SCHROEDER, Chief Judge:
Since 1978, the Outer Continental Shelf Lands Act (“OCSLA”), 43 U.S.C. § 1331 et seq., has set forth the procedures for administrative approval of offshore oil drill[783]*783ing on the Outer Continental Shelf. In this case, six native Alaskans and the environmental organization Greenpeace, Inc., seek review of the Secretary of the Interi- or’s approval of the development and production plan (“DPP”) for the Northstar oil and gas development project, located off the north coast of Alaska in the Beaufort Sea. Petitioners (“Edwardsen”) challenge both the adequacy of the final environmental impact statement (“EIS”) under the National Environmental Policy Act (“NEPA”), 42 U.S.C. § 4321 et seq., and the compliance of the oil discharge prevention and contingency plan (“spill response plan”) with the requirements of § 4202(a) of the Oil Pollution Act of 1990 (“OPA”), codified at 33 U.S.C. § 1321(j). We have jurisdiction to review the NEPA claims. We lack jurisdiction, however, to review the spill response plan, because it was approved in a separate agency action and OPA vests review of such plans in the district court. See 33 U.S.C. § 1321(n).
BACKGROUND
BP Exploration (Alaska), Inc. (“BPXA”) seeks to produce oil from Northstar, an oil and gas reservoir that extends from two to eight miles off the north coast of Alaska in the Beaufort Sea. It is a harsh environment: the average annual temperature is eleven degrees Fahrenheit, and ice covers coastal areas of the sea for at least nine months each year.
Undeterred by the severe climate conditions, a number of animal species make their home in the Beaufort Sea and on its shores. These include caribou and the endangered bowhead whale. For over 4,000 years, the area has also been home to the Inupiat Eskimo, whose traditional subsistence lifestyle includes the hunting of caribou and bowhead whales.
Beneath the Beaufort Sea, the North-star reservoir contains an estimated 158 million barrels of oil reserves. The state and federal governments sold leases to this field in 1979. The federal leases account for about twenty percent of the reserves, and state leases make up the remainder. Exploration began in 1983.
BPXA acquired the rights of the original lessees in 1995 and applied for federal, state, and local approval to begin producing oil from the Northstar reservoir. BPXA proposes to reconstruct and expand Seal Island, an artificial gravel island located near the center of the Northstar reservoir on state submerged lands. BPXA plans to drill oil and gas production wells, gas injection wells, and waste disposal wells from Seal Island. Some wells are to be in the federal portions of the reservoir. A six-mile-long pipeline, buried beneath the sea floor, will carry oil from Seal Island to the shore. A second pipeline, submerged in the same trench as the oil pipeline, will carry natural gas to Seal Island for use as fuel. On shore, the oil pipeline will run for eleven miles above ground to a connection with the Trans-Alaska Pipeline. The Trans-Alaska Pipeline will transport Northstar oil to Valdez, Alaska. From there, tankers will carry the oil to ports in the western U.S. and abroad. Production from the project is expected to last fifteen years.
BPXA applied to the U.S. Army Corps of Engineers (the “Corps”) for permits under Section 404 of the Clean Water Act, 33 U.S.C. § 1344, and Section 10 of the Rivers and Harbors Appropriation Act of 1899, 33 U.S.C. § 403, and to the Minerals Management Service of the U.S. Department of the Interior (“MMS”) for approval of the DPP under OCSLA, 43 U.S.C. § 1351.
In 1995, the Corps determined that the issuance of a permit would constitute a “major Federal action” that would require [784]*784the preparation of an EIS. The MMS and the U.S. Environmental Protection Agency (“EPA”) followed suit in 1996. A single EIS was prepared by the Corps, as lead federal agency, in conjunction with the MMS, the U.S. Fish and Wildlife Service, National Marine Fisheries Service, and the EPA, with the assistance of a third-party contractor funded by BPXA. In July 1999, the MMS adopted the EIS “for use in its decision.” Two months later, the MMS approved the DPP, in accordance with the recommendation of the state of Alaska.
In October 1999, the petitioners sought review of the MMS’ approval of the DPP in this court. Petitioner Greenpeace, Inc., is an international environmental organization. The individual petitioners are Inupi-at Eskimos who maintain that the approval of Northstar threatens their ability to continue hunting, fishing, and gathering traditional subsistence resources.
DISCUSSION
A. NEPA Claims
OCSLA, as amended in 1978, identifies four distinct stages in the development of an offshore oil well on the Outer Continental Shelf. See Secretary of the Interior v. California, 464 U.S. 312, 336-337, 104 S.Ct. 656, 78 L.Ed.2d 496 (1984). “The four stages are: (1) formulation of a five year leasing plan by the Department of the Interior; (2) lease sales; (3) exploration by the lessees; (4) development and production. Id. at 337, 104 S.Ct. 656. The final stage-development and production-is at issue here.”
Before commencing development and production, the lessee must submit a development and production plan to the Department of the Interior for approval. See 43 U.S.C. § 1351(a)(1). The required scope and contents of the DPP are defined by 43 U.S.C. § 1351(c) and 30 C.F.R. § 250.204. The Secretary must forward the DPP to the governor of any affected state and, upon request, to any affected local government. 43 U.S.C. § 1351(a)(3). The Secretary is required to accept the recommendations of the governor upon determining that they “provide for a reasonable balance between the national interest and the well-being of the citizens of the affected State.” 43 U.S.C. § 1345(c). The Secretary’s action to approve, require modification of, or disapprove a DPP is subject to judicial review only in the United States Court of Appeals for the circuit in which the affected state is located. 43 U.S.C. § 1349(c)(2).
Edwardsen contends that in approving the DPP, the Secretary, acting through the MMS, erred by relying upon an environmental impact statement (“EIS”) that did not comply with the requirements of the National Environmental Policy Act (“NEPA”), 42 U.S.C. § 4321 et seq. NEPA requires federal agencies proposing “major Federal actions” that may significantly affect the quality of the human environment to prepare a detailed EIS. See 42 U.S.C. § 4332(2)(C). NEPA does not expressly provide for judicial review. See Daniel R. Mandelker, NEPA Law and Litigation § 4.03(2) (2d ed.2000). Because the alleged NEPA violation arises under OCSLA, which provides for exclusive jurisdiction in the court of appeals, we have original jurisdiction over the NEPA claim. See id.; cf. Nat’l Parks and Conservation Ass’n v. FAA, 998 F.2d 1523, 1527-28 (10th Cir.1993) (exercising original jurisdiction over alleged NEPA violations under the statutory review provision of the FAA Act).
We review an EIS under a rule of reason to determine whether it contains a “reasonably thorough discussion of the significant aspects of probable environmental consequences.” Neighbors of Cuddy Mountain v. United States Forest [785]*785Serv., 137 F.3d 1372, 1376 (9th Cir.1998) (holding an EIS inadequate for failing to comply with the National Forest Management Act, failing to consider cumulative effects, and insufficiently discussing mitigating measures) (quotation marks and citation omitted). We must ensure that the agency took a “hard look” at environmental impacts. Okanogan Highlands Alliance v. Williams, 236 F.3d 468, 473 (9th Cir.2000). In our review, we must not substitute our judgment for that of the agency. Id.
1. Direct and Indirect Effects
NEPA requires an EIS to address:
(i) the environmental impact of the proposed action,
(ii) any adverse environmental effects which cannot be avoided should the proposal be implemented,
(iii) alternatives to the proposed action,
(iv) the relationship between local short-term uses of man’s environment and the maintenance and enhancement of long-term productivity, and
(v) any irreversible and irretrievable commitments of resources which would be involved in the proposed action should it be implemented.
42 U.S.C. § 4332(2)(C)(i)-(v) (2000). “Impacts” and “effects” include direct and indirect effects. See 40 C.F.R. § 1508.8 (2000). Direct effects are those that the proposed action causes and that occur at the same time and place as the action. 40 C.F.R. § 1508.8(a) (2000). Indirect effects are “caused by the action 'and are later in time or farther removed in distance, but are still reasonably foreseeable.” 40 C.F.R. § 1508.8(b) (2000).
Edwardsen argues that the EIS contains an inadequate analysis of direct and indirect effects because it does not include a site-specific oil spill trajectory analysis. The OCSLA regulations the MMS promulgated that relate to the EIS do not require a lessee to conduct a site-specific analysis of the trajectory of spilled oil in determining the environmental impacts. Edwardsen directs us to an MMS regulation that requires “[a]n appropriate trajectory analysis specific to the area in which the facility is located.” 30 C.F.R. § 254.26(b) (2000). That regulation, however, specifies the contents of the worst-case scenario required in a spill response plan, rather than an EIS. Moreover, an EIS need not include a worst-case scenario. See Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 354, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989). See also Mandelker, NEPA Law and Litigation § 10.07[3] at 10-39.
Edwardsen faults the MMS for failing to collect new data from the specific site of Northstar in determining the direct and indirect effects of an oil spill. The MMS used data from the oil spill risk analysis conducted for Lease Sale 170, a federal sale in the Beaufort Sea that occurred on August 5, 1998. Lease Sale 170 encompasses an area that includes a portion of Northstar. The MMS used data from Lease Sale 170 to calculate the probabilities that oil spilled from Northstar would reach land and “ice/sea segments.” The MMS concluded in the EIS that the data from Lease Sale 170 “presents a valid estimate of oil movement and resource areas likely to be contacted. Model input data (such as winds and currents) incorporated into the MMS model is consistent with information that would be used in a site-specific Northstar model.”
The MMS explained that the use of data from Lease Sale 170 in fact approximated a worst-case estimate of the environmental impact of an oil spill:
Seal Island lies close to the center of the two OCS leases.... Use of modeling [786]*786data for an oil spill originating in these lease areas provides a worst case estimate of the maximum areal extent of oil movement and resource areas likely to be contacted. An oil spill originating from the pipeline inside the barrier islands (near shore) would not be exposed to the stronger currents present near Seal Island and, therefore, would not spread as far away from the point of spill.
In using the data from Lease Sale 170, the MMS made a reasoned judgment that the data was relevant and yielded a useful analysis of the extent to which spilled oil would spread under the least favorable conditions.
Edwardsen points out that the U.S. Fish and Wildlife Service (“FWS”) objected to the absence of a site-specific analysis in the EIS. The fact that the FWS would have preferred a site-specific analysis is not sufficient, however, to require a conclusion that the MMS acted unreasonably or in contravention of NEPA by using the Lease Sale 170 data. We hold that the MMS took the required “hard look” in the EIS at the direct and indirect effects of Northstar. See Okanogan Highlands Alliance, 236 F.3d at 473.
2. Cumulative Impacts
The NEPA regulations do not explicitly require an EIS to include a discussion of cumulative impacts. See Mandelker, NEPA Law and Litigation § 10.12, at 10-80. The requirement arises from a regulation directing agencies to consider cumulative impacts in determining the scope of an EIS. See 40 C.F.R. § 1508.25(c)(3) (2000). The cumulative impact of a project is:
the impact on the environment which results from the incremental impact of the action when added to other past, present, and reasonably foreseeable future actions ... Cumulative impacts can result from individually minor but collectively significant actions taking place over a period of time.
40 C.F.R. § 1508.7 (2000).
Edwardsen argues that the EIS contains an inadequate analysis of cumulative impacts. He first contends that the EIS fails to comply with the methodology set forth in the Council on Environmental Quality’s “Guidance” on the matter of cumulative impacts. The handbook that Edwardsen cites does not, however, constitute legally binding agency guidance. Indeed, the preface to the handbook contains the following statement:
The handbook does not establish new requirements for [cumulative effects] analyses. It is not and should not be viewed as formal CEQ guidance on this matter, nor are the recommendations in the handbook intended to be legally binding.
Council on Environmental Quality, Considering Cumulative Effects under the National Environmental Policy Act iii (Jan. 1997), available at http://ceq.eh.doe.gov/nepa/nepanet.htm.
The proper question is therefore whether the EIS’ analysis of consideration of cumulative impacts complies with NEPA and the NEPA regulations. Edwardsen highlights five areas in which, he contends, the EIS is deficient in its analysis of cumulative impacts: freshwater, gravel, air quality, vegetation, and subsistence.
a. Freshwater
Millions of gallons of freshwater will be used to build ice roads for Northstar. Ed-wardsen argues that the EIS inadequately examines the cumulative impacts of this use of freshwater.
The EIS includes a section entitled, ‘Would freshwater withdrawals for ice road construction affect lake levels or wa[787]*787ter quality?” In this section, the EIS estimates the volume of freshwater required for ice road construction. Edward-sen makes three challenges to the EIS’ estimate of the volume of freshwater withdrawals. First, he argues that the EIS’ cumulative effects analysis does not take into account the total amount of freshwater required for road construction. A table contained in Chapter 5 of the EIS estimates that from 13 million to 14.9 million gallons of freshwater would be required for the construction of ice roads, and 5.9 million to 7.8 million gallons of freshwater would be required for ice road “operations” to connect West Dock to Seal Island. The table concludes that each action would have a “[m]inor” impact on lake water levels and water quality, since each quantity represents fifteen percent or less of the permitted usage from Kuparuk Dea-darm mine site, the likely source of freshwater. In Chapter 5, the EIS concludes that the impact to water levels of the withdrawal of 13 to 15 million gallons of freshwater would be minor. In Chapter 6, the EIS reiterates the estimated usage of 13 to 15 million gallons and concludes, “Based on the small amount of drawdown and screened intakes to prevent entrainment of fish, no impacts to freshwater fish are expected.” These statements do not, however, account for the 5.9 million to 7.8 million gallons of freshwater expected to be used for ice road operations.
Edwardsen correctly points out that the EIS’ assessment of the impact of freshwater use appears to have overlooked the 5.9 million to 7.8 million gallons required to connect West Dock to Seal Island. We are not persuaded, however, that such omission is material. The EIS’ conclusion that the impact to water levels will be minor remains supported by the record. The correct amount of total annual freshwater drawdowns, 18.9 million to 22.8 million gallons, is comfortably below the 100 million gallons that Alaska permits to be removed each year from the Kuparuk Dea-darm mine site, the most likely source of freshwater for ice road construction. In addition, the EIS also notes the availability of “several other permitted sources” in the project area. Any freshwater removed is replenished each year during spring breakup.
Edwardsen also faults the EIS for analyzing the projected Northstar drawdowns of freshwater in isolation, without considering drawdowns of freshwater for other ice roads. The relevant cumulative effects analysis of the EIS reads as follows:
Extraction of freshwater for use in the construction of ice roads to support onshore and offshore oil and gas activities would increase as new actions are developed. Water withdrawal from authorized water sources (e.g.lakes, rivers) occurs during the winter in accordance with permit restrictions on water volume. Because freshwater is replenished during the spring and summer months, the cumulative effect on lake water quality due to increased freshwater use for road construction would be negligible.
Although the EIS could have explained existing drawdowns of freshwater or the absence of any such drawdowns, there is no evidence in the record to suggest that there are any other drawdowns of freshwater from the likely sources of freshwater for Northstar.' In the absence of any such evidence, we cannot say that the failure to consider other drawdowns renders this EIS inadequate.
Edwardsen next contends that the EIS inadequately describes the effect of water withdrawals on birds and vegetation. In light of the EIS’ conclusion that the withdrawals would have a minimal effect on water levels, the absence of any further discussion is reasonable. In the EIS, the [788]*788MMS concludes that the impact of freshwater removal on water levels, salinity and alkalinity of the water, and water quality will be negligible. We conclude that the MMS’ analysis was reasonably thorough. See Cuddy Mountain, 137 F.3d at 1376.
b. Gravel
The reconstruction of Seal Island will require 700,000 to 800,000 cubic yards of gravel. Edwardsen contends that the EIS’ analysis of the impacts of this extraction of gravel is “weak” and that it is flawed by the absence of an evaluation of past or ongoing gravel extraction impacts.
Chapter 4 of the EIS contains a thorough analysis of the advantages and drawbacks of various onshore and offshore gravel sites. The EIS explains that mining existing manmade gravel islands or barrier islands could disrupt whale migration as well as the habitats of fish and migratory birds. The EIS notes that seven onshore gravel mine sites exist in North Slope oil fields. It proposes a new gravel mine site that would not interfere with whale migration and would be rehabilitated after being used for a single winter season.
The EIS compares the effects that extraction from each potential source of gravel would have on wildlife. Its cumulative impacts analysis includes the following discussion of the impact of gravel extraction on drainage patterns:
Gravel extraction, fill placement, and other soil disturbances associated with the construction of oil field facilities have the potential to affect surface runoff patterns and modify the soil’s thermal regime. This can result in minor changes to drainage patterns or permafrost, and may cause an expansion of the affected area beyond the original disturbance. The specific details of the foreseeable future actions have not been clearly defined, and the total amount of gravel fill and extraction cannot be determined. Advances in project design based on over 20 years of experience have resulted in the development of successful approaches to help minimize these impacts. The Northstar Project have [sic] been designed to minimize trenching and placement of gravel fill in onshore areas, and the location of the proposed gravel extraction site near the Kuparuk River mouth is expected to prevent the alteration of local drainage patterns.
The EIS contains a sufficient discussion of the environmental impact of gravel extraction.
c. Air Quality
Edwardsen argues that the EIS’ analysis of the impacts of Northstar on air quality is inadequate for three reasons. First, he contends that it fails to examine adequately the contribution North Slope industrial sources may make to Arctic haze, a phenomenon that reduces visibility. Such an analysis is impossible, Edwardsen contends, because the EIS omits baseline data showing air quality levels before the late 1980s and early 1990s.
The EIS notes that Arctic haze was first reported in the 1950s, decades before industrial development began in the region. It cites research attributing Arctic haze to the long-range transport of pollution from industrialized Europe. The EIS contains the following discussion of the potential impact of Northstar on Arctic haze:
Whether ... emissions from Northstar (or combined with reasonably foreseeable future projects) would contribute to arctic haze is not known. Arctic haze is a circumpolar problem with many sources, and Northstar’s contribution would be an incrementally very small addition.
[789]*789It might be interesting to know whether Arctic industrial development has exacerbated Arctic haze, but we cannot say that the omission of such an analysis renders the EIS inadequate. In view of the EIS’ reasoned conclusion that Arctic haze was caused by European pollution, the EIS’ failure to analyze the impact of Arctic development on Arctic haze was reasonable.
Second, Edwardsen maintains that the fact that the North Slope area had complied with regulatory air quality standards has no bearing on potential long-term cumulative impacts. The EIS contains an extensive analysis of the volume of carbon monoxide, nitrogen dioxide, sulfur dioxide, volatile organic compounds, and particulate matter less than 10 microns in diameter that will be emitted during construction, drilling, and operations. The EIS compares pollutant concentrations to the National Ambient Air Quality Standards (“NAAQS”), and rests its conclusion that Northstar will have a minimal effect on air quality on the fact that the area will remain in compliance with the NAAQS after Northstar begins operations. Edwardsen correctly notes that the fact that the area will remain in compliance with the NAAQS is not particularly meaningful, because the ambient air quality in the area presently exceeds NAAQS standards. A more relevant measure would be the degree to which Northstar contributes to the degradation of air quality. EPA regulations regarding Prevention of Significant Deterioration (“PSD”) address this problem, however. The EIS notes the commencement of Northstar construction and drilling operations will trigger a PSD review for nitrogen dioxide, carbon monoxide, ozone, sulfur dioxide, and particulate matter less than 10 microns in diameter. It was not unreasonable for the MMS at this stage to rely upon compliance with the NAAQS.
Finally, Edwardsen faults the EIS’ failure to discuss new PM2.5 or ozone standards. Guidance and rules for phasing in the standards, however, were not promulgated until February 1998. Moreover, the standards were never phased in, since they were challenged in the litigation that culminated in Whitman v. American Trucking Associations, Inc., 531 U.S. 457, 121 S.Ct. 903, 149 L.Ed.2d 1 (2001).
The EIS’ analysis of the effects of the project on air quality supports its conclusion that both short-term and long-term impacts to air quality from the project are “negligible to minor.” The analysis is reasonable.
d. Vegetation
Northstar would consume less than two acres of tundra for the placement of vertical support members and the construction of a gravel pad. Most of this tundra, as well as the tundra disturbed by prior North Slope oil and gas activities, is classified as wetlands under Section 404 of the Clean Water Act, 33 U.S.C. § 1344. Ed-wardsen contends that the EIS’ discussion of Northstar’s impact to wetlands is inadequate because the EIS does not quantify the types of wetlands destroyed to date or discuss in sufficient detail the effect of the wetlands destruction on birds and caribou.
We conclude that the EIS adequately analyzes both the effect of existing and proposed development on wetlands habitat and the effect of wetlands destruction on birds and caribou. The EIS includes the following discussion:
The construction of existing oil field facilities in the Prudhoe Bay-Kuparuk area is estimated to have directly affected over 58 square miles (150 km3) of prime waterfowl wetland habitat, including the destruction of over 14 square miles (36.3 km2) of this habitat. Cumulative habitat losses could affect the [790]*790nesting distribution or density of some species for more than one generation.
The EIS examines the impact of each stage of Northstar on coastal vegetation and animal life. It notes that the construction of ice roads and the onshore pipeline “could affect nesting bird habitat.” Because Northstar’s consumption of two acres of tundra is significantly less than the fourteen square miles of tundra that has already been lost, the EIS reasonably concludes that “the cumulative amount of tundra loss as a result of Northstar, although measurable, would be small when compared to previously disturbed acreage.” The EIS adequately addresses the effect of pipeline construction on the movement of caribou by noting that pipelines will be elevated to permit the passage of caribou and that the construction of permanent roads along pipelines will be minimized. These mitigation measures make reasonable the EIS’ conclusion that the cumulative effects of the pipelines on caribou would be minor.
e. Subsistence
The hunting of caribou and bowhead whales is an integral part of the Inupiat subsistence lifestyle. Edwardsen claims that the EIS fails to adequately discuss the impact of Northstar on this lifestyle. In particular, Edwardsen finds the discussion of Northstar’s impact on onshore subsistence activities to be lacking.
We find the EIS’ analysis of Northstar’s impact on Inupiat subsistence to be reasonable. The EIS examines the impact of each element of the project not only on bowhead whales, but also on caribou. It notes that “[l]ow-level helicopter traffic would cause a short-term disturbance to caribou during insect season as the animals move to the coast.” The EIS’ conclusion that Northstar would not reduce caribou harvests is reasonably supported by the planned mitigation measures, such as the elevation of pipelines.
B. Oil Pollution Act Claims
The Oil Pollution Act of 1990 (“OPA”) requires owners and operators of tank vessels and oil production facilities to submit oil spill response plans for approval by the MMS. See 33 U.S.C. § 1321(j)(5). In June 1999, the MMS approved the spill response plan submitted by BPXA. In a separate administrative action with a separate administrative record, the MMS approved the DPP pursuant to OCSLA in September 1999. That record is not before us.
Edwardsen contends that (1) the spill response plan is defective under OPA, (2) the DPP incorporated this defective spill response plan, and (3) the MMS’ approval of the DPP therefore violated OCSLA. We decline to review Edwardsen’s OPA claims because OCSLA regulations, the special review statute contained in OPA, and the overall regulatory regime created by OPA all make it clear that jurisdiction lies in the district court for actions challenging approval of a spill response plan or modifications to such a plan.
Under the OCSLA regulations, a DPP must be accompanied by “an updated oil-spill response plan as described in part 254 of this chapter or reference to an approved plan.” See 30 C.F.R. § 250.204(b)(3). At the time the DPP was being considered by the MMS, the spill response plan had already been approved by that agency. Therefore, BPXA’s DPP had to be accompanied by no more than a reference to the spill response plan that the MMS approved in June 1999. See id. In determining whether the MMS complied with its regulations in approving the DPP, we ask whether that requirement was satisfied. It was.
OPA itself has jurisdictional provisions governing controversies arising un[791]*791der that statute. OPA confers jurisdiction upon the district court over “any actions ... arising under this section.” See 33 U.S.C. § 1321(n). “This section” includes the provision requiring the owner or operator of a tank vessel or oil production facility to prepare a spill response plan. See 33 U.S.C. § 1321(j)(5). We generally treat a special review statute as a grant of exclusive original jurisdiction. See City of Rochester v. Bond, 603 F.2d 927, 935 (D.C.Cir.1979) (citations and quotation marks omitted) (“[E]ven where Congress has not expressly conferred exclusive jurisdiction, a special review statute vesting jurisdiction in a particular court cuts off other courts’ original jurisdiction in all cases covered by the general statute.”). See generally Note, Jurisdiction to Review Federal Administrative Action: District Court or Court of Appeals, 88 Harv. L.Rev. 980 (1975). Because the district court has exclusive original jurisdiction to review a spill response plan, we lack jurisdiction to do so here.
Under the regulatory regime created by OPA, the spill response plan is subject to change. OPA regulations require the owner or operator of an oil production facility to review the spill response plan at least every two years. Any modifications, or written notice of the absence of any modifications, must be submitted to the MMS. 30 C.F.R. § 254.30. Revisions to the spill plan prompted by significant changes in response capability, the worse case discharge scenario, or the type of oil must be submitted to the MMS for approval within fifteen days of the change. See 30 C.F.R. § 254.30(b). Thus the approval of a spill response plan, and subsequent modifications of it are more transitory administrative actions better suited for the district court review Congress provided. We therefore decline to review Edwardsen’s challenge to the spill response plan under OPA.
CONCLUSION
We have jurisdiction under NEPA to consider the challenges to the EIS, but our review is limited. When we review an EIS, we do not ask whether we would have done things differently had we been in the agency’s shoes. We ask only whether, applying a rule of reason, the agency took a hard look at the environmental impacts of a proposed project. We conclude that the EIS at issue here reasonably documented the environmental effects of Northstar, and that the MMS therefore took the requisite hard look. Accordingly, we deny the petition to review Secretary’s approval of the Northstar DPP. We do not have jurisdiction under OPA to review the earlier approval of the spill response plan that accompanied the DPP. We therefore dismiss the petition to review approval of that plan.
PETITION DENIED IN PART AND DISMISSED IN PART.