Opinion

The City of San Clemente (the City) appeals from a conditional judgment in favor of Avenida San Juan Partnership (the owners). The City imposed an “RVL” or “residential, very low” set of land use restrictions on an undeveloped 2.85-acre parcel in the middle of a residential tract otherwise zoned “Residential, Low Density Zone.” The RVL designation limits parcels to one dwelling per 20 acres. Residential, low (RL) by contrast, allows at least four dwellings per acre. There were two phases of trial. In phase one, the trial court concluded the restrictions constituted spot zoning. It issued *1260 a writ of mandate declaring the resolution denying the owners’ application to develop four houses on the property null and void and ordered the City to adopt a new resolution vacating the resolution denying the owners’ application.

The City asked for a stay of the writ. The stay was granted. In phase two, there was a trial on the owners’ request for damages. The court found a compensable taking, using the “Penn Central factors” test. (See Penn Central Transp. Co. v. New York City (1978) 438 U.S. 104 [ 57 L.Ed.2d 631 , 98 S.Ct. 2646 ] (Penn Central)) The court entered a conditional judgment. The judgment gave the City the choice of either complying with the writ of mandate or paying $1.3 million in damages for the value of the property taken by the RVL restrictions.

We affirm the judgment so far as the trial court gave the City the choice of either complying with the writ of mandate or paying inverse condemnation damages. The City’s refusal to lift the imposition of the RVL restriction on this particular parcel was arbitrary and capricious. As the trial court also found, applying the factors enumerated by Penn Central, that refusal to lift the RVL restrictions imposed specifically on this parcel constituted a taking. Further, this litigation was timely. This action is in substance an “as applied” challenge to the City’s denial of requested changes to the City’s land use scheme which otherwise works a peculiar hardship on the particular parcel at issue here, not a general attack on RVL zoning as such. However, we reverse the judgment to afford the trial court the opportunity to reconsider the fair market value of the property.

FACTS

1. Background

The subject parcel consists of an undeveloped 2.85 acres on a slope that fronts Avenida San Juan. When the owners bought it in 1980, the zoning allowed six dwellings per acre.

In the early 1980’s the owners wanted to develop four houses on the property. Their plans were well within the existing land use restrictions. The City approved plans to subdivide the property allowing for four single family lots. These lots would be connected to the street by a cul-de-sac road. Construction of the road required considerable grading, but the City specifically found that there were no geological obstacles to developing the property with four residences. A city resolution at the time stated that “ ‘all competent evidence before the City Council indicates that the site is developable without danger to adjacent properties.’ ”

*1261 Opposition arose in the neighborhood. In 1983 a landslide occurred in the Verde Canyon area, which is generally located to the southwest of the subject parcel and lies behind the homes which front the southern side of Avenida San Juan. The landslide did not involve the subject parcel. Even so, a group of neighbors petitioned the City to make the subject parcel open space. The City attorney at the time, however, opined that rezoning the parcel as open space would be a compensable taking. The City engineer said there was no reason to reconsider the tentative parcel map approval already given. The property remained zoned six houses per acre. However, the owners did not develop it at the time.

2. The 1993 and 1996 RVL Restrictions

In 1993, the City amended its general plan to create RVL zoning and impose it on several properties, including the subject property. All parcels surrounding the subject property, however, were zoned RL. RVL zoning in San Clemente allows only one residence for every 20 acres. RL zoning allows four dwellings per acre.

The City did not get around to formally rezoning the property from “R-l-B-1” to the one-dwelling-per-20 acres RVL until 1996. In mid-February of that year the City approved changing the zoning for a variety of properties to correspond to the 1993 amended general plan.

The general plan states that the purpose of RVL zoning is to preserve open space in canyons. RVL zoning, under the terms of the City’s ordinance, is intended to apply to cases of “significant acreage.” The subject parcel is 2.85 acres. It is not a canyon. It is a slope. The ordinance says nothing about slopes.

The ordinance specifically recites that the purpose of RVL zoning is to “preserve currently undeveloped canyons which are either geologically unstable, or aesthetic open-space, or biological resources.” There is no dispute that the parcel does not contain “any sensitive biological resources.”

3. The 2006 Variance Application and 2007 Denial

None of the partners in the partnership actually found out about the downzoning from R-l-B-1 to RVL until 2004. That year they hired a civil engineer to help them once again try to develop the property.

*1262 In early September 2006 the owners submitted a development application to build four dwellings on the 2.85 acres. Specifically, the application sought a general plan amendment, zoning amendment, tentative parcel map, site plan permit, conditional use permit, and variance.

Some five months later, in February 2007, the city planning commission recommended denial of the application. The owners sought review by the city council. At a meeting of the city council on July 24, 2007, a resolution was approved denying the application.

4. The Ensuing Litigation

a. Trial level

i. Writ of mandate

Within 40 days, on August 29, 2007, the owners filed suit in federal court. They alleged, among other claims, inverse condemnation based on the spot zoning of the property. A number of city officials were added as nominal defendants. The federal court granted a motion to dismiss, but with leave to amend. After the owners amended their complaint in federal court, they filed this action in state court and dismissed the federal action.

There is no argument on appeal that the dismissal with leave to amend by the federal court has any sort of preclusive effect. The City accepts that the August 29, 2007 filing date in federal court is the one on which the litigation commenced for purposes of the statute of limitations. (See 28 U.S.C. § 1367 (d).)

The parties agreed to an order bifurcating the trial. The trial court was first to consider the owners’ request for administrative mandate. Remaining issues were to be decided in a second phase.

The writ of mandate phase was heard by Judge Stock. Her statement of decision reached three major conclusions;

*1264 The ensuing writ of mandate filed September 10, 2009, (1) declared the City’s resolution denying the owner’s land use application null and void; (2) ordered the City to adopt a new resolution vacating the earlier resolution denying the owners’ application; and (3) ordered the City to file a return to the writ within 60 days showing the City complied with the writ’s terms. The City took the position that it was going to contest the writ in an appeal. The trial court stayed the writ in early December.

ii. Inverse condemnation trial

Phase two commenced in mid-December. Trial included a site visit by Judge McEachen, testimony from 13 witnesses, and numerous exhibits. On February 1, 2010, Judge McEachen issued a statement of decision and judgment finding that the City had deprived the owners of all economically viable use of the parcel.

Judge McEachen specifically rejected the City’s various reasons for zoning the property RVL, finding the parcel is not a canyon, as the City claimed, and there was no evidence of “negative geotechnical data.” In fact the court noted that the land’s “stability” was shown by a nearby cliff that faces the street without a retaining wall, and there was nothing unusual about the parcel’s topography. The vegetation was the same chaparral and brush that is typical of all of Southern California. The trial court found the real reason for the RVL zoning was the City’s desire to keep the property as open space.

Judge McEachen also specifically found “bad faith” in the City’s handling of the owners’ 2006 development application. He wrote: “Even though Plaintiff asked for a general plan amendment and a zoning amendment, the denials were all illogically ‘based’ on the fact that the application did not comply with the general plan and zoning ordinances.” In essence, the City never gave any serious consideration to the owners’ application. Under the heading “City’s Bad Faith Handling of Plaintiff’s 2006 Development Application,” he wrote: “With the September 1, 2006 application, Plaintiff submitted a proposed tentative parcel map which is substantially similar to Tentative Parcel Map 82-832 [(the one conditionally approved back in 1983)]. However, the application was rejected first by Mr. Nicholas [(an associate City planner)], then by the City’s Planning Commission, then by the City Council. Even though Plaintiff asked for a general plan amendment and a zoning amendment, the denials were all illogically ‘based’ on the fact that the application did not comply with the general plan and zoning ordinances. Also, Mr. Nicholas wrote in at least two reports that since his recommendation to deny the application was based on its nonconformity to the general plan and zoning ordinances, ‘staff did not complete any further review,’ meaning no real investigation occurred. Therefore, City had zero evidence for *1265 a finding that the proposed development was somehow physically unsuitable.” (Italics added.) Judge McEachen further found that the City’s land use officials had ignored the City’s own ordinances which seek to promote cooperation with local landowners so as to come to a balanced accommodation of interests.

The judge elaborated on the City’s underlying purpose to keep the property open space: “City targeted the Property for the stated purpose of ‘protecting open space,’ thus forcing Plaintiff to bear a public burden which should be borne by the public as a whole. Although City retained the authority to further amend the general plan and zoning ordinances, it held fast to its RVL zoning decision, instead of correcting this injustice.”

Not surprisingly, Judge McEachen found a regulatory taking of the parcel by the City. He used what is often called the Penn Central factors approach. Judge McEachen enumerated no fewer than nine factors all pointing in the direction of a taking: (1) The major loss of use of the property as the result of the government action; (2) the loss of investment backed expectations from what they had been at the time of acquisition; (3) the arbitrariness of the rezoning; (4) the unequal treatment of the subject parcel in comparison with the surrounding properties; (5) the fact that the City’s action was prompted by the desire of the neighbors to keep the subject parcel as complete open space; (6) the City’s core motivation for the downzoning was to keep the subject parcel as open space; (7) the City’s attitude of being substantively against any development of the property; (8) the absence of any action by the City to mitigate the financial burdens imposed by the downzoning; and finally (9) the fact that the downzoning had clearly prevented the “best use” of the land (which would be four lots, not one).

Judge McEachen then calculated the amount of just compensation. There was evidence its fair market value was $2.8 million if developable with four lots. From that $2.8 million he deducted the cost of the cul-de-sac road. The judge took the cost estimate for the road from the City’s cost expert at about $1.5 million. Subtracting the latter from the former yielded the $1.3 million award.

b. Posttrial and appellate matters

The award resulted in a judgment filed February 1, 2010, for the $1.3 million ($1,316,967 to be exact). A notice of entry of judgment was filed on February 9, 2010. On February 24 the City filed an intention to move for new *1266 trial. The hearing on the new trial motion was scheduled for April 6, 2010. Among the grounds proffered for the new trial motion was that the City had the right, under Hensler v. City of Glendale (1994) 8 Cal.4th 1 [ 32 Cal.Rptr.2d 244 , 876 P.2d 1043 ] (Hensler), “to rescind the RVL standard as an alternative to paying inverse condemnation damages.” The City also made the related argument that by “granting a Writ” and requiring the City to pay damages for a “complete and permanent taking of the Property, the Decision erroneously authorizes a double recovery.”

On March 29, prior to the hearing on the new trial motion, the City filed its first notice of appeal, creating docket No. G043479. The notice of appeal was from “the final judgment in this matter on February 1, 2010 and all orders that are separately appealable.”

About a week later, on April 6, the trial court denied the new trial motion. However, the trial court was obviously concerned with the impact of Hensler . “[R]ather than granting a new trial,” wrote Judge McEachen, Hensler required giving the City the choice of paying damages or rescinding its RVL land use restriction. This “would only require a modification of the judgment pursuant to CCP 662.” Accordingly, the minute order directed that the language of the February 1 judgment would now read: “Defendant City of San Clemente shall have thirty (30) days to comply with the Order Granting and Issuing Writ of Administrative Mandamus filed on 9-10-09, and thereby avoid having to pay compensation for a permanent taking. But if defendant City of San Clemente does not comply with the Order Issuing and Granting Writ of Administrative Mandate within 30 days, on the third cause of action for inverse condemnation in Plaintiff’s complaint, Judgment for Plaintiff and against City for $1,316,967.00 plus interest at the Surplus Money Investment Fund rate as set forth at C.C.P. 1268.350 from July 24, 2007 plus reasonable attorney’s fees and expert fees and costs of suit herein.”

About a week later, on April 13, the City filed its second notice of appeal in this case, creating docket No. G043534. However, it was not until April 19, about a week afterwards, that a formal, signed “modified judgment” was filed.

The terms of the modified judgment provide: The stay on the writ of mandate was lifted. The City would have 30 days (until May 6, 2010) to comply with it. If the City did not comply by then, the owners would recover $1,673,530.52 from the City. The $1,673,530.52 consisted of (1) the $1,316,967 inverse condemnation damages for the taking of the property; (2) $105,769.22 in prejudgment interest;'(3) recoverable costs of $23,644.30; and (4) $227,150 in attorney fees.

*1267 DISCUSSION

1. Appellate Jurisdiction

Generally, the filing of a notice of appeal “divests the trial court of further jurisdiction in the cause.” (In re Estate of Waters (1919) 181 Cal. 584, 585 [ 185 P. 951 ]; see generally Varian Medical Systems, Inc. v. Delfino (2005) 35 Cal.4th 180 [ 25 Cal.Rptr.3d 298 , 106 P.3d 958 ] (Varian) [explicating § 916 of Code Civ. Proc.].) There are, however, a number of exceptions to the rule. (See Varían, supra, 35 Cal.4th at p. 191 [noting collateral matter exception, including new trials and proceedings to expunge a lis pendens].)

Section 662 of the Code of Civil Procedure affords trial judges the option of modifying judgments on new trial motions without actually having to grant a new trial. (See Spier v. Lang (1935) 4 Cal.2d 711, 714 [ 53 P.2d 138 ] [rejecting “contention that the language of the section does not give the court the power to change the findings and modify its judgment so as to make a new or different judgment unless a new trial be granted”].) The option to modify in the course of deciding a new trial motion, can, of course, create the anomaly of a trial court retroactively modifying a judgment already appealed. Under several of the various tests articulated by the high court in Varían (see Varian, supra, 35 Cal.4th at pp. 189-190 [direct or indirect modification of the appealed judgment, interference with appellate court’s ability to conduct appeal, and irreconcilability of original and modified judgment]), the postappeal modification of an already appealed order would appear to impact the effectiveness of the appeal.

The anomaly was solved more than 50 years ago in Neff v. Ernst (1957) 48 Cal.2d 628 [ 311 P.2d 849 ] (Neff). Like the present case, Neff involved the trial court modifying a judgment already the subject of an appeal while the trial court was in the process of denying a new trial motion. Further, as in the present case, in Neff there was a second appeal after the modification of the original judgment. (Id. at pp. 632-633.)

The Neff court solved the anomaly by recognizing the obvious. “The question is: which appeal is properly before this court. Obviously it is one or the other and cannot be both.” (Neff, supra, 48 Cal.2d at p. 633 .) The logic of section 662 dictated, then, that the modified judgment becomes “the final judgment.” ( 48 Cal.2d at p. 634 .) The “appeal from the prior judgment becomes ineffective,” or “nonoperative.” (Ibid.) The Neff court then went on to consider the merits of the modified judgment.

We now follow suit. In the present case we now have only one “operative” appeal, namely the appeal from the modified judgment filed on April 19. In *1268 fact, the relevant notice of appeal, filed April 13, 2010, after the announced modification on April 6 was, in fact, premature. Even so, it is nevertheless effective. (See Cal. Rules of Court, rule 8.104(d).)

The essence of the modified judgment is that the City’s RVL land use restrictions constitute a taking, but the City has the choice of either rescinding its denial of the owners’ application to change those restrictions, as the writ of mandate requires, or paying inverse condemnation damages.

2. The Writ of Mandate

Since the modified judgment gives the City the choice of complying with the earlier writ of mandate issued by Judge Stock we review the core conclusion of that writ, namely that the downzoning of the subject property, and later refusal to change that zoning, was arbitrary and capricious.

The “rezoning of property, even a single parcel, is generally considered to be a quasi-legislative act” thus “subject to review under ordinary mandamus.” The standard for review of a quasi-legislative act is whether the action was “arbitrary or capricious or totally lacking in evidentiary support.” (City of Carmel-by-the-Sea v. Board of Supervisors (1986) 183 Cal.App.3d 229, 238-239 [ 227 Cal.Rptr. 899 ]; see Arnel Development Co. v. City of Costa Mesa (1980) 28 Cal.3d 511, 521 [ 169 Cal.Rptr. 904 , 620 P.2d 565 ]; Consaul v. City of San Diego (1992) 6 Cal.App.4th 1781, 1791 [ 8 Cal.Rptr.2d 762 ] (Consaul).)

Here, the trial court (in the writ of mandate phase presided over by Judge Stock) found the downzoning, embodied in the both the 1993 general plan amendment and the 1996 zoning ordinance, and the City’s 2007 refusal to consider rezoning the property, to be arbitrary and capricious. We agree.

In her statement of decision Judge Stock found a “suggestion of spot-zoning here.” Nevertheless, her reference to the subject parcel being “an isolated area” that had “become an island” of “minimum lot size zoning in a residential ocean of substantially less restrictive zoning,” plus her citation to Hamer v. Town of Ross (1963) 59 Cal.2d 776 [ 31 Cal.Rptr. 335 , 382 P.2d 375 ] (Hamer), a spot zoning case, leave little doubt as to the basis of her decision to grant the writ of mandate.

The essence of spot zoning is irrational discrimination. Arcadia Development Co. v. City of Morgan Hill (2011) 197 Cal.App.4th 1526, 1536 [ 129 Cal.Rptr.3d 369 ] described spot zoning; “ ‘Spot zoning occurs where a small parcel is restricted and given lesser rights than the surrounding property, as where a lot in the center of a business or commercial district is *1269 limited to uses for residential purposes thereby creating an “island” in the middle of a larger area devoted to other uses. . . . Usually spot zoning involves a small parcel of land, the larger the property the more difficult it is to sustain an allegation of spot zoning. . . . Likewise, where the “spot” is not an island but is connected on some sides to a like zone the allegation of spot zoning is more difficult to establish since lines must be drawn at some point. . . . Even where a small island is created in the midst of less restrictive zoning, the zoning may be upheld where rational reason in the public benefit exists for such a classification.’ ” (Citations omitted.)

Justice Mosk observed in a case otherwise devoted to development exactions that because spot zoning implicates discriminatory treatment, it entails a “more rigorous form of judicial review.” (Ehrlich v. City of Culver City (1996) 12 Cal.4th 854, 900 [ 50 Cal.Rptr.2d 242 , 911 P.2d 429 ] (cone. opn. of Mosk, J.) (Ehrlich).)

In the present case, there is no need to rest the determination on either Judge McEachen’s finding of bad faith on the part of the City or the more rigorous form of review advocated by Justice Mosk in his concurrence in Ehrlich . Here, there is no question that the owners’ parcel is a one-house-per-20-acre island in a two-to-six-house-per-acre sea. Zoning maps in the record readily provide substantial evidence of spot zoning. Exhibit 6 is a color photo that shows higher density development on Avenida San Juan encircling the subject property. Even more telling is the City’s own general plan map showing the applicable zoning. The subject parcel is a small RVL spot surrounded by denser land uses. Moreover, Judge McEachen personally visited the site and found that “There is nothing unusual about the Property’s topography.”

Both Hamer and this court’s decision in Ross v. City of Yorba Linda (1991) 1 Cal.App.4th 954 [ 2 Cal.Rptr.2d 638 ] (Ross) found a city guilty of spot zoning when it denied a property owner’s request to rezone the owner’s property to a designation consistent with surrounding properties so the owner could proceed with a development plan. In Hamer , the Supreme Court struck down a minimum one-acre lot size requirement on a 2.2-acre parcel as arbitrary and discriminatory, given the existence of denser construction to the east, north, and west of the property. (See Hamer, supra, 59 Cal.2d at pp. 790-791.) Justice Tobriner emphasized the disparity in size restrictions between the subject property and the more dense surrounding properties: “[Z]oning ordinances which impose a one-acre lot restriction cannot properly apply to property which is virtually surrounded by parcels of lesser size . .. .” (Id. at p. 778; see also id. at p. 781.)

To the same effect is Ross. There a one-acre restriction on a 1.117-acre plot prevented the construction of a second dwelling on the property, which, if the *1270 property had been zoned as the surrounding properties, would have been allowed. (See Ross, supra, 1 Cal.App.4th at p. 961 .)

Indeed, Ross applies directly, if not a fortiori to the case before us. In Ross , this court said: “Even after the proposed division” the development would still be less dense than “many” in the area. (Ross, supra, 1 Cal.App.4th at p. 961 .) Here, the proposed development would be less dense than everything that surrounds it.

In its reply brief, the City asserts its discriminatory downzoning and refusal to meaningfully consider the owners’ 2006 development application was justified because the property is on a slope. Therefore the project requires high retaining walls and what the City characterizes as a “ ‘tunnel-like’ ” access road.

The topography, or “slope,” argument fails for several reasons. First and most basically, the topography rationale in this case does not support discriminatory treatment. Much of the City of San Clemente reflects the same topography, with many houses already built on slopes in the area. Judge Stock’s statement of decision specifically found there were no “unique facts” calling for the parcel to remain a landlocked island. The City’s community development director testified that almost all of San Clemente is on hilly terrain. As exhibit 6 shows, Avenida San Juan itself was carved out of a sloping terrain. The City obviously has no problem with slopes as distinct from canyons. Its own RVL zoning ordinance makes no mention of slopes in regard to RVL zoning.

Second, topography by itself was not a reason to deny the owners’ various applications that would have allowed the City to make an exception to the imposition of RVL (one house per 20 acres) on the subject parcel. The record is full of pictures of houses in San Clemente with high retaining walls. Any disruption from grading would necessarily be temporary and subject to reasonable restrictions to ameliorate any disruption to the neighbors. As Judge McEachen observed, the matter of retaining walls and excavation necessary for the cul-de-sac road were matters that might have been ironed out to “cure” those “specific concerns.” Judge McEachen found that the City ignored its own ordinance requiring the City planning department to cooperate with local landowners so as to achieve a balanced accommodation of interests. And as the City itself recognizes in its opening brief, the specifics of the owners’ application received “no detailed review.” The arbitrariness of the City’s denial is shown by its refusal to consider measures to “cure” the impact of grading or retaining walls.

Even under the City’s 1993 general plan amendments, there were only two points of concern (called “opportunities and constraints” in the text) under the *1271 heading of “topography.” The first was the “potential of increased susceptibility of erosion on steep slopes that have been improperly graded.” (Italics added.) The second was the simple preservation of “the appearance of the natural hillsides and vegetation.” Both of these “constraints,” of course, could be addressed by reasonable mitigation measures. Likewise, language in a 1996 general plan amendment concerning the “[m]aintenance and restoration of significant natural systems and resources” including “steep slopes,” is susceptible to address by reasonable mitigation measures. The 1996 language was framed in the context of allowing development compatible with the preservation of steep slopes, but requiring it to “comply with the Aesthetic Resources Element” including such “[mjaintenance.”

Most fundamental, though, is the contradiction between the basic purpose of RVL zoning under the City’s general plan and zoning laws, which is to preserve canyons, and this property, which, like much of the rest of the City, is on a slope. By its very terms, the rationale for the City’s RVL zoning—protection of canyons—does not apply to the parcel here. This parcel, in this neighborhood, was being singled out for discriminatory treatment independent of the reason for RVL zoning in the first place.

The City has cited us to no case where some aspect of a small parcel susceptible of mitigation has been used to justify what would otherwise be blatant spot zoning. The case the City primarily relies on to say this case is not an instance of spot zoning, Consaul, supra, 6 Cal.App.4th 1781 , is distinguishable. Consaul involved the downzoning of part of a property (an undeveloped lower 1.06 acres of a larger 1.5-acre property, with the upper part already having a single-family residence and two apartments on it) from multifamily to single family, dashing the hopes of the owners to put in a 26-unit condominium project on the lower 1.06 acres. (See id. at p. 1786.) Most of the opinion has nothing to do with spot zoning, but deals with whether the 26-unit project constituted a vested right at the time of the downzoning. (See id. at pp. 1785, 1793-1801; cf. id. at pp. 1811-1813 (dis. opn. of Nares, J.).)

The owners’ spot zoning argument in Consaul was an obvious afterthought. The spot zoning portion of Consaul occupies only three paragraphs of discussion, and the opinion, without going into much detail at all, simply affirms the trial court’s “evaluation on these factual issues concerning the existence of a valid basis for the city’s action” in light of the “topographical and usage characteristics of the land and the surrounding area.” (Consaul, supra, 6 Cal.App.4th at p. 1802 .) While the Consaul majority observed that “lines in zoning must be drawn at some point” (ibid.), that general observation does not mean that any time a local government invokes the word “topography” it can treat a given parcel differently from the surrounding property.

*1272 The clincher is the rezoning map in the appendix to the Consaul opinion. (Consaul, supra, 6 Cal.App.4th at p. 1814 .) Just as the maps in this case clearly show spot zoning, the map in the appendix to the Consaul opinion does not. The property in Consaul was not being arbitrarily singled out for special treatment. One can easily see that there were no “islands.” Moreover, the opinion also noted that in the bottom of the canyon where the property was located there were already a number of single-family homes (id. at p. 1786). The owners in Consaul could hardly argue “spot zoning” when the downzoning merely gave them the same right as similarly situated property owners.

The trial court’s conclusion in granting the writ of mandate that the subject parcel was arbitrarily and capriciously downzoned must therefore be affirmed.

3. The Inverse Condemnation Award

Like Judge Stock before him in phase one, Judge McEachen substantively found spot zoning in phase two. As noted, he specifically noted after his site visit that the subject parcel’s topography was not usual, and he went so far as to find affirmative bad faith on the City’s part. But even without either finding, his conclusion that the City committed a taking was sound.

The City argues that there is no compensatory regulatory taking unless the regulation denies all economically viable use of the property. (Cf. Lucas v. South Carolina Coastal Council (1992) 505 U.S. 1003 [ 120 L.Ed.2d 798 , 112 S.Ct. 2886 ] (Lucas) [taking where all use was denied].) Of course when a regulating authority goes so far as to deprive owners of “all economically beneficial or productive use of land” there is a complete taking under Lucas. (Id. at p. 1015.)

The City’s model, however, is too limited. There are compensable takings that do not involve the deprivation of all economically viable use of land. The United States Supreme Court has declared that a compensable regulatory taking can occur when a regulation goes “ ‘too far,’ ” but stops short of denying all economically viable use. (Palazzolo v. Rhode Island (2001) 533 U.S. 606, 617 [ 150 L.Ed.2d 592 , 121 S.Ct. 2448 ] (Palazzolo), quoting Pennsylvania Coal Co. v. Mahon (1922) 260 U.S. 393, 415 [ 67 L.Ed. 322 , 43 S.Ct. 158 ].)

Whether a regulation goes “too far” is tested under what has been called the “Penn Central factors” approach. Our own Supreme Court has noted that there are three core factors: (1) the economic effect on the landowner; (2) the extent of the regulation’s interference with investment-backed expectations; and (3) the character of the governmental action. (Kavanau v. Santa Monica Rent Control Bd. (1997) 16 Cal.4th 761, 775 [ 66 Cal.Rptr.2d 672 , 941 P.2d 851 ].)

*1273 Although Judge McEachen found a complete taking, he also made findings regarding the Penn Central factors establishing a less than complete taking. We do not agree the City committed a complete taking because substantial evidence does not support Judge McEachen’s findings that (1) the City would not approve any development on the property and (2) development of one home on the property at street level was not financially feasible. We, however, agree that the Penn Central factors establish a partial taking in this case.

The three core Penn Central factors readily apply here: First, the economic effect was dramatic. This was illustrated by the trial court’s $1.3 million difference between the property valued if four houses could be built as per the 1983 City approvals and the zero value the trial judge found under RVL zoning.

Second, the regulation wholly undermined the investment-backed expectations of the owners. The owners purchased their property in 1980 when it was zoned the same as the surrounding properties are currently zoned. As if to emphasize those original expectations, the City itself, in 1983, approved a four-house development along the lines of the one it turned down in July 2007.

Third, the character of the governmental action appears to have been largely motivated to keep the subject parcel open space. The old label for RVL zoning was “OS [open space] 4.” The RVL designation is a virtually impossible one for a residential area, allowing only one house for every 20 acres. Application of RVL to this parcel in the “sea” of RL zoning that allows four houses per acre suggests, as Judge Stock noted, an irrational discrimination.

So-called “minor” Penn Central factors also point to a taking. The subject parcel was singled out for unequal treatment, particularly in light of the land use scheme that governed it in 1980. There was evidence that the neighbors and City simply wanted to keep the parcel as nearby open space. The City did not seriously consider mitigation measures in turning down the application. The best use of the land is consistent with the density of the surrounding neighborhood.

The City makes no meaningful attempt in this appeal to argue that the Penn Central factors were misapplied by the trial court. All its arguments are based on the premise that there must be a total taking—a “Lucas taking”— before there can be any compensable taking.

Preliminarily we reject the City’s assumption that the parcel has an $825,000 value even with the RVL zoning, because one house could still be *1274 built at curbside, and therefore no taking occurred. The $825,000 figure was proffered by the City’s own appraiser. Judge McEachen impliedly rejected it. He found the parcel had no value with RVL zoning.

In any event, the City’s argument is incompatible with Penn Central, as well as the Supreme Court’s relatively recent application in Palazzolo . There the federal Supreme Court held that there was a compensable taking even where an uplands section of the coastal property at issue was valued as high as “$200,000 if developed.” (Palazzolo, supra, 533 U.S. at p. 621 .)

The City also argues that there was no physical invasion or appropriation of the property, ergo there was no taking. That argument contravenes the whole idea of a Penn Central regulatory taking in the first place. A regulation that goes “too far,” even without a physical invasion, may constitute a taking.

Finally, the City also relies on a delay case, Landgate, Inc. v. California Coastal Com. (1998) 17 Cal.4th 1006 [ 73 Cal.Rptr.2d 841 , 953 P.2d 1188 ] (Landgate). But Landgate merely stands for the proposition that a reasonable delay in the processing of a permit is not a compensable taking. The landowner in Landgate did indeed obtain a permit to build a proposed mansion, but only after a two-year delay occasioned by a good faith dispute over whether the California Coastal Commission had jurisdiction over a certain lot line adjustment. (Id. at pp. 1014-1015.) A divided Supreme Court held that the developer could not recover damages for the lost two years because the “regulatory process” necessarily involves the “imposition of certain procedural conditions and substantive requirements on development.” (Id. at p. 1027.)

This is not a delay case, and there is no claim that the time required by the permit process itself constituted a taking. Therefore Landgate has no application here.

4. Statute of Limitations

The City argues that the applicable statute of limitations is section 65009, subdivision (c)(1)(B), which provides for a 90-day statute of limitations from direct attacks on newly enacted land use ordinances. It further argues that under section 65009, subdivision (c)(1)(B), the statute of limitations began to run either (a) in 1993, when the general plan was amended, or (b) in 1996, when the zoning ordinances pursuant to the 1993 general plan amendments were adopted, or at the very latest (c) in 2004, when the owners actually discovered the imposition of the RVL restriction on the subject parcel. The theory is that the owners were required to bring their challenge to the general plan and zoning imposition of RVL land use restrictions at the very time of *1275 those enactments. We disagree. The applicable statutes of limitations are sections 66499.37 and 65009, subdivision (c)(1)(E), and both began to run from the City’s final administrative adjudication in denying the owners’ application in 2007. (See County of Sonoma v. Superior Court (2010) 190 Cal.App.4th 1312, 1324 [ 118 Cal.Rptr.3d 915 ] (County of Sonoma) [to “pinpoint” when the statute of limitations begins to run on a claim against a land use restriction, one must “determine what specific governmental act or acts” are challenged].)

First, there can be no question that insofar as the lawsuit seeks relief from the City’s denial of the owners’ application for a tentative parcel map, the appropriate statute of limitations is section 66499.37. (See Hensler, supra, 8 Cal.4th at pp. 24-26; Friends of Riverside’s Hills v. City of Riverside (2008) 168 Cal.App.4th 743, 751 [ 85 Cal.Rptr.3d 695 ] [“Even our state’s Supreme Court has emphasized, in Hensler . . . that the 90-day statute of limitations under section 66499.37 applies to any type of action seeking review of a legislative or advisory body’s subdivision-related decision under” the Subdivision Map Act (§ 66410 et seq.)].)

Second, as to the challenge to the remainder of the City’s denials, it is section 65009, subdivision (c)(1)(E), which governs challenges to final administrative adjudications, that applies, not section 65009, subdivision (c)(1)(B). (All references to “(c)(1)(B)” or to “(c)(1)(E)” are to subd. (c) of § 65009.) Cases applying (c)(1)(B) have involved direct attacks on ordinances. They have not involved challenges to a final administrative adjudication based on the application of the ordinance to a given case. (County of Sonoma, supra, 190 Cal.App.4th at pp. 1324-1325 [applying (c)(1)(B) to challenge by marijuana dispensary that asserted it did not need even to seek permit as required under local ordinance]; Arcadia Development Co. v. City of Morgan Hill (2008) 169 Cal.App.4th 253, 261 [ 86 Cal.Rptr.3d 598 ] [applying (c)(1)(B) to facial challenge to extension of temporary density restriction ordinance]; Monks v. City of Rancho Palos Verdes (2008) 167 Cal.App.4th 263, 283 [ 84 Cal.Rptr.3d 75 ] [noting previous opinion, now law of the case, had applied (c)(1)(B) to attack on recent ordinance upping “safety factor” requirement]; Vo v. City of Garden Grove (2004) 115 Cal.App.4th 425, 438-439 [ 9 Cal.Rptr.3d 257 ] [applying (c)(1)(B) to direct attack on conditional use permit aspect of cybercafe regulation]; Save Our NTC, Inc. v. City of San Diego (2003) 105 Cal.App.4th 285, 291-292 [ 129 Cal.Rptr.2d 306 ] [applying (c)(1)(B) to attack on recent zoning ordinances allowing greater height limits than those contained in initiative enacted about 30 years previously]; Buena Park Motel Assn. v. City of Buena Park (2003) 109 Cal.App.4th 302, 307-309 [ 134 Cal.Rptr.2d 645 ] [applying (c)(1)(B) to direct attack on city ordinance restricting stays in motels]; Howard Jarvis Taxpayers *1276 Assn. v. City of Los Angeles (2000) 79 Cal.App.4th 242, 247-249 [ 93 Cal.Rptr.2d 742 ] [applying (c)(1)(B) where suit sought to void ordinance imposing city home occupations tax].)

By contrast, cases applying (c)(1)(E) have involved challenges to final administrative decisions, such as the case at bar. (Trinity Park L.P. v. City of Sunnyvale (2011) 193 Cal.App.4th 1014, 1029-1030, 1044-1046 [ 124 Cal.Rptr.3d 26 ] [applying (c)(1)(E), as well as § 66499.37, to challenge to below-market housing requirement imposed on developer as a condition of subdivision approval]; Honig v. San Francisco Planning Dept. (2005) 127 Cal.App.4th 520, 527-528 [ 25 Cal.Rptr.3d 649 ] [applying (c)(1)(E) to appellant’s attack on variance obtained by neighbor]; Royalty Carpet Mills, Inc. v. City of Irvine (2005) 125 Cal.App.4th 1110, 1121-1124 [ 23 Cal.Rptr.3d 282 ] [harmonizing (c)(1)(E) with Pub. Resources Code, § 21167.6, and applying (c)(1)(E) to carpet manufacturer’s challenge to conditional use permit issued to builder of proposed apartments near carpet manufacturer’s plant in industrial section of city].)

The case that is most illustrative of the distinction between direct, facial challenges to ordinances and challenges to the application of those ordinances in final administrative decisions is Travis, supra, 33 Cal.4th 757 . Travis directly rejected the application of (c)(1)(B) in favor of (c)(1)(E).

Because it is dispositive of the City’s argument here, we examine Travis in detail. The enactment of the relevant county ordinance had occurred years previously, in 1981. The ordinance was good news and bad news for property owners. On the one hand, it specifically allowed “ ‘affordable’ ” second dwelling units to be constructed by residential property owners. On the other hand, the ordinance set controls on both the kinds of people to whom the owners could rent their second dwellings, and on the amount of rents that might be charged. (Travis, supra, 33 Cal.4th at p. 763 .) The last substantive amendment of the ordinance was in 1997. ( 33 Cal.4th at p. 765 .)

In 1999, more than 17 years after the 1981 original enactment of the ordinance and about two years after its last substantive amendment, two separate owners each applied for permits to construct a second dwelling unit on their property.

The first owner was granted a permit, but that permit was subject to the restrictive occupancy and rent control rules placed on the property by the 1981 ordinance. He then filed an administrative appeal against those conditions; that appeal was denied in June 1999. The second owner was also granted a permit subject to the same occupancy and rent controls, but did not pursue an administrative appeal. (Travis, supra, 33 Cal.4th at p. 764 .)

*1277 Both owners filed a petition for writ of mandate in early September 1999, attacking the occupancy and rent restrictions as preempted by certain later-enacted state statutes, and also as contrary to various state and federal laws, including the takings clause of the Fifth Amendment. (See Travis, supra, 33 Cal.4th at p. 764 .) For the first owner, the early September action was certainly within 90 days of the denial of his administrative appeal challenging the restrictions. Both sets of owners made “as applied” challenges. (Id. at p. 765.) The trial court ruled that the first owner’s “as applied” regulatory taking claim was timely, but nonmeritorious. The trial court held the second owners’ as applied claim was untimely because the owners did not bring an action within 90 days of the “final decision” on the permit application. The intermediate appellate court, however, said that both sets of owners were subject to the 90-day statute of limitations in (c)(1)(B), and therefore even the first owner’s suit was untimely. ( 33 Cal.4th at p. 765 .)

The California Supreme Court then disagreed with the appellate court’s determination that the first owner’s suit was precluded under (c)(1)(B) on the theory that the statute began running with the 1981 enactment of the land use law. The California Supreme Court pointed out that the owners were making an “as applied” challenge because they challenged the imposition of the conditions on their permits in addition to the original enactment of the ordinance requiring the conditions. ([Travis, supra, 33 Cal.4th at p. 767 .) The court held that a claim does not fall outside (c)(1)(E) “merely because the plaintiff claims the permit or condition was imposed under a facially unconstitutional or preempted law.” ( 33 Cal.4th at p. 768 .)

Next, the Travis court addressed the running of the statute of limitations in the previous Hensler decision. The Travis court said that Hensler was different, because in Hensler there was no doubt that the landowner’s inverse condemnation action was untimely. It was untimely because the action was brought “several years after the city had applied the ordinance to the plaintiff’s property.” (Travis, supra, 33 Cal.4th at p. 768 , italics added.) Hence, said Travis , the action in Hensler was untimely “whether considered as an attack on the ordinance itself or on the city’s application of the ordinance.” ( 33 Cal.4th at p. 768 , italics added.)

We conclude that the present action is not a “facial challenge” but an “as applied” challenge. A facial challenge considers only the text of the challenged law. An as applied challenge considers the application of that law to the particular circumstances of the case. (See Tobe v. City of Santa Ana (1995) 9 Cal.4th 1069, 1084 [ 40 Cal.Rptr.2d 402 , 892 P.2d 1145 ] (Tobe); see also NJD, Ltd. v. City of San Dimas (2003) 110 Cal.App.4th 1428, 1439 [ 2 Cal.Rptr.3d 818 ].) An “as applied” challenge may even seek relief from a facially valid statute. (Tobe, supra, 9 Cal.4th at p. 1084 .) Here, the owners *1278 have sought relief from the first time the 1993 general plan amendment and the 1996 zoning ordinance were ever actually applied to the subject property, which was when the City denied their application to build four houses on the subject parcel in 2007. The owners do not challenge the entire 1993 general plan amendment or the entire 1996 zoning ordinance. Likewise they do not challenge the RVL zoning designation in general nor do they seek relief regarding any parcel other than their own. Instead, the owners challenge only the City’s refusal to even consider changing the RVL restrictions as they apply to the owners’ specific property and its particular circumstances.

The case the City mainly relies on, Traverso v. Department of Transportation (2001) 87 Cal.App.4th 1142 [ 105 Cal.Rptr.2d 179 ] (Traverso), is not a (c)(1)(B) case and in any event is distinguishable. Traverso was an attempt to “revive permits for billboards which were canceled over a quarter-century ago.” (Traverso, supra, 87 Cal.App.4th at p. 1143 .) The plaintiff’s theory, in a complaint filed no earlier than 1997, was that the original cancellation of those permits in the period 1973 to 1975 was invalid. There had been a request for “renewal” of one of the permits in November 1997, but the request had come in the wake of a new statute that provided that any permits not renewed after January 1 were “ ‘deemed revoked.’ ” (Id. at p. 1144.) The Traverso court’s point was that, substantively, the suit sought to set “aside permit revocations that occurred over 25 years ago” in a context where the plaintiff could not obtain “new permits.” (Id. at pp. 1145-1146.) The court said that the plaintiff had no cause of action for denial of the permits because state law precluded new permits. The claim for improper original cancellation back in 1975 was untimely a quarter-century later.

Because we conclude that the owners’ claim is from the 2007 denial of their application to develop the property and not from the 1993 or 1996 enactments, we need not deal with the ramifications, if any, of the trial court’s finding that the City acted in bad faith by effectively hiding the 1993 and 1996 enactments from the owners. Nor is there any need to discuss the “future generations” passages in both Palazzo and Travis that arguably indicate that state statute of limitations laws should not be interpreted to unreasonably preclude the opportunity to present federal law takings claims. (See Palazzolo, supra, 533 U.S. at p. 627 ; Travis, supra, 33 Cal.4th at pp. 770-771.)

5. Ripeness

In direct contrast to its statute of limitations argument (too late), the City argues that any inverse condemnation claim is unripe (too early), because the owners did not apply for the one thing that they might have under the RVL zoning, a single dwelling on their 2.85 acres.

*1279 Ripeness, as explained in Palazzolo , is a matter of final administrative adjudication. The idea of ripeness arises precisely because local authorities should have the chance to change their minds when a local restriction otherwise results in a compensable taking. When they do make up their minds, the case is ripe. Dunn v. County of Santa Barbara (2006) 135 Cal.App.4th 1281 [ 38 Cal.Rptr.3d 316 ] perfectly illustrates the rule. There the court held that the landowner’s regulatory takings claims were ripe for adjudication because the local authority, there a county, had decided it lacked any discretion to allow the subdivision of the property, leaving, as in Palazzolo and the case at hand, the residual possibility of developing a single-family residence.

Judging by the final administrative adjudication standard in Palazzolo , the City’s denial of the owners’ four-dwelling plan here was clearly “final.” In Palazzolo , the fact that the landowner had not applied to build the one house he was otherwise allowed anyway did not make the case unripe. It is the same here.

6. Damages

In its appeal, and in a cross-appeal, both the City and the owners challenge the amount of the $1.3 million inverse condemnation award. The City argues it is too high because the figure does not account for the value of the property with at least one house on it, even one built curbside, without the ocean view that would be otherwise available. The owners claim it is too small because the trial court candidly admitted that it “low-balled” the $1.3 million valuation claim. We agree with both parties and remand the case for a recalculation of damages.

a. The City’s appeal

Judge McEachen ruled the City committed a complete taking based on his finding the City would not allow any development on the property and that developing of the one home the RVL zoning permitted was not economically feasible. Judge McEachen therefore did not deduct the value of the property with one home from the value of the value of the property with four homes when he awarded the owners inverse condemnation damages.

We find no substantial evidence in the record that the City would not at least permit one house, built at curbside without an ocean view. Moreover, the City assured this court in this appeal that it will approve such a dwelling. The City will be judicially estopped in the future to change that position. (See People v. Castillo (2010) 49 Cal.4th 145, 155 [ 109 Cal.Rptr.3d 346 , 230 P.3d 1132 ] [judicial estoppel prevents a “ ‘ “ ‘party from gaining an advantage by *1280 taking one position, and then seeking a second advantage by taking an incompatible position’ ” ’ ”].)

Similarly, we find no substantial evidence to support the finding that developing of the property with one home was not economically feasible. It is simply inconceivable that absolutely no one would be willing to pay at least some money to own these 2.85 acres with the chance to put a house on it. The arguments that the property has no economically viable use even with one curbside house are not availing. The City indeed could, consistent with the RVL zoning in place, allow construction of one house, even under state open space laws. And the fact that neighbors might sue to prevent the construction of even the one curbside house only goes to the residual value of the property with the curbside house, not the absolute preclusion of that house.

A very large taking is not a total taking. Under the Penn Central factors test a total taking is not necessary for compensation. The trial court erred in failing to account for the property’s value with the one house the RVL zoning allows.

b. The owners’ cross-appeal

Like eminent domain, the basic measure of damages in inverse condemnation actions is market value. (City of Ripon v. Sweetin (2002) 100 Cal.App.4th 887, 897 [ 122 Cal.Rptr.2d 802 ]; Housley v. City of Poway (1993) 20 Cal.App.4th 801, 807 [ 24 Cal.Rptr.2d 554 ] (Housley).) Fair market value is determined based on the “ ‘highest and most profitable use to which the property might be put in the reasonably near future.’ ” (City of San Diego v. Neumann (1993) 6 Cal.4th 738, 744 [ 25 Cal.Rptr.2d 480 , 863 P.2d 725 ].) When an incorrect measure of damages is used, the trial court’s award should be reversed and the case remanded for a retrial on the damages issue. (See Housley, supra, 20 Cal.App.4th at pp. 810, 814.)

As explained in its statement of decision, the trial court calculated the damage award by accepting the lowest valuation of the property when zoned for four single-family houses, then deducted the highest cost estimate regarding the construction of the cul-de-sac needed to develop four homes on the property. The trial court did not deduct the property’s value when zoned for one single-family home. At the hearing on the City’s motion for new trial, however, the trial court explained that in calculating the damage award the court took into consideration the value of the property zoned for one single-family home even though it did not rely on any evidence regarding that value. In doing so, Judge McEachen was plain that he guesstimated damages by taking the lower expert opinion for the property’s value zoned *1281 for four single-family homes, then deducted the highest cost estimate for constructing the cul-de-sac: “I didn’t maybe factor that [(i.e., the property’s value when zoned for one single-family home)] in ... my statement of decision; but I did factor it in my thought process that I went through to get to that number. [f] I didn’t state how much a house would be, but I did low-ball what I thought was a big piece of property. I put a lower number in there than I would have normally had they not had some opportunity to build. [][] But quite frankly, the only opportunity that they would have to build was this little area fronting the street of Avenida San Juan which would be financially disastrous for the people to develop property like that. [][] So I actually low-balled a number to factor that in. While I didn’t put it in writing, I did take that into account. . . . And I took the plaintiff’s lower number and deducted the highest cost of the road. And so that was my factor in terms of coming up with a value. And I was looking for the fairest value I could, using my powers of equity to determine the value.”

We must remember that the court calculated the $1.3 million figure while it was under an erroneous impression of law, namely that it was going to enter a judgment directly obligating the City to pay $1.3 million. The court was disabused of that erroneous impression by the City in its new trial motion, when it realized that Hensler requires that the City have a choice of paying inverse condemnation damages or rescinding the land use restriction giving rise to the inverse condemnation damages. In light of the trial court’s candid recognition that it “low-balled” the damages award at a time when it did not realize Hensler required that the City have a choice to pay or rescind, we cannot say that in ascertaining damages the trial court was really focused, as the law requires, on fair market value. Under the circumstances, remand for recalculation in light of this opinion is appropriate. (Cf. Perico’s Enterprises, Inc. v. RUNS Enterprises (1992) 4 Cal.App.4th 238, 245 [ 5 Cal.Rptr.2d 470 ] [“We will, therefore, remand the issue of costs in order to allow the trial court to determine the reasonableness of the filing fee and to exercise its discretion in accordance with this opinion.”].)

On remand, the trial court will have the opportunity to calculate the true market value of the subject parcel, taking into account the fact it must still have some value even if the current land use restrictions remain in place.

A related but minor issue is prejudgment interest. The City makes no argument that prejudgment interest is, all else being equal, inappropriate. It simply says that if the current ascertainment of damages is overturned, the prejudgment interest award must be overturned too. The City is right, at least insofar as the exact figure is concerned. Prejudgment interest will need to be recalculated on remand so the City will know the precise figure it must pay if it elects not to comply with the writ of mandate.

*1282 7. Attorney Fees

a. The city’s appeal

The City’s attack on the attorney fee award is limited. It only argues that in the event of reversal, the fee award to the owners must be reversed. While we do reverse the judgment for a recalculation of damages, by no means does that determination affect the substance of who is the prevailing party. The owners still win. The City’s clear objective in this litigation has been to claim there has been no spot zoning, no regulatory taking and no viable lawsuit. And on those major points the City has failed.

b. The owners’ cross-appeal

There are two parts to the owners’ challenge to the attorney fee award as legally insufficient. The first part is the trial court’s decision not to award any fees at all for legal work done by one of the owners, Patrick O’Keefe, himself an attorney. The second part is the trial court’s perception that it was precluded from applying what it thought was an appropriate 20 percent multiplier for the fees of the owners’ attorney Everett Skillman.

Both issues are settled by the words “actually incurred” used in the governing statute, section 1036 of the Code of Civil Procedure. The statute, in its entirety, provides: “In any inverse condemnation proceeding, the court rendering judgment for the plaintiff by awarding compensation, or the attorney representing the public entity who effects a settlement of that proceeding, shall determine and award or allow to the plaintiff, as a part of that judgment or settlement, a sum that will, in the opinion of the court, reimburse the plaintiff’s reasonable costs, disbursements, and expenses, including reasonable attorney, appraisal, and engineering fees, actually incurred because of that proceeding in the trial court or in any appellate proceeding in which the plaintiff prevails on any issue in that proceeding.” (Italics added.)

Efere, there was no written fee agreement or any evidence that the partnership was billed for O’Keefe’s services. The owners never “actually incurred” fees for his work. And, as to the multiplier, the words “actually incurred” simply preclude application of a multiplier. While the owners argue that a reading of “actually incurred” which precludes a multiplier might mean cash-poor landowners would have trouble finding counsel in inverse condemnation actions, the answer to that is to convince the Legislature to rewrite the statute to make “reasonable fees” recoverable, as distinct from fees “actually incurred.”

Cases relied on by the owners for the possibility of a multiplier are distinguishable. City of Oakland v. Oakland Raiders (1988) 203 Cal.App.3d *1283 78 [ 249 Cal.Rptr. 606 ], was not an inverse condemnation action, but an eminent domain case. Salton Bay Marina, Inc. v. Imperial Irrigation Dist. (1985) 172 Cal.App.3d 914 [ 218 Cal.Rptr. 839 ], was a tour de force on the question of what constitutes reasonable fees in inverse condemnation actions where there is a contingency agreement. (In any event the case reiterated the “actually incurred” standard. (See id. at p. 954.)) And Parker v. City of Los Angeles (1974) 44 Cal.App.3d 556 [ 118 Cal.Rptr. 687 ], was likewise a contingent fee case.

DISPOSITION

We deny the owners’ motions to strike portions of the opening brief and the reply brief.

We affirm the judgment to the extent that it incorporates the provisions of the writ of mandate requiring the City to vacate the resolution denying the owners’ development application. We recognize the conditional nature of the inverse condemnation judgment. If the City complies with the writ of mandate it need not pay the recalculated judgment on remand.

Obviously there is a question of timing. The writ of mandate should remain stayed until after this court’s judgment is final and the trial court has the opportunity to recalculate the fair market value of the property, taking into account the fact that the City has committed in this appeal to allow one house built at curbside. Prejudgment interest should be recalculated. Upon such time as the conditional judgment becomes final, the City will have 30 days to comply with the writ of mandate, i.e., decide, under Hensler , whether to change its mind or not.

Assuming that the City chooses not to change its mind and comply with the writ of mandate, the conditional judgment shall be triggered at the expiration of the those 30 days. That should give the City the opportunity the Hensler decision demands.

Assuming the City does change its mind and complies with the writ of mandate, it obviously may impose reasonable conditions on the construction process to ameliorate the “noise” and “disruption” of any necessary grading. (See Landgate, supra, 17 Cal.4th at p. 1027 ; but cf. id. at p. 1029 [“a government agency may not evade the takings clause by . . . arbitrarily imposing conditions on development in order to delay or discourage that development”].)

We also affirm the judgment to the extent that it awards attorney fees and costs to Avenida San Juan Partnership. The owners remain the prevailing party.

*1284 Avenida San Juan Partnership shall recover its costs and attorney fees on appeal. The trial court shall determine the amount of attorney fees incurred by the partnership on appeal.

Moore, J., and Aronson, J., concurred.

The petition of defendants and appellants for review by the Supreme Court was denied March 14, 2012, S199533.