Save Our Peninsula Committee v. Monterey County Board of Supervisors

Opinion BAMATTRE-MANOUKIAN, J. In this CEQA 1 case, the project applicants, real parties in interest September Ranch Partners, appeal from a judgment granting two petitions for a writ of mandate. The superior court found that the project’s environmental impact report (EIR) was legally inadequate under CEQA and directed that the Monterey County Board of Supervisors (the Board) vacate certification of the EIR and prepare and circulate a legally adequate EIR with respect to specified water and traffic issues. Appellants argue that the Board’s certification of the EIR must be upheld because the Board’s determinations regarding the project’s water and traffic impacts were supported by substantial evidence. After reviewing the record, we conclude that the EIR in this case did not comply with CEQA in its treatment of several critical water issues. Because of these inadequacies, the Board’s action certifying the EIR and approving the project constituted an abuse of discretion. We further conclude, however, that the EIR was adequate in its discussion of traffic impacts and mitigation. We will therefore affirm in part and reverse in part the judgment in favor of petitioners and direct that the trial court issue a new writ of mandate in accordance with the views expressed herein. Background 2 The September Ranch property consists of 891 acres located along Carmel Valley Road approximately 3 miles east of the junction with Highway 1. Most of the property is hilly terrain with south-facing slopes. A level terrace adjacent to Carmel Valley Road of approximately 21 acres contains an *108 equestrian center, including a barn, outside stalls, a training ring, a residence for employees, and pastureland. A regional park and a small county-owned parcel lie to the west and northwest of the property and to the south is a golf resort and lodge. Otherwise the surrounding area is characterized by residential development. The zoning of the September Ranch property is for residential development. The property is governed by the Carmel Valley Master Plan (Master Plan), which is part of the county’s general plan. Under the Master Plan, this amount of acreage would allow for 208 homes. The September Ranch property is located within the Carmel River watershed. The property’s water needs have been served by well water since the early 1930’s. A new well was installed in 1990. Additional wells were installed in 1992 for purposes of data collection. A small aquifer, or “sub-basin,” underlies the 21-acre terrace on the property. It was originally thought by the owners to be a separate aquifer, isolated from the main Carmel Valley aquifer. However testing during the environmental review for this project determined that this sub-basin was not entirely separate and that there was some water exchange between it and the Carmel Valley aquifer. The Carmel Valley aquifer is a primary source of water for the Monterey Peninsula. It is well documented that water availability is a critical problem throughout Monterey County (the County) and in Carmel Valley in particular. In 1988, the County passed Ordinance No. 3310, finding that because of expanded water usage “the potential exists that Monterey County’s allocation of water will be exhausted so as to pose an immediate threat to the public health, safety, or welfare.” In 1995, the State Water Resources Control Board issued Order No. 95-10 and related Decision No. 1632. Order No. 95-10 found that the California-American Water Company (Cal-Am), which was the principal supplier of water to the Monterey Peninsula, had diverted excess water from the Carmel River basin “without a valid basis of right,” causing environmental harm. Cal-Am was ordered to substantially limit its diversions, to mitigate the environmental effects of its excess usage and to develop a plan for obtaining water legally. Decision No. 1632 similarly found that “[ejxisting diversions from the Carmel River have adversely affected the public trust resources in the river.” The Master Plan also recognized the serious water shortage in the Carmel Valley and set the standard for development until a solution was found. In Policy 54.1.7, the Master Plan found that without an additional water supply, such as from a proposed dam project, “development will be limited to vacant lots of record and already approved projects. All development which requires a water supply shall be subject to County adopted water allocation and/or ordinances applicable to lands in the Carmel Valley Master Plan area.” *109 The Morgens family has owned the September Ranch property since the 1960s. In 1995 James Morgens formed a partnership called September Ranch Partners for the purpose of developing the property. The partnership submitted its development application to the County in June of 1995. The proposal was for 100 single-family lots and 17 moderate income housing units. The application included a September Ranch Water Supply Plan, which called for Cal-Am to supply potable water. However, the month after the project application was submitted, the State Water Resources Control Board adopted Order No. 95-10, which cut back Cal-Am’s diversion of water from the Carmel River basin and essentially foreclosed its ability to provide water for new projects. The Draft EIR On August 4, 1995, the County issued its initial study for the September Ranch project, and the notice of preparation of the EIR was filed the same day. The draft EIR was published over two years later, on October 27, 1997. The draft EIR recognized existing policies regarding water resources in the Carmel River valley. It stated that potable water for the project was to be provided by a small mutual water system, independent of the Cal-Am water system, which would supply water pumped from wells on the September Ranch property. It noted that because there was potential groundwater flow between the September Ranch sub-basin and the adjacent Carmel Valley aquifer, “pumping in the September Ranch basin has the potential to affect water levels in areas of the Carmel Valley alluvium.” Furthermore, “any increase in the impacts to the [Carmel Valley] aquifer would be considered an adverse environmental impact given the water supply problems in the Carmel Valley area.” Any impact reducing flow to the Carmel Valley aquifer was “potentially significant.” As mitigation for this impact, the draft stated that water demand for the project must be limited to existing water use on the property. The draft EIR included a discussion of “Existing Water Demand” for the property. It stated that there was “limited historic data” to determine actual water usage over the years; however Monterey Peninsula Water Management District (MPWMD) records from 1991 to 1996 showed that water use on the property ranged from a low of 0.40 acre-feet in 1995 to a high of 40.68 acre-feet in 1993. There was no data prior to 1991. The draft reported that the applicants were “establishing pasture on approximately 21 acres” of the property. Irrigation was an allowable use of well water for the property. Based on the assumption that these 21 acres were irrigated, the draft EIR *110 then determined “for the purposes of assessing impacts” that an estimate of existing water use for the September Ranch property was 45 acre-feet per year. This was based on an estimated 2 acre-feet for each of the 21 acres of pastureland plus 3 acre-feet used by the existing equestrian center and residence. The 2 acre-feet per acre was an estimate for irrigated pastureland taken from MPWMD guidelines for irrigated lands in the area and from a 1985 Pajaro Valley Irrigation Report. Water demand for the project as proposed for 117 residences was calculated at 61.15 acre-feet per year. This resulted in an increase of approximately 16.15 acre-feet per year over the existing estimated usage of 45 acre-feet per year. The draft EIR explained that the groundwater storage in the September Ranch sub-basin was more than adequate to supply the increased water demand during wet or normal weather conditions. However, the sub-basin supply would be vulnerable during a sustained drought of more than five years, which the draft concluded was a significant impact that must be mitigated. Furthermore, increased pumping on the September Ranch property could delay or reduce subsurface groundwater recharge to the Carmel Valley aquifer. Although this reduction would be a “small percentage” of the overall groundwater recharge in the Carmel Valley aquifer, the draft EIR acknowledged that “any impact reducing flow to the Carmel Valley aquifer is potentially significant.” The draft concluded that in order to mitigate the impact of increased pumping, the project applicants would either have to limit water project demand to the baseline of 45 acre-feet per year—either by reducing density or by instituting conservation measures—or they would have to provide an offsetting pumping reduction of 16.2 acre-feet per year elsewhere within the Carmel Valley basin. The draft EIR was circulated for public review and comments were received from agencies, associations and members of the public during the 45-day review period. The comments included numerous responses to the baseline water use figure. Letters from local property owners indicated that the pasturelands on the property had not been irrigated historically, but that the applicants had only recently begun irrigating since the application process had commenced. A comment from the Monterey County Department of Health pointed out that the actual amount of pastureland was significantly less than 21 acres and further that the draft EIR had stated only 11.6 acres were currently irrigated. In their responses to these comments the EIR consultants indicated that the figures regarding water usage were obtained from the project applicants: “This EIR has relied on production information provided by the applicant, *111 well production records available in the recent past and the extrapolation of a reasonable estimate of water use based upon irrigated acres of land on the site.” The responses further explained that the applicants had “stated that this area has been irrigated in the past, although there is no documentation available to confirm this.” The responses acknowledged that “in the recent past only 11.6 acres were irrigated.” The applicants also submitted further information and studies which indicated that irrigated pastureland actually could require as much as 6 acre-feet per year per acre. Furthermore, they represented that they had recently used approximately 23 acre-feet of water to irrigate approximately 11.6 acres of the terrace for only 14 weeks. This, they calculated, would compute to 95 acre-feet per year for the entire 21-acre pasture. However, according to the MPWMD, “this use would be higher than any other documented pasture irrigation in Carmel Valley.” The Final EIR The comments and responses were incorporated into the final EIR, dated March 6, 1998. In its analysis of baseline water usage, the final EIR reiterated that no documentation existed that could confirm historical water usage on the September Ranch. The EIR noted that comments to the draft EIR had suggested both higher and lower amounts than the estimate of 45 acre-feet per year. The final EIR continued to use 45 acre-feet per year as a baseline for purposes of assessing impacts, explaining that “[tjhis EIR attempts to provide a reasonable baseline based upon information of historic use provided by the applicant and a water demand factor for irrigated pastureland accepted by local water agencies (2.0 AF/acre, MPWMD).” However, the EIR then suggested that the Board could accept “additional documentation” and could revise this baseline figure higher or lower. Whether the baseline were set higher or lower, mitigation would require that “[n]o post-project water use will be allowed greater than the baseline (or an acceptable offset for this use [will] be required).” The final EIR included an updated water production data chart compiled from MPWMD records, showing metered water production on the property through 1997. This chart showed that water production had reached a new high of 78.34 acre-feet in 1997. However, the chart explained that approximately 52 of this 78.34 acre-feet were produced during a 47-day period of aquifer testing. Using the 45 acre-feet per year figure that had been determined to be a “reasonable” baseline figure, the final EIR reached the same conclusions as *112 the draft. It found that the project as proposed would result in increased pumping of approximately 16.2 acre-feet over baseline use. Postproject water use greater than identified baseline levels was a significant impact that would require mitigation: either reducing water production for the project to baseline conditions or providing an offsetting pumping reduction within the Carmel Valley basin. The Supplemental Final EIR The County belatedly forwarded the draft EIR to the State Clearinghouse on March 4, 1998, which required a second 45-day review period and generated further comments. The responses to these comments were added as “Volume 2” to the final EIR, dated May 27, 1998. This is also referred to as the “Supplement to Final EIR,” or the supplemental EIR. The supplemental EIR included extensive comments by the State Water Resources Control Board (SWRCB) regarding the EIR’s conclusions about groundwater recharge. These comments indicated that groundwater recovery under normal conditions would be worse than depicted in the EIR and stated that appropriation of water from the aquifer underlying the September Ranch would be subject to the permitting authority of the SWRCB. In response, the applicants then wrote to the SWRCB asserting that they had riparian rights which could be utilized for the project. The SWRCB’s reply indicated the various qualifications under which the project could be considered for riparian rights. The responses in the supplemental EIR addressed, among other things, these asserted riparian rights, which neither the draft EIR nor the final EIR had discussed. The supplemental EIR explained that “[ajlthough the project applicants originally identified that they would be using ‘percolating groundwater’ under the project site, a subsequent letter has clarified their intent to provide water to their proposed project under their ‘riparian’ rights.” The new material went on to explain the differences between groundwater rights, riparian rights and appropriative rights. The supplemental EIR noted that it could not confirm the property’s riparian status and that the SWRCB had not yet made a determination as to the validity of any claimed riparian right. A new mitigation measure was added in the supplemental EIR, requiring that the applicants either provide assurance of a valid riparian claim or secure a permit for an appropriative water right from the SWRCB. On June 22, 1998, after the supplemental EIR was issued, the attorney for the applicants informed the County Planning Department that the applicants had ownership rights to a 10-acre parcel of land along Carmel Valley Road, *113 known as the Berube parcel. The applicants had recently purchased the stipulated right to pump approximately 32 acre-feet of water per year from this property. The attorney asserted that pumping on the Berube parcel could be reduced if mitigation of the impact of water use for the September Ranch project were necessary. An appropriative permit is not required in order to use a reduced pumping offset. Citizen Committees Pursuant to local ordinance, the September Ranch project was presented to the Carmel Valley Citizens Subdivision Evaluation Committee to evaluate the project for compliance with the Carmel Valley Master Plan. On May 18, 1998, the Committee gave the project a failing score of 44 percent in the category of water/hydrology. The county’s land use advisory committee reviewed the project in June of 1998 and voted for denial because it concluded that the project did not comply with Master Plan policies relating to water supply and traffic. Planning Commission Decision On September 30, 1998, the County Planning Commission (Planning Commission) voted to deny the proposed project, based in part on concerns about water impacts. The Planning Commission voted to approve a smaller project with 49 residential units and 7 inclusionary units, which was described as the environmentally superior project in the final EIR. The Planning Commission did not accept the approach used in the EIR to determine baseline use by computing an average estimated use of two acre-feet per year per acre for irrigated pasture. Instead the Planning Commission relied on actual water production records for the September Ranch for the most recent year, namely 1997. It found this figure to be 26.34 acre-feet (a total of 78.34 acre-feet less 52 acre-feet attributed to aquifer testing), and therefore recommended that the project density be reduced accordingly so that there would be no increase in pumping over baseline level. The Planning Commission found that the reduced density project was necessary to ensure that impacts to the Carmel River alluvial aquifer were reduced to a level of insignificance. A hearing for review of the Planning Commission decision was then set before the Monterey County Board of Supervisors for December 1, 1998. Supplemental Information and Errata On November 19, 1998, additional information was submitted by the environmental consultants, entitled “Supplemental Information and Errata *114 for the September Ranch Project Environmental Impact Report.” This supplemental material discussed the reduced density alternative of 49 units adopted by the Planning Commission, and noted that information provided by tiie applicants had indicated that this alternative was economically unfeasible. The errata also contained a further discussion of baseline water usage, recognizing once again that “if the project were to exceed the amount of water used on the site under existing or baseline conditions, a significant unavoidable impact would occur due to potential regional water impacts.” It explained that the EIR had determined the baseline of 45 acre-feet per year by using a “standard water demand factor for irrigated pastureland” based on irrigation formulas and representations by the applicants that “there was an established practice of irrigation on the site.” The MPWMD and the County Environmental Health Department, however, had requested that the EIR consider an alternative that used only “documented past year water use,” which was the approach taken by the Planning Commission. This had resulted in a figure of 26.34 acre-feet per year. The errata concluded that baseline could be established either by using an assigned water demand factor for irrigated pastureland, as the EIR had done, or by relying on recent records of water production. Referring to a newly updated chart of documented water use from 1991 to 1999, the errata then set forth a calculation of baseline water use for various combinations of years: for 1998-1999, average use was approximately 43 acre-feet per year; for 1997-1999, the figure was 51 acre-feet per year; for 1993-1999, average use was approximately 30 acre-feet per year. The supplemental material again emphasized that the EIR required that “post-development water production from the September Ranch aquifer not exceed identified pre-project baseline levels.” The staff report to the Board was prepared the next day, November 20, 1998, and it attached the Supplemental Information and Errata, as well as the supplemental final EIR, and further supplemental information from the applicants regarding the Berube property. The staff prepared a revised Board resolution, dated December 1, 1998. The staff recommended that the Board modify the subdivision evaluation committee’s failing score in the category of water/hydrology and give the project a passing score. This recommendation was based on the fact that the applicants had since identified the Berube property as a source for offset pumping, and the staff had secured evidence from the applicants documenting the availability of water use on the Berube parcel sufficient to provide the necessary mitigation of the impact of pumping water over baseline for the September Ranch property. Because the *115 Supplemental Information and Errata and the new information on the Berube property were made available just prior to the Board hearing, the opportunity for public comment and response was limited. The Decision of the Board of Supervisors On December 1, 1998, the Board conducted a public hearing and decided, on separate three-to-two votes, to certify the EIR, to modify the failing score of the subdivision evaluation committee, and to adopt the findings and conditions of approval for a modified project. Rather than 100 market-rate units and 17 inclusionary units as initially proposed, the Board approved 94 market-rate units and 15 inclusionary units. Recognizing the requirement that project water use be limited to baseline conditions, the Board “selected 51 acre-feet per year as the baseline water use amount.” This figure was derived from an average of water use on the property during the past three reporting years—1997, 1998, and 1999—and was based on the updated chart and information provided in the Supplemental Information and Errata. The Board found that the water demand of the reduced-density project as approved was 57 acre-feet per year. Thus only 6 acre-feet per year were needed to offset the increase over baseline. As a condition of approval of the project, the applicants were to provide an offsetting reduction in pumping on the Berube parcel to ensure that water demand on the Carmel Valley aquifer did not increase as a result of the project. On December 21, 1998, a county clerk published the findings and conditions of the Board in resolution No. 98-500. This resolution contained several changes to the Board’s findings and conditions that were taken from material submitted to the clerk by the attorney for September Ranch after the Board had adjourned. The Mandate Proceeding Two petitions for administrative mandate were filed in superior court, by the Save Our Peninsula' Committee, 3 et al., and by Sierra Club et al., challenging the certification of the EIR and the findings of the Board. The court consolidated the cases for a court trial, which was held on July 1 and July 6, 1999. The court issued a lengthy “Intended Decision” on September 1, 1999, which it adopted as its statement of decision. The court concluded that the Board’s findings as to baseline water conditions were not supported *116 by substantial evidence; that the Board’s findings that there was a long-term water supply in the form of riparian rights were legally inadequate and not supported by the evidence; that the EIR contained no environmental analysis of the use of an off-site water source to offset water usage over baseline; and that the EIR failed to adequately consider mitigation of the traffic impacts of the project at the intersection of Highway 1 and on two other segments of Carmel Valley Road. The court entered judgment in favor of petitioners in both actions and issued a writ of mandate remanding the matter back to the Board and ordering the Board to vacate resolution No. 98-500 and to vacate the certification of the EIR. The Board was ordered to take no further action to approve the project without first preparing, circulating, and considering an EIR that was legally adequate with regard to its analysis of the water and traffic issues delineated in the statement of decision. In light of its ruling on water and traffic issues, the court found the petitioners’ other objections to the project approval and to the EIR were moot, but could be revived depending on the Board’s actions on remand. 4 Attorney fees were awarded to petitioners. Real parties in interest September Ranch Partners and James Morgens appeal. 5 They argue that the EIR was legally sufficient and that the Board’s determinations regarding water supply impacts and mitigation and traffic mitigation were supported by substantial evidence. Real parties also appeal the orders awarding attorney fees. They argue that if the judgment is reversed, the orders awarding attorney fees must also be reversed. The County did not appeal and no cross-appeals were filed by petitioners. Issues Standard of Review In a mandate proceeding to review an agency’s decision for compliance with CEQA, the scope and standard of our review are the same as the trial court’s, and the lower court’s findings are not binding on us. (San *117 Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus (1994) 27 Cal.App.4th 713, 722 [32 Cal.Rptr.2d 704].) We review the administrative record to determine whether the agency prejudicially abused its discretion. (Laurel Heights Improvement Assn. v. Regents of University of California (1993) 6 Cal.4th 1112, 1132-1133 [26 Cal.Rptr.2d 231, 864 P.2d 502].) “Abuse of discretion is established if the agency has not proceeded in a manner required by law or if the determination or decision is not supported by substantial evidence.” (Pub. Resources Code, § 21168.5; Laurel Heights Improvement Assn. v. Regents of University of California (1988) 47 Cal.3d 376, 392, fn 5 [253 Cal.Rptr. 426, 764 P.2d 278]; County of Amador v. El Dorado County Water Agency (1999) 76 Cal.App.4th 931, 944 [91 Cal.Rptr.2d 66].) “Substantial evidence” is defined in the CEQA Guidelines 6 as “enough relevant information and reasonable inferences from this information that a fair argument can be made to support a conclusion, even though other conclusions might also be reached. Whether a fair argument can be made . . . is to be determined by examining the whole record before the lead agency. Argument, speculation, unsubstantiated opinion or narrative [or] evidence which is clearly erroneous or inaccurate . . . does not constitute substantial evidence.” (Guidelines, § 15384, subd. (a).) The agency is the finder of fact and we must indulge all reasonable inferences from the evidence that would support the agency’s determinations and resolve all conflicts in the evidence in favor of the agency’s decision. (Western States Petroleum Assn. v. Superior Court (1995) 9 Cal.4th 559, 571 [38 Cal.Rptr.2d 139, 888 P.2d 1268].) In reviewing an agency’s decision to certify an EIR, we presume the correctness of the decision. The project opponents thus bear the burden of proving that the EIR is legally inadequate. (Al Larson Boat Shop, Inc. v. Board of Harbor Commissioners (1993) 18 Cal.App.4th 729, 740 [22 Cal.Rptr.2d 618]; Barthelemy v. Chino Basin Mun. Water Dist. (1995) 38 Cal.App.4th 1609, 1617 [45 Cal.Rptr.2d 688].) While we are guided by these deferential rules of review, we must also bear in mind that the overriding purpose of CEQA is to ensure that agencies regulating activities that may affect the quality of the environment give primary consideration to preventing environmental damage. (Laurel Heights Improvement Assn. v. Regents of University of California, supra, 47 Cal.3d at p. 390.) CEQA is the Legislature’s declaration of policy that all necessary action be taken “ ‘to protect, rehabilitate, and enhance the environmental quality of the state.’ ” (Id. at p. 392; Pub. Resources Code, § 21000.) “The EIR is the heart of CEQA” and the integrity of the process is dependent on the adequacy of the EIR. (County of Inyo v. Yorty (1973) 32 *118 Cal.App.3d 795 [108 Cal.Rptr. 377]; Sutter Sensible Planning, Inc. v. Board of Supervisors (1981) 122 Cal.App.3d 813 [176 Cal.Rptr. 342].) “ ‘The ultimate decision of whether to approve a project, be that decision right or wrong, is a nullity if based upon an EIR that does not provide the decision-makers, and the public, with the information about the project that is required by CEQA.’ [Citation.] The error is prejudicial ‘if the failure to include relevant information precludes informed decisionmaking and informed public participation, thereby thwarting the statutory goals of the EIR process.’ ” (San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus, supra, 27 Cal.App.4th at pp. 721-722; Galante Vineyards v. Monterey Peninsula Water Management Dist. (1997) 60 Cal.App.4th 1109, 1117 [71 Cal.Rptr.2d 1]; County of Amador v. El Dorado County Water Agency, supra, 16 Cal.App.4th at p. 946.) When the informational requirements of CEQA are not complied with, an agency has failed to proceed in “a manner required by law” and has therefore abused its discretion. (Pub. Resources Code, §§ 21168.5, 21005, subd. (a); County of Amador v. El Dorado County Water Agency, supra, 16 Cal.App.4th at p. 946; Environmental Planning & Information Council v. County of El Dorado (1982) 131 Cal.App.3d 350, 355 [182 Cal.Rptr. 317].) In sum, although the agency’s factual determinations are subject to deferential review, questions of interpretation or application of the requirements of CEQA are matters of law. (Galante Vineyards v. Monterey Peninsula Water Management Dist., supra, 60 Cal.App.4th 1109, 1117; County of Amador v. El Dorado County Water Agency, supra, 16 Cal.App.4th at pp. 952-956; San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus, supra, 27 Cal.App.4th at pp. 728-729.) While we may not substitute our judgment for that of the decision makers, we must ensure strict compliance with the procedures and mandates of the statute. (Citizens of Goleta Valley v. Board of Supervisors (1990) 52 Cal.3d 553, 564 [276 Cal.Rptr. 410, 801 P.2d 1161].) Water Issues The EIR in this case recognized the serious water concerns in the Carmel Valley and acknowledged the state and local policies seeking to limit any new development that would result in increased water pumping affecting the Carmel Valley alluvial aquifer. In consideration of these concerns, the analysis of water issues in the EIR rested on the premise that any increase in water pumping above preproject levels would constitute an adverse and significant environmental impact, mandating mitigation. No one disputes this general premise. Rather, it is the determination of the preproject or *119 baseline water use, against which the water demands of the project are to be measured, that is at the center of the controversy here. We turn to this issue first and to several questions which must necessarily be resolved along with it. Is the determination of baseline water use a policy decision, properly addressed to the discretion of the decisionmaking agency, or does CEQA require that baseline use be established in the EIR? Was the EIR’s estimate of baseline water use for irrigated pastureland supported by the evidence? Was the Board’s determination that baseline water use in this case was 51 acre-feet per year supported by evidence in the record? And what is the time at which a baseline for water use is properly determined? Is it at the beginning of the environmental review process or at the end when the project is approved? We next address two additional and related water issues: whether the EIR adequately analyzed off-site pumping reduction on the Berube property as mitigation of any increased water usage over baseline, and whether the EIR adequately discussed the applicants’ asserted riparian rights as a long-term water source. Baseline Appellants argue that the determination of a baseline condition is a matter of policy to be resolved by the agency, based on the information and analysis provided in the EIR. Appellants remind us that the EIR is only an informational document and that the agency is the decision maker. (County of Inyo v. City of Los Angeles (1977) 71 Cal.App.3d 185, 189 [139 Cal.Rptr. 396].) Here the preparers of the EIR ultimately found that the question of “the establishment of a baseline use and mitigations based upon this baseline” raised policy implications best addressed to the Board’s discretion. Appellants argue that this was proper because the EIR contained an array of evidence regarding baseline and a variety of suggested formulas for determining baseline. The Board’s choice of a particular formula was therefore within its discretion and was supported by the evidence. Respondents argue that the baseline environmental conditions must be established in the EIR itself. Without a determination and description of the existing physical conditions on the property at the start of the environmental review process, the EIR cannot provide a meaningful assessment of the environmental impacts of the proposed project. (Pub. Resources Code, §§ 21100, subd. (a), 21060.5; Environmental Planning & Information Council v. County of El Dorado, supra, 131 Cal.App.3d at p. 354.) “Before the impacts of a project can be assessed and mitigation measures considered, an EIR must describe the existing environment. It is only against this baseline *120 that any significant.environmental effects can be determined.” (County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at p. 952; Guidelines, §§ 15125, subd. (a), 15126.2, subd. (a).) There is some merit in both of these positions. Because the chief purpose of the EIR is to provide detailed information regarding the significant environmental effects of the proposed project on the “physical conditions which exist within the area,” it follows that the existing conditions must be determined, to the extent possible, in the EIR itself. (Pub. Resources Code, § 21060.5; Environmental Planning & Information Council v. County of El Dorado, supra, 131 Cal.App.3d at p. 354; Galante Vineyards v. Monterey Peninsula Water Management Dist., supra, 60 Cal.App.4th at p. 1122.) On the other hand, the agency has the discretion to resolve factual issues and to make policy decisions. If the determination of a baseline condition requires choosing between conflicting expert opinions or differing methodologies, it is the function of the agency to make those choices based on all of the evidence. {Barthelemy v. Chino Basin Mun. Water Dist., supra, 38 Cal.App.4th 1609, 1617.) If an EIR presents alternative methodologies for determining a baseline condition, however, we believe CEQA requires that each alternative be supported by reasoned analysis and evidence in the record so that the decision of the agency is an informed one. We further find that the EIR must set forth any analysis of alternative methodologies early enough in the environmental review process to allow for public comment and response. This is particularly important in a case such as this, where water issues were a matter of widespread public concern, and where the determination of the figure for baseline water usage dictated the density of the project. Here the draft EIR initially established a baseline of 45 acre-feet per year, based on the representation by the owners that 21 acres were irrigated, although the EIR acknowledged that the record contained “no documentation” showing any substantial irrigation prior to 1997. Furthermore, having estimated a baseline figure and having used that figure throughout the EIR to assess the project’s impacts, the EIR consultants ultimately referred the baseline determination to the Board, to be decided as a matter of “policy.” At the very end of the environmental review process, the Board was invited to choose among various calculations compiled from updated water meter readings on the property. But some of these figures, although generated from recent pumping on the property, did not reflect water actually used for irrigating the property. We conclude, as explained more fully below, that this treatment of baseline water use violated the basic *121 principles of CEQA, which require that an EIR start with a description of “the existing environment.” (County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at p. 952.) Respondents argue that since there was no documentation to support the EIR’s threshold determination that the September Ranch property was irrigated pastureland, baseline water use should properly have been set at a figure that more closely represented water actually used historically on the property. The evidence was indeed sparse on this subject. There was some evidence that the property had been farmed prior to 1950. After that time, the equestrian uses began. However, accounts from neighbors in the area indicated that the pasturelands were not regularly irrigated during this time. Although the MPWMD has required well reports since 1980, there were no reports on this property. The applicants indicated at trial that the old well had not been used for at least 10 years before 1990, when a new well was installed. Records starting in 1991 show a temporary aquifer test was conducted in 1991 and produced 1.20 acre-feet. In the following year 40.68 acre-feet were pumped. However this too was all for aquifer testing. Over the next three years prior to the submission of the development application in this case, water production totals were 11.58 acre-feet, 0.40 acre-feet, and 1.08 acre-feet. We have no objection to the EIR’s methodology of estimating historical water use on property where no documentation is available to verify actual use. But estimating water used for irrigation where there was no substantial evidence to show that the property was in fact irrigated does not accurately reflect existing conditions. Appellants’s argument that it was entitled to use this amount of water for irrigation is not the same as actual use. As various courts, including this one, have held, the impacts of the project must be measured against the “real conditions on the ground.” (City of Carmel-by-the Sea v. Board of Supervisors (1986) 183 Cal.App.3d 229, 246 [227 Cal.Rptr. 899]; Environmental Planning & Information Council v. County of El Dorado, supra, 131 Cal.App.3d at p. 354; County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at p. 952; Galante Vineyards v. Monterey Peninsula Water Management Dist, supra, 60 Cal.App.4th at p. 1122.) We are mindful that judicial review does not allow for a reweighing of the evidence and that “[determinations in an EIR must be upheld if they are supported by substantial evidence.” (Barthelemy v. Chino Basin Mun. Water Dist, supra, 38 Cal.App.4th 1609, 1620.) However, “[a]n EIR must focus on impacts to the existing environment, not hypothetical situations.” (County of *122 Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at p. 955.) And “unsubstantiated opinion or narrative . . . does not constitute substantial evidence.” (Guidelines, § 15384, subd. (a).) Here it would appear that the only evidence that the terrace on the September Ranch property was irrigated pasture was the representation of the applicants themselves, who clearly had a vested interest in establishing a water use baseline high enough to allow the project to go forward. On this record, we must question the premise accepted in the EIR, that pre-project water usage on the September Ranch property was for irrigating the pastureland. Furthermore, in response to public comments that the draft EIR’s estimated water use did not reflect the actual use, the EIR stated that “[t]he request for documentation for historic use is referred to decision makers.” We are concerned by this apparent delegation of duty to the decision makers to gather the necessary information to support a determination of baseline water use. We believe CEQA requires that the preparers of the EIR conduct the investigation and obtain documentation to support a determination of preexisting conditions. (See, e.g., San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus, supra, 27 Cal.App.4th 713, 727-729.) This is a crucial function of the EIR. If further investigation would have uncovered documentary evidence regarding the historical use of water on the property, that was the province of the EIR and not the Board. And while the Board is entitled to accept or reject evidence or to adopt one methodology over another, the EIR’s estimate of baseline by using a standard formula for irrigated pastureland must be based on substantial evidence that this property could be characterized as irrigated pastureland. Even if we were to accept the EIR’s initial premise that an estimate of water used for irrigable lands was appropriate in this case, in the absence of documentary evidence to establish actual use, the EIR’s baseline analysis reveals further, and in our view more critical, inadequacies. After determining a “reasonable baseline” of 45 acre-feet per year, and after using this figure throughout the draft and final EIR “for the purposes of assessing impacts,” the EIR ultimately retreated from this estimate and deferred to the Board to determine baseline usage based on an entirely different methodology. In the Supplemental Information and Errata, which was submitted to the County just prior to the Board meeting, the EIR consultants suggested for the first time that a baseline determination of water use could be established either by using a “standard water demand factor for irrigated pastureland,” as the EIR had done, or by using documented water meter records showing water production in recent years. *123 The water production chart for the property showed that after the development application was submitted in this case in the summer of 1995, water production on the property increased substantially. In 1996 and 1997, extensive aquifer testing was done. For 1997, water production was measured at 78.34 acre-feet. In 1998, water production was 34.04 acre-feet and for the partial reporting year of 1999, just before the Board hearing, it was up to 41.14 acre-feet. The Supplemental Information and Errata then suggested several possible combinations and averages of these production numbers, one of which, 51 acre-feet per year, was the figure eventually selected by the Board. This figure was a departure, both numerically and methodologically, from the 45 acre-feet per year that had been developed as the baseline figure by the consultants and had been used throughout the EIR process. And since it first appeared in supplemental information supplied to the County shortly before the Board convened, there was little opportunity for public comment and meaningful response as to either the methodology or the evidence to support the figures used. Furthermore, the supplemental information contained little meaningful analysis as to why any of the suggested calculations might represent a reasonable determination of baseline water usage for irrigating this property. Indeed it appears that several of the figures on the water production chart do not represent water actually used for irrigating the property. For example, the 51 acre-feet per year figure selected by the Board was an average of water meter readings in the past three years, including 1997. The figure for 1997 is 78.34 acre-feet. However, the chart clarifies that “[o]f this total, about 52 acre-feet were produced during a 47 day period of aquifer testing .... The remainder, 26.34 acre-feet is the amount accepted by the MPWMD as the water production for irrigation in RY [reporting year] 1997.” (Italics added) Even though only 26.34 acre-feet was actually used for irrigation, the EIR advised that the Board “could accept the actual water production amount, the full 78.34 AF/yr, or deduct the amount of water used for aquifer testing (52 AF), as requested by the MPWMD to account for the anomaly of the aquifer testing.” This reasoning is clearly faulty. A baseline figure must represent an environmental condition existing on the property prior to the project. There is simply no justification for using a total of 78.34 acre-feet of water as part of a baseline calculation for this property, when the evidence was that 52 acre-feet of this amount was pumped for the purpose of aquifer testing and was discharged into the Carmel River. By inviting the Board to pick from an array of numbers to determine an important aspect of the baseline environmental setting, the EIR failed to *124 fulfill its function of providing information and analysis of environmental impacts. In a recent case involving a massive water project that proposed to divert 17,000 acre-feet of water from three high Sierra lakes, the court found the EIR’s baseline analysis to be inadequate, on similar facts. (County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at 953.) In County of Amador, the EIR’s discussion of baseline conditions consisted of a recitation of month-end lake levels for the three lakes. It failed to explain how those lake levels were maintained, the historical duration and timing of the water releases, and the impacts on fishery resources and recreational uses. The court found that the lake level figures alone were insufficient to describe the existing water release program. The court noted that “this is not a case involving conflicting expert opinions about historical operation.” (Id. at p. 954.) Rather the EIR simply presented data without meaningful analysis. The court in County of Amador underscored the “importance of an adequate baseline description, for without such a description, analysis of impacts, mitigation measures and project alternatives becomes impossible.” (Id. at p. 953.) The court concluded that “[a]n adequate EIR requires more than raw data; it requires also an analysis that will provide decision makers with sufficient information to make intelligent decisions.” (Id. at p. 955; see also Guidelines, § 15151.) The EIR in this case similarly provided raw data, in the form of recent water meter figures for the September Ranch property, and then invited the Board to select a baseline from among several suggested combinations of these figures. As in County of Amador, this was not a case where the Board was called upon to perform its discretionary function of resolving a factual dispute or choosing from conflicting expert opinions or methodologies regarding water usage. Instead this was an arbitrary process, involving arithmetic rather than analysis. The Board was permitted to make the crucial determination of baseline water use by choosing from a selection of numbers, some of which did not represent water actually used to irrigate the property. And this occurred at the very end of the environmental review process, thus avoiding public scrutiny and precluding the meaningful comparison of preproject and postproject conditions required by CEQA. This brings us to the question whether it was proper in any event to rely on water production figures generated at the end of the environmental review process, rather than at the beginning, to determine a baseline figure. The relevant Guideline at the time of the environmental review for the September Ranch project was section 15125, which provided: “An EIR must include a description of the environment in the vicinity of the project, as it exists before the commencement of the project, from both a local and regional *125 perspective.” (Guidelines, former § 15125, subd. (a), italics added.) Appellants take the italicized words to mean immediately before the project is approved and permits are issued. Respondents contend that existing conditions must be evaluated as closely as possible to the date the notice of preparation of the EIR is filed, as that is the date the project is officially commenced within the meaning of CEQA. They maintain that an EIR cannot adequately analyze the impacts on the environment if it does not start with a description of the physical conditions existing on the property at the beginning of the environmental review. A subsequent amendment to section 15125 of the Guidelines supports respondents’ interpretation. Section 15125, subdivision (a), now provides: “An EIR must include a description of the physical environmental conditions in the vicinity of the project, as they exist at the time the notice of preparation is published, or if no notice of preparation is published, at the time environmental analysis is commenced .... This environmental setting will normally constitute the baseline physical conditions by which a lead agency determines whether an impact is significant.” (Italics added.) Furthermore, section 15126.2 now provides as follows: “In assessing the impact of a proposed project on the environment, the lead agency should normally limit its examination to changes in the existing physical conditions in the affected area as they exist at the time the notice of preparation is published, or where no notice of preparation is published, at the time environmental analysis is commenced.” These amendments reflect and clarify a central concept of CEQA, widely accepted by the courts, that the significance of a project’s impacts cannot be measured unless the EIR first establishes the actual physical conditions on the property. (County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at p. 953; Environmental Planning & Information Council v. County of El Dorado, supra, 131 Cal.App.3d at p. 354; City of Carmel by-the-Sea v. Board of Supervisors, supra, 183 Cal.App.3d 229.) In other words, baseline determination is the first rather than the last step in the environmental review process. We adopt this general rule. We also agree with appellants, however, that the date for establishing baseline cannot be a rigid one. Environmental conditions may vary from year to year and in some cases it is necessary to consider conditions over a range of time periods. In some cases, conditions closer to the date the project is approved are more relevant to a determination whether the project’s impacts will be significant. (See Mira Monte Homeowners Assn. v. County of Ventura (1985) 165 Cal.App.3d 357 [212 Cal.Rptr. 127].) For instance, where the issue involves an impact on traffic levels, the EIR might necessarily take into account the normal increase in *126 traffic over time. Since the environmental review process can take a number of years, traffic levels as of the time the project is approved may be a more accurate representation of the existing baseline against which to measure the impact of the project. (See, e.g. Fairview Neighbors v. County of Ventura (1999) 70 Cal.App.4th 238 [82 Cal.Rptr.2d 436] [maximum estimated traffic was appropriate baseline].) Even in the case before us, if the more recent water production figures could be shown to represent a continuation of preproject water usage, such figures might be relevant to a determination of baseline water conditions. However, here the more recent figures consisted primarily of aquifer testing where water was pumped and released into the river. Water which was pumped for irrigation in 1997, 1998, and 1999 was a significantly higher amount than in the previous six recorded years. Thus these recent figures do not appear to represent a normal fluctuation in usage over time, as appellants suggest. Furthermore, there are sound reasons for determining baseline water use in this particular case as of the time of the commencement of the environmental review. Here the environmental review process spanned three and a half years. During that time it became apparent that the water supply for this project was a critical issue. A state water board decision precluded a hookup with the local water company. State and local policy restricted development that would increase pumping in the Carmel Valley basin. And pumping tests established that the sub-basin underlying the property was not separate from the Carmel Valley aquifer. Because any water used by the project in excess of baseline would constitute a significant adverse impact, it was clear that the baseline figure would dictate the amount of allowable density for the project. Production of water on the property during the lengthy environmental review process was controlled by the applicants. It was in their interests to elevate water production figures in order to establish as high a baseline as possible. While we do not speculate as to whether this occurred, we believe water production figures generated towards the end of the environmental review process must be regarded with some caution in these circumstances. Their relevance to baseline conditions would depend on whether they are representative of the amount of water historically produced for use on the property. The better approach, however, would be to follow the general rule expressed in the Guidelines and cases that baseline conditions are normally to be determined as of the time environmental review is begun. This most closely describes the environment “as it exists before the commencement of the project.” (Guidelines, former § 15125, subd. (a).) Cases cited by appellants do not support the proposition that baseline is determined at the end rather than at the beginning of the environmental *127 review. In Riverwatch v. County of San Diego (1999) 76 Cal.App.4th 1428 [91 Cal.Rptr.2d 322], the court found that the EIR did not need to consider a baseline date some 12 years prior to the commencement of the project, in order to account for previous unlawful activity by the owners that had degraded the property. Riverwatch does not address the question raised here, whether the baseline conditions should be established as of the beginning or the end of the environmental review process. The court in Riverwatch did state as a general principle that environmental impacts should be examined “in light of the environment as it exists when a project is approved.” (Riverwatch v. County of San Diego, supra, 76 Cal.App.4th at p. 1453.) However, in context it appears the court was simply rejecting the notion that the baseline should be set a number of years earlier than the commencement of the current project. Moreover, the authorities relied on in Riverwatch do not support the view that baseline should be determined as of the date of project approval. Bloom v. McGurk (1994) 26 Cal.App.4th 1307 [31 Cal.Rptr.2d 914] did not involve preparation of an EIR but rather addressed the question of baseline for purposes of determining a categorical exemption from CEQA. That case in turn relied on City of Carmel-by-the-Sea v. Board of Supervisors, supra, 183 Cal.App.3d 229. In City of Carmel we stated that “[i]n assessing the impact of [a] rezoning, it is only logical that the local agency examine the potential impact on the existing physical environment.” {Id. at p. 246.) In the context of that case our meaning was that the agency must examine the impact of the project as against the physical conditions on the subject property, as opposed to measuring the potential impact against a draft general plan. We said nothing expressly about whether the existing conditions are to be determined at the beginning or at the end of the environmental review process. However our statement in City of Carmel clearly implies that meaningful environmental review must proceed at the outset from a determination of the property’s existing physical conditions. We believe that this is the correct interpretation of CEQA as applied to this case. This view is supported by the courts and by the Guidelines, and is consistent with the central function of the EIR, to inform decision makers about the impacts of the proposed project on the existing environment. (County of Amador v. El Dorado County Water Agency, supra, 76 Cal.App.4th at pp. 952-956; County of Inyo v. City of Los Angeles (1981) 124 Cal.App.3d 1, 9 [177 Cal.Rptr. 479]; Environmental Planning & Information Council v. County of El Dorado, supra, 131 Cal.App.3d at p. 354; City of Carmel-by-the-Sea v. Board of Supervisors, supra, 183 Cal.App.3d at p. 246.) An EIR in which a baseline water use determination is elastic and can be *128 modified by the Board at the end of the environmental review process without benefit of analysis or public participation does not fulfill this function. If an EIR fails to include relevant information and precludes informed decisionmaking and public participation, the goals of CEQA are thwarted and a prejudicial abuse of discretion has occurred. (Sierra Club v. State Bd. of Forestry (1994) 7 Cal.4th 1215, 1236 [52 Cal.Rptr.2d 19, 876 P.2d 505]; Fall River Wild Trout Foundation v. County of Shasta (1999) 70 Cal.App.4th 482, 492 [82 Cal.Rptr.2d 705]; County of Amador v. El Dorado County Water Agency, supra, 16 Cal.App.4th at p. 954; Pub. Resources Code, § 21005, subd. (a).) “Our role here, as a reviewing court, is not to decide whether the board acted wisely or unwisely, but simply to determine whether the EIR contained sufficient information about a proposed project, the site and surrounding area and the projected environmental impacts arising as a result of the proposed project or activity to allow for an informed decision . . . .” (San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus, supra, 27 Cal.App.4th at p. 718.) Based on these guiding principles, we conclude here that the EIR was inadequate in its baseline discussion in several respects: by failing to investigate and present evidence to support the assumption that the preproject use of water on the property was for irrigation; by introducing a new methodology for baseline determination at the end of the environmental review process without any informational discussion or opportunity for public review; and by inviting the Board to select a baseline among water production figures with no meaningful analysis and no showing that the figures represented water actually used on the property consistent with historical use. Because of these inadequacies, the Board’s decision setting baseline water use at 51 acre-feet per year was not supported by the evidence and was an abuse of discretion. Off-site Pumping Reduction on the Berube Property Although the EIR had indicated that any increased water pumping over baseline would have to be mitigated either by reducing the project density or by reducing pumping elsewhere within the Carmel Valley basin, the applicants did not identify an offsetting pumping location until well after the comment periods had closed. In June of 1998, the attorney for the applicants informed the County that the applicants had recently acquired pumping rights to approximately 32 acre-feet of water per year on the 10-acre Berube parcel. The Berube property was located further up Carmel Valley Road approximately two miles away from the September Ranch property. The information about the Berube parcel was contained in the Supplemental *129 Information and Errata, which was submitted to the Board just prior to the hearing along with staff recommendations. It was on the basis of the identification of the Berube parcel that staff recommended that the Board modify the failing score given to the project by the subdivision evaluation committee in the category of water/hydrology. As a condition of approval of the project, the Board required that the applicants reduce pumping on the Berube property in order to offset project water demand over baseline. All that was required of the applicants was to show proof of control of the water rights on the offset parcel, and evidence of a deed restriction mandating reduction, subject to approval by the MP-WMD and the director of environmental health. No permit would be necessary to secure this offset mitigation. Comments received during the circulation of the draft EIR expressed concerns about the precedent-setting impacts of using offset water credits at another location in the Carmel Valley to mitigate increased pumping at the site of the project. Among other things, such a policy would take water from property capable of being irrigated for agricultural purposes. The Monterey County Environmental Health Department commented that “if [water credit transfers] will be used in the final EIR, then the EIR should also analyze the precedent setting impacts throughout the valley for all properties that are capable of being irrigated for pasture, grapes, crops etc.” The health department noted that it would be “crucial” to analyze the specifics and enforcement mechanisms of any off-site pumping offset to make sure the reduction property was situated so that there was a nexus between the offset and the increased pumping for the project. The health department urged that the site be identified as soon as possible so that it could be analyzed for feasibility and the necessary findings could be made. In response to these comments, the EIR agreed that there must be a “nexus” between the impact and the mitigation. If off-site pumping were to be used as mitigation, the reduction must be “an actual reduction in documented current water use, not simply a reduction on potential future pumping.” After the applicants had identified the Berube property as an offset pumping reduction site, the County’s chief environmental health officer wrote to the planning director. He pointed out that there had been no discussion of this property in the EIR. He also noted that “offsets do not necessarily provide water ‘savings’ ” and may not be sufficient to provide proof of a long-term water supply. The supplemental material for the EIR provided no response and contained no further discussion of the effects of this offsetting pumping reduction on the Berube property. Other concerns *130 were expressed as to the validity of the water rights on the Berube property, and the question whether the impacts of overpumping at one site are in fact balanced out by refraining from pumping at a different site miles away. There was no analysis of the historic usage at the Berube property or whether the offset would result in an actual reduction of pumping or would simply be a “paper credit.” The trial court found that the Board’s approval of this mitigation measure was not supported by the evidence because there was no environmental analysis in the EIR of the impacts of the pumping reduction on the Berube parcel and no analysis of the broader issues that were raised in numerous comments as to whether this offsetting mitigation resulted in potential cumulative growth-inducing impacts. Appellants argue that the EIR is not required to discuss the environmental effects of mitigation measures. They contend that substantial evidence supports the Board’s determination that the pumping offset would mitigate the impacts of any increased pumping without causing any new significant impacts. We disagree with these contentions. An EIR is required to discuss the impacts of mitigation measures. At the time of the environmental review in this case, former section 15126 of the Guidelines provided that “if a mitigation measure would cause one or more significant effects in addition to those that would be caused by the project as proposed, the effects of the mitigation measure shall be discussed[,] but in less detail than the significant effects of the project as proposed.” (Guidelines, § 15126, former subd. (c).) 7 Furthermore, section 15126, former subdivision (g), provided that the growth-inducing impact of the proposed action must be discussed in the EIR, including “the ways in which the proposed project could foster economic or population growth, or the construction of additional housing, either directly or indirectly, in the surrounding environment.” 8 (See also San Joaquin Raptor/Wildlife Rescue Center v. County of Stanislaus, supra, 27 Cal.App.4th at p. 734 [EIR inadequate for failing to address off-site impacts of a project].) Appellants argue that sufficient information about the Berube property was provided with the errata, shortly before the Board meeting. This documentation, however, does not make up for the lack of analysis in the EIR. (See Environmental Defense Fund, Inc. v. Coastside County Water Dist. (1972) 27 Cal.App.3d 695, 706 [104 Cal.Rptr. 197].) As county counsel conceded at trial, there was no discussion in the EIR of the impacts of *131 transferring water credits “because the issue of the water transfer came towards the end of the process.” If, subsequent to the period of public and interagency review, the lead agency adds “significant new information” to an EIR, the agency must issue new notice and must “recirculate” the revised EIR, or portions thereof, for additional commentary and consultation. (Pub. Resources Code, § 21092.1; Guidelines, § 15088.5, subd. (a); Laurel Heights Improvement Assn. v. Regents of University of California, supra, 6 Cal.4th 1112.) The revised environmental document must be subjected to the same “ ‘critical evaluation that occurs in the draft stage,’ ” so that the public is not denied an “ ‘ “opportunity to test, assess, and evaluate the data and make an informed judgment as to the validity of the conclusions to be drawn therefrom.” ’ ” (Sutter Sensible Planning, Inc. v. Board of Supervisors, supra, 122 Cal.App.3d 813, 822.) In light of the atmosphere of public concern about the water shortage in the Carmel Valley, and the focused concerns expressed in the comments calling for an analysis of the feasibility of any specific offset pumping site to provide actual mitigation, we believe the identification of the Berube parcel late in the environmental review process warranted further discussion and analysis and an opportunity for public response. Although the Board may exercise its discretion as to the viability of a policy allowing for off-site water credits as mitigation for increased pumping in the valley, and as to the feasibility of the Berube property in particular for this purpose, it must do so on the basis of information collected and presented in the EIR and subjected to the test of public scrutiny. A revised EIR must include a discussion of the Berube parcel, the history of water pumping on this property and its feasibility for providing an actual offset for increased pumping on the September Ranch property, as well as the growth-inducing effect of a policy of offset pumping reduction in the Carmel Valley. Riparian Rights The issue whether the September Ranch property had valid riparian rights and could utilize them to support a private water system for the subdivision also arose late in the environmental review process and suffers from a similar lack of analysis. During the second period for circulation and comment the SWRCB wrote that the applicants would need an appropriative rights permit to pump water because “the alluvium underlying the September Ranch is part of the Carmel River subterranean stream.” The applicants then asserted for the first time in a letter dated May 2, 1998, that the property had a riparian right, which ran with the land and entitled them to use water from the subterranean stream without an appropriative permit. Neither the draft *132 EIR nor the revised EIR had mentioned such a right. The SWRCB responded that a valid riparian right could be utilized for project purposes, if such a right existed, but that no determination had yet been made as to such a right. The supplemental EIR (vol. 2) added a discussion of riparian rights. A valid riparian right can be established if: 1) the property is contiguous to the water course; 2) the property is within the watershed of the water course; and 3) the riparian right has not been severed through subdivision or separate conveyance. The supplemental EIR concluded that the September Ranch was “at least partially contiguous to the water course,” namely the Carmel River subterranean stream flow, and that the property was located within the Carmel River watershed. A title search indicated, and county counsel later confirmed, that the 891-acre September Ranch was a single lot of record. Thus there had been no severing of riparian rights. An early deed showed, however, that September Ranch’s riparian rights may have been subordinated to a predecessor utility of Cal-Am. The supplemental EIR reported that riparian rights entitle the owner to use “the amount of water that can be reasonably and beneficially used on the riparian parcel” without applying for a permit. In times of shortage a riparian owner must share water with other riparian users, but its rights are superior to the rights of appropriators. The supplemental EIR clarified that whether the water right was riparian or appropriative, any increase of water use over preproject use would be a significant environmental impact requiring mitigation. In the final changes and corrections to the EIR, mitigation measure 7b was added, which required “either the assurance of a valid riparian claim or the requirement that the applicants secure a permit for an appropriative water right from the State Water Resources Control Board.” But this mitigation measure was not included in the conditions of approval in the Board’s resolution certifying the EIR. The trial court pointed out numerous factual and legal issues, as well as policy concerns, that the court believed remained to be resolved before any determination could be made that the property owners have riparian rights sufficient to guarantee a long-term water supply for this project. Even if a riparian right were established, the court found that the approval of a private water system for a large subdivision, based on a subterranean riparian right under only one portion of the property, could set an undesirable precedent and have a growth-inducing effect. This, the court found, was a potential cumulative impact which should have been considered and discussed in the EIR. The court concluded that “the failure of the EIR to consider potential growth inducing and/or other cumulative impacts of the use of alleged *133 subterranean riparian rights” was error. Consequently, the Board’s findings approving a long-term water supply for the project, to the extent those findings were based on the existence of valid subterranean riparian rights, were not supported by substantial evidence. The judgment granting the writ of mandate directed the preparation of an EIR that properly analyzed whether water rights existed for the project. Appellants argue that the court erred in ordering that the EIR analyze the legalities of their riparian water rights, contending that CEQA does not require any such analysis. Appellants maintain that as a matter of water law, their land has riparian rights to the subterranean streamflow without having to obtain a permit. Furthermore, they argue, the EIR explained that whether the water use is based on an appropriative right or a riparian right, the physical impact is still the same. In either case if the project’s water use exceeds the preproject use, mitigation is required. Finally, they claim that the petitioners in this case waived any water rights claims by failing to brief them before the trial court. First, there is no basis for finding that petitioners in this case waived claims regarding water rights issues. These issues were adequately raised in briefing and argument before the trial court. Any failure to fully develop arguments can be attributed in part to the fact that the applicants asserted their intent to utilize their riparian rights very late in the environmental review process. As in the previous section, the late introduction of this theory and new information resulted in an incomplete analysis in the EIR. Furthermore, there was no opportunity for meaningful public comment and response. “The purpose of requiring public review is ‘ “ ‘to demonstrate to an apprehensive citizenry that the agency has, in fact, analyzed and considered the ecological implications of its action.’ ” . . .’ . . . Public review permits accountability and ‘ “informed self-government.” ’ . . . ‘[Pjublic review and comment. . . ensures that appropriate alternatives and mitigation measures are considered, and permits input from agencies with expertise’ . . . Thus[,] public review provides the dual purpose of bolstering the public’s confidence in the agency’s decision and providing the agency with information from a variety of experts and sources.” (Schoen v. Department of Forestry & Fire Protection (1997) 58 Cal.App.4th 556, 573-574 [68 Cal.Rptr.2d 343], citations omitted.) The primary reason that public comment is solicited is so that potential significant adverse effects of the project can be identified “at the earliest possible time.” (Laurel Heights Improvement Assn. v. Regents of University of California, supra, 6 Cal.4th at p. 1129.) The requirement in *134 Public Resources Code section 21092.1 that an EIR be recirculated when “significant new information” is added is not intended “to promote endless rounds of revision and recirculation of EIR’s. Recirculation is intended to be [the] exception, rather than the general rule.” (Laurel Heights Improvement Assn. v. Regents of University of California, supra, 6 Cal.4th at p. 1132.) We believe the exception applies in all of the circumstances of this case. The supplemental EIR presented new and significant information regarding the applicants’ asserted riparian right, which raised important water issue questions. If the validity of such a right were determined, would this entitle the applicants to rights superior to those of appropriative water users? How would these rights be superior? How would this affect other riparian water users in the area during times of drought? If the exercise of a riparian right would not require a permit, but would be subject only to a rule of “reasonable use,” how is water use regulated and controlled? Can a riparian right underlying one portion of the property be the basis for a private mutual water company providing water to the entire subdivision? Does the exercise of such a right create a precedent for other subdivisions and thus result in a growth-inducing impact? Is the exercise of a riparian right, which may justify an expanded use of water, consistent with local policies limiting water for new development? Were further mitigation measures warranted? For example, the supplemental EIR added a mitigation measure requiring that the applicants either provide assurance of a valid riparian claim or secure an appropriative permit from the SWRCB. The fact that this mitigation measure was not carried over into the Board’s final resolution only illustrates the difficulties presented by adding significant changes late in the EIR process. In sum, we believe the addition of this new information regarding the asserted riparian right as a basis for long-term water supply for this project changed the EIR “in a way that deprive[d] the public of a meaningful opportunity to comment upon a substantial adverse environmental effect of the project or a feasible way to mitigate or avoid such an effect.” (Laurel Heights Improvement Assn. Inc. v. Regents of University of California, supra, 6 Cal.4th at pp. 1129-1130; Sierra Club v. Gilroy City Council (1990) 222 Cal.App.3d 30 [271 Cal.Rptr. 393].) We agree with appellants that the final decision determining county policy on this issue is a matter of the Board’s discretion. However, the EIR must provide sufficient information to make the exercise of this discretion an informed one. *135 Traffic Issues Traffic issues center around the EIR recommending, and the Board adopting, the payment by the applicants of in-lieu fees into county traffic impact fee programs as mitigation for traffic increases attributed to the project. The Carmel Valley Road traffic impact fee program is designed to respond to cumulative growth in traffic by generating the funds needed for construction of improvements along Carmel Valley Road. The road is divided into segments with assigned traffic thresholds. Projected traffic increases that will cause a threshold to be crossed trigger the need for improvements designed to return the segment to an acceptable level of service. The fee impact program thus enables the County to collect fees and add roadway improvements as new development increases traffic to unacceptable levels. The traffic analysis in the draft EIR indicated that on two segments of Carmel Valley Road, segments 6 and 7, the projected traffic increase from the September Ranch project, plus traffic from already approved projects, would exceed the threshold, thus triggering the need for improvements. As to segment 7, which included the frontage along the September Ranch property, the threshold would be exceeded with existing traffic and projected traffic from projects already approved but not yet built out. The draft found that the traffic increase over the threshold was a significant impact, which could be reduced through the implementation of Carmel Valley Road improvements. As mitigation, the project applicants would be required to pay fees to the County, as established in the traffic impact fee program for Carmel Valley Road. The Carmel Valley Road traffic impact fees imposed on the project were based on a traffic impact fee ordinance adopted by the Board in 1992. The fee program was enacted to enable the County to fund improvements to Carmel Valley Road on a “pay-as-you-go basis” and to avoid a moratorium affecting development within the Carmel Valley area. Prior to the issuance of any building permit, a traffic mitigation fee was to be paid into a separate interest-bearing account, to be used “for road and street improvements to Carmel Valley Road generally consistent with the Carmel Valley Master Plan . . . .” In a 1995 resolution the County adopted a traffic mitigation fee schedule for all new development along Carmel Valley Road. New development was to be assessed $16,000 per unit, plus annual increases tied to the construction cost index. The traffic mitigation program calls for regular monitoring of Carmel Valley Road traffic conditions to determine when *136 traffic thresholds along the various segments are reached. The draft EIR found that it was up to the County “to determine the nature and timing of the required improvements to Carmel Valley Road.” A second problem area for traffic involved the intersections along Highway 1 in the vicinity of Carmel Valley Road. The draft EIR found that the level of service at several of these intersections was currently substandard during peak hours. The County, in conjunction with the California Department of Transportation (CalTrans), had prepared a program of interim improvements to address these deficiencies. According to one study, these operational improvements were designed to maintain an acceptable level of service or better at four intersections along Highway 1 and to support a 27 percent growth in peak hour traffic. The EIR found that unless these proposed interim improvements to Highway 1 were implemented, the traffic increase from this project and other approved projects in the area would “exacerbate unacceptable levels of service of roadways and intersections in the vicinity of Carmel Valley Road and Highway 1 . ...” As mitigation, the project applicants were to pay to the County, prior to the issuance of building permits, a pro rata share toward the cost of 12 interim Highway 1 improvements. The draft further found, however, that cumulative impacts would eventually require long-range solutions, such as the proposed Hatton Canyon Freeway or the widening of Highway 1. The final EIR included updated traffic counts, which did not change the statistics significantly. The previous conclusions regarding the two segments of Carmel Valley Road were still valid. Recommended mitigation, as before, involved the payment of fees to the County pursuant to its traffic impact fee program. The intersections along Highway 1 continued to operate at unacceptable levels. Comments from CalTrans expressed “great concerns” over the project generating additional traffic along Highway 1, a corridor that already operated at an unacceptable level of service. According to CalTrans, the level of service in that area was not likely to improve significantly until the Hatton Canyon Freeway was built. CalTrans urged that the September Ranch project not be approved until this freeway was completed. 9 The EIR’s response to these comments indicated that interim improvements would provide short-term congestion relief pending the construction of the Hatton Canyon Freeway. The EIR provided further that as the decisionmaking body *137 “it is up to the Board of Supervisors to decide when the improvements are scheduled to be completed.” The final EIR noted that the Board and the Transportation Agency for Monterey County had developed a “Deficiency Plan” calling for 12 operational improvements along Highway 1. The EIR acknowledged that the additional traffic generated by the September Ranch project would cause a significant impact on traffic volumes at these intersections unless the proposed interim improvements to Highway 1 were in place. State funding for these improvements was to be supplemented with county funds pursuant to the traffic impact fee program. The final EIR recommended that traffic impacts be mitigated by payment by the developer of a pro rata share of the 12 interim improvements to Highway 1 prior to the issuance of building permits. The Board adopted these fee payment mitigation measures as conditions of approval and also required that the applicants install various circulation improvements on Carmel Valley Road at the entrance to the project, provide a safe transit stop convenient to the entrance, dedicate a right-of-way for future widening of the road, and implement a trip-reduction program. The Board determined that because of the delay in the construction of the Hatton Canyon Freeway, the 12 interim improvements in the vicinity of Carmel Valley Road and Highway 1 would be implemented and would be funded through collection of Carmel Valley Road traffic impact fees to supplement CalTrans funds. In addition, the Board determined that the project would be phased so that no more than 50 lots could be developed prior to the completion of Highway 1 interim road improvement No. 5, “or another traffic solution for Highway 1 is approved.” Improvement No. 5 was the planned construction of dual right-turn lanes onto Highway 1. Petitioners argued that the mitigation proposed by the EIR and adopted by the Board was inadequate in that the in-lieu fees did not readily translate into actual improvements. They contended that the fees were not likely to result in improvements, considering that the traffic problems were long standing and that the County had failed to act to implement improvements in the past, despite assurances that new projects would not be approved unless the infrastructure was in place to support such projects. Furthermore, allowing the County to determine “the nature and timing” of the improvements was no guarantee that the fees would go to the improvements needed in the areas where the project caused significant impacts. Petitioners argued that the EIR failed as an informational document because it failed to tie the fee mitigation plan to the actual physical impacts of the *138 project on the environment. They claimed the EIR mitigation plan must identify the nature of specific improvements and their timing and how the improvements would mitigate the impact of the increased traffic. And finally they claimed that the Board’s approval of the project with the adoption of these mitigation measures created an inconsistency with the traffic policy in the Master Plan. The trial court agreed with these arguments. The court acknowledged that in-lieu fees are appropriate in some cases, but reasoned that after the critical threshold is reached or surpassed and the improvements have still not been implemented such fees are no longer adequate mitigation. The court focussed on the County’s previous interpretation of policy No. 39.1.6 of the Master Plan, as represented by county counsel in prior litigation involving the Master Plan. Policy No. 39.1.6 of the Master Plan, adopted in 1986, provides that “[ejvery effort should be made to obtain funding and proceed with construction of the Hatton Canyon Freeway at the earliest possible date.” However, if after five years of allocation the freeway has not been built, “the Board shall limit further development until the freeway is under construction.” In litigation challenging the approval of the Master Plan, county counsel represented that this policy meant that “ ‘if . . . the infrastructure is not available to support growth, growth will not be permitted.’ ” Specifically, if the Hatton Canyon Freeway were not funded and other mitigation measures were not implemented the County’s alternative would be “ ‘not to approve development unless there is infrastructure to support it.’ ” The trial court noted that 12 years had passed since the approval of the Master Plan and that the time for “action, not words” had come. The court concluded: “With respect to the intersection of Highway One and the other two segments of Carmel Valley Road which have reached the ‘threshold’ trigger, the EIR should have specifically considered when in fact the improvements are to be done and whether that time period is feasible. The County should have made specific findings as to whether they are going to be done and when. If the improvements are not to be done in the immediate future, then, in accordance with the [Master Plan], development must be limited or action taken to amend the plan.” Appellants argue that the EIR’s traffic analysis and mitigation measures complied with CEQA, that substantial evidence supported the Board’s conclusion that traffic impacts would be mitigated, and that the Board’s interpretation of Master Plan policy No. 39.1.6 was within its discretion and was reasonable. We agree with appellants. *139 First, we restate our standard of review here. Our task is to determine whether the agency prejudicially abused its discretion either by not proceeding in the manner required by law or by making a decision not supported by substantial evidence. (Pub. Resources Code, § 21168.5; Laurel Heights Improvement Assn. v. Regents of University of California, supra, 47 Cal.3d at p. 392.) We presume the correctness of the agency’s decision and the petitioners thus bear the burden of proving that the EIR is legally inadequate or that the record does not contain substantial evidence to support the agency’s decision. (Al Larson Boat Shop, Inc. v. Board of Harbor Commissioners, supra, 18 Cal.App.4th atp. 740; Barthelemy v. Chino Basin Mun. Water Dist., supra, 38 Cal.App.4th at p. 1617.) The substantial evidence rule does not require certainty; substantial evidence is “enough relevant information and reasonable inferences from this information that a fair argument can be made to support a conclusion, even though other conclusions might also be reached.” (Guidelines, § 15384, subd. (a).) Where the dispute is whether adverse affects could be better mitigated, we do not weigh the evidence and determine who has the better argument. (Laurel Heights Improvement Assn, v. Regents of University of California, supra, 47 Cal.3d at p. 392-393.) “We have neither the resources nor scientific expertise to engage in such analysis, even if the statutorily prescribed standard of review permitted us to do so.” (Id. at p. 393.) CEQA requires that an EIR indicate the ways in which a project’s significant effects can be mitigated, by setting forth “[mjitigation measures proposed to minimize significant effects on the environment.” (Pub. Resources Code, §§21100, subd. (b)(3), 21002.1, subd. (a), 21061.) The discussion should identify mitigation measures which “could reasonably be expected to reduce adverse impacts if required as conditions of approving the project.” (Guidelines, former § 15126, subd. (c), now § 15126.4, subd. (a)(1)(A).) We believe the EIR adequately fulfilled these requirements. It contained a comprehensive traffic analysis that compared the total projected traffic from this project, and from other projects in the area that were approved but not built, against an established capacity threshold for each road segment along Carmel Valley Road and the intersections with Highway 1. It identified problem areas and described the programs designed to address these areas of concern. And it recommended mitigation in the form of pro rata fees paid to a traffic impact fee program established by county ordinance and designed to implement road improvements as needed. Further recommended mitigation was construction of safe transit stops, implementation of a trip reduction program, installation of circulation improvements at the entrances to the project site, and dedication of a right-of-way for the widening of Carmel Valley Road. *140 Fee-based infrastructure mitigation programs have been found to be adequate mitigation measures under CEQA. (See, e.g., Russ Bldg. Partnership v. City and County of San Francisco (1988) 44 Cal.3d 839, 845 [244 Cal.Rptr. 682, 750 P.2d 324] [upholding transit impact development fee]; San Franciscans for Reasonable Growth v. City and County of San Francisco (1989) 209 Cal.App.3d 1502 [258 Cal.Rptr. 267].) The CEQA Guidelines also recognize that when an impact is not unique to a single project, but is instead the result of cumulative conditions, the only feasible mitigation may involve adoption of ordinances or other regulations designed to address the cumulative impact. (Guidelines, § 15130, subd. (c).) Section 15130 of the Guidelines now specifically provides that an EIR may determine that a project’s contribution to a cumulative impact may be mitigated by requiring the project “to implement or fund its fair share of a mitigation measure or measures designed to alleviate the cumulative impact.” (Guidelines, § 15130, subd. (a)(3).) The trial court recognized that the payment of fees and phased improvements was appropriate, at least with respect to traffic impacts which have not yet reached the threshold trigger. Of course a commitment to pay fees without any evidence that mitigation will actually occur is inadequate. (Kings County Farm Bureau v. City of Hanford (1990) 221 Cal.App.3d 692 [270 Cal.Rptr. 650].) In the City of Hanford case, the city had found that certain impacts on groundwater were insignificant, in reliance on a “mitigation agreement” with the water district by which the project applicant agreed to pay the district to purchase water supplies to make up for amounts used by the project. However, the record contained no evidence indicating that any such water supplies were or would be available. Consequently, the developer’s promise to pay the fees bore no connection to actual mitigation of impacts. The court found that the EIR was inadequate in this respect Here, however, the collection of fees was not an idle act. The EIR reported that the County had adopted the traffic impact fee program in order to fund improvements to Carmel Valley Road. A citizens advisory committee, the Carmel Valley Road Improvement Committee, had studied potential road improvements and had reported to the Board. Studies in the EIR indicated that existing traffic levels at all segments along Carmel Valley Road were below the threshold at the time the EIR was completed. Therefore, the requirement for improvements to bring the service back to an acceptable level had not yet been triggered. However, traffic projected from projects already approved but not yet built would exceed the threshold on segment 7. And both segments 6 and 7 would be exceeded when all approved projects plus the September Ranch project were built out. Planned *141 improvements included intersection channelization and passing lanes on segments 6 and 7, the two segments most affected by the project in this case. As to the intersections along Highway 1, where the level of service was unacceptable at peak hours, the EIR reported that the County had adopted, and the Monterey County Transportation Agency had endorsed, a deficiency plan to resolve congestion problems. Twelve interim improvements were proposed. At the time of the final EIR one of the scheduled improvements had been completed and another, improvement No. 5, which was specifically identified in the Board’s resolution, was funded and scheduled for construction. Thus with respect to the problem areas for traffic identified in the EIR, the evidence indicated that road improvement plans were in place and in some cases construction was proceeding. A time schedule for improvement was inherent in the County’s traffic impact program, in that it provided for improvements to be constructed as the traffic triggering the need for the improvements exceeded a projected threshold and the funds to pay for the improvements were generated by the new development. We are not unsympathetic to concerns, voiced by the trial court, about the County’s failure to act in the past to implement road improvements. We do not believe, however, that CEQA requires that the EIR set forth a time-specific schedule for the County to complete specified road improvements. All that is required by CEQA is that there be a reasonable plan for mitigation. (Sacramento Old City Assn. v. City Council (1991) 229 Cal.App.3d 1011 [280 Cal.Rptr. 478]; see also Laurel Heights Improvement Assn. v. Regents of the University of California, supra, 47 Cal.3d 376, 418.) Furthermore, we must presume and expect that the County will comply with its own ordinances, and spend the fees it collects on the appropriate improvements to the affected road segments. (See, e.g., Erven v. Board of Supervisors (1975) 53 Cal.App.3d 1004, 1012 [126 Cal.Rptr. 285].) On this record we find that the traffic impact mitigation fees were sufficiently tied to the actual mitigation of the impacts of increased traffic. We therefore conclude that the EIR’s discussion of traffic mitigation measures was adequate and the Board’s adoption of the conditions of approval was supported by the evidence. Furthermore, we find that the Board’s determination that the project was consistent with policy No. 39.1.6 of the Master Plan was not an abuse of discretion. The relevant portion of the policy stated that the Board “shall limit further development” until the Hatton Canyon Freeway was under construction. The EIR did not find an inconsistency with this policy *142 because interim improvements were planned to maintain an acceptable level of service pending the construction of the Hatton Canyon Freeway, oranother long-term plan, and because the policy required only that further development be limited, not that it was prohibited. The Board’s resolution did in fact provide limitations, requiring that development of the project be phased to coincide with completion of identified interim improvements. When we review an agency’s decision for consistency with its own general plan, we accord great deference to the agency’s determination. This is because the body which adopted the general plan policies in its legislative capacity has unique competence to interpret those policies when applying them in its adjudicatory capacity. (City of Walnut Creek v. County of Contra Costa (1980) 101 Cal.App.3d 1012, 1021 [162 Cal.Rptr. 224].) Because policies in a general plan reflect a range of competing interests, the governmental agency must be allowed to weigh and balance the plan’s policies when applying them, and it has broad discretion to construe its policies in light of the plan’s purposes. (Sequoyah Hills Homeowners Assn. v. City of Oakland (1993) 23 Cal.App.4th 704 [29 Cal.Rptr.2d 182]; Greenebaum v. City of Los Angeles (1984) 153 Cal.App.3d 391, 407 [200 Cal.Rptr. 237].) A reviewing court’s role “is simply to decide whether the city officials considered the applicable policies and the extent to which the proposed project conforms with those policies.” (Sequoyah Hills Homeowners Assn. v. City of Oakland, supra, 23 Cal.App.4th at pp. 719-720.) Here, the EIR discussed the Master Plan, including policy No. 39.1.6, and the Board expressly found that the project was consistent with that policy. We find no abuse of discretion. The purpose of policy No. 39.1.6. was to prevent unacceptable increases in congestion at the intersection of Highway 1 and Carmel Valley Road due to new development until a long-term plan such as the Hatton Canyon Freeway could be implemented. Notwithstanding the representations of counsel during litigation in 1987, the policy did not prohibit all further development until the freeway was built. We believe the Board was entitled to exercise its discretion to determine what limitations were appropriate in light of its review of current levels of service, approved development and planned interim improvements. 10 *143 Disposition The judgment granting a peremptory writ of mandate is reversed in part and affirmed in part. The matter is remanded to the superior court with directions that the court issue a new writ of mandate ordering the Board to vacate resolution No. 98-500, including the approval of any permits or entitlements for the project described in that resolution, and to vacate the certification of the EIR prepared in regard to the project. The Board shall be ordered not to take any further action to approve the project without the preparation, circulation and consideration under CEQA of a legally adequate EIR with regard to the water issues discussed in this opinion. The revised EIR is to investigate and analyze the baseline water conditions on the property at or around the time of the commencement of the environmental review process for this project. Baseline water figures shall reflect actual water use on the property, where possible, and methodologies for determining baseline shall be supported by evidence of actual water use on the property or, where no documentation is available, by good faith estimates of actual historical use. The revised EIR is to discuss and analyze the growth-inducing impact of mitigating increased pumping over baseline with off-site pumping reduction, including the loss of agricultural lands, and specifically the feasibility of a pumping offset on the Berube parcel, including water availability and pumping history on the Berube parcel and whether there is an actual nexus between reduced pumping on that property and increased pumping on the September Ranch property. The revised EIR is to discuss and analyze the asserted riparian right of the applicants, including whether such a right has been established, whether it entitles the applicants to an expanded use of water in derogation of the rights of other water users in the area, whether such a right may support a mutual water system serving the entire subdivision, and whether the utilization of riparian rights may result in a growth-inducing impact. The portion of the superior court’s judgment granting a writ of mandate and directing that the Board prepare • a revised EIR to include further discussion regarding mitigation of traffic impacts is reversed. The superior court’s order awarding attorney fees is hereby vacated. Upon remand, the court may issue a new order, in light of our disposition herein, or may reinstate the same order. *144 The parties are to bear their own costs on appeal. Premo, Acting P. J., and Wunderlich, 1, concurred. Footnotes [1] California Environmental Quality Act (CEQA), Public Resources Code section 21000 et seq. [2] This discussion is confined to water issues. We will include the background of the traffic issues in the discussion in that section. [3] Two parties in this action, Ed Leeper and Save Our Peninsula Committee, were dismissed following a demurrer sustained without leave to amend. The remaining petitioner, Responsible Consumers of the Monterey Peninsula, is still a party and is the respondent in appeal No. H020900. [4] As to the asserted changes made to the Board’s findings after the Board had adjourned, the trial court noted that the record revealed “numerous instances” where the applicants’ attorney had prepared critical documents for county planners. The court disapproved such a practice and pointed out that the County had indicated it had “recognized the problem and taken appropriate action.” [5] The two petitions were consolidated only for administrative purposes at trial. Therefore, two separate appeals were filed. The two appeals have been consolidated here for the limited purposes of filing the administrative record, oral argument and decision. [6] The CEQA Guidelines are found at California Code of Regulations, title 14, section 15000 et seq. (hereafter Guidelines). [7] This same language now appears in Guidelines section 15126.4, subdivision (a)(1)(D). [8] This language now appears in Guidelines section 15126.2, subdivision (d). [9] The Hatton Canyon Freeway has not gone forward due to local opposition. At oral argument, respondents represented that state funding for this project has been diverted to other uses. [10] Respondents have raised several further arguments challenging other aspects of the EIR and the Board’s action. The trial court determined that its judgment granting a peremptory writ of mandate mooted any additional challenges, which could be raised again depending on the Board’s action on remand. Respondents have not cross-appealed and these further issues are not before us at this time.