COWPASTURE RIVER PRESERVATION ASSOCIATION; Highlanders for Responsible Development; Shenandoah Valley Battlefields Foundation ; Shenandoah Valley Network; Sierra Club; Virginia Wilderness Committee; Wild Virginia, Inc., Petitioners, v. FOREST SERVICE, an Agency of the U.S. Department of the Agriculture; Kathleen Atkinson, in Her Official Capacity as Regional Forester of the Eastern Region; Ken Arney, in His Official Capacity as Acting Regional Forester of the Southern Region, Respondents, Atlantic Coast Pipeline LLC, Intervenor., 911 F.3d 150

THACKER, Circuit Judge: In this case, we address whether the United States Forest Service ("Forest Service") complied with the National Forest Management Act ("NFMA"), the National Environmental Policy Act ("NEPA"), and the Mineral Leasing Act ("MLA") in issuing a Special Use Permit ("SUP") and Record of Decision ("ROD") authorizing Atlantic Coast Pipeline, LLC ("Atlantic"), the project developer, to construct the Atlantic Coast Pipeline ("ACP" or "the pipeline") through parts of the George Washington and Monongahela National Forests ("GWNF" and "MNF," respectively) and granting a right of way across the Appalachian National Scenic Trail ("ANST").

For the reasons more fully explained below, we conclude that the Forest Service's decisions violate the NFMA and NEPA, and that the Forest Service lacked statutory authority pursuant to the MLA to grant a pipeline right of way across the ANST. Accordingly, we grant the petition for review of the Forest Service's SUP and ROD, vacate those decisions, and remand to the Forest Service for further proceedings consistent with this opinion. I. A. Background The ACP is a proposed 604.5 mile, 42-inch diameter natural gas pipeline that would stretch from West Virginia to North Carolina.

The ACP route approved by the Federal Energy Regulatory Commission ("FERC") -- and for which the Forest Service issued the SUP, ROD, and right of way challenged in this case -- crosses 21 miles of national forest land (about 16 miles in the GWNF and five miles in the MNF) and crosses the ANST in the GWNF. Construction would involve clearing trees and other vegetation from a 125-foot right of way (reduced to 75 feet in wetlands) through the national forests, digging a trench to bury the pipeline, and blasting and flattening ridgelines in mountainous terrains.

Following construction, the project requires maintaining a 50-foot right of way (reduced to 30 feet in wetlands) through the GWNF and MNF for the life of the pipeline. Pursuant to NEPA, when a federal agency proposes to take a "major Federal action[ ] significantly affecting the quality of the human environment," the agency must prepare a detailed environmental impact statement ("EIS") describing the likely environmental effects, "adverse environmental effects which cannot be avoided," and potential alternatives to the proposal. 42 U.S.C. § 4332(C).

On April 27, 2015, the Forest Service provided scoping comments on FERC's Notice of Intent to prepare an EIS for the ACP project. The scoping comments stated, among other concerns, that the EIS must analyze alternative routes that do not cross national forest land, and that the EIS must address the Forest Service's policy that restricts special uses on national forest lands to those that "cannot reasonably be accommodated on non-National Forest System lands." J.A. 3593; see also Forest Serv. Manual, Addendum to Pet'rs' Br. 65-66.

The Forest Service's comments further identified concerns about landslides, slope failures, sedimentation, and impacts to groundwater, soils, and threatened and endangered species that it believed would result from the ACP project. On September 18, 2015, Atlantic filed its formal application with FERC to construct, own, and operate the pipeline. On November 12, 2015, Atlantic applied for the SUP from the Forest Service to construct and operate the pipeline across the MNF and GWNF. This application was amended in June 2016. B.

Review and Comment As FERC prepared the EIS, the Forest Service reviewed and commented on draft environmental resource reports, construction designs, biologic evaluations, and the first draft of Atlantic's Construction, Operation, and Maintenance ("COM") Plan filed with FERC. Additionally, in a letter to Atlantic dated October 24, 2016, the Forest Service requested ten site-specific stabilization designs for selected areas of challenging terrain to demonstrate the effectiveness of Atlantic's proposed steep slope stability program, which Atlantic called the "Best in Class" ("BIC") Steep Slopes Program.

As the Forest Service explained: Both the [GWNF and MNF] contain Forest Plan standards that limit activities in areas that are at high risk for slope and soil instability. To facilitate the acceptance of ACP's [SUP] application for further processing, the Forests need to be able to determine that the project is consistent or can be made consistent with this Forest Plan direction. J.A. 3379.

The letter further noted that the ten selected sites were "merely representative sites that have been selected to demonstrate whether stability can be maintained for the purpose of making a preliminary determination of Forest Plan consistency. Should the ACP Project be permitted, multiple additional high hazard areas will need to be addressed on a site-specific basis." Id. In a meeting between Atlantic and the Forest Service on November 21, 2016, Atlantic presented the first two of these site-specific stabilization designs (identified as MNF01 and GWNF02 in the October 24, 2016 letter).

According to the meeting notes, the MNF Forest Supervisor noted: [W]hile the BIC program [Atlantic] is proposing is laudable [the MNF Forest Supervisor] is skeptical the techniques will work; the Forest Service has seen slope failures on lesser slopes and would be able to provide examples. [Atlantic] needs to be able to demonstrate that the techniques will work in extreme conditions. ... The [Forest Service] wants to know beforehand that these examples have a reasonable chance of working. J.A. 3319.

Additionally, the Forest Service observed that the MNF01 and GWNF02 "drawings are a step in the right direction but more detail is needed for site specific design, the Forest Service needs to see how this lays out on the land."Id. at 3320. Thereafter, beginning in December 2016, Atlantic circulated a timeline of "FERC and Forest Service Reviews" to the Forest Service, which set the following deadlines for the agency's decisions (as proposed by Atlantic):

(1) FERC's Draft Environmental Impact Statement ("DEIS") to be issued in December 2016;

(2) FERC's Final Environmental Impact Statement ("FEIS") to be issued in June 2017; (3) the Forest Service's draft ROD to be issued also in June 2017; (4) a "Federal Agency Decision Deadline" of September 2017 (for issuance of the FERC Certificate of Convenience and Public Necessity and the Forest Service's SUP and ROD);

(5) Forest Plan amendments completed in October 2017; and (6) the pipeline in service by 2019. See J.A. 3252-53. In line with Atlantic's deadlines for the agencies' decisions, FERC issued the DEIS on December 30, 2016. Regarding its analysis of alternative routes, the DEIS explicitly stated that the ACP was routed on national forest lands in order to avoid the need for congressional approval for the pipeline to cross the ANST: A significant factor in siting ACP was the location at which the pipeline would cross the ANST .

In the general project area, the ANST is located on lands managed by either the [National Park Service ("NPS") ] or [the Forest Service]. The NPS has indicated that it does not have the authority to authorize a pipeline crossing of the ANST on its lands. Instead, legislation proposed by Congress and signed into law by the President would be necessary to allow the NPS the authority to review, analyze, and approve a pipeline crossing of the ANST on its lands.

Because of this legislative process, Atlantic considered locations where the ANST was located on lands acquired and administered by the [Forest Service] , which significantly constrained the pipeline route and severely limits opportunities for avoiding and/or minimizing the use of [National Forest System] lands. J.A. 3207-08 (emphasis supplied).

Regarding the environmental impact on forest resources, the DEIS further stated: [W]e acknowledge that a shorter pipeline route could conceptually have significantly greater qualitative impacts to sensitive resources than a longer route, which could make the longer route preferable. In this instance, we have not identified or received any information that suggests the shorter pipeline route through the National Forests has significantly greater impacts to sensitive resources than the alternative, but acknowledge that ground resource surveys have not been conducted . Id. at 3208 (emphasis supplied).

On February 17, 2017, Atlantic and the Forest Service met again to discuss the ten requested site-specific stabilization designs. During this meeting, Atlantic informed the Forest Service that the two earlier site designs were for demonstration purposes, and the remaining eight sites were not currently being designed. The Forest Service stated that it was "not comfortable" with not seeing the remaining designs, and that it was the Forest Service's understanding that specific designs for all ten sites were still needed. J.A. 2939.

Significantly, the Forest Service stated, it "want[ed] to see actual information, including specs on the actual controls and protocol on how they will be installed, not conceptual drawings." Id. On April 6, 2017, the Forest Service provided comments on FERC's DEIS. In multiple places, the Forest Service's comments stated that FERC's conclusions in the DEIS were premature given the incomplete information used to make them -- this was particularly the case regarding the extent of impacts to national forest resources and the effectiveness of mitigation techniques.

See, e.g. , J.A. 2444 ("This statement [in the DEIS] acknowledges deficiencies in information needed to conduct an appropriate effects analysis for at least some sensitive species.

Given this, the [Forest Service] has serious reservations about the conclusions of the analyses up to this point because those conclusions have been reached prior to acquiring the necessary information to substantiate what must otherwise be presumed to represent judgments based on incomplete information."); id. at 2445 ("There will be irreversible impacts to the soil and vegetation resources from construction of the ACP pipeline on [National Forest System] lands.

No matter how [Atlantic] plans to implement measures to reduce these impacts, there will still be an unavoidable irreversible dedication of the soil resource as defined by NEPA .... The [COM] Plan is currently not complete, and substantial work remains to develop and refine measures to avoid, minimize, and mitigate impacts to a variety of resources on [National Forest System] lands, including steep slopes/sensitive soils; threatened, endangered, and sensitive species; and management indicator species.").

Further, regarding the DEIS's analysis of non-national forest alternative routes, the Forest Service commented: No analysis of a National Forest Avoidance Alternative has been conducted, and environmental impacts of this alternative have not been considered or compared to the proposed action. Therefore, the Forest Service cannot support the recommendation that the National Forest Avoidance Alternative be dropped from consideration.

In our scoping comments, we requested that all alternatives, including a National Forest Avoidance Alternative, be fully addressed in regard to their feasibility and environmental effects. We hereby reiterate that request. J.A. 2454 (emphasis supplied). The Forest Service's comments on Atlantic's draft biologic evaluation, issued on April 24, 2017, paint a similarly grim picture of the ACP project's effects on erosion and on threatened and endangered species.

For example, Atlantic's draft biologic evaluation contained the following statement: "Construction activities may displace certain sensitive species from within and areas adjacent to the right-of-way, but the impact is expected to be short-term and limited to the period of construction. After construction, Atlantic will restore the right-of-way as near as practicable to preconstruction contours and conditions ...." J.A. 2324.

In response, the Forest Service stated: Restoration will consist of erosion control, some NNIS [non-native invasive species] control, and some native plant re-introduction, so it will create habitat of some sort, but the impact to sensitive species should be expected to be long-term . Restoration plantings will take many years to establish and flourish, will in most cases consist of different species than were present before, and will in many cases not re-create the conditions sensitive species need to survive.

NNIS introductions, given the current lack of plans to conduct treatment along access roads, likely will create long-term negative impacts to the ecosystem, including potentially to sensitive species . Id. (emphasis supplied).

Additionally, in response to a statement in the draft biologic evaluation that the loss of potential roosting habitat for the little brown bat (caused by construction of the pipeline and the resulting permanent right of way) would be "offset," since the species could use the right of way as foraging habitat, the Forest Service stated: A potential increase in foraging habitat (which is not really proven here) does not offset the long-term loss of good roosting habitat -- they apply to different life history needs and an increase in one does not offset loss of the other.

Also, the loss of forested habitat would be a long-term impact given the time period required for recovery. J.A. 2333. The Forest Service further noted, "Bats utilizing the more open areas (such as the [right of way] and road corridors) for foraging are also more vulnerable to predators. This offset is counteracted by an increase in potential predation, which negates the [right of way] and roads as potentially beneficial to the bat." Id. at 2332. C.

Change of Course Despite the Forest Service's clearly stated concerns regarding the adverse impacts of the ACP project, as Atlantic's deadlines for the agency's decisions drew closer, its tenor began to change. On May 14, 2017, the Forest Service sent a letter to FERC and Atlantic in which it stated -- for the first time -- that it would not require the remaining eight site-specific stabilization designs before authorizing the project.

Specifically, the letter stated: "If the ACP project is authorized, the site-specific designs for the remaining eight sites identified in our October 24, 2016 letter must be reviewed and approved by the [Forest Service] before construction at those locations could begin." J.A. 2307. The letter did not acknowledge that the agency was changing its position from its original request for all ten site designs prior to granting approval for the ACP nor did it provide any further explanation regarding the reason for the Forest Service's change in position.

On July 5, 2017, the Forest Service sent a letter to Atlantic "acknowledg[ing]" that the two site-specific stabilization designs that had so far been provided (MNF01 and GWNF02) and the subsequent information about those sites provided by Atlantic "w[ere] adequate for the purposes of disclosing the environmental effects" associated with the ACP project. Id. at 1881. The letter did not provide any explanation as to why the two plans were "adequate." On July 21, 2017, FERC released the FEIS.

On the very same day, and in line with Atlantic's timeline, the Forest Service released its draft ROD proposing to adopt the FEIS, grant the SUP, and exempt Atlantic from several forest plan standards. The FEIS's "National Forest Avoidance Route Alternatives" section, which the Forest Service commented on previously (as explained above), is identical to the DEIS.

Regarding the alternatives analysis, the Forest Service's draft ROD states: "FERC's evaluation concluded that the major pipeline route alternatives and variations do not offer a significant environmental advantage when compared to the proposed route or would not be economically practical." Id. at 1411. Regarding the COM Plan, on October 6, 2017, the Forest Service sent a letter to Atlantic stating that Atlantic's June 30 responses to the Forest Service's second draft COM Plan comments "largely addressed our comments except for a limited number of items needing further explanation or clarification." J.A. 847.

The letter requested an updated COM Plan incorporating these responses. Atlantic filed this third (and final) draft of the COM Plan on October 27, 2017. FERC issued the Certificate of Convenience and Public Necessity to ACP for construction of the pipeline on October 13, 2017. Shortly after, on October 27, 2017, the Forest Service filed its responses to objections to the draft ROD.

In response to an objection regarding the range of non-national forest route alternatives, the Forest Service stated that FERC "adequate[ly] consider[ed] the route across the National Forests" and "concluded these alternatives would not provide a significant environmental advantage over a shorter route that passes through National Forests." J.A. 676. On November 16, 2017, the Forest Service sent a letter to Atlantic regarding Atlantic's updated biologic evaluation, which had been filed on August 4, 2017.

That biologic evaluation stated that the ACP project was likely to result in a "loss of viability" for three Regional Forester Sensitive Species ("RFSS") in the MNF, a conclusion which, we note, was in line with the Forest Service's April 24, 2017 comments on the draft biologic evaluation. Nonetheless, in an about-face, the Forest Service's letter amended the updated biologic evaluation to conclude that, in fact, the project was not likely to result in a loss of viability to the three RFSS.

This conclusion is significant, because the Forest Service cannot authorize uses of national forests that are likely to result in a loss of viability for a species. See J.A. 64 ("Per [Forest Service Manual] 2670.32, activities or decisions on [National Forest System] lands 'must not result in a loss of species viability or create significant trends towards federal listing.' "). However, as noted above, the Forest Service had already issued its draft ROD proposing to authorize the SUP before the updated biologic evaluation was filed.

The Forest Service issued its final ROD on November 17, 2017, and it issued the SUP and granted the right of way across the ANST on January 23, 2018. Cowpasture River Preservation Association, Highlanders for Responsible Development, Shenandoah Valley Battlefields Foundation, Shenandoah Valley Network, Sierra Club, Virginia Wilderness Committee, and Wild Virginia, Inc. (collectively, "Petitioners") filed this challenge on February 5, 2018. We possess jurisdiction pursuant to the Administrative Procedure Act ("APA"), 5 U.S.C. §§ 701 - 06, and the Natural Gas Act, 15 U.S.C. § 717r(d)(1). II.

We may " 'hold unlawful and set aside [a federal] agency action' for certain specified reasons, including whenever the challenged act is 'arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law.' " Sierra Club, Inc. v. U.S. Forest Serv. , 897 F.3d 582, 589-90 (4th Cir. 2018) (quoting 5 U.S.C. § 706(2)(A) ).

An agency's decision is arbitrary and capricious if: the agency relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. Id. at 590 (quoting Defs. of Wildlife v. N.C. Dep't of Transp. , 762 F.3d 374, 396 (4th Cir. 2014) ). III.

Petitioners assert that the Forest Service violated three federal Acts in issuing the ROD and SUP: the NFMA, NEPA, and the MLA. We address each of these Acts and alleged violations in turn. A. National Forest Management Act The NFMA sets forth substantive and procedural standards that govern the management of national forests. See 16 U.S.C. § 1604. As this court recently explained in Sierra Club v.

Forest Service , the NFMA establishes a procedure for managing National Forest System lands using "Forest Plans," which "provide a framework for where and how certain activities can occur in national forests." Sierra Club, Inc. v. U.S. Forest Serv. , 897 F.3d 582, 600 (4th Cir. 2018) (quoting Am. Wild Horse Pres. Campaign v. Perdue , 873 F.3d 914, 919 (D.C. Cir. 2017) ; 16 U.S.C. § 1604(a) ).

First, the NFMA directs the Forest Service to "develop, maintain, and, as appropriate, revise" Forest Plans; second, it directs the Forest Service to ensure that all activities on national forest lands -- specifically, all "resource plans and permits, contracts, and other instruments for the use and occupancy of National Forest System lands" -- are consistent with the Forest Plans. Id. (quoting Perdue , 873 F.3d at 919 ; 16 U.S.C. § 1604(i) ).

The NFMA also charges the Department of Agriculture (through the Forest Service, see 36 C.F.R. § 200.3(b) ) with "promulgating guidelines for Forest Plans, which should, inter alia, 'insure consideration of the economic and environmental aspects of various systems of renewable resource management' and 'provide for diversity of plant and animal communities based on the suitability and capability of the specific land area.' " Sierra Club , 897 F.3d at 600 (quoting 16 U.S.C. § 1604(g)(3)(A)-(B) ).

At issue in this case are two Forest Service regulations issued pursuant to this authority: the 2012 Planning Rule and the 2016 Amendment to the 2012 Planning Rule, both of which deal with amendments to Forest Plans.

Petitioners assert that the Forest Service violated the NFMA by: (1) determining that amendments to the GWNF and MNF Plans' standards to accommodate the ACP were not "directly related" to the 2012 Forest Planning Rule's ("2012 Planning Rule's") substantive requirements; (2) failing to meet public participation requirements in amending forest plans; and (3) failing to analyze whether the ACP project's needs could be reasonably met off of national forest land. 1. 2012 Planning Rule Petitioners assert that the Forest Service violated the NFMA by failing to apply the substantive requirements of the 2012 Planning Rule to the amendments of the GNF and MNF Plans' standards.

Specifically, Petitioners assert that the amendments are directly related to the substantive requirements both in their purpose and their effects. a. Background In 2012, the Forest Service updated its Forest Planning Rule, which superseded the 1982 rule and set forth new, substantive requirements for Forest Plans. See 2012 Planning Rule, 77 Fed. Reg. 21,162 (U.S. Dep't of Agric. Apr. 9, 2012). The updated substantive requirements in the 2012 Planning Rule apply to Forest Plans developed under the 1982 rule in certain circumstances. See 36 C.F.R. §§ 219.8 - 219.11 ; Sierra Club , 897 F.3d at 600-01.

Specifically, as the 2016 Amendment to the 2012 Planning Rule clarified, a substantive requirement from the 2012 Planning Rule applies to a Forest Plan amendment if that requirement is "directly related to the plan direction being added, modified, or removed by the amendment." Sierra Club , 897 F.3d at 601 (quoting 36 C.F.R. § 219.13(b)(5) (emphasis supplied in Sierra Club ) ). If the substantive requirement is directly related to the amendment, then the responsible official must "apply such requirement(

s) within the scope and scale of the amendment." Sierra Club , 897 F.3d at 601 (quoting 36 C.F.R. § 219.13(b)(5) ). Conversely, if the substantive requirement from the 2012 Planning Rule is not directly related to the amendment, the responsible official is not required to apply it to the amended Forest Plan. See id. Thus, Petitioners' arguments on this point turn on whether the requirements in the 2012 Planning Rule are directly related to the Forest Service's amendments to the GWNF and MNF Plans.

A substantive requirement is directly related to the amendment when the requirement "is associated with either the purpose for the amendment or the effects (beneficial or adverse) of the amendment." Sierra Club , 897 F.3d at 602 (quoting 2016 Amendment to 2012 Rule, 81 Fed. Reg. 90,723, 90,731 (U.S. Dep't of Agric.

Dec. 15, 2016) ); see also 36 C.F.R. § 219.13(b)(5)(i) ("The responsible official's determination must be based on the purpose for the amendment and the effects (beneficial or adverse) of the amendment, and informed by the best available scientific information, scoping, effects analysis, monitoring data or other rationale.").

Further, regarding the adverse effects of an amendment, "[t]he responsible official must determine that a specific substantive requirement is directly related to the amendment when scoping or NEPA effects analysis for the proposed amendment reveals substantial adverse effects associated with that requirement, or when the proposed amendment would substantially lessen protections for a specific resource or use." 36 C.F.R. § 219.13(b)(5)(ii). b.

GWNF and MNF Plan Amendments: Purpose Analysis In its ROD, the Forest Service decided to apply project-specific amendments to a total of 13 standards in the GWNF and MNF Plans for the purpose of construction and operation of the ACP. The amendments exempt the ACP project from four MNF Plan standards and nine GWNF Plan standards that relate to soil, water, riparian, threatened and endangered species, and recreational and visual resources. Petitioners assert that the Forest Service violated the NFMA and the 2012 Planning Rule because it skipped the "purpose" prong of the "directly related" analysis.

Consistent with our decision in Sierra Club , we conclude that Petitioners are correct. Although the ROD states the rule correctly, see J.A. 36 ("[W]hether a planning regulation requirement is directly related to an amendment is based upon the amendment's purpose or its effect (beneficial or adverse)."), it fails to analyze the purpose of the amendments and instead moves directly to analyzing the amendments' effects, see id. at 36-48.

This omission is particularly striking because the Forest Service specifically identified the purpose and need for the amendments in the ROD: The purpose of the amendments are [sic] to meet the requirements of the NFMA and its implementing regulations that projects authorized on [National Forest System] lands must be consistent with the LRMP.

Without the MNF and GWNF project-specific Forest Plan amendments the ACP project would not be consistent with some Forest Plan standards related to soil, riparian, threatened and endangered species, utility corridors, the ANST, an Eligible Recreational River Area, and scenic integrity objectives. Id. at 31. Indeed, this purpose and need is repeated several times throughout the ROD.

See, e.g. , J.A. 27 ("The project-specific amendments to MNF and GWNF LRMP's [sic] approved by this decision are needed to allow the ACP Project to be consistent with LRMP standards."); id. at 37 ("[T]he purpose of the plan amendments is to ensure consistency of the ACP Project with the provisions of the two Forest Plans."). There would be no need to amend the Forest Plans to "ensure consistency" if the ACP project could meet the Forest Plan standards in the first place.

In other words, the ROD makes clear that the purpose of the amendments was to lessen certain environmental requirements in the GWNF and MNF Plans because the ACP project could not meet those Plans' existing requirements. Accordingly, by failing to analyze whether the substantive requirements of the 2012 Planning Rule are directly related to the purpose of the amendments, the Forest Service "entirely failed to consider an important aspect of the problem." Defs. of Wildlife v. N.C. Dep't of Transp. , 762 F.3d 374, 396 (4th Cir. 2014) (quoting Motor Vehicle Mnfs. Ass'n v. State Farm Mut. Auto. Ins.

Co. , 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) ).

This failure is significant, because it is clear that the amendments (intended to lessen protections for soils, riparian areas, and threatened and endangered species in the GWNF and MNF Plans) are directly related to the 2012 Planning Rule's substantive requirements for these same categories: "soil and soil productivity" ( 36 C.F.R. § 219.8(a)(2)(ii) ); "water resources" ( id. § 219.8(a)(2)(iv) ); "ecological integrity of riparian areas" ( id. § 219.8(a)(3)(i) ); "ecological integrity of terrestrial ... ecosystems" ( id. § 219.8(a)(1) ); "appropriate placement and sustainable management of ... utility corridors" (id. § 219.10(a)(3) ); and "recovery of federally listed ... species" (id. § 219.9(b) ). c.

Ex Post Facto Statements of Purpose Notwithstanding the Forest Service's statements of purpose and need in the ROD, in its briefing and at oral argument the Forest Service attempted to recharacterize the purpose of the amendments as "to relax thirteen planning standards just enough to 'authorize [Atlantic] to use and occupy [National Forest System] lands for the [ACP] Project' consistent with the forest plans." Resp't's Br. 18.

Meanwhile, Atlantic asserts that the Forest Service did "explicitly evaluate[ ] the purpose of the proposed amendments" and determined that "the purpose of ACP is not directly related to any of [the 2012 Planning Rule's] management guidelines." Intervenor's Br. 25. Instead, according to Atlantic, "the purpose of ACP is to 'serve the growing energy needs of multiple public utilities and local distribution companies, and Virginia and North Carolina' and the 'purpose and need' of the 'proposed action' is to 'respond to Atlantic's application for a special use permit.' " Id. (quoting J.A. 10, 37).

Quite the contrary -- the ROD does not analyze whether the amendments' purpose is directly related to the 2012 Planning Rule's substantive requirements. Rather, the ROD lists the purpose and need of the amendments but analyzes only the amendments' effects. See J.A. 36-48. The Forest Service's and Atlantic's attempts to recharacterize the purpose of the amendments (despite the clear statements of the amendments' purpose in the ROD) are without merit.

First, the Forest Service asserts that the true purpose of the amendments was just to authorize the ACP project -- not to lessen environmental protections for certain resources -- and that "not every amendment with an effect on a particular resource has the purpose of adjusting the forest plan's direction for that resource." Resp't's Br. 18-19 (emphasis in original). But this contradicts the Forest Service's own description of the amendments' purpose in both the ROD and in its brief, which begins with the phrase "to relax thirteen planning standards." Id. at 18.

Relaxing, lessening, loosening -- regardless of the Forest Service's verb preference, the purpose of the Forest Plan amendments is to reduce the Plans' environmental protections for certain resources. Further, this is not a situation where a proposed project-specific amendment may have an incidental effect on a Forest Plan standard; rather, the amendments' entire purpose is to weaken existing environmental standards in order to accommodate the ACP, which cannot meet the current standards.

To say that a 2012 Planning Rule requirement protecting water resources (as one example) is not "directly related" to a Forest Plan amendment specifically relaxing protection for water resources is nonsense.

Meanwhile, Atlantic conflates the purpose of the amendments to the Forest Plans with, first, the overall purpose of the ACP project (to "serve the growing energy needs of multiple public utilities and local distribution companies, and Virginia and North Carolina," Intervenor's Br. 25), and second, the Forest Service's reason for taking action at all (to "respond to Atlantic's application for a special use permit," id. ). Both

interpretations of "purpose" are facially incorrect applications of the 2012 Planning Rule's "directly related" analysis, and neither address the Forest Service's purpose for amending the GWNF and MNF Plans. First, the purpose of the plan amendment , not the ACP project, is the focus of this analysis. Second, the Forest Service's need to respond to Atlantic's application for the SUP is overly broad and does not address the need for amending the Forest Plans -- clearly, the Forest Service could have "responded" to Atlantic's application without the amendments.

Finally, both the Forest Service and Atlantic suggest that only amendments changing a management standard for the forest as a whole -- and not project-specific amendments -- can trigger the substantive requirements of the 2012 Planning Rule. See Resp't's Br. 18-20 ("A substantive requirement is directly related to the purpose for an amendment when the amendment's objective is to adjust the management of the corresponding forest resource."); Intervenor's Br. 26 ("[T]he proposed amendments for ACP did not change any of the generally applicable standards or guidelines in the forest plans.").

Neither party offers authority to support this assertion, which is contrary to the purpose of the 2012 Planning Rule: to promote consistency in the protections for national forest resources across Forest Plans. See 2012 Planning Rule, 77 Fed. Reg. at 21,162. If the Forest Service could circumvent the requirements of the 2012 Planning Rule simply by passing project-specific amendments on an ad hoc basis, both the substantive requirements in the 2012 Planning Rule and the NFMA's Forest Plan consistency requirement would be meaningless. Accordingly, in line with our decision in Sierra Club v.

Forest Service , we conclude that the 2012 Planning Rule requirements for soil, riparian resources, and threatened and endangered species are directly related to the purpose of the Forest Plan amendments. The Forest Service acted arbitrarily and capriciously in concluding otherwise. d.

Effects Analysis Although we need not reach the "effects" prong of the analysis in light of our conclusion that the purpose of the amendments is directly related to the 2012 Planning Rule's substantive requirements, the Forest Service's assertion that the Plan amendments will not have substantial adverse effects warrants additional discussion. As noted above, a substantive requirement is directly related to a Forest Plan amendment when the requirement "is associated with ... the effects (beneficial or adverse) of the amendment." Sierra Club , 897 F.3d at 602 (quoting 2016 Amendment to 2012 Rule, 81 Fed.

Reg. at 90,731 ); see also 36 C.F.R. § 219.13(b)(5)(i). The Forest Service asserts that an adverse effect must be "substantial" in order to be directly related to a substantive provision in the 2012 Planning Rule. When asked at oral argument how the Forest Service defines "substantial adverse effects," counsel for the Forest Service responded: COUNSEL: [T]he best guidance for that issue can be found in the

preamble to the 2012 [Planning] Rule where the Forest Service says that rarely, if ever, will a project-specific amendment rise to the level of having a substantial adverse effect on these resources. ... COURT: How can that be, rarely if ever will something rise to have a substantial adverse effect on the forest? How many trees do you cut down before it is a substantial adverse effect? Maybe not one. All of them? COUNSEL: The way the Forest Service stated it in the 2012

preamble to [the Planning] Rule was that it was going to look at the impact of the resource over the entire forest. Oral Argument at 22:55-24:04, Cowpasture River Preservation Ass'n v. Forest Serv. , No. 18-1144 (4th Cir. Sept. 28, 2018), http://www.ca4.uscourts.gov/oral-argument/listen-to-oral-arguments (hereinafter "Oral Argument").

It is nothing short of remarkable that the Forest Service -- the federal agency tasked with maintaining and preserving the nation's forest land -- takes the position that as a bright-line rule, a project-specific amendment, no matter how large, will rarely, if ever , cause a substantial adverse effect on a national forest. And it is even more remarkable that the agency is unable to say what would constitute a substantial adverse effect on the forest.

Indeed, counsel's response did not answer the court's question, and the Forest Service has never explained (in its briefing nor at argument) what makes an adverse effect "substantial." Even more telling, however, is that the "rarely, if ever" language used by counsel is nowhere to be found in the

preamble to the 2012 Planning Rule, nor in any other Forest Service guidance that the court could find. The closest language to counsel's assertion that the court could identify is in the

preamble to the 2016 Amendment to the 2012 Planning Rule, which states, "[i]t is unlikely that a change in land allocation for a small area would have substantial adverse effects." 2016 Amendment to 2012 Rule, 81 Fed. Reg. at 90,728. This language was a response by the Forest Service to a public comment which was concerned that the proposed rule (the 2016 Amendment) might impose a burden on small changes to land allocation. The Forest Service's full response was as follows: The 2012 rule did not require that every resource or use be present in every area.

The Department clarifies in this final rule that directly related specific substantive requirements within §§ 219.8 through 219.11 apply within the scope and scale of the amendment. Changes in land allocation for a small area would likely require a similarly narrow application of the directly related substantive requirements, depending on the purpose and effects of the changes. It is unlikely that a change in land allocation for a small area would have substantial adverse effects. Id. Even assuming that this language from the 2016 Amendment's

preamble is what counsel was referring to during argument, it still does not provide any support for the Forest Service's

interpretation of "substantial adverse effects." A "change in land allocation for a small area" is plainly not the same as generalizing to any project-specific amendment, and "unlikely" is a far cry from "rarely, if ever." Perhaps this is why counsel struggled to define what "rarely, if ever" would mean in this context. Thus, we find no basis in the law for the Forest Service's assertion that "rarely, if ever, will a project-specific amendment rise to the level of having a substantial adverse effect" on the natural forests.

In any event, the Forest Service's application of the "effects" prong of the directly related test was still flawed. In each instance in the ROD where the Forest Service concluded that the 2012 Planning Rule's substantive requirements were not "directly related" to the Plan amendments, the ROD states that the amendment "will not cause substantial long-term adverse effects." J.A. 39, 41, 43 (emphasis supplied).

But nowhere do the regulations (nor does the ROD, nor does the Forest Service's brief) state that a substantial adverse effect must be long term for the substantive requirement in the 2012 Planning Rule to be "directly related" to the amendment. The Forest Service's strained and implausible

interpretations of "substantial adverse effects" are especially striking in light of the significant evidence in the record that the GWNF and MNF Plan amendments would cause substantial adverse effects on the forests.

See, e.g. , J.A. 25 ("Sedimentation modeling indicates annual soil loss will be 200 to 800 percent above baseline erosion during the first year of construction, returning to pre-construction levels within 5 years following restoration"); id. at 2320 ("Full recovery of forested sites would take many decades."); id. at 2351 ("It is unsubstantiated as to how [erosion] increases of that magnitude are considered moderate and impacts will be temporary and minimal.").

The lengths to which the Forest Service apparently went to avoid applying the substantive protections of the 2012 Planning Rule -- its own regulation intended to protect national forests -- in order to accommodate the ACP project through national forest land on Atlantic's timeline are striking, and inexplicable. Accordingly, we conclude that the Forest Service's determination that the GWNF and MNF Plan amendments would not have substantial adverse effects on the forests was arbitrary and capricious. e.

Remand to the Forest Service Because the 2012 Planning Rule requirements for soil, riparian resources, and threatened and endangered species are directly related to the purpose and effect of the GWNF and MNF Forest Plan amendments, the Forest Service must "apply [those] requirement[s] within the scope and scale of the amendment." Sierra Club , 897 F.3d at 603 (quoting 36 C.F.R. § 219.13(b)(5) (alterations in Sierra Club ) ).

Accordingly, we remand to the Forest Service for proper application of the Planning Rule requirements for soil, riparian resources, and threatened and endangered species to the Forest Plan amendments. The Forest Service contends that remand is unnecessary because the Plan amendments already meet the substantive requirements of the 2012 Planning Rule. Thus, the Forest Service asserts, any error in applying the 2012 Planning Rule was harmless. We find no basis to support such a conclusion.

In fact, the ROD suggests just the opposite is true: in its analysis of the amendments' compliance with the 2012 Planning Rule's substantive requirements, the Forest Service explicitly stated when an amendment met the applicable substantive requirement. For example, regarding the GWNF Plan amendment for utility corridors, the ROD states: The FEIS evaluated a variety of options to transport natural gas and adequately analyzed the appropriate placement and sustainable management of the ACP. Consequently, I find this amendment meets the 36 CFR 219.10(a)(3) planning rule requirement .

Since the amendment meets the rule requirement, there is no need to make a further determination as to whether the rule requirement is directly related to it. J.A. 41-42 (emphasis supplied); see also id. at 44, 46, 47, 48, 103 S.Ct. 2856 (similarly concluding that the Plan amendments for the ANST, scenic integrity objectives, road reconstruction, and management of old growth, respectively, meet the 2012 Planning Rule's substantive requirements and thus "there is no need" to determine whether the substantive requirement is directly related to the amendment).

Yet, tellingly, the Forest Service specifically did not conclude that the GWNF and MNF Plan amendments for soils, riparian areas, and threatened and endangered species met the applicable 2012 Planning Rule's substantive requirement. Instead, it concluded (incorrectly) that in each case, the substantive requirements were not directly related to the applicable Plan amendment.

According to the ROD, conducting the directly related analysis would have been unnecessary if the amendment in fact satisfied the substantive requirement: where "the amendment meets the rule requirement, there is no need to make a further determination as to whether the rule requirement is directly related to it ." J.A. 41-42 (emphasis supplied) ). Accordingly, the case must be remanded. 2. Public Participation Requirements Petitioners further assert that the Forest Service violated the NFMA because it provided no opportunity for public comment for four of the amended forest plan standards.

Even assuming Petitioners are correct (a point the Forest Service disputes), Petitioners do not attempt to demonstrate "that the outcome of the process would have differed in the slightest had notice been at its meticulous best." Friends of Iwo Jima v. Nat'l Capital Planning Comm'n , 176 F.3d 768, 774 (4th Cir. 1999). Without even an allegation of prejudice, Petitioners fail to carry their burden to prove that any notice-related deficiency was prejudicial. Accordingly, we reject this argument. 3.

Accommodation of the ACP Project on Non-National Forest Land Petitioners assert that the Forest Service violated NEPA by failing to consider alternatives that avoid national forest land. Relatedly, Petitioners argue that the Forest Service violated the GWNF and MNF Plans and the NFMA because it failed to demonstrate that the ACP project's needs could not be reasonably met on non-national forest lands. The GWNF Plan limits "Special Use Authorizations" to "needs that cannot be reasonably met on non-[National Forest System] lands or that enhance programs and activities." J.A. 4068 (emphasis supplied).

Similarly, an MNF Plan goal states: "[p]roposed special uses of [National Forest System] lands ... are considered that meet public needs, are consistent with direction for other Forest resources and management prescriptions, and cannot be accommodated off the National Forest." J.A. 4069 (emphasis supplied). Finally, the Forest Service's regulations state: "[a]n authorized officer shall reject any proposal ... if, upon further consideration, the officer determines that: ... the proposed use would not be in the public interest." 36 C.F.R. § 251.54(e)(5)(ii).

The Forest Service Manual provides further guidance on § 251.54(e)(5)(ii), directing that a proposed use should be authorized as "in the public interest" "only if ... the proposed use cannot reasonably be accommodated off of National Forest System lands." Forest Serv. Manual, Addendum to Pet'rs' Br. 65-66 (emphasis supplied). The Forest Service Manual further directs, "[d]o not authorize the use of National Forest System lands solely because it affords the applicant a lower cost or less restrictive location." Id. at 66.

We agree that the Forest Service violated its obligations under the NFMA and its own Forest Plans because it failed to demonstrate that the ACP project's needs could not be reasonably met on non-national forest lands. The Forest Service's ROD adopted and incorporated FERC's alternative routes analysis in the EIS, but the EIS applied a different standard than the one imposed on the Forest Service by the NFMA and its own Forest Plans. In the EIS, FERC considered only whether a route alternative "confers a significant environmental advantage over the proposed route." J.A. 1533.

This is a significantly different standard than whether the proposed use "cannot reasonably be accommodated off of National Forest System lands." Forest Serv. Manual, Addendum to Pet'rs' Br. 65-66 (emphasis supplied); cf. Sierra Club , 897 F.3d at 604-05 (concluding that the Bureau of Land Management violated its MLA obligations where it failed to analyze whether alternative pipeline routes were "impractical," as required by the Bureau's regulations, and instead adopted an EIS that considered only whether an alternative route offered a "significant environmental advantage").

Accordingly, adopting FERC's EIS was not sufficient for the Forest Service to fulfill its obligations under the Forest Service Manual and its own Forest Plans, and the Forest Service did not purport to undertake this required analysis anywhere else in the ROD. The Forest Service asserts that it "determines project consistency only 'with respect to standards and guidelines,' not general forest planning 'goals' like Monongahela LS17." Resp't's Br. 24 (quoting 2012 Planning Rule, 77 Fed. Reg. at 21,241 ).

As an initial matter, the Forest Service regulations and the Forest Service Manual apply to both the GWNF and the MNF, so even if the court were to disregard the MNF goal cited by Petitioners, the proposed use of national forest land must still fit the Forest Service Manual's definition of "in the public use," which contains essentially the same requirement as the MNF goal: that the proposed use cannot be reasonably accommodated outside of the national forest. See Forest Serv. Manual, Addendum to Pet'rs' Br. 65-66.

However, the Forest Service's assertion about forest planning goals and objectives deserves additional discussion. The regulatory guidance quoted by the Forest Service -- from the

preamble to the 2012 Planning Rule, 77 Fed. Reg. at 21,241 -- is a response by the Forest Service to a public comment regarding the 2012 Planning Rule's consistency requirement, which states: The Forest Service policy was that consistency could only be determined with respect to standards and guidelines, or just standards, because an individual project alone could almost never achieve objectives and desired conditions. ...

The Department continues to believe that the consistency requirement cannot be interpreted to require achievement of the desired conditions or objectives of a plan by any single project or activity , but we believe that we can provide direction for consistency to move the plan area toward desired conditions and objectives, or to not preclude the eventual achievement of desired conditions or objectives , as well as direction for consistency with the other plan components. 77 Fed. Reg. at 21,241 (emphasis supplied).

In other words, even if the Forest Service is not required to conclude that an individual project alone meets a forest planning goal, it is not free to disregard the goal entirely -- as the Forest Service apparently wishes to do here. The Forest Service was aware of its obligation to determine that the ACP project could not be reasonably accommodated on non-national forest land from the beginning of the project. Indeed, the Forest Service specifically cited to the Forest Service Manual and Forest Plan requirements in its initial scoping comments in response to FERC's Notice of Intent to Prepare an EIS.

See J.A. 3593 ("[T]he analysis must address Forest Service Manual direction that restricts special uses to those that cannot reasonably be accommodated on non-National Forest System lands (FSM 2703.2)."); id. at 3593-94 (stating that the GWNF Plan requires special use authorizations be "[l]imit[ed] to needs that cannot be reasonably met on non-[National Forest System] lands or that enhance programs and activities"). The Forest Service's failure to undertake this analysis violated the NFMA.

Accordingly, we remand to the Forest Service for proper analysis of whether the ACP project's needs can be reasonably met on non-national forest lands, in compliance with the NFMA and the GWNF and MNF Plans. B. National Environmental Policy Act As this court recently explained in Sierra Club v. Forest Service , Congress enacted NEPA "to reduce or eliminate environmental damage." 897 F.3d at 590 (quoting Dep't of Transp. v. Pub. Citizen , 541 U.S. 752, 756, 124 S.Ct. 2204, 159 L.Ed.2d 60 (2004) ).

" 'NEPA itself does not mandate particular results in order to accomplish these ends,' but rather, 'imposes only procedural requirements on federal agencies with a particular focus on requiring agencies to undertake analyses of the environmental impact of their proposals and actions.' " Id . (quoting Dep't of Transp. , 541 U.S. at 756-57, 124 S.Ct. 2204 ). NEPA requires that agencies consider alternatives to the proposed action, 40 C.F.R. § 1502.14, and "take a hard look at environmental consequences," Robertson v.

Methow Valley Citizens Council , 490 U.S. 332, 350, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989) (internal quotation marks omitted). To that end, whenever a federal agency proposes to take a "major Federal action[ ] significantly affecting the quality of the human environment," the agency must prepare a detailed EIS describing the likely environmental effects of the proposal, any unavoidable adverse environmental effects, and potential alternatives. 42 U.S.C. § 4332(2)(C). Consideration of alternatives "is the heart of the [EIS]." 40 C.F.R. § 1502.14.

In this case, FERC was the lead agency charged with issuing the EIS, and the Forest Service acted as a cooperating agency by assisting FERC to analyze the environmental impacts to 430 acres of national forest lands on the proposed ACP route. As a cooperating agency, the Forest Service may adopt FERC's EIS only if it undertakes "an independent review of the [EIS]" and "concludes that its comments and suggestions have been satisfied." 40 C.F.R. § 1506.3(c) ; see also Sierra Club , 897 F.3d at 590. It must also ensure that the EIS is "adequate" under NEPA regulations. 40 C.F.R. § 1506.3(a).

In reviewing an EIS, the court's responsibility is to "determine whether the [agency] has considered the relevant factors and articulated a rational connection between the facts found and the choice made." Sierra Club , 897 F.3d at 594 (quoting Balt. Gas & Elec. Co. v. Nat. Res. Def. Council, Inc. , 462 U.S. 87, 105, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983) ). Petitioners assert that the Forest Service violated NEPA by (1) failing to study alternative off-forest routes, and (2) adopting a FEIS that failed to take a hard look at landslide risks, erosion, and degradation of water quality. 1.

Study of Alternative Off-Forest Routes As noted above, an agency may only adopt an EIS if it "meets the standards for an adequate statement" under the applicable regulations. 40 C.F.R. § 1506.3(a). One applicable regulation provides: If a [DEIS] is so inadequate as to preclude meaningful analysis , the agency shall prepare and circulate a revised draft of the appropriate portion. The agency shall make every effort to disclose and discuss at appropriate points in the draft statement all major points of view on the environmental impacts of the alternatives including the proposed action.

Id. § 1502.9(a) (emphasis supplied). Petitioners assert that FERC's FEIS was inadequate because it failed to sufficiently study alternative pipeline routes for the ACP that avoided national forest lands. According to Petitioners, the Forest Service violated NEPA because it adopted FERC's inadequate EIS without undertaking the required "independent review," and because the FEIS did not satisfy the Forest Service's earlier comments and suggestions on the DEIS. Id. § 1506.3(c).

In counter, the Forest Service asserts that once FERC had issued the Certificate of Convenience and Public Necessity, the choice before the Forest Service was simple: either approve the pipeline route as it was authorized by FERC or deny the right of way. According to the Forest Service, since FERC was responsible for analyzing alternative pipeline routes, the Forest Service reasonably relied on that alternatives analysis in adopting the FEIS.

The Forest Service frames Petitioners' argument as an impermissible collateral attack on FERC's actions, but that ignores the Forest Service's obligation to "independent[ly] review" the EIS and ensure its comments and suggestions to the lead agency were satisfied before adopting it. 40 C.F.R. § 1506.3(c). Neither the Forest Service nor Atlantic points to evidence in the record to demonstrate that the Forest Service undertook the required independent review. To the contrary, the record suggests that they did not.

Instead, the record reflects that at first the Forest Service strenuously objected to the lack of non-national forest route alternatives in the DEIS, but it eventually reversed course and adopted the FEIS even though the analysis of non-national forest alternatives was unchanged from the DEIS -- all in an effort to prevent Atlantic from having to obtain congressional approval for the project to cross the ANST.

From the beginning, the Forest Service made clear through its comments to FERC and Atlantic that the EIS would need to analyze non-national forest alternative routes and justify the necessity of any proposed route crossing of national forest lands. The Forest Service's scoping comments for the ACP project noted: It is ... necessary to understand why any proposed routes (preferred or alternative) crossing [National Forest System] lands are selected over those not crossing [National Forest System] lands.

Therefore, the EIS should contain a comparison of project effects for routes crossing [National Forest System] lands versus routes not crossing [National Forest System] lands. Discussions and other relevant information should also be provided to justify the necessity of any proposed route crossing [National Forest System] lands. ... Comparisons of the alternatives should be based on analyses of site-specific impacts to resources potentially affected by the proposed project, which may not necessarily be correlated with the footprint of the proposed project. J.A. 3593.

Then, FERC's DEIS indicated that "[a] significant factor in siting ACP was the location at which the pipeline would cross the ANST." J.A. 3207. As the DEIS stated, crossing the ANST on NPS lands would require congressional approval.

"Because of this legislative process " -- that is, to avoid obtaining congressional approval to cross the ANST on NPS lands -- "Atlantic considered locations where the ANST was located on [Forest Service lands], which significantly constrained the pipeline route and severely limits opportunities for avoiding and/or minimizing the use of [National Forest System] lands." Id. at 3207-08 (emphasis supplied).

Because of this, and even though ground resource surveys had not been conducted, FERC concluded that it "ha[d] not identified or received any information that suggests the shorter pipeline route through the National Forests has significantly greater impacts to sensitive resources than the alternative" that avoided national forest lands. Id. at 3208.

In response to this analysis of off-forest routes in the DEIS, the Forest Service commented: No analysis of a National Forest Avoidance Alternative has been conducted, and environmental impacts of this alternative have not been considered or compared to the proposed action. Therefore, the Forest Service cannot support the recommendation that the National Forest Avoidance Alternative be dropped from consideration. In our scoping comments, we requested that all alternatives, including a National Forest Avoidance Alternative, be fully addressed in regard to their feasibility and environmental effects.

We hereby reiterate that request. Id. at 2454. Further, in response to the DEIS's assertion that in general, as the length of a pipeline route increases, the environmental impacts also increase, the Forest Service commented: "Miles of line do not necessarily equate to severity of the environmental impact. The nature of the resources to be impacted needs to be considered. The Forest Service has previously requested that such comparative information on impacts be obtained and considered for alternatives to the proposed action." Id. at 2451.

Despite the Forest Service's concerns regarding the lack of study of off-forest alternatives, the "National Forest Avoidance Route Alternatives"

section in the FEIS is identical to the DEIS. Nevertheless, on the very same day that FERC issued the FEIS, the Forest Service released its draft ROD, which proposed adopting the FEIS (and, consequently, the unchanged alternatives analysis). Without explaining the Forest Service's change of position from the scoping comments or its comments on the DEIS, the draft ROD states: "FERC's evaluation concluded that the major pipeline route alternatives and variations do not offer a significant environmental advantage when compared to the proposed route or would not be economically practical." J.A. 1411.

The Forest Service's discussion on this point was essentially identical in its response to objections filed to the draft ROD and in its final ROD. The Forest Service asserts, "Petitioners present no record evidence that FERC did not" continue to analyze non-national forest alternatives following the Forest Service's comments on the DEIS. Resp't's Br. 39.

But no such analysis is apparent anywhere in the record, and most tellingly, neither the Forest Service nor Atlantic even attempt to identify evidence to demonstrate that FERC did anything to address the Forest Service's concerns about off-forest alternative routes. What is apparent from the record is that: (1) the Forest Service repeatedly expressed concerns about the need to analyze alternative pipeline routes that avoided the national forests (particularly in the scoping comments, comments on the draft resource reports, and the DEIS);

(2) FERC's analysis of alternative pipeline routes remained unchanged from the DEIS to the FEIS, and there is no other evidence apparent from the record that FERC addressed the Forest Service's concerns about off-forest alternative routes; and (3) the Forest Service never explains, in the ROD or elsewhere, how its concerns about off-forest alternative routes were assuaged. The chain of events surrounding the Forest Service's sudden acquiescence to the alternatives analysis in the FEIS is similar to that in Sierra Club v.

Forest Service , where we determined that the Forest Service had acted arbitrarily and capriciously in adopting the sedimentation analysis in the FEIS for a different pipeline project. See Sierra Club , 897 F.3d at 594-96. Here, like in Sierra Club , "[g]iven the circumstances, we simply cannot conclude that the Forest Service undertook an independent review and determined that its comments and concerns were satisfied" when it seemingly dropped its demand that off-forest alternative routes be studied before the ACP was authorized without any further analysis. Id. at 595.

In light of this, and particularly considering the Forest Service's earlier skepticism that location decisions for the ACP were made solely to avoid congressional approval, we hold that adopting the unchanged alternatives analysis in the FEIS was arbitrary and capricious. 2. Analysis of Landslide Risks, Erosion, and Degradation of Water Quality Petitioners further contend that the Forest Service's deficient analysis of landslide risks, erosion impacts, and water quality degradation from the ACP project violated NEPA.

Specifically, Petitioners assert that the Forest Service abandoned its request for ten site-specific stabilization designs prior to granting the SUP, which it previously stated were necessary to evaluate effects under NEPA, and instead accepted the two that Atlantic provided as "adequate" without explanation for this change in position. Additionally, Petitioners assert that Atlantic's erosion and sedimentation mitigation plan had not been determined at the time the FEIS and ROD were issued.

Thus, the Forest Service did not know if the mitigation measures it relied on to approve the project would actually be successful. As a result, Petitioners argue that the FEIS does not provide "a thorough investigation into the environmental impacts of [the] agency's action." Pet'rs' Reply Br. 29 (quoting Nat'l Audubon Soc'y v. Dep't of Navy , 422 F.3d 174, 185 (4th Cir. 2005) ).

For its part, the Forest Service contends that it thoroughly analyzed the impacts of the proposed route on national forest lands, and that NEPA does not require an agency to formulate and adopt a complete mitigation plan before it can act. As noted above, NEPA does not require the Forest Service to ensure "environment-friendly outcomes." Nat'l Audubon Soc'y , 422 F.3d at 184.

Rather, "an agency decision is acceptable even if there will be negative environmental impacts resulting from it, so long as the agency considered these costs and still decided that other benefits outweighed them. 'NEPA merely prohibits uninformed -- rather than unwise -- agency action.' " Id. (quoting Robertson , 490 U.S. at 350-51, 109 S.Ct. 1835 (citations omitted) ). Nevertheless, an EIS must still "contain a detailed discussion of possible mitigation measures." Robertson , 490 U.S. at 351, 109 S.Ct. 1835.

Further, NEPA requires "particular care" "when the environment that may be damaged is one that Congress has specially designated for federal protection," such as national forests. Nat'l Audubon Soc'y , 422 F.3d at 186-87. We conclude that the Forest Service violated NEPA by failing to take a hard look at the environmental consequences of the ACP project.

The Forest Service expressed serious concerns that the DEIS lacked necessary information to evaluate landslide risks, erosion impacts, and degradation of water quality, and it further lacked information about the effectiveness of mitigation techniques to reduce those risks.

Specifically, the record reflects that the Forest Service voiced concerns about (1) authorizing the SUP without ten site-specific stabilization designs to demonstrate the effectiveness of Atlantic's BIC program; (2) the overly high efficiency rate of erosion control devices used in the sedimentation analysis (96 percent); (3) relying on the use of water bars as a mitigation technique, when Atlantic had not analyzed whether water bars would mitigate or exacerbate erosion effects during construction; and

(4) Atlantic's use of averaged versus episodic sediment calculations to analyze the water resource impacts from increases in sedimentation due to the ACP project. However, the FEIS did not address any of these concerns; rather, it made clear that this incomplete and/or inaccurate analysis in the DEIS remained incomplete.

The FEIS stated (among other examples): "slope instability/landslide risk reduction measures have not been completed or have not been adopted," J.A. 1615; "[Atlantic is] currently working to provide documentation of the likelihood that their proposed design features and mitigation measures would minimize the risk of landslides in the project area," id. at 1616 (emphasis supplied); "specific [erosion] effects are unknown" and "it is unclear if erosion control and rehabilitation measures would meet the standards of the Forest Plan[s]," id. at 1659; and "water resource impacts from sedimentation are largely uncertain," id. at 1663.

Accordingly, the FEIS could not have satisfied the Forest Service's concerns that the DEIS lacked necessary information to evaluate the environmental consequences of the pipeline. Indeed, the FEIS conceded that the Forest Service's concerns remained unresolved. Nevertheless, as Atlantic's deadlines drew near, the Forest Service disregarded these concerns and adopted the FEIS -- including its conclusions that landslide risks, erosion impacts, and degradation of water quality remained unknown -- the very same day FERC issued it.

To support its decision to approve the project and grant the SUP, the Forest Service relied on the very mitigation measures it previously found unreliable. This was insufficient to satisfy NEPA, and did not constitute the necessary hard look at the environmental consequences of the ACP project. a.

Landslide Risks The Forest Service clearly explained its concerns about landslides, erosion, and pipeline safety and stability in its October 24, 2016 letter requesting the ten site-specific stabilization designs: The route for the [ACP project] proposed by [Atlantic] would cross some very challenging terrain in the central Appalachians. Potentially difficult situations include steep slopes, presence of headwater streams, geologic formations with high slippage potential, highly erodible soils, and the presence of high-value natural resources downslope of high hazard areas.

These hazards are exacerbated by high annual rates of precipitation and the potential for extreme precipitation events. Similar hazards on other smaller pipeline projects in the central Appalachians have led to slope failures, erosion and sedimentation incidents, and damage to aquatic resources. Therefore, the [Forest Service] is concerned that crossing such challenging terrain with a much larger pipeline could present a high risk of failures that lead to resource damage. J.A. 3379.

In addition to highlighting these concerns, the Forest Service's October 24, 2016 letter made clear that the ten selected sites were "merely representative sites," required for the Forest Service to determine whether the ACP project could be permitted in the GWNF and MNF. J.A. 3379. In other words, the site designs were needed to aid the Forest Service in its decision whether to permit the pipeline at all. Accordingly, the Forest Service's later decision to only require the designs prior to construction was not simply a question of timing.

It meant the Forest Service approved the pipeline without information it previously determined was necessary to making its decision, and it did so without acknowledging, much less explaining, its change in position. The Forest Service's reversal is particularly puzzling considering the reason it requested the site-specific stabilization designs in the first place: to demonstrate that Atlantic's BIC program could actually work in particular conditions, rather than simply being a "cookbook with generalities." J.A. 2514.

The Forest Service also conducted a literature review on Atlantic's BIC incremental controls to attempt to determine the effectiveness of these measures. Far from proving the effectiveness of the BIC program, the literature review concluded: "[T]he majority of these BIC incremental controls are either too new to provide any real insight to the effectiveness on erosion control, especially on steep slopes, or there has not been any research to prove the effectiveness of these incremental controls for adequate erosion control." Id. at 3703.

Thus, despite its own well-documented concerns with Atlantic's mitigation plans, the Forest Service abandoned its request for the eight site-specific stabilization designs and adopted the FEIS, all without science-based evidence of the BIC program's effectiveness. This falls far short of NEPA's hard look requirement, and the Forest Service's brief, conclusory letter stating that the information provided by Atlantic was "adequate" is insufficient to show that the Forest Service's concerns had been addressed as NEPA requires. J.A. 1881.

Perhaps nothing demonstrates the dangers of the Forest Service's insufficient analysis of landslide risks clearer than the FEIS's use of the Columbia Gas Transmission pipeline as an example of an existing pipeline in the Appalachian Mountains that safely crosses karst terrain. See, e.g. , J.A. 1589, 1609 ("There are differences between ACP and corridor and the Columbia pipeline project and corridor, and so, there can be more potential for project-induced slope failures in the ACP corridor.

But the decades of slope stability performance of the Columbia pipeline corridor on slopes generally similar to those along the ACP pipeline route is relevant information to consider."). Significantly, during the briefing of this case, a landslide in Marshall County, West Virginia, caused the Columbia pipeline -- highlighted by the Forest Service for its safety and stability -- to rupture and explode.

Clearly, the Forest Service's concerns about landslide risks and pipeline safety highlighted in its October 24, 2016 letter deserve serious consideration, for the protection of both the environment and the public. b. Erosion Impacts and Degradation of Water Quality In adopting the FEIS and approving the pipeline, the Forest Service concluded that because of "mitigation measures, impacts on groundwater and surface waters will be effectively minimized or mitigated." J.A. 25.

However, as explained above, the Forest Service had previously expressed serious concerns about the extensive erosion and sedimentation that the ACP project could cause, and it additionally questioned the mitigation techniques that Atlantic relied on to reduce those impacts. This is particularly true regarding the overly high efficiency rate of erosion control devices used in the sedimentation analysis (96 percent), the use of water bars as a mitigation technique, and the use of averaged versus episodic sediment calculations to analyze water resource impacts in the sedimentation analysis.

Despite these concerns, and the FEIS's conclusion that "specific [erosion] effects [remained] unknown," id. at 1659, the Forest Service nevertheless relied on the incomplete analysis in the FEIS and disregarded its concerns about the effectiveness of the mitigation techniques. For example, in the draft biologic evaluation, Atlantic asserted that installation of erosion control devices would "reduce erosion by about 96 percent." J.A. 2633.

The Forest Service criticized this conclusion in its March 10, 2017 comments to the draft biologic evaluation, stating, "Use of lab testing and efficiency rates are inappropriate for steep slope pipeline construction. Update model with more conservative assumptions about containment efficiencies. Document the literature references that apply to efficiencies in the field, particularly mountainous terrain in WV and VA." Id. at 2357.

However, Atlantic did not comply with the Forest Service's request, and the 96 percent erosion control efficiency rate remained in Atlantic's August 2017 Soil Erosion and Sedimentation Modeling Report. See J.A. 909 ("Installation of [erosion control devices] was predicted to reduce erosion by about 96 percent.").

We note that this report was issued five months after the Forest Service directed Atlantic to update its erosion efficiency rate, one month after the Forest Service issued its draft ROD, just two months before the final version of the COM Plan was issued, and only three months before the Forest Service issued the final ROD.

Accordingly, we see no evidence in the record that the Forest Service's concerns regarding the 96 percent erosion control efficiency rate were ever resolved; nonetheless, the Forest Service ultimately relied on this figure to determine that Atlantic's proposed mitigation measures would effectively reduce erosion and sedimentation impacts from the ACP project.

During oral argument, Atlantic claimed that the Forest Service's concern about the 96 percent efficiency rate was resolved because Atlantic agreed not to use silt fences as a mitigation technique in certain areas, which it claims were the cause of the "overly optimistic" efficiency rate. Oral Argument at 37:50-39:41. As counsel for Atlantic stated: The Forest Service never accepted the 96 percent efficiency. Indeed, that model was predicated on a standard erosion and sediment control device called the silt fence.

Instead of debating ... over the percent effectiveness of the silt fence, the Forest Service made a much more direct and compelling move, which was to prohibit the use of silt fences in the areas over which it had concern ... Atlantic committed not to use the silt fences that were the subject of the overly optimistic erosion sediment model. Id. As an initial matter, we note that the Soil Erosion and Sedimentation Modeling Report attributes the 96 percent erosion control efficiency rate to all erosion control devices "such as silt fences, waterbars, and mulch application," not just silt fences. J.A. 929.

Additionally, the final draft of the COM Plan is riddled with uses of silt fences as proposed mitigation techniques. See, e.g. , id. at 303, 409, 473, 475, 586, 587. However, even if Atlantic is correct that it committed not to use silt fences in certain areas, this is beside the point. The use of silt fences was not the problem. The problem, as the Forest Service itself pointed out, was assuming that these devices would function nearly perfectly to reduce erosion and sediment, despite a wealth of evidence to the contrary. This assumption remained in the August 2017 Soil Erosion and Sedimentation Modeling Report.

See J.A. 908 n.2 ("The effectiveness predicted by the model is influenced by slope, soil, groundcover, and type of erosion control device; the model assumes perfect installation, soil retention, and maintenance ." (emphasis supplied) ). This assumption infected the sedimentation model -- the model that produced the "200 to 800 percent above baseline erosion" estimate cited in in the ROD. Id. at 25. Crucially, we can identify no other more conservative efficiency rate used to correct the sedimentation model which drove the Forest Service's erosion and sedimentation analysis.

Indeed, the use of the 96 percent efficiency rate in the August 2017 Soil Erosion and Sedimentation Modeling Report, which was issued only three months before the Forest Service's final ROD, suggests that the Forest Service's concern with Atlantic's overly high efficiency rate for erosion control devices was never resolved. See J.A. 908-09 ("Installation of [erosion control devices] was predicted to reduce erosion by about 96 percent."). Additionally, the FEIS relied on the use of water bars as a mitigation technique that would reduce the environmental impacts of the ACP project.

See J.A. 1662 ("The use of water bars (i.e., slope breakers) was assumed on long slopes ...."). The Forest Service had previously stated in its comments on Atlantic's updated biologic evaluation that further analysis was needed to determine whether water bars would be effective: "Slope breaker locations relative to pertinent habitat features need to be disclosed[.] It is important to be sure that they are not potentially directing water into habitats (in which case they would actually do more harm than good)." Id. at 2337.

Nevertheless, the FEIS candidly acknowledged that this further analysis was never done: [W]ater bars create concentrated flows where they discharge adjoining off right-of-way areas. The [Forest Service] has stated that Atlantic has not assessed how or whether the adjoining areas can receive concentrated flows, or whether measures would be implemented to allow these areas to safely receive and convey the concentrated flows . In addition, the slopes to be encountered in the MNF and GWNF would require several water bars to be "stacked" along their length, creating multiple points of discharge.

The [Forest Service] has stated the potential impacts of multiple points of concentrated discharges onto the adjoining areas has not been assessed . Id. at 1663 (emphasis supplied).

Once again, the Forest Service adopted the FEIS (including its use of water bars as a mitigation technique), issued its ROD, and granted the SUP based on an erosion and sedimentation analysis using water bars as a mitigation technique, despite the clear evidence in the record that (1) the Forest Service had concerns with this technique; (2) the Forest Service's concerns were not resolved in the FEIS; and (3) the effectiveness of water bars for this project was never analyzed.

Finally, the record further reflects that the Forest Service believed Atlantic used an incorrect calculation to analyze how sedimentation from the ACP project would impact aquatic species.

In its draft biologic evaluation, Atlantic analyzed the total sediment that would erode a stream in a year divided by the volume of water that would flow through the stream in a year -- to create an average sediment level over an entire year -- rather than analyzing sediment levels in terms of discrete episodic events, where the sediment levels vary based on precipitation events that cause larger amounts of erosion to enter the stream. In other words, Atlantic employed a simplistic (and unrealistic) calculation that made in-stream sedimentation levels look much lower than they would be during construction.

Of note, the Forest Service sharply criticized this approach in its comments on the draft biologic report: This entire paragraph has false rationale and needs to be deleted or modified extensively. Erosion and sediment transport to streams cannot be averaged evenly over a year, rather it happens in discrete episodic events. It is not appropriate to minimize impacts by making a comparison of total load evenly spread over time. The point of the load calculation is to address impacts to sensitive aquatic species which are impacted by flow and timing of sediment during these erosion events. J.A. 2358.

However, despite the Forest Service's concerns with Atlantic's calculations in the sedimentation analysis, the record does not indicate that Atlantic ever updated its calculation to reflect actual conditions. Nevertheless, the Forest Service adopted Atlantic's updated biologic report and the FEIS, and it concluded that erosion and sedimentation from the ACP project would not substantially adversely affect sensitive aquatic species. The Forest Service argues -- correctly -- that NEPA does not req

[…]