are entitled to a full evidentiary hearing under Sections 556 and 557 of the Administrative Procedure Act (APA), 5 U.S.C. §§ 556, 557 (1975). The courts must review the decision reached as a result of this hearing by applying the "substantial evidence" standard set forth in the APA, 5 U.S.C. § 706(2)(E) (1976). Appellants go on to suggest that if no adjudicatory hearing is provided at the agency level, then they are entitled to de novo review in the courts under Section 706(2)(F) of the APA, 5 U.S.C. § 706(2)(F) (1976). We find that these claims are without merit. We agree that the Corps' decision to implement an authorized project should be subject to judicial review, since it is a final agency decision. 5 U.S.C. § 704 (1976). We do not believe, however, that either the Constitution or the APA entitles appellants to a full adjudicatory hearing at the agency level. Nor do we believe that appellants are entitled to de novo judicial review of the agency decision, or even to review under the "substantial evidence" standard.

The Constitution plainly does not require a full hearing either before the Corps or in the courts. The protections of the Due Process Clause are extended only when a "property" or "liberty" interest has been threatened. But generalized environmental concerns do not constitute a property or liberty interest. See, e. g., Ely v. Velde, 451 F.2d 1130 (4th Cir. 1971); Federal Employees for Non-Smokers' Rights v. United States, 446 F. Supp. 181, 184-185 (D.D.C.1978), aff'd, 598 F.2d 310 (D.C.Cir.), cert. denied, 444 U.S. 926, 100 S. Ct. 265, 62 L. Ed. 2d 182 (1979). Nor can the railroads' interest in avoiding competition be viewed as a property or liberty interest creating a right to procedural due process. See, e. g., Wells Fargo Services Corp. v. Georgia Public Service Comm'n, 547 F.2d 938 (5th Cir. 1977); see also United States v. Dixie Highway Express, 389 U.S. 409, 88 S. Ct. 539, 19 L. Ed. 2d 639 (1967) (per curiam); Bowman Transportation, Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 95 S. Ct. 438, 42 L. Ed. 2d 447 (1974).36

As for their rights under the APA, nothing in that act or any other statute entitles appellants to a full adjudicatory hearing at the agency level. Under the APA, a formal adjudication is necessary under Sections 556 and 557 only when some statute requires a determination "on the record after opportunity for an agency hearing." 5 U.S.C. § 554 (1976). There is no statute setting forth such a requirement for the Corps. Although we hold below, see Part III-B infra, that the Corps must conduct a public meeting at some point during the post-authorization stage, that meeting is nonadjudicatory. Post-authorization decisions to implement a project should be viewed as an example of "informal" adjudication. Cf. Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 91 S. Ct. 814, 28 L. Ed. 2d 136 (1971) (decision to build highway is not formal adjudication).37

As for the scope of judicial review under the APA, appellants would have us ignore the Supreme Court's decision in Citizens to Preserve Overton Park, Inc. v. Volpe, supra.38 In that case the Supreme Court held that Section 706(2) (F) requires de novo review in only two circumstances: "when the action is adjudicatory and the agency factfinding procedures are inadequate," or "when issues that were not before the agency are raised in a proceeding to enforce nonadjudicatory agency action." Id. at 415, 91 S.Ct. at 823. Neither situation exists here. Moreover, the "substantial evidence" standard of Section 706(2)(E) does not apply. Review under that section is necessary only when the agency action is "taken pursuant to a rulemaking provision of the Administrative Procedure Act itself, * * * or when the agency action is based on a public adjudicatory hearing." Id. at 414, 91 S.Ct. at 822. But the Corps is not exercising its rulemaking powers. And as we have already stated, no adjudicatory hearing is required. We believe that the Corps' decision to go forward with the project should be reviewed under the "arbitrary and capricious" standard of review set forth in Section 706(2)(A), 5 U.S.C. § 706(2)(A) (1976). Cf. Overton Park, supra (decision to build highway is subject to "arbitrary and capricious" standard of review).

To summarize, the Corps makes a post-authorization decision to implement the project after determining whether there have been significant changes in circumstances. This decision, which constitutes informal adjudication, is reviewable under the "arbitrary and capricious" standard. In our view, the Corps' conclusion that there had been no major change in circumstances warranting reevaluation, and that the authorized project should be reaffirmed and implemented, was not arbitrary or capricious. Like the District Court, see Atchison V, supra, 480 F.Supp. at 987, 1001, we do not believe that any material changes affecting the economic or environmental soundness of the replacement facility occurred after passage of P.L. 95-502. Appellants point to three "events" they believe constitute material changes. First, they note a 1979 report estimating the capacity of replacement Locks and Dam 26 to range from 96 to 101 million tons per year. See Waterway Experiment Station Report (PX 300). They argue that congressional approval was based on a proposed capacity of 86 million tons. But appellants fail to recognize that these revised capacity figures are well within the range of capacity estimates disclosed to Congress in the FER and the RDSEIS. See FER Vol. 3 at G-1 to G-40, G Att-1 to G Att-180; RDSEIS Vol. 1 at 6-24, 6-144, Tables 6-5 and 6-32 (capacity up to 109 million tons); see also, e. g., FEIS Vol. 2 at 235, 279-284, 310-322, 338-355, 367-371, 443-450 (critiques of capacity estimates); Senate Hearings 1977 at 696-714, 794-805. Second, appellants note that the GDM Supp. No. 2 adjusted the cost-benefit ratio to account for 1978 prices. See GDM Supp. No. 2, 20-11 to 20-14.39 But these revisions were based on data developed at the time of or prior to enactment of the 1978 authorizing legislation. Moreover, even after the revisions were made the benefit-cost ratio remained above unity.40 Finally, appellants point to a 1979 report discussing in detail the possibility of rehabilitating existing Locks and Dam 26. GDM Foundation and Test Program (DX 80d). As the Government explains, however, this report was originally commissioned prior to the enactment of P.L. 95-502. After Congress authorized the replacement project, the Corps concluded that the rehabilitation alternative had been mooted. It nonetheless decided to permit completion of the report, because it would be of general interest to members of the engineering profession. See Chief of Engineers Memorandum, January 10, 1979, at JA 2406. In any event, the Corps was able to produce ample evidence supporting its decision not to pursue rehabilitation. See Part IV-D infra.

At the outset of post-authorization planning, the Chief of Engineers decided that a post-authorization public meeting would not be necessary since no significant changes had occurred after the enactment of P.L. 95-502. See Chief of Engineers Memorandum, January 10, 1979, at JA 2406-2409.41 The District Court found that this decision violated Corps regulations. Atchison V, supra, 480 F.Supp. at 993-994. The court ruled that the Corps should have held an informal nonadjudicatory public meeting in order to inform the community about the proposed project and to elicit views on implementation. It decided not to enter an order requiring such a meeting, however. We agree with the District Court that a public meeting should have been held, but disagree with its decision not to require a meeting.

The general policy of the Corps is to conduct their programs in "an atmosphere of public understanding * * *." 33 C.F.R. § 209.405(c) (1979). Consistent with this policy, Corps regulations provide that "at least one public meeting will be held in connection with the preconstruction planning of authorized projects." 33 C.F.R. § 209.405(f)(5) (1979). The Corps and AIMR argue that this regulation is precatory only42; they point out that 33 C.F.R. § 209.405 purports to do no more than set forth "policy, responsibility and guidance for holding * * * public meetings in connection with all Civil Works planning activities * * *."43 But as the District Court observed, the interest in public participation, which is reflected in several other regulations explicitly stating that meetings should be held, strongly supports the conclusion that the requirement is mandatory. The Corps' NEPA regulations explicitly state that "(p)ublic meetings * * * will be held during post-authorization planning studies to insure that views of interested parties will be considered in the development of the plan and that all interested parties will be kept informed of study progress." 33 C.F.R. § 209.410(n)(ii) (1979). And the Corps' internal regulations state that "one or more public meetings or workshops should be held during post-authorization planning studies to insure that views of interested parties will be considered in the development plan and that all interested parties will be kept informed of study progress."44 ER 1110-2-1150 at SA 162.45

Instead of requiring the Corps to conduct such a meeting before proceeding with construction of Locks and Dam 26, the District Court simply entered a declaratory judgment in favor of appellants. Atchison V, supra, 480 F.Supp. at 1002-1003. After noting that equitable relief is committed to the discretion of the court, it provided several reasons which led it to conclude that the Corps should not be required to hold a meeting. First, Congress had authorized the project after carefully considering all objections, and had decided that construction should begin. An order requiring the Corps to hold a public meeting might unduly delay implementation of this legislative decision. Second, appellants would not benefit from a public meeting, since they had already participated vigorously at the administrative and legislative levels. Third, the public had already received an opportunity to comment on the project through their representatives in Congress. Under the circumstances, reasoned the court, a public meeting might be "vain or useless." Id. at 1002-1003. However, the court did urge the Corps to "conduct their own evaluation of the situation and to hold a meeting if they find that one may be reasonably scheduled." Id. at 1003.

In our view, the District Court should have required the Corps to hold a public meeting. It is true that the decision whether to grant equitable relief ordinarily rests in the discretion of the trial court. Cf. Realty Income Trust v. Eckerd, 564 F.2d 447, 457 (D.C.Cir.1977) (injunctive relief need not be provided for every violation of NEPA); State of Alaska v. Andrus, 580 F.2d 465, 485-487 (D.C.Cir.1978) (same). But we do not believe that this discretion was properly exercised here. When an agency has clearly violated its own public meeting regulations, there should be a presumption in favor of an order requiring a meeting. Cf. Jones v. District of Columbia Redevelopment Land Agency, 499 F.2d 502, 513 (D.C.Cir.1974) (presumption in favor of injunction in NEPA cases); Realty Income Trust v. Eckerd, supra, 564 F.2d at 456 (same). This presumption should be particularly strong in cases like this one, where the project in question will have a wide-ranging impact. As the District Court observed, the economic and environmental impact of replacement of Locks and Dam 26 "may eventually be felt in every home and farm in the Midwest. Moreover, at stake is the soundness, utility and beauty of perhaps the greatest water resource on the continent the Upper Mississippi and its tributaries." Atchison V, supra, 480 F.Supp. at 994. The public has not had an opportunity to participate in the decisionmaking process since 1975, when the last public meetings were held. Indeed, although four public meetings were held prior to that time, they focused on the two-lock proposal; no meeting has ever been held regarding the single-lock proposal. If the fact that the public was represented in Congress vitiates the need for a post-authorization meeting, then such meetings would never be required. It may be that the benefit provided by the meeting will be relatively small. But the Corps' regulations require such a meeting, and there is no good reason why one should not be held. Thus we remand to the District Court so that it may amend its judgment to require a public meeting. This meeting should be held not later than 30 days after the time the judgment, as amended, becomes final.46

We emphasize that this meeting should be more than a mere formality. The Corps must receive any comments with an open mind. As the District Court noted, "A public meeting after an agency has already made its decision would be an empty gesture." Atchison V, supra, 480 F.Supp. at 1003. See Nat'l Tour Brokers Ass'n v. ICC, 591 F.2d 896 (D.C.Cir.1978). To ensure that the meeting is more than a formality, the District Court should also enter an order requiring the Corps to respond in writing to any objections made at the meeting. This response should be prepared within 30 days of the date the meeting is held. Such a response will not constitute a substantial departure from ordinary Corps practice; responses have been prepared in the past. See RDSEIS Vol. 3 § 9.5.3; see also 33 C.F.R. § 209.405(a) (1979) (requiring preparation of record of public meetings).

The appellants argue that the FEIS prepared by the Corps pursuant to Section 102 of NEPA, 42 U.S.C. § 4332 (1976), is deficient in several respects. They also argue that the Corps failed to fulfill its substantive obligations under Section 101 of NEPA, 42 U.S.C. § 4331 (1976). After deciding that it had jurisdiction to review these claims, Atchison IV, supra, 480 F.Supp. at 974-975, and after considering the evidence submitted by the parties, the District Court concluded that the Corps had complied with NEPA, Atchison V, supra, 480 F.Supp. at 993-1001. We agree with this determination.

Appellant AIMR suggests that the question whether the Corps has complied with NEPA is insulated from review. It makes two closely related arguments. The first pertains only to the claims of the environmental and railroad appellants that the FEIS is deficient. AIMR suggests that the FEIS, like the pre-authorization cost-benefit analysis, was prepared solely for the benefit of Congress. Thus, once Congress reviewed the project and decided that authorization should be granted, judicial review is unnecessary and intrusive. The second argument applies to both the FEIS and the Section 101 claims. AIMR suggests that Congress repealed NEPA by implication when it passed the authorizing legislation. After carefully reviewing the evidence on environmental impacts, Congress decided that construction should begin immediately. Judicial review to determine compliance with NEPA would be inconsistent with this decision. We believe that both of these arguments are without merit.

With respect to the first argument, we do not believe that the FEIS should be viewed simply as a legislative aid. Unlike the pre-authorization cost-benefit analysis, the Corps' environmental impact statements are not prepared solely for the benefit of Congress. Section 102(2)(C) of NEPA establishes the environmental impact statement requirement for proposals for legislation in part to ensure that the public has an opportunity to participate meaningfully in decisionmaking at the administrative and legislative levels. See, e. g., Sierra Club v. Morton, 510 F.2d 813, 819 (5th Cir. 1975); Trout Unlimited v. Morton, 509 F.2d 1276, 1281 (9th Cir. 1974); Silva v. Lynn, 482 F.2d 1282, 1285 (1st Cir. 1973); Atchison IV, supra, 480 F.Supp. at 774-775; see also Atchison III, supra, 459 F.Supp. at 726-730.47 Thus the argument that supported dismissal of appellants' claims that the pre-authorization cost-benefit analysis failed to comply with the Water Resources Planning Act, various other statutes, the Water Resources Council Principles and Standards, and the Corps' cost-benefit regulations, does not by itself justify dismissal of claims that the Corps' FEIS failed to comply with NEPA.

AIMR's claim that Congress impliedly repealed NEPA when it enacted P.L. 95-502 deserves somewhat more discussion. To support its argument AIMR notes that Congress gave considerable attention to the environmental claims raised by the appellants and concluded that they were without merit:

(T)he environmental impact statements have been reviewed very carefully by the Committee * * *. This final environmental impact statement setting forth the impacts of a single 1200 foot lock and new dam is considered by the Committee to provide adequate information for Congress to make its decision with respect to the recommended project. * * *

H.R.Rep. No. 95-545, supra, at 9. The Committee then concluded that "(t)he impact of the single lock proposal will be minor." Id. at 18. AIMR also points out that although P.L. 95-502 authorizes immediate construction of the new Locks and Dam 26,48 it prohibits further expansion of the waterways system until a study of the environmental and economic effects of this expansion has been completed. AIMR suggests this provision shows that Congress intended to exempt the Locks and Dam 26 project from any further environmental studies. In effect, argues AIMR, Congress reached a compromise between the environmentalists and the proponents of waterways development. See brief for AIMR at 13-14.49 The court should not now tamper with this compromise by taking steps to determine whether the Corps has complied with NEPA.

The Supreme Court has stated that repeal by implication is disfavored. In Tennessee Valley Authority v. Hill, 437 U.S. 153, 98 S. Ct. 2279, 57 L. Ed. 2d 117 (1978), it confronted the argument that Congress had impliedly repealed the Endangered Species Act, 16 U.S.C. § 1531 et seq. (1976), when it appropriated funds for construction of Tellico Dam even though it knew this project would destroy the habitat of an endangered species of fish, the snail darter. The Court rejected this claim, stating that to find Congress had repealed the Endangered Species Act by passing an appropriations measure would do violence to " 'the cardinal rule ... that repeals by implication are not favored.' " Id. at 189, 98 S.Ct. at 2299, quoting Morton v. Mancari, 417 U.S. 535, 549, 94 S. Ct. 2474, 2482, 41 L. Ed. 2d 290 (1974) (ellipsis in original). As the Supreme Court has stated elsewhere, "Only a clear repugnancy between the old law and the new results in the former giving way * * *." Georgia v. Pennsylvania R. Co., 324 U.S. 439, 457, 65 S. Ct. 716, 726, 89 L. Ed. 1051 (1945). In the absence of some affirmative showing of an intention to repeal, the only permissible justification for finding repeal by implication is when the earlier and later statutes are irreconcilable. Morton v. Mancari, supra, 417 U.S. at 550, 94 S. Ct. at 2482.

This court has applied the principles set forth by the Supreme Court to hold that an appropriations bill will not operate to relieve an agency from compliance with NEPA. See Realty Income Trust v. Eckerd, supra, 564 F.2d at 458 n.38; see also Committee for Nuclear Responsibility v. Seaborg, 463 F.2d 783, 785-786 (D.C.Cir.1971); National Audubon Society v. Andrus, 442 F. Supp. 42, 45-47 (D.D.C.1977).50 These decisions do not directly control the result in this case, however; P.L. 95-502 is not an appropriations measure, but substantive legislation authorizing a specific project. Other courts have found repeal by implication in cases involving substantive legislation. In Texas Committee on Natural Resources v. Bergland, 573 F.2d 201 (5th Cir.), cert. denied, 439 U.S. 966, 99 S. Ct. 455, 58 L. Ed. 2d 525 (1978), the Fifth Circuit ruled that a statute giving the Forest Service authority to permit immediate commencement of clearcutting in national forests precluded further inquiry under NEPA. The court reasoned that the legislative history showed that "Congress gave full attention to environmental and economic concerns inherent in managing the nation's renewable timber resources," and "clearly recognized the significant impact of clearcutting in national forests." Id. at 209. Thus the authorization provision represents a congressional decision precluding further judicial inquiry. This decision would not be subject to indirect review by requiring the Forest Service to prepare an environmental impact statement. Id. at 209-210. See Kansas ex rel. Stephan v. Adams, 608 F.2d 681 (10th Cir. 1979) (authorization bill approving termination of certain AMTRAK routes vitiates need for preparation of environmental impact statement examining consequences of route termination).51

We recognize that the arguments in favor of repeal by implication will ordinarily be stronger when Congress has passed substantive legislation authorizing a specific project than when it has enacted an appropriations bill. We note, however, that NEPA itself states that all government action must be taken in accordance with the goals set forth in the Act. 42 U.S.C. § 4332 (1976). Moreover, Congress has shown that it is fully capable of expressing its desire to exempt projects from NEPA. For example, the Trans-Alaska Pipeline Act, Public Law 93-153, provides that "actions * * * shall be taken without further action under the National Environmental Policy Act * * *." Given Congress' clearly expressed desire to ensure that all government actions are taken in accordance with NEPA, and its ability to expressly override the requirements of the Act, we believe that, even when substantive legislation is involved, repeal by implication should be found only in the rarest of circumstances. Absent very strong evidence in the legislative history demonstrating a congressional desire to repeal NEPA, or a direct contradiction between that Act and the new legislation, claims under NEPA should be reviewed. See Flint Ridge Development Co. v. Scenic Rivers Ass'n, 426 U.S. 776, 785-793, 96 S. Ct. 2430, 2436-40, 49 L. Ed. 2d 205 (1976) (clear conflict must be found before NEPA gives way); Environmental Defense Fund, Inc. v. EPA, 489 F.2d 1247 (D.C.Cir.1973); see generally Note, The Environmental Impact Statement Requirement in Agency Enforcement Adjudication, 91 Harv.L.Rev. 815, 825 (1978).

Here we are unable to find evidence in the legislative history conclusively demonstrating a congressional desire to repeal NEPA. It is true that Congress considered the environmental impact of Locks and Dam 26, and that at least some legislators believed that the FEIS was adequate.52 But as the District Court emphasized, see Atchison IV, supra, 480 F.Supp. at 975, other members of Congress believed that the commands of NEPA would continue to apply.

One, with respect to the court case of Judge Richey, that case has basically two components or two hurdles that need to be surmounted. One was the lack of authorization of locks and dam 26 and two was the adequacy of the environmental impact statement on locks and dam 26.

The bill at least clears the initial hurdle, that is, a lack of authorization of the facility. The Department of Transportation has in addition said that it will do all it can to expedite the decision on the second hurdle. So I believe that at least we have taken a step in the right direction.

124 Cong.Rec. S18050 (daily ed. Oct. 10, 1978) (statement of Senator Schmitt). See also 123 Cong.Rec. S10477 (daily ed. June 22, 1977) (statement of Senator Nelson); 124 Cong.Rec. H2700 (daily ed. Oct. 13, 1978) (statement of Representative Edgar); id. S6603 (daily ed. May 1, 1978) (statement of Senator Long). We are also unable to conclude that NEPA and P.L. 95-502 are directly contradictory. The fact that Congress did not require further environmental or economic studies of the Locks and Dam 26 project is not irreconcilable with our conclusion that judicial review of appellants' NEPA claims should be available.53

The appellants claim that the administrative record fails to include information vital to an evaluation of the NEPA claims. They argue that the Corps has failed to disclose the data, methodology, and calculations it used in preparing its cost-benefit analysis. As we noted earlier, see Part II supra, and explain in more detail below, see Part IV infra, the cost-benefit analysis is relevant because the Corps relies in part on it to conclude that the project could be justified under NEPA. See, e. g., FEIS Vol. 2 at 1.54 Like the District Court, we find that appellants' claims regarding the adequacy of the record are without merit.

Appellants are clearly correct in suggesting that the administrative record must disclose the studies and data used in compiling environmental impact statements. Moreover, any methodologies relied upon should be carefully described. The impact statement must be "sufficient to enable those who did not have a part in its compilation to understand and consider meaningfully the factors involved." Environmental Defense Fund, Inc. v. Corps of Engineers, 492 F.2d 1123, 1136 (5th Cir. 1974). See 40 C.F.R. §§ 1502.23-1502.24 (1980) (NEPA regulations requiring disclosure).55 See generally W. Rodgers, Environmental Law 725-728 (1977). NEPA clearly contemplates that the public should have an opportunity to challenge the adequacy of environmental impact statements. See Sierra Club v. Morton, supra. But without full disclosure the public would not be able to make independent judgments about the agency's action. Moreover, disclosure is necessary if the courts are to review environmental impact statements for compliance with NEPA. Additional support for the disclosure requirement is provided by the Administrative Procedure Act, 5 U.S.C. § 704 (1976). An environmental impact statement is a "final agency action" within the meaning of the APA, id. Under the APA judicial review of final agency action must "be based on the full administrative record that was before the (decisionmaker) at the time he made his decision." Citizens to Preserve Overton Park, Inc. v. Volpe, supra, 401 U.S. at 420, 91 S. Ct. at 825.56

We find no basis, however, for the appellants' contention that the Corps has failed to disclose the methodology on which it relied in computing the cost-benefit ratio; this claim is disingenuous. A step-by-step description of the methodology used by the Corps is contained in the FER, see FER Vol. 1 at 7-1 to 7-12; in the SED, see SED Vol. 1 at 7-1 to 7-20; and in the GDM Supp. No. 2, see GDM Supp. No. 2 at 20-1 to 20-20. Moreover, Corps economists apparently described the methodological basis of the Corps' studies to appellants and their technical experts. See Eickhorst Affidavit at 2-3, Appellees' Opposition to Appellants' Motion for Stay.57 In fact, they helped appellants adapt the Corps' computer programs for use on appellants' computer system. Id.58 Appellants' economic expert, Dr. Robert Haveman, used the adapted programs to produce alternative forecasts of the impact of the Locks and Dam 26 proposal. Haveman Testimony at JA 1018-1029, 1034, 1096-1100. His testimony at trial revealed a thorough understanding of the Corps' methodology.

We also find that the Corps fulfilled any obligation it had to disclose the raw data on which its cost-benefit analysis was based. Appellants claim, in particular, that the Corps has never released the data it used to determine the rate savings shippers would receive if the capacity of Locks and Dam 26 were expanded. The rate savings figures are quite important; the Corps' determination that rate savings would be significant strongly influenced its conclusion that the economic benefits of the project exceeded its cost. See, e. g., FER Vol. 1 at 5-8 (Table 5-3).59 Rate savings were determined by comparing the actual cost of shipping by barge with the rate on an alternative mode. The difference between the two was used to calculate the amount shippers would save if the capacity of Locks and Dam 26 was expanded and they could ship by barge goods currently being shipped by other modes.60 The underlying data were gathered by Charles R. Donley & Associates, a private consulting firm hired by the Corps. But as the District Court explained, Atchison V, supra, almost all of this information is already available to appellants. They possess data on the rates on alternative modes of transportation.61 They also know the tonnage, the commodity, and the origin, destination, and average rate for each of the 231 barge movements examined in compiling the rate data. See FER Vol. 3 at App. I & J. In fact, appellants have admitted that the only data they do not possess are the identities of the barge carriers interviewed by Donley & Associates. Haveman Testimony at JA 1084-1092.62 The consulting firm never released these names to the Corps, claiming that they were confidential proprietary information and therefore privileged. Chimento Deposition at JA 2942-2943, 3008-3011, 3078.

The District Court upheld the claim of privilege. We are persuaded by its reasoning. The information provided by the unidentified barge carriers was derived primarily from long-term shipping contracts. FER Vol. 3 at App. 1, attachment I-16. Thus requiring Donley & Associates to disclose the identities of the carriers might harm the competitive positions of those firms. Disclosure would also undermine Donley's ability to carry out its business; it would no longer be able to obtain information from carriers. See Chimento Deposition at JA 2942-2943, 3008-3011, 3078. It seems clear that this is legitimate proprietary information, and that a privilege should apply. It is well established that an agency may claim a privilege with respect to documents that may have influenced a particular decision. See Nat'l Courier Ass'n v. Board of Governors of Federal Reserve System, 516 F.2d 1229, 1241-1242 (D.C.Cir.1975).63

We would not uphold the claim of privilege if failure to disclose the identities of the barge carriers would substantially impair the rights of appellants. We do not believe, however, that any such impairment will occur. Appellants argue that the identities of the barge carriers are crucial if they are to challenge the accuracy of the barge rate data. They suggest that the carriers would have had an incentive to understate barge rates because they would have wanted the project to be approved. Given this possibility of distortion, appellants claim they should be given an opportunity to cross-examine the carriers. See reply brief for appellants at 41 n.9; Haveman Testimony at JA 1087, 1089-1090. But two independent verification studies establish the reliability of the Donley & Associates data. At the Corps' request the Tennessee Valley Authority independently obtained rate information and found little deviation from the Donley study.64 Second, the St. Louis District Engineer conducted a barge rate verification survey and also found little deviation.65 Moreover, if they had so desired the appellant railroads could have conducted their own verification study. Because they are competitors of the barge companies, railroads closely monitor current barge rates. See Senate Hearings 1977 at 781, 786-787; Haveman Testimony at JA 1094-1095.

Before analyzing appellants' specific objections, we will briefly discuss the standard of review to be applied in determining whether an agency has complied with NEPA. Appellants claim that the District Court should have done more than simply ensure that the agency's conclusions were based on the relevant factors and were not arbitrary and capricious. Instead, the court should have resolved any disputed factual issues de novo and by a preponderance of the evidence. Brief for appellants at 14-26. This argument is clearly incorrect; such intrusive review would be directly contrary to well settled principles.

In reviewing compliance with NEPA, courts must first determine whether the agency has complied with its "procedural" obligations under Section 102, which establishes the environmental impact statement requirement: it must ensure that the statement contains sufficient discussion of the relevant issues and opposing viewpoints to enable the decisionmaker to take a "hard look" at environmental factors, and to make a reasoned decision. Kleppe v. Sierra Club, 427 U.S. 390, 410 n.21, 96 S. Ct. 2718, 2730, n.21, 49 L. Ed. 2d 576 (1976); Sierra Club v. Adams, 578 F.2d 389, 393-396 (D.C.Cir.1978). Second, reviewing courts must determine whether the agency has complied with its "substantive" obligations under Section 101; it must ensure that the agency's conclusions are not irrational or otherwise "arbitrary and capricious."66 See Calvert Cliffs' Coordinating Committee, Inc. v. AEC, 449 F.2d 1109 (D.C.Cir.1971); County of Suffolk v. Sec'y of Interior, 562 F.2d 1368, 1383 (2d Cir. 1977), cert. denied, 434 U.S. 1064, 98 S. Ct. 1238, 55 L. Ed. 2d 764 (1978); Strycker's Bay Neighborhood Council v. Karlen, 444 U.S. 223, 226-228, 100 S. Ct. 497, 499-500, 62 L. Ed. 2d 433 (1980).67 In making these determinations the courts must be governed by a "rule of reason." Concerned About Trident v. Rumsfeld, 555 F.2d 817, 827 (D.C.Cir.1977). They should not substitute their judgment for that of the agency. Kleppe v. Sierra Club, supra, 427 U.S. at 410 n.21, 96 S. Ct. at 2730 n.21. In particular, they should not attempt to resolve conflicting scientific opinions. County of Suffolk v. Sec'y of Interior, supra, 562 F.2d at 1383. So long as the agency's conclusions have a substantial basis in fact, the mandate of NEPA has been satisfied.

NEPA requires agencies to include in their environmental impact statements a discussion of "alternatives to the proposed action." 42 U.S.C. § 4332(2)(C) (iii) (1976). Appellants contend that the FEIS prepared by the Corps fails adequately to address three alternatives to the proposed Locks and Dam 26 project: (1) rehabilitation of the existing structure; (2) investment in alternative transportation modes; and (3) improved congestion control. The District Court correctly found that these contentions lack merit. See Atchison V, supra, 480 F.Supp. at 995-996.68

When Congress has enacted legislation approving a specific project, the implementing agency's obligation to discuss alternatives in its environmental impact statement is relatively narrow. It is true that the "rule of reason" still applies. But as we explained in Sierra Club v. Adams, supra, 578 F.2d at 396:

While subsequent congressional action does not vitiate the need for an environmental impact statement or a discussion of alternatives (unless explicitly stated in the statute), such action does have a bearing on what is considered a reasonable alternative, and a reasonable discussion. * * *

In Sierra Club v. Adams the appellants argued that the Federal Highway Administration had failed to give sufficient attention to alternatives to the construction of a highway project. We approved "somewhat brief" discussions of those alternatives, "in view of the fact that Congress has authorized United States assistance in constructing a highway * * *." Id. (emphasis in original).

In this case extensive discussion of the rehabilitation alternative listed by the appellants was made unnecessary and would be inconsistent with the congressional decision to approve replacement of Locks and Dam 26. In passing P.L. 95-502 Congress was clearly motivated by a desire to increase the capacity of Locks and Dam 26. See P.L. 95-502 § 102(a); see also H.R.Rep.No.95-545, supra, at 10-11; S.Rep.No.95-215, supra, at 4-5, 14-15, 16-17. Moreover, in Section 101(j) of the statute Congress explicitly stated that the new structure should be designed "to provide for possible future expansion" of the entire inland waterways system. Rehabilitation would not fulfill this congressional mandate. P.L. 95-502 § 101(j).69 But even if there was no conflict with the congressional mandate, we believe the Corps took a sufficiently "hard look" at the rehabilitation alternative. Rehabilitation is discussed at length in the FEIS, see FEIS Vol. 1 at 70-71, 74-78, and in the RDSEIS, see RDSEIS Vol. 1 at xviii, 6-6 to 6-10, 6-52 to 6-78, which was an appendix to the FEIS. See generally FER Vol. 1 at JA 2066-2067, 2071 (economic analysis of rehabilitation).70 The appellant railroads' views on rehabilitation are printed verbatim in the FEIS and are responded to by the Corps. FEIS Vol. 2 at 274-279, 286-290. The FEIS also contains other conflicting viewpoints on rehabilitation. FEIS Vol. 2 at 104-106, 133-136, 142-145, 162-181, 186-188, 213-219, 223-224, 233-236, 239-240, 269, 451-453, 469-701. The Corps determined that massive rehabilitation would be necessary to ensure that Locks and Dam 26 remained safe over the long run. The costs of rehabilitation without interrupting traffic were believed to be comparable to or greater than the costs of replacement. FEIS Vol. 1 at ii, 75, 77. Moreover, the Corps found an unacceptably high risk that rehabilitation would encounter insoluble engineering problems and that the construction process itself might damage the facility. The Corps also reasoned that a rehabilitated facility would never be as structurally sound as a new facility. RDSEIS Vol. 1 at 6-8 to 6-10.

At trial appellants disputed the Corps' resolution of the rehabilitation issue. They presented an engineering expert, Mr. William A. Wahler, who questioned the Corps' judgment that the present structure was unsafe and that rehabilitation was not a prudent alternative. He proposed various rehabilitation schemes that he claimed would be safe and economical. Wahler Testimony at JA 1337-1473. Appellees responded by producing their own engineering expert, Mr. Homer Willis, who addressed each of the claims made by Mr. Wahler. Willis Testimony at JA 1560-1615.71 The District Court began its discussion of this evidence by stating that it should not resolve conflicting scientific judgments. It then concluded that appellants' evidence did not demonstrate that the Corps had failed to give adequate attention to the rehabilitation alternative, or that it had acted arbitrarily in rejecting that alternative; the Willis testimony and the discussion in the FEIS showed that the Corps had a reasonable basis for its decision. Atchison V, supra, 480 F.Supp. at 995-996. We see no reason to disturb this conclusion.

Appellants also suggest that the Corps failed to give adequate consideration to the possibility of increasing investment in an alternative mode of transportation specifically, railroads. In our view, the Corps fulfilled its obligation to consider this alternative. Again, extensive discussion was not necessary since Congress expressed a desire to favor development of water transportation when it passed P.L. 95-502. Despite this limited obligation, the Corps' discussion of investment in railroads is fairly detailed. See FEIS Vol. 1 at 76, 77, 80-81; id. Vol. 2 at 104-106; RDSEIS Vol. 1 at 6-11 to 6-12, 6-14 to 6-15, 6-93 to 6-134. See also FER Vol. 1 at S-3 to S-4, 3-11 to 3-29, 6-71 to 6-105. It noted that a large investment would be needed to expand the capacity of the railroads. See FEIS Vol. 2 at 104-106.72 Even if this investment were not necessary, reliance on railroads would be less desirable from an economic standpoint. See FER Vol. 1 at S-14 to S-17. The Corps also pointed out that investment in railroads would have to be accompanied by rehabilitation of Locks and Dam 26; failing to take any action would not solve the safety problem that the existing structure posed. But rehabilitation had already been shown to be undesirable. FEIS Vol. 1 at 71, 76, 77, 80, 81.

Finally, appellants claim that the Corps failed to give adequate attention to methods of controlling congestion at the current structure. They suggest, for example, that imposition of a "congestion toll" would have solved the problem by deterring inefficient tow operators from using the locks. Although congestion controls are examined in the Corps' economic analysis, see FER Vol. 1 at 6-159, 6-160; SED Vol. 1 at 6-154 to 6-155, they are not discussed in the FEIS. But congestion controls would not fulfill the congressional mandate to expand the capacity of the waterways system. Moreover, they would not solve the safety problem posed by the existing structure. Finally, as appellants have conceded, the Corps lacks the authority to impose a congestion toll. Haveman Testimony at JA 1107. As the District Court stated, "It would certainly be a waste of agency resources to test the efficiency of an alternative which is not only beyond the agency's power, but also incapable of either fully solving the problem at hand or fulfilling the mandate of Congress." Atchison V, supra, 480 F.Supp. at 996.

E. Preparation of Programmatic Environmental Impact Statement

Appellants argue that the Corps should have prepared a "programmatic" environmental impact statement that discussed the authorized project as part of a master plan to deepen the entire system's navigation channel from nine to 12 feet. According to appellants, the new facility is designed so that it can accommodate 12-foot draft traffic. The old structure could only accommodate nine-foot draft traffic. They state that the Corps also plans to deepen channels elsewhere, and that an environmental impact statement discussing the impact of this systemwide deepening must be prepared. The District Court rejected these arguments, finding that there was no secret plan to expand the waterways system, and that a programmatic environmental impact statement was not required. See Atchison V, supra, 480 F.Supp. at 996-998. Once again, we agree with the District Court's conclusion.

An environmental impact statement must be prepared whenever an agency makes a definite recommendation for federal action. Andrus v. Sierra Club, 442 U.S. 347, 350 n.2, 99 S. Ct. 2335, 2337 n.2, 60 L. Ed. 2d 943 (1979); Kleppe v. Sierra Club, supra, 427 U.S. at 399-403, 96 S. Ct. at 2725-2727; Defenders of Wildlife v. Andrus, 627 F.2d 1238, 1243-1244 (D.C.Cir.1980). The scope of the environmental impact statement depends on the scope of the proposal actually made. Kleppe v. Sierra Club, supra, 427 U.S. at 405-410, 410 n.20, 96 S. Ct. at 2728-2730 n.20. Thus, a programmatic statement is appropriate only where the proposal itself is regional or systemic in scope, or where the proposal is one of a series of interrelated proposals that will produce cumulative systemwide effects that can be meaningfully evaluated together. Id. at 403-415, 96 S.Ct. at 2727-2733; Joseph v. Adams, 467 F. Supp. 141, 157 (E.D.Mich.1979). See also National Wildlife Federation v. Appalachian Regional Comm'n, (D.C.Cir. No. 79-2349, decided March 19, 1981).73 These criteria have not been met here.

The District Court found, based on all the evidence produced at trial, that there is no pending proposal for a 12-foot channel. Atchison V, supra, 480 F.Supp. at 997. The District Court also found that the new facility will have a deeper floor than the existing structure, not so that it can accommodate 12-foot draft traffic, but so that it is more efficient in handling nine-foot draft traffic. Id. We find no reason to overturn these findings. They are overwhelmingly supported by the record. FEIS Vol. 1 at 4-5, 14-19; id. Vol. 2 at 486; RDSEIS Vol. 1 at 1-30 to 1-35, 1-55 to 1-57, 9-95 to 9-99; see also GDM Supp. No. 2 at 9-1 to 9-4; BERH Rep. at V. Appellants have been unable to identify any current proposal to deepen the channel.74 Moreover, the matter has been mooted by Congress. Section 101(i) of P.L. 95-502 expressly precludes any project that would expand the capacity of the waterways system until the master development plan required by Section 102(a) has been completed.

F. Environmental Impact of the Replacement Project

The heart of every environmental impact statement is its examination of the environmental consequences of the proposed agency action. Appellants claim that the Corps' FEIS fails to assess the systemwide biological impact of either the present barge traffic or the increment in traffic that will be caused if the new Locks and Dam 26 is built.75 Indeed, they suggest that this case represents "the most blatant disregard of the obligation to investigate, evaluate and balance environmental damage in the history of NEPA * * *." Brief for appellants at 119. This hyperbole lacks any creditable foundation. Like the District Court, we believe the FEIS adequately examines the environmental impact of the replacement project. See Atchison V, supra, 480 F.Supp. at 998-999.

As the District Court explained, see id. at 998, determining the systemic environmental impact of increased river usage is a two-step process. First, it is necessary to measure the physical effects of the increase in barge traffic. Then, once the physical effects are known, the biological impact can be determined. At least in a very general way, the physical and biological impacts of tow traffic are well understood. Barges stir up and resuspend bottom sediment; a plume of resuspended sediment may stretch behind a tow as far as two or three miles. Once the sediment has been resuspended, it may be carried by the main channel flow into backwater areas of the Illinois and Mississippi Rivers. Id.76 Increased river turbidity and sedimentation can interfere with fish habitats and upset the ecology of backwater areas. See generally Claflin Testimony at JA 1174; Sparks Testimony at JA 1275.

In order to measure more carefully the physical effects of increased tow traffic, the Corps conducted four sedimentation studies before issuing its FEIS.77 In these studies it assumed that the 18 percent expansion in the capacity of Locks and Dam 26 resulting from construction of the replacement facility would lead to roughly six more tows per day. See FEIS Vol. 1 at ii, vi, 71-73; RDSEIS Vol. 7 at 6-178 to 6-179. The Corps concluded that there would be statistically significant but minor increases in the turbidity level on the Upper Mississippi above Locks and Dam 26. FEIS Vol. 1 at 55; RDSEIS Vol. 1 at 4-14 to 4-23, 5-1 to 5-2. The impact on the Mississippi River below Locks and Dam 26 would be insignificant; turbidity levels on that stretch of the river were already extremely high. FEIS Vol. 1 at 55-56; RDSEIS Vol. 1 at 4-14 to 4-20, 5-1 to 5-2. Increases in barge traffic would have virtually no effect on turbidity levels on the Illinois River during periods of high river flow. Turbidity levels would be increased during periods of low river flow, but not above the levels present during high flow. FEIS Vol. 1 at 56-58; RDSEIS Vol. 1 at 4-14 to 4-23, 5-1 to 5-2. The studies concluded that backwater sedimentation attributable to the increases in turbidity would be insignificant. They found that (1) very large amounts of sediment are deposited in backwater areas by natural processes; (2) the amount of sedimentation caused by existing tow traffic is very slight in relation to these natural processes; and (3) the amount of sedimentation that would be caused by a several-tow-per-day traffic increase would not be substantial. FEIS Vol. 1 at vi, 23-24, 26-27, 34-35, 37, 40-41, 53-60, 62, 80; RDSEIS Vol. 1 at xv, xvii, 2-31 to 2-36, 2-333, 4-55 to 4-72, 8-2 to 8-3. As the FEIS states, increased tow traffic "cannot be blamed for the continuing, processional loss of backwater areas." FEIS Vol. 1 at 54.78

Having concluded that the physical impact of the replacement project would be minor, the Corps chose not to attempt to quantify the biological impact of the project. The Corps conceded that it would have been possible to determine the biological impact stemming from the slight increase in backwater sedimentation and turbidity. But it argued that such a study would be unjustified since it would take from five to ten years and would cost at least one million dollars.79 However, the Corps did commission a brief examination of the environmental impact on the ecosystem of the Illinois River. This study, which relied on currently available data, was unable to conclude that the biological impact would be significant. See DSEIS Vol. 3, App. J (DX 46); see also RDSEIS Vol. 2 at 9-89 to 9-94. Moreover, the FEIS does set forth potential biological problems. It explains that increased turbidity may interfere with fish, waterfowl, and various aquatic organisms, and that increased sedimentation can alter the ecology of backwater areas.80 It also carefully describes the gaps and uncertainties in the Corps' approach. FEIS Vol. 1 at vi, 59-63; see also id. at 24-43; RDSEIS Vol. 7 at 4-55 to 4-107.

The appellants claim that the Corps should have made an effort to quantify the biological impact of the Locks and Dam 26 project. More precisely, they argue that the Corps should have delayed construction of the new facility for five or ten years and conducted a multimillion dollar environmental study. See Claflin Testimony at JA 1761-1771. But appellants have failed successfully to attack the Corps' conclusion that the physical impact of the project would be minor. They produced two expert biologists who stated that the physical effects might be significant. See id. at JA 1198-1199, 1261-1271; Sparks Testimony at JA 1288-1290, 1303-1307. But they were unable to identify any other experts who supported their position. See, e. g., Claflin Testimony at JA 1262. An independent sedimentation expert supported the Corps at trial. See Simons Testimony at JA 1686-1692, 1709-1711, 1719.

Particularly in light of the fact that reviewing courts should not resolve conflicting scientific opinions, we cannot conclude that the Corps' determination that the physical effects would be minor lacked a substantial basis. And given the Corps' conclusion that the physical effects would be minor, its decision not to conduct a major biological study was clearly justified. NEPA does not require federal agencies to examine every possible environmental consequence. Detailed analysis is required only where impacts are likely. Carolina Environmental Study Group v. United States, 510 F.2d 796, 799 (D.C. Cir. 1975). Where adverse environmental impacts are not likely, expensive and time-consuming studies are not necessary. So long as the environmental impact statement identifies areas of uncertainty, the agency has fulfilled its mission under NEPA. As the Ninth Circuit stated in Jicarilla Apache Tribe of Indians v. Morton, 471 F.2d 1275, 1280 (9th Cir. 1973), "If we were to impose a requirement that an impact statement can never be prepared until all relevant environmental effects were known, it is doubtful that any project could ever be initiated." Here the Corps explained its conclusions and identified areas of uncertainty. Appellants' claim that the FEIS did not adequately address potential environmental problems must fail.

G. Balancing of Environmental and Economic Costs and Benefits

Under Section 101 of NEPA, 42 U.S.C. § 4331 (1976), agencies must consider whether the environmental costs of a project outweigh its economic and technical benefits. "A rather finely-tuned and 'systematic' balancing analysis" should be performed. Calvert Cliffs' Coordinating Committee v. AEC, supra, 449 F.2d at 1113. The appellants claim that the Corps did not fulfill its substantive duties under Section 101, because its analysis of the economic costs and benefits of the replacement project was flawed. In particular, they claim that the Corps: (1) failed properly to quantify congestion on the Illinois River; (2) omitted the costs of systemwide maintenance; and (3) failed to treat certain railroad losses as costs. According to appellants, if the Corps had included these economic costs in its analysis, it might have been less willing to conclude that the economic benefits of the replacement project outweighed any adverse environmental impact. After noting that a substantive decision should not be reversed unless it can be shown that the balance of costs and benefits struck was arbitrary or clearly gave insufficient weight to environmental values, Atchison V, supra, 480 F.Supp. at 999 (citing Calvert Cliffs', supra), the District Court reviewed appellants' claims and found that they lacked merit. Id. at 999-1001. We see no reason to overturn this finding. We note, at the outset, that appellants' contention that the Corps failed to give sufficient weight to environmental values would not be very persuasive even if the Corps did overstate the economic benefits of the project. As we held above, see Part IV-F supra, the Corps reasonably concluded that the adverse environmental consequences of the project would be minimal. Under the circumstances, it is very difficult to argue that the Corps violated the substantive mandate of NEPA. But in any event we do not believe that the flaws identified by appellants support a finding that the cost-benefit analysis was arbitrary.

First, appellants argue that the Corps failed to recognize that if the capacity of Locks and Dam 26 is expanded, any benefits obtained by alleviating delays at that facility will be offset by increased delays on the Illinois River; that is, it failed to understand that the construction project will do no more than move the bottleneck upstream. This argument is not persuasive. It is clear that the Corps did give some attention to the possibility that expansion of Locks and Dam 26 might cause delays in the rest of the system. It computed its cost-benefit ratio by developing a supply-and-demand model that enabled it to predict how a change in capacity at Locks and Dam 26 would affect usage of the entire waterways system. The possibility that an 18 percent expansion of capacity at Locks and Dam 26 would result in bottlenecks elsewhere in the system, and in particular on the Illinois River, was considered. See FER Vol. 1 at 5-1 to 5-5; SED Vol. 1 at 5-1 to 5-5; Daniel Testimony at JA 1778-1783.81 The Corps concluded that expansion of Locks and Dam 26 would not substantially increase delay costs on the Illinois River. See, e. g., FEIS Vol. 1 at 5 ("no system improvements are needed to move (the new traffic); therefore, there are no system costs attributable to the proposal"); see generally FER Vol. 1; SED Vol. 1.

In an attempt to bolster their argument, appellants contend that the Corps has underestimated the growth of local traffic on the Illinois River traffic that will not pass through Locks and Dam 26. They point to data in a General Design Memorandum for the Corps' Illinois River Duplicate Locks Project ("Illinois Duplicate Locks GDM") (PX 519), Tables B-26 and B-32, which they claim show that local coal and sludge traffic will grow substantially over the next 50 years. According to appellants, the existing Illinois locks will not have the capacity to handle both this traffic and the projected increase in long-haul traffic passing through Locks and Dam 26. But as the District Court carefully explains, see Atchison V, supra, 480 F.Supp. at 1000, appellants misrepresent the contents of the Illinois Duplicate Locks GDM. The figures relied on by appellants were computed without regard to the capacity of the Illinois River.82 The memorandum indicates that unless the capacity of the Illinois River is drastically expanded as a result of construction of duplicate locks, local coal and sludge traffic will remain at current levels. See Illinois Duplicate Locks GDM at Table B-38.83 Recent data support this view, showing that coal and sludge traffic on the Illinois has been stagnating for several years. See Daniel Testimony at JA 1782, 1873.

Appellants complain that the Corps failed to recognize losses that will be incurred by the railroads if the replacement facility is built. As appellants concede, the railroads will not lose any of their current traffic if the project goes forward. Their share of traffic will not increase, however; they will not receive roughly $135 million dollars in future revenues they have been projected to receive if the new lock is not constructed. FEIS Vol. 1 at 67-68; FER Vol. 1 at 6-98; see also FEIS Vol. 1 at 76-81. According to appellants, loss of these potential revenues should have been taken into account in computing the cost-benefit ratio. The Corps justified exclusion of these losses on two grounds. First, it claimed that there is no real loss because the revenues are merely potential gains that may never accrue. Second, it argued that the potential revenues will merely be transferred from the railroads to the barge carriers, and that these transfer payments do not represent a real loss to the national economy. The first argument is not persuasive: all of the future costs and benefits of a project are speculative. In our view, however, the second argument is sufficient.

Appellants attempt to rebut the Corps' argument that the railroads' loss of future revenues does not represent a loss to the national economy by pointing out that some portion of those future revenues would have covered the railroads' fixed costs. According to appellants, that percentage should be treated as a loss to the economy, since fixed costs must be defrayed by charges to shippers regardless of volume. They further contend that as much as 40 percent of revenues would have been used to cover fixed costs. See Haveman Testimony at JA 1017; FEIS Vol. 2 at 429, 437. The Corps adequately responds to these arguments, however. In its FER it states that railroad fixed costs will become variable costs in the long run, and thus should not be treated as a loss to the economy. See FER Vol. 1 at 6-141 to 6-147.84 The Corps has also stated that even if the percentage of future revenues that would have covered fixed costs should be treated as a real loss, that percentage is closer to 10 percent than 40 percent. See FEIS Vol. 2 at 429, 437; Daniel Testimony at JA 1783.85 There is support for the Corps' position. See Department of Transportation 1977 Report at JA 2500. Under the circumstances, we cannot conclude that the approach adopted by the Corps was arbitrary and capricious.

Third, appellants claim that appellees should not have ignored system maintenance costs. But the Corps has explained that its cost-benefit analysis focused solely on the incremental costs and benefits attributable to the replacement project. It assumed that the costs of operating and maintaining the present waterways system will be paid regardless of future construction. The benefits derived from the system as a whole were also excluded from the analysis. Daniel Testimony at JA 1802-1804. The Corps always uses this incremental approach in performing its cost-benefit analyses. Id. The incremental approach is consistent with the principles set forth by the Water Resources Council in its Principles and Standards at SA 235.

We uphold the District Court's conclusion that it was without jurisdiction to review the pre-authorization cost-benefit analysis prepared by the Corps pursuant to the Water Resources Planning Act, the Water Resources Council's Principles and Standards, the Department of Transportation Act, the River and Harbor and Flood Control Act of 1970, and its own cost-benefit regulations. Since these statutes and regulations require a cost-benefit analysis primarily to aid Congress in evaluating water resources projects, a congressional decision to approve a particular project insulates the analysis from review. We also affirm the District Court's conclusion that it could review the question whether the Corps had complied with NEPA, and that the Corps fulfilled its obligations under that statute.

At the outset of its post-authorization planning the Corps decided that since no significant changes in external conditions had occurred since the enactment P.L. 95-502, the Locks and Dam 26 project should be implemented. We hold that this decision was not arbitrary and capricious. However, the Corps violated its own regulations when it decided not to conduct a public meeting during the course of its post-authorization planning; this meeting, which would have been nonadjudicatory, would have provided the community with an opportunity to comment on GDM Supp. No. 2 and other matters relating to the replacement project. The Corps should have been required to hold such a meeting. This project is extremely important; the public should have a final chance to state its views. We remand so that the District Court may amend its judgment to require a meeting to be held within 30 days of the time the judgment, as amended, becomes final. The judgment should also be amended to require the Corps to respond to the comments made at the meeting within 30 days of the time the meeting is held.

So ordered.