*3Opinion
BLEASE, J.We are called upon a second time to determine whether the City of Los Angeles and its department of water and power have complied with the California Environmental Quality Act and the writ of mandate issued by this court in 1973 (County of Inyo v. Yorty (1973) 32 Cal.App.3d 795, 814-816 [108 Cal.Rptr. 377]) directing them to prepare an environmental impact report (EIR) covering their extraction of subsurface water in the Owens Valley. In 1977 this court said that the city’s EIR, submitted in its return to the writ, did not comply with the writ because it failed to provide an accurate, stable and finite project description in accordance with the 1973 decision and also failed to describe reasonable alternatives to the project. (County of Inyo v. City of Los Angeles (1977) 71 Cal.App.3d 185 [139 Cal.Rptr. 396].) The city has submitted a second EIR by way of a return to the writ, to which the county objects inter alia on grounds paralleling those advanced by this court in rejecting the legal sufficiency of the first EIR. We shall sustain the county’s objection on the grounds set forth below.
I
These protracted proceeedings are chronicled in County of Inyo v. Yorty, supra, 32 Cal.App.3d 795; County of Inyo v. City of Los Angeles (1976) 61 Cal.App.3d 91 [132 Cal.Rptr. 167]; County of Inyo v. City of Los Angeles, supra, 71 Cal.App.3d 185; and County of Inyo v. City of Los Angeles (1978) 78 Cal.App.3d 82 [144 Cal.Rptr. 71], to which readers must repair for a complete understanding of this decision. We emphasize here those portions of the opinions necessary to this decision.
Beginning in 1900 the city acquired land and water rights in the Owens Valley which “at the present time [aggregate] 300,000 acres in Inyo and Mono Counties, comprising roughly 97 percent of the available privately held land.” (32 Cal.App.3d at p. 799.)
In 1913 the city completed a surface aqueduct, the first aqueduct, between the Owens Valley and the city and began receiving Owens Valley water. The capacity of the aqueduct is a constant average flow of 430 cubic feet per second (cfs).
*4At a relatively early date, as an auxiliary to the natural precipitation in the area, the city commenced the drilling of a large number of wells to tap the subsurface pools of underground water in Owens Valley. These wells were heavily used during dry years to assure, as a supplementary source, a continuous and adequate flow through the first aqueduct. (32 Cal.App.3d at p. 799.)
In 1959 the city proposed “the construction of a second aqueduct to carry water from Owens Valley to the City for the purpose of completing ‘the development of our Inyo-Mono supply that began over fifty years ago’ and to insure that City would not lose certain water rights on which it had filed primarily in the Mono Basin development. [H] Historically, the first aqueduct conveyed most of the available surface runoff from Owens Valley. It was planned that the second aqueduct would be filled largely from three sources: increased surface diversion from Mono Basin, reduced irrigation of City-owned lands in the affected counties and increased pumping of groundwater reservoirs in Owens Valley.” (32 Cal.App.3d at p. 800.)
The design capacity of the two aqueducts is a constant average flow of 666 cfs, amounting to an increase in capacity occasioned by the second aqueduct of 236 cfs.
This action began in 1972 when the county sought injunction against the extraction of subsurface water for export from the Owens Valley, against the utilization of subsurface water in place of surface water within Inyo County and to require an EIR. “At that stage of the litigation, the city insisted that exportation of increased groundwater was an inseparable part of its second aqueduct, an ‘ongoing project’ completed prior to the effective date of CEQA and thus immune from the demands of CEQA.” (71 Cal.App.3d at p. 193.)
We rejected the argument concluding, in sum, that “‘while the capacity of the second aqueduct was fixed and known for a number of years before CEQA, the effect of its construction on subsurface water extraction has been a variable but steady escalation, dependent in large part, no doubt, upon the extent of seasonal rain and snowfall from year to year. Thus the ecological impact of the second aqueduct, viewed in conjunction with the underground pumping and measured by the quantity of extraction, has not been fixed but has substantially increased in severity in the period before, during and after its construction.... [IT] We conclude from the foregoing that the legislative intent so strongly ex*5pressed in CEQA can be met only by considering the expanded groundwater extraction as a “project” separate and divisible from the second aqueduct, and we so treat it.’ (32 Cal.App.3d at p. 806.)” (71 Cal.App.3d at pp. 194-195.)
In its first EIR (May 1976) the city seized upon the “last quoted statement” to excise the part of the increased pumping rate, which it claimed was destined for export to the city (89 cfs) from “the CEQAsubject side of the line and place[d] it on the exempt side of the line” thus creating a CEQA-subject project of pumping which, “by a process of verbal transmutation, will now be devoted to in-valley use and not exported at all.” (71 Cal.App.3d at p. 195.) We said that “the impermissibly truncated” project definition severely distorted not only the actual project but the alternatives to the project. (71 Cal.App.3d at p. 201.)
We also said that the first EIR impermissibly created a no-project baseline for measurement of alternatives to the project different than the conditions preexisting the project. We concluded that “[bjecause the final EIR does not include a genuine ‘no project’ alternative, because its list of alternatives is not tied to a reasonably conceived or consistently viewed project, the Los Angeles EIR does not comply with CEQA’s demand for meaningful alternatives. This lack results in an EIR which does not meet CEQA’s goal of ensuring that ‘the long-term protection of the environment shall be the guiding criterion in public decisions.’ (§ 21001, subd. (d).)” (71 Cal.App.3d at p. 203.)
II
The Writ
This litigation began as pleadings, unaltered following the 1973 decision of this court, which define the scope of the action as the continued extraction of subsurface waters from the Owens Valley. (County of Inyo, supra, 71 Cal.App.3d at pp. 190, fn. 4, 193.) The scope of this court’s 1973 mandate was fashioned in view of the issues tendered.
This court has labored to make clear the requirements of its 1973 writ of mandate. In 1977 we explicated the 1973 decision in these words: “In any objective view the outlines of the ‘project’ conceived by our 1973 decision were quite clear. They were clear in 1973 and they *6are clear now. Unfortunately there is a limit to the precision of words. Judicial opinion writers cannot always armor their language against wishful misinterpretation. At the risk of future misinterpretation we shall attempt the following reformulation of our 1973 formulation: The project which forms both the scope of this litigation and the subject of the EIR mandated by this court is the department of water and power’s program for increasing the average rate of groundwater extraction and use (both for export and in-valley use) above a baseline rate reasonably representing the average rate of groundwater extraction and use (both for export and in-valley use) preceding the second aqueduct’s availability for use.” (Fn. omitted.) (71 Cal.App.3d at p. 196.)
In similar measure, we set out the “baseline” for consideration of alternatives to the project. We said: “In order to mark out a ‘no project alternative,’ the EIR should describe what condition or program preceded the project.” (Italics added.) (71 Cal.App.3d at p. 201.) Thus, we held that for purposes of defining the project and measuring its impact and “‘no project alternative,”’ a baseline representing the conditions upreceding the second aqueduct’s availability for use” must be utilized. (Italics added.) {Id., at p. 196.)
We turn to examine the EIR’s compliance with this mandate.
Ill
The Project Description
The resolution of the Board of Water and Power Commission of the City of Los Angeles, dated June 28, 1979, approving the final environmental impact report (FEIR) and approving the project reviewed by the EIR, describes the project as follows: “The Project, which is described in the EIR and is the subject of this Resolution, is to increase the pumping from the Owens Valley Groundwater Basin beyond historic pumping rates (long-term and maximum annual average): to supply water for use within the City of Los Angeles; to make the water available for agriculture, recreation, wildlife, and other uses on City lands within Inyo and Mono Counties; and for other uses in Owens Valley. The Project area is the Owens Valley Groundwater Basin in Inyo County.”
Although an avowed purpose of the project is “to make .. . water available for agriculture, . .. and other uses . .. within [both] Inyo and Mono Counties,” the resolution limits the project area to the ground*7water basin in Inyo County and1 asserts that “the diversion, export and occasional use of surface waters on City-owned lands in Inyo and Mono Counties is not a part of the Project.”2 (Italics added.)
The FEIR states that project increase in groundwater pumping will amount to 145 cfs “on an average annual basis,” of which 66 cfs will be exported to city and 79 cfs “would be supplied directly or indirectly to uses in the Owens Valley.”3 There is then an asterisk and accompanying footnote which explicates “indirectly” as “the pumped water designated for irrigation or other use will flow into the Aqueduct and an equal amount of surface water will be diverted from a stream or ditch convenient to the location of use.”4
Counsel for the city conceded on oral argument that as a physical matter virtually all of the additional pumped water will flow into the aqueducts and to Los Angeles.5 Thus, the water to be “made available” for in-valley irrigation use is surface water. Yet the project definition *8excludes such water from the FEIR and consequently there is no express discussion whether surface water would in fact be available for in-valley uses.
There is considerable doubt on the matter. As we pointed out in an earlier opinion: “Historically, the first aqueduct conveyed most of the available surface runoff from Owens Valley. It was planned that the second aqueduct would be filled largely from three sources: increased surface diversion from Mono Basin, reduced irrigation of City-owned lands in the affected counties and increased pumping of groundwater reservoirs in Owens Valley.” (32 Cal.App.3d at p. 800.) We were there speaking of an earlier plan. But that plan must have considered the physical realities of surface water availability.
We are told that the second aqueduct increases the capacity for export from 430 cfs to 666 cfs or 236 cfs and also that the projected annual average increase in pumping occasioned by the project is 145 cfs. If we assume that all of this pumped water is to be exported to Los Angeles, which, indeed, it largely is, as a physical matter, there still must be found an additional 91 cfs of surface water to meet the 666 cfs total export requirement, wholly without regard to in-valley use.
We are not told whether or how this deficiency is to be made up.6 In fact, an examination of table Dl-6 of FEIR, volume I, which displays the long-term surface runoff figures for Owens Valley, shows that the entire historical flow would have to be diverted in order to meet project demands for surface water (average demand 611 cfs). This is consistent with the declaration of Duane L. Georgeson, one of the city’s aqueduct engineers, that: “From earliest planning to the present day utilization of increased ground water in the Owens Valley was a substantial part of the water required to fill the Second Barrel. Since the First Los Angeles Aqueduct conveys to the City of Los Angeles most of the available surface [runoff] from the Owens Valley, the Second Los Angeles [Aqueduct] had to be filled largely by other means.”
*9An EIR may not define a purpose for a project and then remove from consideration those matters necessary to the assessment whether the purpose can be achieved, Since the FEIR removes the availability of surface water from examination it fails the legal duty and the mandate of this court to provide an informed and accurate analysis of the project and its impacts.
The deficiencies of the project description also show in the EIR’s alternatives to the project, a subject to which we now turn.
IV
Pre-project and No-project Alternative
As we have said, “[a]n accurate, stable and finite project description is the sine qua non of an informative and legally sufficient EIR.” As a corollary to this requirement, the project must be compared with its pre-project conditions in order, inter alia, to provide a uniform baseline for the measurement of its impact and to “assess the advantage of terminating the proposal.” (71 Cal.App.3d at p. 193.) This is called a “no-project” alternative and is required by law. (71 Cal.App.3d at p. 203; Cal. Admin. Code, tit. 14, § 15143, subd. (d).)
A.
We found the previous EIR deficient because an “impermissibly truncated” project definition severely distorted not only the critical project but the alternatives to the project.
The county claims that a truncated project description was again adopted in order to conceal the fact that virtually all of the pumped wa*10ter is destined for export to Los Angeles. The county claims: “(1) The EIR’s project description arbitrarily and artificially defines the increased pumping as 79 cfs for uses in the Owens Valley and 66 cfs for export to Los Angeles, whereas almost all the new water supply made available by pumping of 155 cfs actually is intended for export to Los Angeles.”
The claim was made the subject of a detailed comment by the county in response to the draft EIR. The county said: “Throughout the EIR, LADWP [city] attempts to disguise this unjustified and unannounced policy decision through the mask of arbitrarily dividing the putative 145 cfs of increased groundwater pumping between 66 cfs for export and 79 cfs for in-valley uses presently served by surface supplies. Yet a detailed and thorough evaluation of information contained in this EIR and other LADWP documents show that in fact virtually all of the 145 cfs is intended for export.”
The county calculated the actual increase in export to Los Angeles by subtracting the capacity of the first aqueduct (430 cfs) from the combined capacity of both aqueducts (666 cfs), then deducted the increased surface water flow from Mono County (87 cfs) to arrive at an export figure of 149 cfs.7
The city responded to this criticism exclusively as follows: “This is an erroneous conclusion. Under no-project conditions, 600 cfs will be exported, which includes only 10 cfs of pumping, leaving 66 cfs, not 145 cfs to be met from pumping.” The city here introduces a no-project export figure of 600 cfs. This figure is not the pre-project capacity of the first aqueduct (430 cfs), but is a wholly different figure.8
The establishment of this no-project alternative, predicated upon conditions different than those which preceded the project, criti*11cally distorts the relation of the project to the export of water to the City of Los Angeles.
The 600 cfs figure is greater than the export of water using the baseline period “preceding the second aqueduct’s availability for use” (fn. omitted) (71 Cal.App.3d at p. 196) (430 cfs average) and less than the long-term average capacity of both first and second aqueducts (666 cfs).
The difference between the no-project figure (600 cfs) and the project figure (666 cfs) is 66 cfs (which converts to 48,000 AF/yr), the precise amount which the city claims is the share of water from increased pumping which is to be exported to Los Angeles.9
Nowhere in the FEIR is there an explanation of how the 600 cfs is arrived at and it appears to be the product of subtraction of the amount of pumped water (66 cfs) which the EIR states is destined for Los Angeles from the long-term capacity of both aqueducts (666 cfs). The creation of a “synthetic” no-project alternative by such arithmetic means makes a sham of the EIR process and invalidates the EIR. We again hold that the project alternatives must be measured against the conditions preceding the second aqueduct.
B.
In 1977 we similarly found that the first EIR was defective in providing a “synthetic” no-project alternative regarding the irrigation of valley lands (alternative 4).
We said: “Although alternative 4 is not labeled as the ‘no project alternative,’ the city asserts that it is actually tendered for that purpose. It fails to fulfill that purpose. In order to mark out a ‘no project alternative,'' the EIR should describe what condition or program preceded the project. Alternative 4 does not portray the pre-project stage; does not describe what quantities of water from what surface or subsurface *12sources were supplied to what lands. The litigation record supplies some helpful data. For many years before the construction of the second aqueduct the city had supplied irrigation water (surface and subsurface) to ranchers on city-owned lands. The 1963 report on feasibility of the Second Los Angeles Aqueduct stated that total Inyo-Mono acreage (inclusive of Indian lands) then supplied with city irrigation water was 40,117 acres, of which 31,817 acres were supplied on an interruptible basis. (1963 Report, vol. VI, p. 6-13.) In 1973, 10 years later, [Duane L.] Georgeson filed an affidavit . .. stating that approximately 19,000 acres of city-leased land (11,000 in Inyo County and 8,000 in Mono County) were being supplied with irrigation water. The project described in the EIR would supply subsurface water for uses ‘not anticipated’ in 1963, when the second aqueduct was planned. As a purported ‘no project’ alternative to fulfilling these unanticipated uses, alternative 4 would deny water to both anticipated and unanticipated uses, including the bulk of land supplied in 1963. Alternative 4 offers a synthetically conceived election between the synthetically conceived project and destruction of the Owens Valley cattle industry. The EIR’s project description and its purported ‘no project’ alternative simply do not match.
“Because the final EIR does not include a genuine ‘no project’ alternative, because its list of alternatives is not tied to a reasonably conceived or consistently viewed project, the Los Angeles EIR does not comply with CEQA’s demand for meaningful alternatives. This lack results in an EIR which does not meet CEQA’s goal of ensuring that ‘the long-term protection of the environment shall be the guiding criterion in public decisions.’ (§ 21001, subd. (d).)” (Italics added.) (71 Cal.App.3d at pp. 201-202, 203.)
The first EIR was defective precisely because it created a no-project alternative based upon conditions different from those which “preceded the project.” As a result, the first EIR made it appear that the project was necessary to avoid something worse, the “destruction of the Owens Valley cattle industry,” a condition which did not precede the project.
It can readily be seen that any project can be made to look good by posing a “synthetic” “no-project” alternative consisting of the absence of the project plus some additional condition which is much worse than the project. That is what the first EIR did and that is what the present N *13EIR does in a repetition of the exact deficiency upon which w¿ rejected the first EIR.
The current project description advances an irrigation “policy” which is claimed to provide a “secure supply” of water to irrigate certain lands within the Owens Valley. The EIR states: “Historically, groundwater was not available in the Owens Valley area on Los Angeles lands leased for agriculture. Surface water to irrigate up to 30,000 acres was available only after export requirements to Los Angeles were fulfilled. The feast-or-famine water supply caused ranchers to make significant economic decisions with high degrees of uncertainty. Groundwater pumping will provide a secure supply to irrigate the best 17,100 acres of the 30,000 acres irrigated during normal years under pre-project conditions. Total productivity would not be reduced significantly but economic stability is increased.”
The EIR sets forth the no-project alternative as follows: “Alternative A is no-project. The groundwater basin would be pumped during dry periods to maintain flow in the First Aqueduct and to supply 600 acres of dairy land. The no-project pumping rates are 7,000 AF/yr. (10 cfs) long-term (1935-36 through 1965-66), and 97,000 AF/yr. (135 cfs) maximum (1960-61). This alternative necessarily involves dry year reduction in total water exported to Los Angeles and also involves continuation of historic interruptible agricultural water” (Italics added.)
The italicized language appears to measure the no-project alternative by pre-project (“historic”) conditions involving supply of the First Aqueduct. But the explication of alternative A in the EIR belies the appearance.10
*14The details are shown in a chart and description in the Final Environmental Impact Report, volume I, pages Bll-4 to Bll-5.11
The chart makes plain that the no-project alternative is not only different than the pre-project “historic” irrigation use, but that it utilizes both the first and second aqueducts for export. The chart reveals that the no-project alternative reduces the “historic” (pre-project) water delivery for irrigation by at least half and contemplates 18 dry years, as opposed to 5 historic dry years, out of a 31-year period. The EIR gives no coherent description of the impact of the no-project alternative upon Owens Valley agriculture, although the fact that in 18 years out of 31 no water would be provided (other than for 600 acres of dairy farming) makes the impact palpably disastrous for Owens Valley agriculture.
Because the EIR does not include genuine no-project alternatives, because its project description impermissibly removes an essential element from the matters required to be considered, because the alternatives are not tied to a consistently viewed project, the second EIR does not comply with mandate of this court and the requirements of CEQA.
Disposition
We hold that the city’s return to the writ of mandate issued as a result of our June 1973 decision fails to comply with the writ. This court has continuing jurisdiction to enforce the writ until it is fully satisfied. *15(Code Civ. Proc., § 1097; County of Inyo v. City of Los Angeles, supra, 61 Cal.App.3d at p. 95.) The writ is not discharged but remains in force; the City of Los Angeles and its department of water and power are directed to take reasonably expeditious action to comply with it. Interim restraints upon the rate of extraction of groundwater and upon the decrease of water supplied to Owens Valley uses will remain in effect.
Puglia, P. J., concurred.
"Mono Basin Operations. The Project does not involve pumping or diversions of water for export from the Mono Basin. A benefit of the Project is stabilitation of irrigation on City lands in Mono Basin by substitution or replacement of water. Such substitution or replacement in no way increases, decreases or affects the amount of water which may hereafter reach Mono Lake.”
“Surface Diversions. The increased groundwater pumping which constitutes the Project will be used to supplement surface flows, but the diversion, export and occasional use of surface waters on City-owned lands in Inyo and Mono Counties is not a part of the Project. Such surface diversions and the exercise of its diversion rights by the City of Los Angeles are ongoing exercises of practices and rights undertaken and fully implemented prior to the effective date of CEQA in November, 1970.”
“The increase in pumping, on an average annual basis, to make available water for project uses is 105,000 AF/yr. In wet years the increase in minimum pumping to provide for uses in Owens Valley (fish hatcheries and towns) is 33,000 AF/yr (45 cfs). In dry years the increase in maximum pumping to provide for all uses is 130,000 AF/yr. Of the 105,000 AF/yr (145 cfs) average annual increase in pumping, approximately 48,000 AF/yr. (66 cfs) would be used in Los Angeles and 57,000 AF/yr. (79 cfs) would be supplied directly or indirectly to uses in the Owens Valley.” (Italics added.)
“The supply resulting from increased pumping planned for use in Inyo and Mono Counties will be supplied directly or indirectly depending on the location of the use relative to the location of existing wells. In some cases the water will be supplied directly from wells such as for irrigation in the Bishop area. In other cases, where there are no wells, the supply will be indirect, i.e., the pumped water designated for irrigation or other use will flow into the Aqueduct and an equal amount of surface water will be diverted from a stream or ditch convenient to the location of use.”
Most of the wells are located in the lower end of the Owens Valley while most of the land to be irrigated is located in the upper end of the valley. There is no means for pumping the water up-valley.
We are left to speculate. Perhaps the additional surface water for both export (91 cfs) and in-valley use (79 cfs) is to come from the Mono Basin, but the FEIR excludes that subject, as well, from the scope of the FEIR. The FEIR does show an increase in flow from Mono Basin of about 89 cfs (336 cfs to 425 cfs). This increase is never explained or discussed beyond the fact that the figure appears in a chart. There is no analysis of whether the 89 cfs will be sufficient to meet increased demand and there is no discussion of the long-term realiability of this source. Reduction in evapotranspira*9tion (ET) might increase water availability for use by the project. Table Dl-9 shows a reduction in ET loss of approximately 146 cfs:
Pre-Project ET: 822 Project ET: 676
However, table A6-1 (DEIR Appen. 6) shows that approximately 140 cfs of the reduced ET will be due to lowering of groundwater levels. ET from present high groundwater areas will drop from 330 cfs to 190 cfs. So only a small portion (approximately 6 cfs) of the reduction in ET will be reflected in increased surface flow. Even including the 89 cfs of surface water from Mono County and the increase of 6 cfs by reduction in ET, there is still missing some 75 cfs needed for in-valley use.
“Project export 666
(less) Pre-project export -430
Total increase in export 236
(less) Increase in export from Mono - 87
Increase in export from Owens Valley 149 cfs”
At two other places in the FEIR, it is made clear that the no-project alternative is based upon use of both aqueducts at a flow rate in excess of the first aqueduct (430 cfs). The FEIR states that, under the no-project alternative the “export through the Aqueducts would be approximately 434,000 AF/yr (600 cfs).” In discussing the no-project irrigation alternative, the use of both aqueducts is premised. (See post, fn. 10.)
“The essential element of the export component of this project is the increase in the average annual Second Aqueduct flow of 48,000 acre-feet of groundwater. Variations in annual pumping for export from the operational pattern of stable annual export are possible. Such variations may prove desirable to mitigate unanticipated short-term impacts of groundwater extractions in Owens Valley or to respond to operational problems within the Los Angeles water system. This flexibility in operations for the export requirements of the project is coupled with some flexibility in adjusting pumping patterns to mitigate unanticipated impacts in Owens Valley.”
The EIR’s right hand does not appear to know what its left hand is doing.
In response to the County of Inyo’s claim that “[t]he statement that the ‘no project alternative describes water use policies prior to 1970’ is completely untrue and is contradicted by the description given on Page 11-4,” the city said: “Disagree. There is no inconsistency between the two descriptions. The irrigation records and lease provisions are indications of policies prior to 1970.”
However, elsewhere the city states: “The water supplies for irrigation in the Owens Valley for pre-project, no-project, and project conditions are provided in Section B, page 11-5 of this Final EIR” (FE1R, vol. II, p. II-7) and also states: "The no-project alternative describes First and Second Aqueduct operations, without increased pumping. This condition did not exist in the pre-project period. Refer to Section B, Pages 3-2 and 11-5 of this Final EIR.” (Italics added.)
“The following table provides data on the No-Project operation described in this alternative, the actual historical irrigation experience during which surface water was exported only through the First Los Angeles Aqueduct, and the proposed project. The comparison is based on runoff data for the 31-year hydrologic study period (1936-1966).
(1) Maximum 30,000 acres interruptible irrigation, no groundwater pumping for irrigation, First Aqueduct export. Irrigation data not available for 1936. (2) Maximum 30,000 acres interruptible irrigation, no groundwater pumping for irrigation, First and Second Aqueduct export. (3) Maximum 17,000 acres regular irrigation, groundwater pumping for irrigation, First and Second Aqueduct.”