OPINION

PAEZ, Circuit Judge:

WildEarth Guardians, Montanans for Quiet Recreation, Inc., and Friends of the Bitterroot, Inc. (collectively, "WildEarth"), challenge the United States Forest Service's decision to designate over two million acres of public land in the Beaverhead-Deerlodge National Forest ("Forest") for use by winter motorized vehicles, principally snowmobiles. WildEarth alleges that the Forest Service's review of the environmental impacts of snowmobiles under the National Environmental Policy Act ("NEPA")1 was inadequate in several material respects. WildEarth also alleges that the Forest Service failed to comply with the minimization requirements of Executive Order 11644. We affirm in part, reverse in part, and remand for further proceedings.

I.

At 3.35 million-acres, the Forest is the largest national forest in the state of Montana. The island mountain ranges within the Forest provide a variety of habitats spanning from cold desert to alpine peaks. Over 300 terrestrial land species live in the Forest, including grizzly bears, wolves, wolverines, lynx, and a broad variety of "big game" species, such as mule deer, white-tailed deer, black bear, moose, elk, bighorn sheep, mountain goat, and antelope. The Forest is also nationally renowned as a recreation destination. Recreational opportunities include non-motorized activities such as fishing, hunting, hiking, skiing, and mountain biking, as well as motorized activities, including motorcycle riding and snowmobiling.

In 2002, the Forest Service issued a notice of intent to revise the Land and Resource Management Plan ("forest plan") for the Forest pursuant to the National Forest Management Act, 16 U.S.C. § 1604. 67 Fed.Reg. 22,396 (May 3, 2002). The purpose of a forest plan is to guide decisions regarding natural resource management and other activity over a period of ten to fifteen years. Because a forest plan may have a significant impact on the environment, NEPA requires the Forest Service to prepare an environmental impact statement.

In January 2009, after considering various alternative plans, the Regional Forester signed and released a Record of Decision ("ROD") approving the Environmental Impact Statement2 ("EIS") and adopting the Beaverhead-Deerlodge Revised Forest Plan ("Revised Forest Plan" or "Revised Plan"). The Revised Plan, which adopts "modified Alternative Six," covers eight "revision topics," including "Recreation and Travel Management," which governs snowmobile access within the Forest. The Revised Forest Plan divides the Forest into twelve different "landscape areas," which are, in turn, divided into multiple "management areas." In 2010, the Forest Service issued a second ROD ("2010 ROD") implementing the travel management decisions in the Revised Plan.

At issue in this case is the designation in the Revised Forest Plan of over two million acres, or 60%, of the Forest for snowmobile use. As compared to prior forest plans,3 the Revised Plan decreased the area open to snowmobiles. The revision, however, will not necessarily result in a reduction of snowmobile impacts. There has been a sharp increase in snowmobile use since the 1980s, and advances in technology allow snowmobiles to reach altitudes and terrain not previously accessible.

The Regional Forester acknowledged in the ROD that "the unmanaged expansion of motorized uses[, including snowmobiles,] has resulted in resource damage, wildlife impacts, and competition and conflict between user groups." Snowmobiles affect wildlife in part because they stress animals and provoke a flight response during the winter season, when the animals are particularly vulnerable to depletion of their energy reserves. Because some species avoid all motorized vehicles, snowmobiles can effectively reduce the amount of available habitat. There is also evidence that snowmobiles can disturb reproduction cycles of wildlife species such as the wolverine. In addition to disturbing wildlife, snowmobiles can interfere with non-motorized winter recreation activities because of the noise and pollution they generate.

WildEarth and other groups filed a number of administrative appeals challenging the EIS and ROD. In October 2009, the Reviewing Officer for the Forest Service consolidated and rejected the appeals. WildEarth subsequently filed suit in the United States District Court for the District of Montana. Relevant to this appeal, WildEarth alleged that: (1) the Forest Service violated NEPA because it failed to analyze adequately the site-specific impacts of snowmobile use on big game winter habitat and conflicting recreational uses; (2) the Forest Service violated Executive Order 11644, 37 Fed.Reg. 2877 (Feb. 8, 1972), and Executive Order 11989, 42 Fed.Reg. 26,959 (May 24, 1977), because it failed to apply specified criteria when designating areas open to snowmobile use; and, (3) Subpart C of the 2005 Travel Management Rule ("TMR"), 36 C.F.R. §§ 212.80-81, which exempts over-snow vehicles ("OSVs") from compliance with the minimization criteria in Executive Order 11644 and 11989, is invalid. The Montana Snowmobile Association and the Idaho State Snowmobile Association intervened as Defendants.

The parties filed cross-motions for summary judgment, which the district court granted in part, and denied in part. Wildlands CPR, Inc. v. U.S. Forest Serv., 872 F.Supp.2d 1064 (D.Mont.2012). The court concluded that, although the Forest Service's environmental analysis of snowmobile impacts on wildlife "lack[ed] clarity," the analysis was nevertheless adequate given the deference afforded to agencies by the Administrative Procedure Act ("APA"), 5 U.S.C. § 704, and NEPA. Id. at 1078. Turning to Executive Order 11644, the court concluded that the Forest Service met the Order's requirements in designating the general areas to close to snowmobile use, but not in making designations at the route-specific level. Id. at 1082. Finally, the court ruled that WildEarth's challenge to the exemption for over-snow vehicles in Subpart C of the TMR was not ripe because the Forest Service did not rely on Subpart C to justify its actions under the Revised Forest Plan. Id. at 1083.

WildEarth timely appealed. We have jurisdiction pursuant to 28 U.S.C. § 1291.

II.

We review de novo the district court's grant of summary judgment. Lands Council v. Powell, 395 F.3d 1019, 1026 (9th Cir.2005). A final agency action "for which there is no other adequate remedy in a court" is subject to judicial review under the APA. 5 U.S.C. § 704; W. Radio Servs. Co. v. U.S. Forest Serv., 578 F.3d 1116, 1122 (9th Cir.2009). We may set aside an agency's action if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A); Pauly v. U.S. Dep't of Agric., 348 F.3d 1143, 1148 (9th Cir.2003). "We must uphold agency decisions so long as the agenc[y] ha[s] considered the relevant factors and articulated a rational connection between the factors found and the choices made." City of Sausalito v. O'Neill, 386 F.3d 1186, 1206 (9th Cir.2004) (internal quotations omitted). We have also said that an EIS is adequate if it "contains a reasonably thorough discussion of the significant aspects of the probable environmental consequences." Id. (internal quotations omitted).

III.

NEPA serves two fundamental objectives. First, it "ensures that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts." Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 349, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989). And, second, it requires "that the relevant information will be made available to the larger audience that may also play a role in both the decisionmaking process and the implementation of that decision." Id. NEPA does not impose substantive obligations on the action agency, but it does establish "procedural requirements designed to force agencies to take a `hard look' at environmental consequences." Earth Island Inst. v. U.S. Forest Serv., 351 F.3d 1291, 1300 (9th Cir.2003). NEPA and the Council on Environmental Quality's ("CEQ") regulations implementing NEPA, 40 C.F.R. §§ 1500-1508, prescribe the procedures that must be followed in conducting environmental review. Churchill Cnty. v. Norton, 276 F.3d 1060, 1071 (9th Cir.2001). "We must ... strictly interpret the procedural requirements in NEPA and the CEQ regulations to the fullest extent possible consistent with the policies embodied in NEPA." Id. at 1072 (internal quotations omitted).

WildEarth raises two NEPA arguments: first, that the Forest Service did not adequately analyze the site-specific impact of snowmobile use on big game wildlife, and second, that the Forest Service's analysis of conflicts between snowmobiles and other recreational uses was insufficient. We examine these arguments in order.

A. Impact of snowmobile use on big game wildlife

As required by NEPA, the EIS is structured around alternatives that provide varying degrees of protection for big game wildlife by managing vehicle access.4 See 42 U.S.C. § 4332(C)(iii). In the Wildlife Habitat section of the EIS, Table 176 compares the alternatives in terms of the percentage of big game winter range closed to snowmobiles throughout the Forest and in each landscape area. In the same section, Table 175 compares the "open road density for wildlife" in each alternative. In addition to this quantitative data, the EIS includes for each alternative a short qualitative discussion of the effects that snowmobiles would have on wildlife habitat, particularly the big game winter range.

WildEarth argues that the EIS fails to comply with NEPA's procedural requirements because it does not: (1) identify the location of the winter range for big game animals; (2) establish where snowmobiles impact that range; and (3) discuss what options are available to avoid the concomitant impacts. We agree. The information provided in the EIS meets neither the public disclosure purpose of NEPA nor the specific requirements in the CEQ regulations.

The CEQ regulations state that, to comply with NEPA, an agency "must insure that environmental information is available to public officials and citizens before decisions are made and before actions are taken. The information must be of high quality. Accurate scientific analysis, expert agency comments, and public scrutiny are essential to implementing NEPA." 40 C.F.R. § 1500.1(b). To fulfill NEPA's public disclosure requirements, the agency must provide to the public "the underlying environmental data" from which the Forest Service develops its opinions and arrives at its decisions. See Idaho Sporting Cong. v. Thomas, 137 F.3d 1146, 1150 (9th Cir.1998), overruled on other grounds by Lands Council v. McNair, 537 F.3d 981 (9th Cir.2008) (en banc). Alternately, the agency may incorporate publicly available data underlying the EIS by reference. 40 C.F.R. § 1502.21; see Jones v. Nat'l Marine Fisheries Serv., 741 F.3d 989, 998 (9th Cir.2013). To incorporate underlying data by reference, the agency must cite the source in the EIS and briefly describe the content. 40 C.F.R. § 1502.21. A source may be incorporated by reference only if "it is reasonably available for inspection by potentially interested persons within the time allowed for comment." Id.; see also 40 C.F.R. § 1502.24 (requiring the agency to "make explicit reference by footnote to the scientific and other sources relied upon for conclusions in the [EIS]").

Here, the Wildlife Habitat section of the EIS lists the percentage of big game winter range protected in each landscape area, but provides virtually no information about where the big game winter range is actually located, nor the concentration of game in each area. In other words, the EIS does not make public the "underlying environmental data," nor specifically reference any documentary source that the Forest Service relied upon in making its determinations on snowmobile access. See Idaho Sporting Cong., 137 F.3d at 1150; 40 C.F.R. §§ 1502.21, 1502.24.

The Forest Service points to several parts of the EIS to argue that it provided or referenced data adequately for the public to assess snowmobile impacts on the big game winter range. However, none are sufficient to satisfy NEPA's requirements.

First, the Forest Service refers to a "wolverine habitat prediction" map in the EIS. This map uses the big game winter range as an indicator of wolverine habitat because wolverines depend on big game carrion for food. Notably, the map is contained in an appendix that discusses impact on wolverine denning habitat, not big game. The EIS does not mention that the wolverine habitat map identifies the big game winter range. Nor does the EIS explain anywhere that the wolverine habitat prediction map serves as a proxy for a map of the big game winter range. An interested person, without more, would not be able to discern that a map entitled "wolverine habitat prediction" provides the baseline data for the tables depicting the big game winter range in the EIS. See 40 C.F.R. § 1502.21. And even if someone did manage to make this connection, that wouldn't be enough to access the Forest Service's baseline data, as the Forest Service concedes that this map does not accurately depict the big game winter range. The Forest Service states that it remedied this error by using updated maps provided by Montana Fish, Wildlife & Parks ("MFWP") in its final analysis. But those maps are neither included, nor referenced, in the EIS. "To take the required `hard look' at a proposed project's effects, an agency may not rely on incorrect assumptions or data in an EIS." Native Ecosystems Council v. U.S. Forest Serv., 418 F.3d 953, 964 (9th Cir.2005) (citing 40 C.F.R. § 1500.1(b)). It surely follows that the data the Forest Service provides to the public to substantiate its analysis and conclusions must also be accurate. If the wolverine habitat prediction map does not accurately depict the big game winter range, and the Forest Service ultimately worked from a different, accurate map, then it is the accurate map that must be disclosed to the public.5

Second, the Forest Service states that the information WildEarth demands in the form of a map was "otherwise provided" in the tables and accompanying qualitative discussion in the EIS. As WildEarth acknowledges, the Forest Service was not required to present the data in any particular format. See Ecology Ctr. v. Castaneda, 574 F.3d 652, 667 (9th Cir.2009) ("We defer to an agency's choice of format for scientific data."). The issue, however, is one of substance, not format. Without data on the location of the big game winter range, the public was severely limited in its ability to participate in the decision-making process.

For instance, MFWP's comments show why geography matters. It submitted extensive comments to the Forest Service on the impact snowmobile use would have on moose, one of the big game species, in several specific management areas within the Boulder River Landscape. One comment states that "[m]oose occur commonly throughout [the Boulder River-Sheepshead Management Area] and their needs should be specifically addressed, including the importance of not approaching or stressing them during the winter.... [I]t is apparent that snowmobiles are driving cross-country through willow communities, likely ... reducing moose forage." There is virtually nothing in the EIS responsive to the MFWP's comments.

The Forest Service maintains that it nonetheless did adequately discuss impacts on moose. In support, the Forest Service refers only to Table 176 in the EIS, showing the percentage of big game winter range closed to snowmobiles, and a one-sentence statement that winter non-motorized "allocations are designed to protect low elevation winter range for deer, elk, and moose." This paltry information does not allow the public to determine where the range for moose is located, whether the areas open to snowmobile use will affect that range, or whether the Forest Service considered alternatives that would avoid adverse impacts on moose and other big game wildlife. In other words, the EIS does not provide the information necessary to determine how specific land should be allocated to protect particular habitat important to the moose and other big game wildlife. Because the Forest Service did not make the "relevant information" available, Methow Valley Citizens Council, 490 U.S. at 349, 109 S.Ct. 1835, the public was limited to two-dimensional advocacy — interested persons could argue only for the allocation of more or less land for snowmobile use, but not for the protection of particular areas. As a result, the Forest Service effectively stymied the "public's ability to challenge agency action." Ecology Ctr., 574 F.3d at 667.

Third, the Forest Service argues that it adequately considered impacts on big game wildlife because it acknowledged that "motorized winter recreation can adversely affect wildlife by causing them to move away when demands on their energy reserves are highest," and provided illustrative data. This data is contained in Table 179 of the EIS showing the comparative probability that elk and mule deer would take flight from all-terrain vehicles, bicycle riders, horse riders, and hikers passing by at different distances. There is no basis for concluding that this table provides probative evidence of how big game wildlife would respond to snowmobiles in winter.

The study from which Table 179 is drawn is specific to mule, deer, and elk, not to big game species generally. Michael J. Wisdom, et al., Effects of Off-Road Recreation on Mule Deer and Elk, Transactions of the 69th North American Wildlife and Natural Resource Conference 531-50 (2004). Further, the study measures flight response to four-wheel all-terrain vehicles, not snowmobiles. Id. at 534. And, notably, the study measures flight response in spring, summer, and fall, but not winter. There is no discussion in the EIS, nor the study itself, whether this information is probative of how big game, generally, would respond to snowmobiles in winter. Id. Nor is there any acknowledgment or explanation in the EIS of the absence of data on snowmobile disturbance of specific species. See 40 C.F.R. § 1502.22 (establishing that if data is "incomplete or unavailable," then "the agency shall always make clear that such information is lacking").

We have stated that NEPA "emphasizes the importance of coherent and comprehensive up-front environmental analysis to ensure informed decisionmaking to the end that `the agency will not act on incomplete information, only to regret its decision after it is too late to correct.'" Blue Mountains Biodiversity Project v. Blackwood, 161 F.3d 1208, 1216 (9th Cir.1998) (quoting Marsh v. Or. Natural Res. Council, 490 U.S. 360, 371, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989)). Here, the Forest Service asks us to assume the adequacy and accuracy of partial data without providing any basis for doing so. NEPA requires more.

In sum, we conclude that the EIS does not provide the public adequate access to information about the impact of snowmobiles on big game wildlife and habitat. The information included in and referenced by the EIS therefore does not allow the public to "play a role in both the decisionmaking process and the implementation of that decision." See Methow Valley Citizens Council, 490 U.S. at 349, 109 S.Ct. 1835. We reverse the district court's grant of summary judgment to the Forest Service on this issue.

B. Conflicts between snowmobiles and other recreational uses

WildEarth next argues that the Forest Service violated NEPA because it did not adequately address how the snowmobile allocations in the Revised Plan affect other winter recreational activities, such as cross-country skiing and snowshoeing. We disagree and affirm the district court's ruling on this issue.

Under the Multiple-Use Sustained-Yield Act ("MUSYA"), the Forest Service must administer national forests in a manner that balances different uses and interests. 16 U.S.C. §§ 528-31. In allocating land for different uses, the Forest Service was bound by this "multiple-use mandate." See 16 U.S.C. § 1604(e). As the ROD explains, "[d]ecisions about the type and location of public recreation access are among the most difficult for land managers because of the delicate balance between competing public values and natural resource protection."

The Revised Forest Plan accomplishes this delicate task by creating five categories of recreational opportunities: (1) areas emphasizing motorized recreation; (2) areas where motorized use is permitted in winter, but not in summer; (3) areas where only non-motorized use is allowed, "providing for quiet recreation year-round"; (4) "semi-primitive backcountry" areas with a "wide mix" of motorized and non-motorized designations; and (5) designated wilderness areas where motorized use and as well as mountain biking are prohibited. Although snowmobile use is permitted in roughly 60% of the forest, 100% of the forest is open to at least some non-motorized winter recreation activities. The Forest Service explained in the EIS that it allocated recreational opportunities to meet "different user expectation[s]" and to "create logical areas where recreational settings could effectively be managed."

WildEarth's argument that the Forest Service did not adequately review the consequences of its recreation allotments is not supported in the record. The EIS includes a section devoted to "recreation and travel management," which covers both summer and winter recreation activities. This section addresses the results of a survey estimating visitation levels and the type of recreation in which visitors engaged; discusses application of the "Recreation Opportunity Spectrum," a method used to "categorize, evaluate, and monitor settings and opportunities based on the natural, managerial, and social environment"; presents a discussion of forestwide recreation trends; and presents an extensive comparison of recreational opportunities at the landscape level under each alternative. The ROD and EIS illustrate that the Forest Service collected information and, based on that information, adopted guidelines that it applied in its decision-making process. The Forest Service made that information available to the public so that interested persons could effectively participate in the process. See Methow Valley Citizens Council, 490 U.S. at 349, 109 S.Ct. 1835.

WildEarth specifically complains about the Forest Service's allocation decisions in the Mt. Jefferson management area. The ROD notes that management of snowmobile use will be difficult in this area because there is not an "effective topographical barrier to illegal motorized entry" into non-motorized areas. As we understand WildEarth's argument, the mere possibility of "illegal motorized entry" triggered a responsibility under NEPA to address in the EIS the possibility of non-compliance. NEPA, however, does not require that the Forest Service affirmatively address in an EIS every uncertainty. Cf. Lands Council, 537 F.3d at 1001. Here, the Forest Service aimed to balance recreational uses, acknowledged that the effectiveness of its decision relies upon voluntary compliance, created monitoring protocols for the area, and stated that it plans to re-evaluate its decision if noncompliance occurs.

In sum, the Forest Service provided sufficient information to establish that it took a "hard look" at the impacts of snowmobile use on non-motorized recreation in these particular management areas and throughout the Revised Plan area. See Earth Island Inst., 351 F.3d at 1300.

IV.

We next turn to WildEarth's argument that the Forest Service failed to comply with the minimization requirements in Executive Order 11644.6 In 1972, President Richard Nixon issued Executive Order 11644, directing agencies to promulgate regulations that require that all "areas and trails" allowing off-road vehicles ("ORVs")7 on public lands be located in areas that:

Exec. Order 11644 § 3(1)-(3).

Before 2005, the Forest Service permitted each national forest, or other administrative unit, to designate areas and trails open to ORV use on an ad hoc basis. See 36 C.F.R. § 295.2 (repealed 2005); George Cameron Coggins & Robert L. Glicksman, 3 Pub. Nat. Resources L. § 31:8 (2nd ed.) (2010). Recognizing that a sharp increase in ORV use, coupled with "advances in [ORV] power, range, and capabilities," had made it impossible to fulfill the intent of Executive Orders 11644 and 11989 without taking a more active approach to regulation, Final Rule: Travel Management; Designated Routes and Areas for Motor Vehicle Use, 70 Fed.Reg. 68264-01, 68265 (Nov. 9, 2005), the Secretary of Agriculture8 promulgated the 2005 Travel Management Rule ("TMR") to improve implementation of the executive orders and establish a national system of roads, trails, and areas with restricted ORV use. Id.; 36 C.F.R. §§ 212.50-212.57. The relevant section of the TMR requires that,

36 C.F.R. § 212.55(b), (b)(1)-(3) ("minimization criteria").

Notably, WildEarth does not frame its argument under the TMR, but rather challenges the Forest Service's implementation of Executive Order 11644. Wildlands, 872 F.Supp.2d at 1080-81. Where the Forest Service has placed "restrictions or prohibitions" on snowmobile use within a forest plan area, as it has here, the Forest Service must comply with the TMR, including the section implementing the criteria in Executive Order 11644. 36 C.F.R. § 212.81(c).9 Thus, by challenging the Forest Service's implementation of Executive Order 11644's minimization criteria, WildEarth is necessarily challenging the implementation of the TMR.

The district court concluded that Executive Order 11644 could be enforced through a private right of action. We need not address that issue, however, because we construe WildEarth's claim as seeking to enforce the TMR. Under the Administrative Procedure Act, an aggrieved person may challenge an agency's implementation of its own regulation, see 5 U.S.C. §§ 702, 706; Gifford Pinchot Task Force v. U.S. Fish & Wildlife Serv., 378 F.3d 1059 (9th Cir.2004).

The Forest Service argues that its analysis of snowmobile allocations in the EIS satisfied the TMR.10 Specifically, the Forest Service cites to generalized statements in the EIS that it designed snowmobile allocations to "protect low elevation winter range for deer, elk, and moose; ... [to] protect[] high elevation secure habitat for mountain goat and wolverine;[] and to provide quiet winter recreation opportunities in locations people can drive to" as evidence that it considered the minimization criteria. The district court agreed, concluding that the NEPA analysis, with its comparison of the environmental impacts in each of the six alternatives, adequately demonstrated that the Forest Service complied with the minimization criteria in designating areas open to snowmobile use.

We disagree. The EIS's reference to plan-wide data and general decision-making principles is inadequate under the TMR. There is nothing in the TMR, or anywhere else, that allows the Forest Service to designate multiple areas for snowmobile use on the basis of a single forest-wide analysis and general decisionmaking principles. Instead, the TMR requires the Forest Service to apply the minimization criteria to each area it designated for snowmobile use.

True, the TMR refers to a designated area as "an area on National Forest System lands that is designated for motor vehicle use," 36 C.F.R. § 212.1, without specifying how narrowly an area must be drawn. But it is apparent that the Forest Service must provide a more granular minimization analysis to fulfill the objectives of Executive Order 11644, which the TMR was designed to implement. See Proposed Rules: Travel Management; Designated Routes and Areas for Motor Vehicle Use, 69 Fed.Reg. 42381-01, summary (July 15, 2004); 70 Fed.Reg. 68264-01, summary (Nov. 9, 2005). Executive Order 11644 directs affected agencies to promulgate rules requiring application of the minimization criteria for "designation of the specific areas and trails on public lands on which the use of off-road vehicles may be permitted." Exec. Order 11644 § 3 (emphasis added).

Our conclusion does not require the Forest Service to area and trail it designates for snowmobile use. The TMR does not prevent the Forest Service from conducting an analysis of multiple areas and trails at once, nor from integrating NEPA and TMR compliance into a single process.11 Indeed, the Forest Service has contemplated such efficiencies in its rulemaking and guidance. See, e.g., 70 Fed.Reg. at 68279 (explaining that "public involvement associated with the NEPA process will often fulfill" the public participation requirements in the TMR). If there is data available pertinent to compliance with both NEPA and the TMR, the Forest Service can certainly use it for both purposes. What is required is that the Forest Service document how it evaluated and applied the data on an area-by-area basis with the objective of minimizing impacts as specified in the TMR. There is nothing in the record to suggest that the Forest Service did so.

In fact, the EIS and ROD demonstrate that the Forest Service neglected to consider the minimization criteria in the TMR at all. At the end of each of sixteen sections in the principal substantive chapter of the EIS devoted to "environmental consequences analysis," the Forest Service identified the "[l]egal and [a]dministrative framework" for the preceding section. For instance, the Recreation and Travel Management section identifies seven "laws and executive orders," and five "regulation[s] and polic[ies]" that form the legal and administrative framework for the analysis in that section. There is not a single citation to the TMR, Executive Order 11644, or Executive Order 11989 in the Recreation and Travel Management Section, or any of the other fifteen sections.

It is not clear why the Forest Service omitted the TMR from its analysis. But, in one of the few references to the TMR in the record, the ROD explains that the "Forest Supervisor will issue a second ROD ... making site-specific decisions based on the Revised Forest Plan ... that will include further analysis to designate routes for motorized travel under 36 C.F.R. [§] 212." The second ROD, issued in 2010, states that it "enacts the allocations and standards set forth in the 2009 Revised Forest Plan" and incorporates all its underlying analysis. The 2010 ROD acknowledges that "[t]he 2005 Travel Management Rule (36 C.F.R. [§] 212) prescribed a new process for making site-specific decisions" for route and area designations. Yet the 2010 ROD again defers compliance with the TMR when it states that "[t]he next stage of travel planning will include further analysis to formally designate routes for motorized travel in areas where motorized use is permitted under 36 C.F.R. [§] 212 Subpart B." This discussion in the 2010 ROD illustrates that, upon implementation of the allocations in the Revised Forest Plan, the Forest Service had not complied with § 212.55, but was waiting until the "next stage of travel planning." The Forest Service does not explain when — or whether — this "next stage" of planning occurred, nor whether it ever considered the TMR for purposes of winter motorized travel designations. There is no evidence in the record that it did.

Moreover, as various district courts have held, mere consideration of the TMR's minimization criteria is not sufficient to comply with the regulation. In Idaho Conservation League v. Guzman, for example, the district court determined that, although matrices included in an EIS showed that the Forest Service "met its duty to consider the minimization criteria," 766 F.Supp.2d 1056, 1071 (D.Idaho 2011), the Forest Service nonetheless failed to comply with the TMR because it did not include a "description of how the selected routes were designed `with the objective of minimizing' impacts," id. at 1073 (quoting 36 C.F.R. § 212.55(b)). As a result, there was "no way to know how or if the Forest Service used [the information in the matrices] to select routes with the objective of minimizing impacts." Id. at 1072. As another district court explained, the Forest Service is under an "affirmative obligation... to actually show that it aimed to minimize environmental damage when designating trails and areas." Cent. Sierra Envtl. Res. Ctr. v. U.S. Forest Serv., 916 F.Supp.2d 1078, 1096 (E.D.Cal.2013); see also Defenders of Wildlife v. Salazar, 877 F.Supp.2d 1271, 1304 (M.D.Fla.2012).

Center for Biological Diversity v. United States Bureau of Land Management, 746 F.Supp.2d 1055, 1079-81 (N.D.Cal. 2009) similarly noted that, "`[m]inimize'... does not refer to the number of routes, nor their overall mileage ... [but] to the effects of route designations, i.e. the [Bureau of Land Management] is required to place routes specifically to minimize `damage' to public resources, `harassment' and `disruption' of wildlife and its habitat, and minimize `conflicts' of uses." Id. (quoting 43 C.F.R. § 8342.1(a)-(c)).12

We agree with the approach taken by these district courts. First, mere "consideration" of the minimization criteria is not enough to comply with the TMR. Rather, the Forest Service must apply the data it has compiled to show how it designed the areas open to snowmobile use "with the objective of minimizing" "damage to ... forest resources," "harassment of wildlife," and "conflicts [with other] recreational uses." 36 C.F.R. § 212.55(b), (b)(1)-(3). Second, the Forest Service cannot rely upon a forest-wide reduction in the total area open to snowmobiles as a basis for demonstrating compliance with the minimization criteria. The TMR is concerned with the effects of each particularized area and trail designation. The minimization criteria must be applied accordingly.

In sum, the Forest Service's designation of areas open to snowmobile use was "not in accordance with law." 5 U.S.C. § 706(2)(A). We therefore reverse the district court's summary judgment ruling on this claim as it relates to area designations and remand for further proceedings.

V.

We turn to WildEarth's final argument, that the district court erred when it concluded that the challenge to Subpart C of the TMR was unripe. As explained above, Subpart B of the TMR implements the minimization criteria in Executive Order 11644. 36 C.F.R. § 212.55. Subpart C, however, exempts over-snow vehicles ("OSV") from compliance with Subpart B if the Forest Service does not "propose[] restrictions or prohibitions on use by over-snow vehicles." 36 C.F.R. § 212.81(c) ("Subpart C exemption").13 WildEarth argues that the Subpart C exemption is invalid because there is no distinction between motor vehicles and OSVs in Executive Order 11644. The district court determined that because the Forest Service did not apply the Subpart C exemption to justify its actions in this case, the issue was not ripe for review. Wildlands, 872 F.Supp.2d at 1082-83. Indeed, the Forest Service argued in the district court, and continues to argue before this court, that it satisfied the minimization criteria.

The ripeness doctrine serves to "prevent courts from entangling themselves in abstract disagreements over administrative policies, and also to protect the agencies from judicial interference until an administrative decision has been formalized and its effects felt in a concrete way by the challenging parties." Abbott Labs. v. Gardner, 387 U.S. 136, 148-49, 87 S.Ct. 1507, 18 L.Ed.2d 681 (1967), abrogated on other grounds by Califano v. Sanders, 430 U.S. 99, 97 S.Ct. 980, 51 L.Ed.2d 192 (1977). To determine ripeness, a reviewing court considers: (1) whether the issues are fit for judicial decision, and; (2) the hardship to the parties of withholding review. Id. at 149, 87 S.Ct. 1507. "Agency action is fit for review if the issues presented are purely legal and the regulation at issue is a final agency action." Ass'n of Am. Med. Colleges v. U.S., 217 F.3d 770, 780 (9th Cir.2000) (internal quotations omitted). In Lujan v. National Wildlife Federation, 497 U.S. 871, 873, 110 S.Ct. 3177, 111 L.Ed.2d 695 (1990), the Supreme Court cautioned against engaging in judicial review before the "controversy has been reduced to manageable proportions, and its factual components fleshed out, by concrete action that harms or threatens to harm the complainant."

Although the TMR itself was a final agency action, the Forest Service has not applied subpart C of the TMR in this case. WildEarth's challenge to the exemption is therefore a purely abstract disagreement at this juncture. Moreover, because the Forest Service has not invoked the exemption, the record contains no facts concerning the exemption's impact to inform our decisionmaking. Finally, because we fully review compliance with the minimization criteria under the TMR, there is no hardship to WildEarth in declining to exercise jurisdiction over the validity of the Subpart C Exemption. In sum, we conclude that the district court did not err in ruling that WildEarth's challenge to the Subpart C exemption is unripe.

VI.

We affirm the district court's ruling that the EIS sufficiently analyzed the conflicts between snowmobiles and other recreational uses in the Revised Forest Plan. Further, we agree that WildEarth's challenge to the Subpart C exemption in the TMR is not ripe for review. We reverse the district court's NEPA ruling, in part, because the Forest Service did not properly disclose the information underlying its analysis of snowmobile impacts on big game wildlife in the EIS. We also reverse the district court's ruling that the Forest Service adequately applied the minimization criteria in the TMR. We remand for further proceedings consistent with this opinion. The parties shall bear their own costs on appeal.

AFFIRMED IN PART, REVERSED IN PART, AND REMANDED.