¶1 — This case concerns the adequacy of State Environmental Policy Act (SEPA)1 phased review and mitigation for the proposed expansion of a sockeye salmon hatchery on the Cedar River. Appellant Roz Glasser contends that the Seattle Hearing Examiner (Hearing Examiner) wrongly prohibited her from presenting evidence or argument challenging the continued validity of the analyses in the programmatic SEPA document underlying the city of Seattle’s (City) decision to proceed with an expanded hatchery. We affirm.

I

¶2 Since the 1900s, the City has operated the Landsburg Project on the Cedar River in King County to provide a *733municipal water supply to the greater Seattle area. The Cedar River watershed also serves as habitat for a number of species, including several that are currently listed under the Endangered Species Act (ESA).2 Until recently, the Landsburg Project’s diversion dam prevented anadromous fish from migrating to spawning habitat in the upper watershed.

¶3 In the 1930s, a run of sockeye salmon was successfully introduced into the Cedar River watershed. After building to robust levels in the 1960s, the run experienced significant declines. The Washington State Legislature responded in 1989 by enacting Senate Bill 51563 to provide for construction of a spawning channel designed to mitigate Landsburg Project’s impacts on sockeye in the Cedar River. The legislature specified that “[t]he channel shall be designed to produce, at a minimum, fry comparable in quality to those produced in the Cedar river and equal in number to what could be produced naturally by the estimated two hundred sixty-two thousand adults that could have spawned upstream of the Landsburg diversion.”4

¶4 In 1991, the State constructed an “interim” hatchery capable of producing up to 17 million fry annually. However, in 1993 the Cedar River Sockeye Policy Committee decided to postpone construction of the spawning channel and continue operating the interim hatchery as an emergency measure to reverse the precipitous decline of the sockeye fishery. The interim hatchery has been operated by the Washington Department of Fish and Wildlife ever since.

¶5 Eventually the City entered into a lengthy planning process with state and federal agencies regarding management of the Cedar River watershed. The end result was the Cedar River Watershed Habitat Conservation Plan (HCP), a 50-year comprehensive plan that addresses watershed management, instream flows, fish passage at Landsburg, *734and incidental take permits for ESA-listed species, including Chinook salmon and bull trout. In 1999, a joint NEPA5/ SEPA programmatic document was prepared to broadly analyze all aspects of the proposed HCP, including fishery resources. The HCP environmental impact statement (EIS) evaluated a number of sockeye mitigation alternatives, including one for the proposed expanded hatchery and another emphasizing habitat protection and restoration.

¶6 In 2000, the parties signed the Cedar River HCP in conjunction with an implementation agreement, the Lands-burg mitigation agreement, and the instream flow agreement. These agreements provided for construction of a fish ladder to allow all salmon species except sockeye to spawn above the dam because of concerns that the large number of returning sockeye would pose a risk to the City’s drinking water supply. Sockeye mitigation would instead be accomplished by construction of a new, expanded hatchery with a production capacity of up to 34 million fry per year, which was assumed to be the number of fry needed to achieve the equivalent of 262,000 spawners required by Senate Bill 5156 for mitigation purposes. The parties also agreed to implement an adaptive management plan designed to “address critical questions as they arise and make changes in management based on the results of monitoring.”

¶7 In 2003, the programmatic HCP EIS was followed by a supplemental EIS (SEIS) to evaluate specific project alternatives for constructing the expanded sockeye hatchery. Consistent with the goals and objectives referenced in the HCP and Landsburg mitigation agreement, the SEIS alternatives focused on various design and siting considerations for the expanded sockeye hatchery.

¶8 In May 2003, Glasser challenged the adequacy of the SEIS before the Hearing Examiner. The Hearing Examiner granted the City’s motion to limit the appeal to issues *735concerning the hatchery EIS, ruling that issues that appear to challenge the programmatic HCP EIS were dismissed. In November 2003, the Hearing Examiner again refused to consider any arguments that went to the adequacy of the programmatic HCP EIS but accepted Glasser’s claim that the hatchery EIS failed to include a worst-case analysis for significant impacts that have not been adequately addressed, and failed to adequately disclose the impacts of the proposed mitigation/adaptive management plan. The Hearing Examiner remanded to the City for preparation of a revised SEIS.

¶9 The City prepared a revised hatchery SEIS that focused on the specific deficiencies noted in the Hearing Examiner’s decision, and Glasser appealed. In September 2005, the Hearing Examiner reiterated that the scope of the appeal did not include a challenge to the revised hatchery EIS based on its failure to reanalyze programmatic alternatives in the HCP EIS, and in December 2005 the Hearing Examiner ruled that the revised hatchery SEIS was “adequate by all applicable standards.” Glasser obtained a writ of review in superior court, which ruled in July 2006 that Glasser failed to prove that the Hearing Examiner’s decisions were erroneous. Glasser now appeals that decision.

¶10 In addition, Glasser recently filed suit in federal court directly challenging the 1999 programmatic HCP EIS.6 On March 1, 2007, the City asked the National Oceanic and Atmospheric Administration’s Fisheries National Marine Service and the United States Fish and Wildlife Service to remove Cedar River sockeye from the list of species included in the Cedar River HCP and to remove the proposed Cedar River sockeye hatchery from the mitigation agreement in order to protect the overall HCP from litigation that has been focused on the hatcheries.

*736II

¶11 Glasser argues that the Hearing Examiner erred in refusing to accept any evidence or argument concerning the continuing validity of analysis and assumptions in the 1999 programmatic HCP EIS that underlay the subsequent hatchery EIS. Glasser argues that the programmatic alternatives for sockeye mitigation were developed based on an erroneous analysis of the City’s legal mitigation obligations under Senate Bill 5156 and erroneous assumptions about the number of fry needed to meet those requirements. These alleged errors led to the rejection of nonhatchery alternatives in the HCP EIS and to the adoption of the expanded hatchery alternative. Therefore, the programmatic HCP EIS is not competent to support the project-level hatchery EIS. Glasser insists that she is not asking the City to redo the original HCP EIS or to consider new programmatic alternatives; rather, she wants this court to remand the revised hatchery EIS to the City for reevaluation of the alleged errors in the HCP EIS.

¶12 The Hearing Examiner never actually evaluated the adequacy of the EIS on this issue. Rather, she dismissed this portion of Glasser’s arguments as a matter of law based on her interpretation of SEPA’s phased review regulations and parallel provisions in the Seattle Municipal Code (SMC). Accordingly, the issue concerns the interpretation of statutes and regulations, which is a question of law reviewed de novo.7

¶13 The principal purpose of SEPA is to provide decision makers and the public with information about potential adverse impacts of a proposed action.8 SEPA provides that environmental review may occur in phases. “ ‘Phased review’ means the coverage of general matters in broader environmental documents, with subsequent nar*737rower documents concentrating solely on the issues specific to the later analysis.”9 This process “assists agencies and the public to focus on issues that are ready for decision and exclude from consideration issues already decided or not yet ready. Broader environmental documents may be followed by narrower documents, for example, that incorporate prior general discussion by reference and concentrate solely on the issues specific to that phase of the proposal.”10 “Phased review is appropriate when: (i) The sequence is from a nonproject document to a document of narrower scope such as a site specific analysis . . . ; or (ii) The sequence is from an environmental document on a specific proposal at an early stage (such as need and site selection) to a subsequent environmental document at a later stage (such as sensitive design impacts).” 11 “When a project is then proposed that is consistent with the approved nonproject action, the EIS on such a project shall focus on the impacts and alternatives including mitigation measures specific to the subsequent project and not analyzed in the nonproject EIS. The scope shall be limited accordingly.”12

¶14 Recognizing that SEPA documents can become outdated over time, the regulations provide standards and processes for ensuring the continuing validity of the analysis. “When preparing a project EIS ... the lead agency shall review the nonproject EIS to ensure that the analysis is valid when applied to the current proposal, knowledge, and technology. If it is not valid, the analysis shall be reanalyzed in the project EIS.”13 Previously prepared environmental documents may be used “provided that the information in the existing document(s) is accurate and reasonably *738up-to-date.”14 A supplemental EIS is required if there are substantial changes to a proposal so that the proposal is likely to have significant adverse environmental impacts, or if there is new information indicating a proposal’s probable significant adverse environmental impacts, including discovery of misrepresentation or lack of material disclosure. A new SEIS is not required if probable significant adverse environmental impacts are covered by the range of alternatives and impacts analyzed in the existing environmental documents.15

115 Glasser argues that the City failed to review the assumptions and analyses in the HCP EIS to ensure their validity as applied to the hatchery proposal. She implies that this review, if performed, would have shown that the alternatives analysis in the 1999 programmatic HCP EIS was not valid, which would call into question the hatchery EIS’s analysis of the expanded hatchery as a foregone conclusion. Glasser further argues that a new SEIS is required because probable significant adverse environmental impacts were not covered by the range of alternatives in the hatchery EIS.

¶16 We disagree. The SEPA phased review regulations are designed to streamline environmental review as a proposal progresses from broad planning to narrow, site-specific implementation. The scope of the project EIS is to be limited accordingly.16 The requirement to review a nonproject EIS to “ensure that the analysis is valid” refers to the validity of the environmental impacts analysis in the programmatic EIS, not the purpose and need or the range of alternatives. Allowing opponents to use a project EIS to collaterally attack previous programmatic policy decisions would disrupt the finality of the decision and eliminate any benefits of phased review. Parties who are dissatisfied with *739a programmatic EIS may challenge it directly — which is precisely what Glasser is doing in federal court.

¶17 Here, the initial programmatic SEPA analysis was conducted in conjunction with the Cedar River HCP. The alternatives in the HCP EIS presented a broad range of policy choices for mitigating anadromous fish impacts, including a no-hatchery/habitat restoration alternative favored by Glasser. However, the agencies did not choose this alternative; instead, they selected the expanded hatchery alternative. Following this programmatic decision, the City then prepared a project-level hatchery EIS that focused its alternatives analysis on a narrower set of considerations relating to the siting and construction of the hatchery. The Hearing Examiner ruled that the City’s revised hatchery EIS did consider whether assumptions and analyses in the 1999 HCP EIS were still accurate, at least with respect to issues within the scope of that document (worst-case analysis). This process properly followed the letter and spirit of the SEPA regulations.

¶18 Glasser also argues that the hatchery EIS is inadequate because the adaptive management plan (AMP) proposed as one of the mitigation measures fails to provide reasonable assurances that adaptive management will actually occur, and that the Hearing Examiner erred in applying the “clearly erroneous” standard of review to this issue. We agree that the Hearing Examiner erred in applying the “clearly erroneous” standard to Glasser’s EIS adequacy challenge. EIS adequacy refers to the legal sufficiency of the environmental data contained in the document.17 The adequacy of an EIS is a question of law, reviewed de novo.18 The appellate court reviews the agency’s decision regarding EIS adequacy, not the decision of the *740Hearing Examiner or the trial court.19 Although review is de novo, the court gives “substantial weight” to the agency’s determination that the EIS is adequate under SEPA.20 The adequacy of an EIS is tested under the “rule of reason.”21 Under the rule of reason, the EIS must present decision makers with a “ ‘reasonably thorough discussion of the significant aspects of the probable environmental consequences’ ” of the agency’s decision.22 In contrast, threshold determinations and decisions regarding whether a supplemental EIS is required involve the application of law to facts and are reviewed under the “clearly erroneous” standard set forth in RCW 34.05.570(3)(d).23 A decision is clearly erroneous if the court is “ ‘left with the definite and firm conviction’ that a mistake has been committed.”24 The City cites Preserve Our Islands v. Shoreline Hearings Board25 and Thornton Creek Legal Defense Fund v. City of Seattle26 in arguing for application of the clearly erroneous standard to Glasser’s claim.27 However, because those cases involved challenges both to the adequacy of the EIS and to the agency’s decision regarding whether a supplemental EIS was required, both standards of review came into play. Here, Glasser’s adaptive management claim goes only to the adequacy of the EIS. Therefore, we review the City’s *741decision regarding the adequacy of the EIS de novo, giving substantial weight to the agency’s decision and applying the rule of reason.

¶19 WAC 197-ll-440(6)(a) provides that an EIS must “discuss reasonable mitigation measures that would significantly mitigate” the significant impacts. The EIS shall “[c]learly indicate those mitigation measures . . . that could be implemented or might be required, as well as those, if any, that agencies or applicants are committed to implement.”28 The EIS must “[indicate what the intended environmental benefits of mitigation measures are for significant impacts”; however, it “need not analyze mitigation measures in detail unless they involve substantial changes to the proposal causing significant adverse impacts, or new information regarding significant impacts, and those measures will not be subsequently analyzed under SEPA.”29

¶20 Glasser’s argument is primarily substantive: that the AMP is unlikely to result in management changes that will actually mitigate for adverse impacts. Glasser claims that SEPA requires “reasonable assurances” that adaptive management will actually occur, citing Port of Seattle v. Pollution Control Hearings Boardl30 and Miller v. City of Port Angeles31 Neither case supports Glasser’s theory. Miller held that “a city may properly require an improvement outside of its territorial jurisdiction if it conditions that requirement on annexation or the consent of the government having jurisdiction,” observing that impacts should be mitigated if possible.32 And Port of Seattle concerned whether the adaptive management plan satisfied the “reasonable assurance” requirement of Clean Water Act *742§ 401 (33 U.S.C. § 1341).33 In contrast, SEPA is primarily a procedural statute that requires the disclosure of environmental information.34 “ ‘SEPA does not demand a particular substantive result in government decision making’; rather, it ensures that environmental values are given appropriate consideration.”35

¶21 Therefore, the issue is whether the AMP is described in sufficient detail to allow for a reasonable evaluation of the proposal’s impacts. Glasser claims that the AMP should have specified a statistical power analysis and that it lacks defined triggers for action. However, the EIS explained that the parameters will be developed once the specific monitoring studies are designed, and the AMP does identify numeric thresholds triggering review and responses that should be taken. Glasser also argues that the AMP is defective because it lacks a dispute resolution process, but the Hearing Examiner properly concluded that a more extensive process was not legally required under SEPA. The revised AMP meets SEPA requirements for display of impacts.

¶22 Affirmed.

Agid and Cox, JJ., concur.

Review denied at 163 Wn.2d 1033 (2008).