OPINION

Lake Roosevelt in eastern Washington state serves a variety of purposes, including irrigation, navigation, flood control, power generation, recreation, and fish management. We consider here a challenge by environmental groups to a proposed incremental drawdown of water from the lake. As always, our review under the National Environmental Policy Act (“NEPA”), 42 U.S.C. § 4321 et seq., is limited to determining whether the agency, in this case the United States Bureau of Reclamation (“Reclamation”), took a “hard look” and genuinely scrutinized the environmental consequences of its proposed action. Our own close look at the record persuades us that Reclamation was keenly aware of, and appropriately discharged, this duty when it prepared the drawdown project analysis.

Background

The Columbia River forms the centerpiece of a basin that spans over 250,000 square miles of the United States and Canada. The river, which is home to numerous species of salmon and other fish, is controlled by a “complex and highly regulated system of ... dams and reservoirs.” These include the Grand Coulee Dam, which is located in eastern Washington, nearly 600 miles from the mouth of the Columbia River and approximately 150 miles from the Canadian border. The portion of the Columbia River between the Grand Coulee Dam and the border is known as Lake Roosevelt.

Lake Roosevelt is operated by Reclamation and other federal agencies in cooperation with state agencies and in accord with a hodgepodge of treaties, statutes, and contracts. The lake typically holds about 5 million acre-feet of water. Water levels in the lake are, however, routinely lowered twice a year — once in early spring for flood control, and again in summer to increase downstream flow in the Columbia River. Water levels also fluctuate on a daily basis as a result of power-generating operations at the Grand Coulee Dam. In addition, the government diverts 2.65 million acre-feet of water from Lake Roosevelt every year to irrigate farmland in Washington state.

Faced with a variety of water needs in the Columbia River Basin, Reclamation, the State of Washington Department of Ecology (“Ecology”), and other agencies agreed to pursue various strategies to increase water supplies. Those undertakings are set out in a 2004 Memorandum of Understanding (“MOU”). Significant to this appeal, the parties to the MOU decided to divert a total of 82,500 acre-feet of water per year from Lake Roosevelt for three purposes — municipal and industrial use, groundwater replacement in the so-called Odessa Subarea in eastern Washington, and increased downstream flow to benefit fish populations. The parties also agreed to pursue an additional diversion of 50,000 acre-feet per year in drought years to alleviate water shortages. These agreements were preliminary: by its terms, the MOU did not “create a legally binding contract,” and it expressly “acknowledge^] that Reclamation’s actions are subject to federal reclamation law[s],” including NEPA.

In 2006, the Washington state legislature passed the Columbia River Water Management Act (“Water Management Act”), Wash. Rev.Code § 90.90.005 et seq., “to aggressively pursue the development of water supplies to benefit both instream and out-of-stream uses.” The Water Management Act, like the MOU, provided for “new releases of Lake Roosevelt water of approximately eighty-two thousand five hundred acre feet of water, increasing to no more than one hundred thirty-two thousand five hundred acre feet of water in drought years.” Id. § 90.90.060.

Consistent with the Washington State Environmental Policy Act, Ecology prepared a Preliminary Environmental Impact Statement (“PEIS”) that evaluated the impact of “the major components of the [Act’s Water] Management Program.” After describing the various projects in the Management Program and the direct environmental effects of those projects, the PEIS briefly discusses the cumulative impacts of the various parts of the program.

One year after completing the PEIS, Ecology issued a Supplemental Environmental Impact Statement (“SEIS”) specifically addressing the drawdown from Lake Roosevelt envisioned by the MOU and the Water Management Act. The SEIS considers a number of methods of releasing the water as well as alternatives to a draw-down and discusses at some length the impacts of the proposed drawdown on various aspects of the environment. The SEIS also includes a brief analysis of the cumulative impacts of the drawdown.

Shortly before Ecology released its final SEIS, Reclamation applied to Ecology for two secondary use water permits that would allow it to withdraw water from Lake Roosevelt. Ecology issued the permits on December 1, 2008. Taken in combination, the permits allow Reclamation to withdraw 82,500 acre-feet of water per year from Lake Roosevelt for municipal, industrial, agricultural, and instream use.

Soon after receiving the permits, Reclamation published a draft Environmental Assessment (“EA”) for the drawdown project; the final EA issued in June 2009. The final EA analyzes the project as envisioned in the MOU, the Water Management Act, and Ecology’s PEIS and SEIS. It also briefly discusses a no-action alternative. Like the SEIS, the EA describes the potential impacts of the project in detail. Also like the SEIS, the EA includes a short section expressly evaluating the project’s cumulative impacts, including three future projects involving the Columbia River. Along with the EA, Reclamation issued a Finding of No Significant Impact (“FONSI”) memorializing its finding that “implementation of [the drawdown project] and associated environmental commitments would have no significant impact on the quality of the human environment or the natural resources in the affected area.”

The Center for Environmental Law and Policy and other groups (collectively, “CELP”) originally filed this action as a challenge to the timeliness of Reclamation’s preparation of an EA. After Reclamation released its final EA, CELP amended its complaint to allege that the EA was not only untimely but also inadequate with respect to cumulative effects, indirect effects, and reasonable alternatives. Ecology and the East Columbia Basin Irrigation District intervened as defendants, and the parties filed cross-motions for summary judgment. The district court granted summary judgment to the defendants, holding that “the NEPA documents at issue” — including Ecology’s PEIS and SEIS as well as Reclamation’s EA— “thoroughly account for the history of development in the region and the project’s cumulative impacts thereto,” that the agencies’ “analysis of indirect impacts complies with NEPA,” that the EA’s discussion of alternatives was sufficient in light of the “long collaborative process between [various] stakeholders” that led to the drawdown project, and that because Reclamation “retained the discretionfin the EA] to move forward with the project or not,” its NEPA review was timely.

We review de novo the district court’s grant of summary judgment to Reclamation. See Lands Council v. McNair, 629 F.3d 1070, 1074 (9th Cir. 2010). However, because “judicial review of agency decisions under ... NEPA” is governed by Section 706 of the Administrative Procedure Act, we will uphold the agency’s action “unless it is ‘arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.’ ” Id. (quoting 5 U.S.C. § 706 (2)(A)). Our review is “limited to the question of whether the agency took a ‘hard look’ at the proposed action as required by a strict reading of NEPA’s procedural requirements.” Bering Strait Citizens for Responsible Res. Dev. v. U.S. Army Corps of Eng’rs, 524 F.3d 938, 947 (9th Cir.2008). We must defer to any agency decision that is “fully informed and well-considered,” but we will not overlook “a clear error of judgment.” Blue Mountains Biodiversity Project v. Blackwood, 161 F.3d 1208, 1211 (9th Cir. 1998) (internal quotation marks omitted).

CELP’s most significant challenge on appeal is to the cumulative effects analysis in the EA. Although we agree with CELP that the portion of the EA exclusively devoted to cumulative effects is conclusory and unenlightening, reading the EA as a whole reveals that Reclamation understood and accounted for the cumulative effects of past projects. The EA does not discuss the cumulative impact of one of the reasonably foreseeable projects known as the Odessa Subarea Special Study. Significantly, however, Reclamation has committed itself to scrutinizing the cumulative effects of the Special Study with the draw-down project before implementing any action resulting from the Special Study. Under our precedents and the circumstances presented here, this procedure does not violate NEPA. Our review reveals no other deficiencies in the substance of the EA, and although Reclamation took several steps toward implementing the drawdown project before drafting the EA, it scrupulously adhered to NEPA’s timing requirements. We therefore affirm the district court.

Analysis

Congress enacted NEPA in recognition of “the ‘profound impact’ of human activities, including ‘resource exploitation,’ on the environment.” Ctr. for Biological Diversity v. U.S. Dep’t of the Interior, 623 F.3d 633, 642 (9th Cir.2010) (quoting 42 U.S.C. § 4331 (a)). NEPA guarantees that environmental aspects of a policy are considered, see Bering Strait Citizens, 524 F.3d at 947 , and ensures that agencies do not make environmentally significant decisions in a vacuum, see Dep’t of Transp. v. Public Citizen, 541 U.S. 752, 768 , 124 S.Ct. 2204 , 159 L.Ed.2d 60 (2004).

The statute, however, is purely procedural: NEPA “does not contain substantive environmental standards.” Bering Strait Citizens, 524 F.3d at 947 . To implement its goals, NEPA “requires an agency to prepare an environmental impact statement (‘EIS’) for ‘major Federal actions significantly affecting the quality of the human environment.’ ” Ctr. for Food Safety v. Vilsack, 636 F.3d 1166, 1169 (9th Cir.2011) (quoting 42 U.S.C. § 4332 (2)(C)). “As a preliminary step, an agency may prepare an [EA] in order to determine whether a proposed action may significantly affect the environment and thereby trigger the requirement to prepare an EIS.” Ctr. for Biological Diversity v. NHTSA, 538 F.3d 1172, 1185 (9th Cir.2008) (citing 40 C.F.R. § 1508.9 (a)(1)) (internal quotation marks and alteration omitted).

Because an EA is intended to “serve practically as an important contribution to the decisionmaking process,” the EA for any “project[] directly undertaken by [a] Federal agenc[y]” must “be prepared at the feasibility analysis (go-no go) stage” of the project. 40 C.F.R. § 1502.5 . The EA must “include brief discussions of the need for the proposal, of alternatives [to the proposed action], [and] of the environmental impacts of the proposed action and alternatives.” Id. § 1508.9(b). As part of the analysis, the agency must consider “the direct, indirect, and cumulative impacts of the action.” Or. Natural Desert Ass’n v. Bureau of Land Mgmt., 625 F.3d 1092, 1100 (9th Cir.2010) (citing 40 C.F.R. §§ 1508.7 -.8). After analyzing the environmental effects and alternatives, “[i]f the agency concludes there is no significant effect associated with the proposed project, it may issue a FONSI in lieu of preparing an EIS.” Envtl. Prot. Info. Ctr. v. U.S. Forest Serv., 451 F.3d 1005, 1009 (9th Cir.2006).

I. Timing of the EA

To avoid post hoc agency rationalizations, “[p]roper timing is one of NEPA’s central themes.” Save the Yaak Comm. v. Block, 840 F.2d 714, 718 (9th Cir.1988). The agency must complete an EA before the “go-no go” stage of a project, Metcalf v. Daley, 214 F.3d 1135, 1142 (9th Cir. 2000) (internal quotation marks omitted), which is to say before “making an irreversible and irretrievable commitment of resources,” id. at 1143 .

Reclamation satisfied this requirement, because it undertook no irreversible commitments vis-a-vis the drawdown before the EA was completed. Although CELP argues that the agency irretrievably committed water from the lake to the drawdown project when it procured water use permits from Ecology in December 2008, simply obtaining permits was not tantamount to an irreversible commitment. Rather, Reclamation “retain[ed] absolute authority to decide whether” water from Lake Roosevelt would be committed to the draw-down project until after it published the final EA and FONSI. Friends of S.E.’s Future v. Morrison, 153 F.3d 1059, 1063 (9th Cir.1998) (internal quotation marks omitted).

Under Washington law, secondary use permits authorize Reclamation to use the allocated water without requiring it to do so. See Wash. Rev.Code § 90.03.250 (forbidding any person from “us[ing] or diverting] ... waters until he has received a permit from [Ecology]”). The permits allowed Reclamation to keep its options open; the agency remained free not to divert water from Lake Roosevelt if it determined that the drawdown project entailed significant environmental effects. The permits also did not give any other entity the unilateral right to insist on use of the water. See id. § 90.03.310 (noting that, subject to certain conditions, Reclamation possessed the option to assign its water rights). If Reclamation chose not to use the water, rights to the water would revert to the state, see id. § 90.14.130, and any entity wishing to use the water would then need to apply for a new permit.

In short, Reclamation “clearly retained the authority to change course or to alter the plan it was considering implementing” even after receiving the permits, WildWest Inst. v. Bull, 547 F.3d 1162, 1169 (9th Cir.2008), and did not “surrender [its] right to prevent” diversion of the water, Metcalf, 214 F.3d at 1144 . The permits accordingly did not irretrievably commit the agency to use water from Lake Roosevelt for the drawdown project.

II. Cumulative Effects

“NEPA requires an agency to consider” cumulative effects, which “result! ] from the incremental impact of the action when added to other past, present and reasonably foreseeable actions regardless of what agency ... or person undertakes such other actions.” NRDC v. U.S. Forest Serv., 421 F.3d 797, 814 (9th Cir. 2005) (internal quotation marks omitted); see also 40 C.F.R. § 1508.7 (defining cumulative impacts in this way). “Consideration of cumulative impacts requires some quantified or detailed information” that results in a “useful analysis,” even when the agency is preparing an EA and not an EIS. See Kern v. Bureau of Land Mgmt., 284 F.3d 1062 , 1075 (9th Cir.2002) (internal quotation marks omitted). “[G]eneral statements about possible effects and some risk do not constitute a hard look absent a justification regarding why more definitive information could not be provided.” Id. (internal quotation marks omitted). An agency may, however, characterize the cumulative effects of past actions in the aggregate without enumerating every past project that has affected an area. See, e.g., League of Wilderness Defenders Blue Mountains Biodiversity Project v. Allen, 615 F.3d 1122, 1135 (9th Cir.2010).

CELP characterizes the EA as deficient in two ways. It contends that the EA’s discussion of past projects is both conclusory and impenetrably filled with cross-references. It also asserts that the EA must, but does not, account for the cumulative impact of the Odessa Subarea Special Study. Because the factual underpinnings of these claims are distinct, we consider past and future effects separately. Although we acknowledge that CELP has pinpointed certain aspects of the EA that are less than ideal, these deficiencies do not rise to the level of a NEPA violation.

A. Cumulative Effects of Past Projects

In assessing Reclamation’s cumulative impacts analysis, some description is in order. As CELP rightly points out, only three paragraphs of the cumulative effects section deal with past projects, and none of those paragraphs rises above vague generalities.

The first of the three paragraphs simply parrots the potential categories of impacts listed in Ecology’s PEIS. Reclamation’s repetition of these possible effects recognizes that the PEIS concerned various “water projects in the Columbia River Basin,” but Reclamation neither discusses which effects are likely to result from the drawdown project nor provides any detail about those impacts. Instead, the EA rests on the truism that “[t]he cumulative impacts could cause ... more severe impacts than if a single project were constructed in a less disturbed environment.”

The second paragraph is no better; it merely restates the conclusions in Ecology’s SEIS that the drawdown “could incrementally increase the impacts described in the [PEIS]” but that “the incremental impact would not be expected to be significant.” Finally, the third paragraph— again like the SEIS — states that “[t]he increased stream flows [from the project] are expected to provide cumulative benefits to fish in most months,” without more. In short, these paragraphs include nothing more than repetition of general and vague information culled from other documents. Such a superficial analysis is a far cry from the requirements we set out in Kern and, were that the sum of the analysis, such statements would “not constitute a ‘hard look’ ” at the cumulative effects of the drawdown. Kern, 284 F.3d at 1075.

The perfunctory discussion in the “Cumulative Impacts” section of the EA is not, however, reflective of Reclamation’s overall approach. The analysis of various effects in other portions of the EA displays sensitivity to, and consideration of, the multitude of changes previously wrought by mankind on the Columbia River Basin. For example, much of the first section of the EA that addresses environmental consequences for the land is devoted to the risk of landslides in the Columbia River basin; the section also considers alluvial deposits and erosion. The discussion of landslides opens by noting that “[t]he Upper Columbia River has an extensive history of landslides, both prior to and after completion of the Grand Coulee Dam” and then goes on to recognize that landslides both upstream and downstream of the dam often occur when water levels are unusually low. Specifically, for the area around Lake Roosevelt, the EA identifies “[t]he concern for landslides ... as minor for lake levels above 1,260 feet, moderate for lake levels between 1,240 and 1,260 feet, and major for lake levels below 1,240 feet.” The EA then states that “[l]andslide potential would not change as a result of the proposed action,” in large part because “[t]he additional drawdown during the period” when the lake is less than 1,240 feet “would be minimal (less than 1 inch).”

Although contained in the portion of the EA devoted to direct effects, this discussion of the risks of landslides is indicative of the manner in which the EA satisfies NEPA’s requirement that Reclamation consider the cumulative effects of the drawdown project with past actions. It explains both “the existing condition of [the] area” and “what the effects of the project would be,” Ecology Ctr. v. Castaneda, 574 F.3d 652, 667 (9th Cir.2009), mentions past projects “necessary to describe the cumulative effect of all past actions combined,” Allen, 615 F.3d at 1135 , notably including construction of the Grand Coulee Dam, assesses the impact of the drawdown project against the risk of landslides resulting from the aggregate of all past projects, and relies on “quantified or detailed information” to support its conclusion that the project will not increase that risk, Kern, 284 F.3d at 1075 (internal quotation marks omitted).

Crucially, the EA’s detailed scrutiny is not limited to landslides. The section on surface water, for instance, describes in detail the human-driven fluctuations in Lake Roosevelt’s level, known information about the determinants of water quality in the lake, the timing and severity of downstream water quality concerns, the effect of existing dams on instream flow, and existing diversions of water in the Odessa Subarea. That section goes on to detail precise information about the effects of the additional drawdown. The EA also includes similarly detailed accounts of the cumulative effects of the drawdown with past projects on, among other things, groundwater, water rights, fish populations, aquatic plants, and prehistoric sites.

In other words, “[t]he record includes extensive evidence that [Reclamation] considered the relevant prior ... actions and took the requisite hard look before approving the” drawdown project. Ecology Ctr., 574 F.3d at 667 . Although this evidence is not presented in the cumulative effects section of the EA, it would impermissibly elevate form over substance to hold that Reclamation must replicate its entire analysis under the heading of cumulative effects. See id. (holding that an EIS’s discussion of cumulative effects was adequate because, even though “the cumulative effects section ... referred] generally to ‘past and proposed activities,’ ... other parts of the EIS g[ave] extensive history about past actions in the area”); see also Envtl. Prot. Info. Ctr., 451 F.3d at 1014 (approving a cumulative effects analysis based on a model that pervaded the EA). We therefore hold that the EA’s discussion of the cumulative effects of the drawdown project and past projects satisfies NEPA. In reaching this conclusion, we do not countenance an agency shirking its obligation to take a “hard look” and fully consider the cumulative effects of past projects. Nor do we suggest that agency discussion of impacts can be so diffuse, scattered, or opaque that a court must play Humpty Dumpty to put the pieces together in a coherent fashion.

B. Cumulative Effects of Future Projects

CELP also disputes the EA’s treatment of the cumulative effects of three future projects. As to two of those projects — one of which would channel already diverted water by different routes and one of which involves developing conservation programs — CELP has not carried its initial burden to show “the potential for cumulative impact.” Te-Moak Tribe of W. Shoshone of Nev. v. U.S. Dep’t of Interior, 608 F.3d 592, 605 (9th Cir.2010). In particular, CELP challenges neither the EA’s statement that the rechanneling project will not involve additional, and therefore cumulative, diversions of water from the Columbia River nor Reclamation’s assertion that the conservation project is too nascent to have reasonably foreseeable cumulative effects. CELP has accordingly not shown that the EA’s discussion of these projects fails to comport with NEPA.

CELP’s challenge to the EA’s treatment of the third future project, the Odessa Subarea Special Study, is considerably more consequential. The cumulative effects section of the EA mentions the Special Study, in which Reclamation is “investigating] the possibility of” delivering additional water from the Columbia River “to lands currently using groundwater in the Odessa Subarea.” The EA nonetheless concludes that “[n]one of the actions under study as part of the ... Special Study” is reasonably foreseeable.

This conclusion, which the district court accepted and which is pressed by both Reclamation and Ecology on appeal, cannot be harmonized with our precedents. We have “define[d] ... reasonably foreseeable actionfs], for which cumulative impacts must be analyzed, to include proposed actions.” N. Alaska Envtl. Ctr. v. Kempthorne, 457 F.3d 969, 980 (9th Cir. 2006) (internal quotation marks omitted). “Furthermore, we have previously held that an action is not too speculative” to qualify as a proposed action when the agency issues a notice of intent to prepare an EIS. Id.; accord City of Tenakee Springs v. Clough, 915 F.2d 1308 , 1313 (9th Cir.1990). Here, in August 2008, almost a year before Reclamation published its final EA, the agency issued a notice of intent “to prepare an [EIS] for the Odessa Subarea Special Study.” See 73 Fed. Reg. 49487 -02 (Aug. 21, 2008). The Special Study was therefore reasonably foreseeable at the time Reclamation prepared the EA for the drawdown project.

It does not, however, follow that the EA violated NEPA. In Northern Alaska Environmental Center, we held that an agency did not violate NEPA even though it failed to consider the cumulative effects of an action for which it had previously issued a notice of intent. We reasoned that “[t]he issue is one of timing.” 457 F.3d at 980 . Because the notice of intent had “in effect given notice that [the agency would] consider all impacts,” including cumulative impacts, as part of the future EIS, and because the courts could hold the agency to that implied promise, it was sufficient for the agency to “address [the impacts of the future project] at a later stage.” Id.; see also Salmon River Concerned Citizens v. Robertson, 32 F.3d 1346, 1357-58 (9th Cir.1994) (“Having persuaded the district court that it understands its duty to follow NEPA in reviewing future site-specific programs, judicial estoppel will preclude the [agency] from later arguing that it has no further duty to consider the cumulative impact of [those] programs.”). In other words, having acknowledged the significance of a future project through a notice of intent, the agency cannot escape its downstream obligation to consider the impact of the project.

We face the same situation as in Northern Alaska Environmental Center. By issuing the notice of intent to prepare an EIS for the Special Study, the government impliedly promised to consider the cumulative effects of the special study with the drawdown project in that EIS. In fact, Reclamation expressly made the same promise to this court, stating in its opening brief that “there is no danger that [actions taken as a result of the Special Study] would escape NEPA review.” The im port of the government’s promises is that Reclamation will prepare a “document [that] explores the collective impact of’ the drawdown project and the Special Study. See Blackwood, 161 F.3d at 1215 (rejecting a cumulative effects analysis because no such document was in the offing). As a result, CELP cannot argue that Reclamation is attempting to initiate multiple draw-downs from Lake Roosevelt “without ever having to evaluate the ... cumulative environmental impacts” of those actions. Native Ecosys. Council v. Dombeck, 304 F.3d 886 , 897 (9th Cir.2002).

We also note that the drawdown project discussed in the EA involves much smaller diversions of water than the Special Study, which could result in a drawdown of well over 300,000 acre-feet per year. To direct that Reclamation must account for the cumulative effects of the drawdown project and the Special Study in the drawdown project EA would thus be to direct the agency to wag the dog by its tail.

For these reasons, and in accordance with Northern Alaska Environmental Center, we conclude that Reclamation’s failure to consider the cumulative effects of the Special Study and the drawdown project in the drawdown project EA does not violate NEPA.

III. Indirect Effects

Agencies conducting NEPA review must also consider the indirect effects of the proposed project. Indirect effects are those effects “caused by the [agency] action [that] are later in time or farther removed in distance, but are still reasonably foreseeable.” 40 C.F.R. § 1508.8 (b). Such effects “include growth inducing effects and other effects related to induced changes in the pattern of land use, population density or growth rate, and related effects on air and water and other natural systems, including ecosystems.” Id.; see also, e.g., Ocean Advocates v. U.S. Army Corps of Eng’rs, 402 F.3d 846 , 867-70 (9th Cir.2005) (holding that an agency violated NEPA by failing to account for the environmental effects of the additional tanker traffic that would be caused by a proposed dock expansion).

CELP’s indirect effects challenge is rooted in the expansion of the Weber Siphons, which form two portions of a canal system near Interstate 90 in eastern Washington. Specifically, CELP relies on the fact that the expansion will increase the siphons’ capacity by 1,950 cubic feet per second (“cfs”), even though the draw-down project itself will only use 181 cfs of that capacity. It is undisputed both that the expansion will remove one barrier to further diversions of water from Lake Roosevelt and that the EA does not account for the effects of diverting the additional 1,769 cfs that the expanded siphons could handle.

The effects of full usage of the expanded siphons are not indirect effects of the drawdown project. Agencies need not account for potential growth effects that might be caused by a project if the project is exclusively intended to serve a much more limited need. Seattle Cmty. Council Fed’n v. FAA, 961 F.2d 829, 835-36 (9th Cir.1992). That rule applies with force here. To be sure, the siphon expansion will make it easier to divert up to 1,950 cfs from Lake Roosevelt, but it is uncontested that the current project — as approved — will only use a fraction of that capacity.

In fact, the causal tie between the project and growth is significantly more attenuated than it was in Seattle Community Council. There, “air traffic ... [was] expected to increase” as the result of private actions once the agency implemented its proposal. See 961 F.2d at 835 . Here, by contrast, expansion of the siphons is not sufficient to allow additional diversions from the Columbia River. Even with the expansion, those diversions cannot occur unless at least three other events take place — a decision by Reclamation to utilize the expanded capacity, the expansion of other canals in the area, and NEPA review of the additional drawdowns. The use of the expanded capacity remains both firmly in the control of Reclamation and is subject to review in a future EA or EIS. Any environmental effects from additional diversions that flow through the Weber Siphons are not indirect effects of the current drawdown project.

IV. Analysis of Alternatives

“NEPA requires federal agencies to ‘study, develop, and describe appropriate alternatives to recommended courses of action in any proposal which involves unresolved conflicts concerning alternative uses of available resources.’ ” Native Ecosys. Council, 428 F.3d at 1245 (quoting 42 U.S.C. § 4332 (E)). Although an EA need only include a “brief discussion” of alternatives, that discussion must follow “full and meaningful consideration” of the alternatives by the agency. Id. (internal quotation marks omitted). Nevertheless, “an agency’s obligation to consider alternatives under an EA [remains] a lesser one than under an EIS.” Id. at 1246 .

Reclamation’s EA discusses two alternatives — the proposed drawdown and a no action alternative. CELP challenges this selection of alternatives as overly constrained. Insofar as CELP simply argues that two alternatives is too few, its challenge must be rejected. As we have previously recognized, there is no “numerical floor on alternatives to be considered.” Id.

CELP’s argument also rests on the more precise premise that Reclamation did not consider the alternatives CELP prefers, such as conservation and water rights marketing. Albeit colorable, this argument is not persuasive. In the FONSI, Reclamation twice states that it “developed the proposed action alternative with reference to a wider range of alternatives considered in the Department of Ecology’s SEIS.” The SEIS, meanwhile, considered and rejected five other alternatives, including an all-conservation approach and water marketing. Ecology decided not to proceed with conservation alone because most potential conservation projects would have benefits that were too limited in time and geographic area. It rejected water marketing on the basis of “concerns about the impact to local economies from the transfer of the needed volumes of water.”

CELP does not argue that these explanations, if contained in the EA, would violate NEPA. But under the circumstances presented here, the fact that the discussions instead appear in Ecology’s SEIS is of no moment. We have previously held that “the absence of a more thorough discussion in [an] EIS of alternatives that were discussed in and rejected as a result of prior state studies does not violate NEPA.” Laguna Greenbelt, Lie. v. U.S. Dep’t of Transp., 42 F.3d 517 , 524 n. 6 (9th Cir.1994). Laguna Greenbelt, like this appeal, involved alternatives rejected by pri- or state environmental impact reviews of the same joint project that was subject to later federal review. See id. at 522. Rec lamation thus discharged its duty to consider the discarded alternatives by explaining in the FONSI that it rejected them because of Ecology’s SEIS. See id.; cf. N. Idaho Cmty. Action Network v. U.S. Dep’t of Transp., 545 F.3d 1147, 1153-54 (9th Cir.2008) (approving an EA that discussed only two alternatives because a prior EIS had evaluated the project’s environmental effects in depth). We accordingly reject CELP’s challenge to the analysis of alternatives in the EA. See Native Ecosys. Council, 428 F.3d at 1249 (holding that an EA that discussed only the agency’s preferred alternative and a no-action alternative satisfied NEPA).

AFFIRMED.

. CELP further contends that, by preparing the EA only after signing the MOU and securing permits, Reclamation performed an irretrievably biased analysis consistent with its already-established preferences. The potential for agency bias does not impose a timing constraint separate from the irretrievable commitment rule. See, e.g., Pit River Tribe v. U.S. Forest Serv., 469 F.3d 768 , 782 (9th Cir. 2006) ("[A]n EIS is not required in cases where the government has not irretrievably committed resources.”); Metcalf, 214 F.3d at 1144 (discussing the probability that an EA was biased only after concluding that it was prepared after the go-no go stage of the project). In any event, we are not persuaded that any of Reclamation’s pre-EA actions left it unable to "objectively evaluate[]” the draw-down project. Metcalf, 214 F.3d at 1142 . . CELP suggests that the permits gave Ecology newfound veto power over Reclamation’s options, because once the permits issued, Reclamation could no longer commit a smaller amount of water to the drawdown project without Ecology's approval. This argument rests on a false dichotomy. CELP assumes that Reclamation previously possessed unilateral power to divert a lesser amount of water. With very narrow exceptions, however, Ecology must approve any diversion of any magnitude. See, e.g., Wash. Rev.Code §§ 90.03.010, -.250, -.252. Thus, any lesser diversion was subject to Ecology’s approval even before Reclamation applied for permits in a specific amount. . Reclamation’s other pre-EA actions also did not involve the irretrievable commitment of resources. The MOU signed by Reclamation, Ecology, and others is contingent on NEPA compliance. So, too, is the drawdown project-related funding Reclamation received in the American Recovery and Reinvestment Act. See Pub. L. No. 111-5, § 1609 , 123 Stat. 304 (2009). . As a result, we do not consider either Reclamation's assertion that the EA's discussion of cumulative effects may properly incorporate sections of the PEIS and SEIS or CELP's argument that this attempted incorporation ”send[s] the reader” on a fruitless journey through "a series of cross-references.” . We grant CELP’s uncontested motion to take judicial notice of the draft EIS for the Special Study, which was published during the pendency of this appeal. See, e.g., Or. Nat'l Desert Ass'n, 625 F.3d at 1112 n. 14 (taking judicial notice of a Bureau of Land Management Handbook as a public document); City of Las Vegas v. FAA, 570 F.3d 1109 , 1113 n. 1 (9th Cir.2009) (granting a motion to take judicial notice of events that occurred "after the close of ... administrative proceedings"). Although we may not and do not decide whether the draft EIS complies with NEPA, we note that it appears to follow through on the government's promise by incorporating the drawdown project into its baseline view of the existing situation in the Columbia River basin. See, e.g., Odessa Subarea Special Study Draft Environmental Impact Statement 1-18, 2-2 (October 2010), available at: http://www.usbr.gov/pn/ programs/eis/odessa/drafteis/draft-eis-odessa. pdf (last accessed June 20, 2011). . We also note that the choice to expand the siphons was not made on a blank slate. The siphons were originally designed to include "two pipes next to each other,” but "only one pipe was installed as part of the construction in the early 1950s.” As part of the drawdown project, Reclamation is planning to expand the siphons according to their original design. . The EIS at issue in Laguna Greenbelt was itself jointly prepared by state and federal agencies. See 42 F.3d at 522.