2026-06-29

The Same River, Twice:

A Watershed, Two Rulings, and the Cost of Waiting

A Water Policy Series — June 2026

Introduction

Two judges, two bodies of law, one watershed, one lesson.

In June 2026, a Los Angeles judge upheld the endangered listing of the Southern California steelhead and left United Water Conservation District constrained on the Santa Clara River — the case this series examined last week.¹ Eight months earlier, in October 2025, a Santa Barbara judge had decided the companion fight over the water beneath that same watershed, after a four-day trial: she fixed the safe yield of the Pleasant Valley groundwater basin at a number the City of Camarillo says will choke the $66 million desalination plant it built to wean itself off imported water.² The two cases share a river, an aquifer system, and a defendant. They also share an ending. In each, the losing party staked its case on attacking a computer model it found inconvenient; in each, the court declined to second-guess the model and pronounced the challenger’s alternative “speculative.” But the model fight was never the real contest. It was the closing act of a problem each agency had spent years declining to face — and losing it in court was the most expensive way left to lose.

The Ventura County water world is often described as a set of separate disputes — a fish case, a groundwater case, a desalter, a brine line. It is better understood as one hydrological story told in several courtrooms. The Santa Clara River that the steelhead needs is the same river United Water diverts to recharge the Oxnard Plain; the salt those recharge operations hold back from the coast is the same salt the Pleasant Valley adjudication turns on; and the desalter at the center of Camarillo’s distress is, in the end, a machine for pumping that salt back out to sea. This article follows that single thread — and arrives at a conclusion as applicable to a city over an aquifer as to a district on a river: under the law that governs both, once a model is understood to be accurate, it becomes the reality it describes — and you cannot litigate against reality. By the time you are reduced to disputing a model’s answer while conceding its accuracy, you have already lost the larger fight — the one that was winnable years earlier, when the problem the model describes was still yours to get in front of.

I. Two Rulings, One Watershed

Start with the geography, because it is the argument. The Oxnard and Pleasant Valley sub-basins are not freestanding aquifers; both are components of the Santa Clara River Valley Groundwater Basin — the steelhead’s river, underground. The Oxnard sub-basin spans roughly 58,000 acres along the coast and is actively intruded by seawater; Pleasant Valley covers about 19,840 acres inland and is, for now, intrusion-free but hydrologically tied to its coastal neighbor. Both were designated by the state as critically overdrafted in 2018.³ United Water sits at the hinge of both stories: it operates the Santa Clara River diversion the steelhead listing constrains, and it spreads that diverted river water to recharge these very basins.^{3 4} The agency this series last described as hemmed in by a fish is, in the groundwater case, pushing hard for tight pumping limits — fighting a two-front war over one river’s worth of water.

The Pleasant Valley ruling itself is precise. After a bench trial in September 2025, Judge Donna Geck issued a Phase 1 Statement of Decision setting the basins’ “safe yield” — the quantity that can be withdrawn each year on a sustainable basis. For the Oxnard basin, she found a total safe yield of 43,500 acre-feet per year and a native yield of 12,006 acre-feet per year; those numbers were undisputed. For Pleasant Valley, she found a total safe yield of 13,750 acre-feet.² A comprehensive adjudication like this one runs in stages: Phase 1 settles the size of the pie — the basin’s total safe yield; the phases that follow divide that pie into individual pumping rights, and ultimately impose a “physical solution,” the watermaster-run management regime the basin will operate under for good. October’s decision resolved only the first stage. But the size is destiny: a small pie, filled by existing users, leaves little for anyone who needs more. Camarillo took the loss up the appellate ladder on an emergency basis — a writ petition to the Court of Appeal in December 2025, then a petition for review in the California Supreme Court in April 2026 — and in May the Supreme Court declined to hear it, leaving the 13,750 figure standing while the case grinds on.^{2 5 6}

II. The Same Water

What makes this one story rather than two is salt, and the Pleasant Valley ruling is where the connection stops being rhetorical and becomes a finding of fact. For years the parties had argued over whether pumping in inland Pleasant Valley had anything to do with the seawater creeping into the coastal Oxnard basin; a consultant to the basin’s own sustainability agency had once called the data inconclusive.³ Judge Geck resolved it. On the weight of the expert testimony, she found that “groundwater pumping in the Pleasant Valley Basin causes lower groundwater levels in the Oxnard Basin, which in turn exacerbates seawater intrusion in the Oxnard Basin,” and she invoked the state regulation forbidding one basin’s management from undermining a neighbor’s sustainability.² The inland pumping and the coastal salt are now legally one system.

That is the same salt-versus-fresh balance the steelhead case runs on, viewed from below the ground instead of along the riverbed. United Water’s whole conservation mission — the river diversions, the spreading grounds — is a wall of fresh water built to hold the ocean out of the Oxnard Plain; pump too hard inland, and you lower the wall. The desalter completes the irony. Camarillo’s North Pleasant Valley Desalter, online since 2023, takes brackish groundwater the fresh-water users don’t want, drives it through greensand filters and reverse-osmosis membranes, and turns roughly 4,500 acre-feet of pumped water a year into about 3,800 acre-feet of drinking water — now the city’s largest single source of supply.⁷ The salt it strips out — the roughly fifteen percent that comes off the membranes as reject, on the order of 700 acre-feet a year of concentrated brine — does not vanish; it is piped into the Calleguas regional brine line along Somis Road and carried some twenty-two miles to an ocean outfall at Port Hueneme, and discharged into the Pacific.^{7 8} One agency spreads river water to keep the sea’s salt out of the aquifer; another pumps the aquifer, extracts the salt, and ships it back to the sea. They are managing opposite ends of a single problem, and the adjudication has now ruled that the two ends are connected.

III. The Stakeholders and the Number

Sort the parties by the safe-yield number, and the case will organize itself. A low number is a small, finite pie that today’s pumpers already fill — which suits the people doing the pumping. The case was brought in 2021 not by a city but by the OPV Coalition, a bloc of large agricultural landowners overlying the two basins — names like A. Duda & Sons’ Oxnard entities and Marathon Land — represented by O’Melveny & Myers and, in particular, Russell McGlothlin, a prominent groundwater-rights litigator whose firm’s work on the landmark Antelope Valley adjudication this article relies on elsewhere.^{4 9} The coalition did not stumble onto the model of a negotiated “physical solution”; it retained lawyers steeped in it. Tellingly, the suit names not the cities as defendants but the Fox Canyon Groundwater Management Agency — the basin’s own state-created sustainability regulator — and asks a court to override the agency’s chosen management of the water.⁴

By August 2025, the coalition had reached a settlement on the Phase 1 numbers, and the list of signatories is the real map of power in the basin: United Water, the Pleasant Valley County Water District, the City of Oxnard, and the large landowners all backed the low yield.^{2 5} Camarillo stood essentially alone against it — and not by accident, because Camarillo is the party most exposed. Roughly two-thirds of the city’s water now comes from local groundwater, the desalter chief among them; a constrained basin is, for Camarillo, uniquely an existential threat rather than a manageable irritation.^{7 5} That asymmetry is the same one this series noted in the steelhead case, where United Water carried the fight alone precisely because it held the concrete injury no one else did. In both basins, the party with the most to lose is the party that litigates hardest and stands most isolated.

There is a quieter constituency the number will bind without consulting: the small farmers and disadvantaged landowners who lack the resources to intervene in a decade-long adjudication. A comprehensive adjudication determines everyone’s rights, including those of pumpers who never appear; as a UC Davis legal clinic warned the region’s small growers, the choice is to join an expensive lawsuit or risk having your water right diminished without your input.³ Layered onto the doctrine that subordinates “dormant,” unexercised rights to near-zero, the result is a process that rewards whoever is already pumping at scale and quietly disadvantages whoever is not.

IV. When the Model Becomes Reality

Here is where the two cases rhyme so closely it reads like a moral. In Pleasant Valley, the safe yield was calculated with United Water’s groundwater model — and the decisive fact is that every testifying expert, Camarillo’s own included, agreed it was well calibrated.² That agreement was the whole case. Once a model is understood to be accurate, it ceases to be a description of the basin and becomes the basin itself — the reality the court reasons from. Camarillo had granted the model’s accuracy and then refused its answer. It complained that the calculation leaned on outdated streamflow data and ignored the salty inflows and the salt-removal its desalter performs, but it never built a competing analysis or put a different number in evidence. And fighting a description you have already agreed is accurate is not a fight; it is an objection the record overruled in advance. The judge treated it as one: she found the city’s evidence “speculative” and held that Camarillo “failed to offer any admissible or credible expert opinion… on the ultimate factual issues” — the actual yields.²

The collapse ran through Camarillo’s expert. On voir dire, the court limited his testimony on the yields after he conceded in deposition that his own opinions on the point were speculative; his trial testimony, the judge wrote, “did not square with these candid admissions,” and his “credibility was called into doubt.” He admitted he had never before worked on a groundwater model where the question was the advance of seawater intrusion, and acknowledged that lowering Pleasant Valley’s water levels “could cause seawater intrusion at the coast” — conceding, in effect, the causal link his client was contesting. Even his complaint about the “outdated data,” he allowed, would have only a “speculative” effect on the yield. “Argument is not evidence,” the court noted, declining to credit the alternative numbers the city floated in closing.²

Now read that against the steelhead ruling. There, United Water staked its case on the Cramer Life-Cycle Model, which it held out as a better estimate of the fish’s extinction risk than the agency’s science — but it never supplied its underlying mathematics or source code, so the agency’s staff could not evaluate it, and the court deferred to the science the agency could actually examine.¹ The pattern is identical: a party that dislikes the official model attacks it, but never does the work to replace it, and discovers that pointing at flaws is not the same as proving a different answer.

A careful reader should note where the analogy strains. These were different proceedings under different standards — the steelhead case was deferential appellate review of an agency’s administrative record, while Pleasant Valley was a trial judge weighing live expert testimony — and one should not pretend they applied the same legal test.^{1 2} But the underlying discipline is the same in both forums, and it is the practical lesson worth carrying out of both: when a decision turns on a technical model the experts have accepted, the burden is not on the model to be perfect; it is on the challenger to unseat the consensus that it is accurate — and you cannot unseat a consensus you have joined. Short of a more accurate one — not a friendlier one, a truer one — the challenger is not rebutting the science; he is disputing a description of the world everyone else in the room has already accepted, and no court can credit that.

Which raises the obvious question: why didn’t Camarillo, a city that found $66 million to build a desalter, do more than criticize the number that now threatens it? Its complaints were not frivolous — a safe-yield calculation that omits the salty inflows from upstream wastewater plants, or gives a desalter no credit for the salt it physically hauls out of the basin, is one a serious hydrologist can quarrel with. But the answer is not that Camarillo should have hired its own modelers to produce a friendlier figure. A model built to reach a number a party has chosen in advance is advocacy, not science; there is one best account of how much water the basin holds, and it is drawn from the data, not from the answer anyone needs — and a court that set aside United Water’s own preferred model in the steelhead case, for want of support, would have done the same to Camarillo’s. The real move was never to manufacture a rival reality but to help build the true one early — to have been in the room when the basin’s official science was assembled, where a seat is its own kind of power — and then to spend the years and the money not on litigation but on what an honest safe-yield number leaves no choice but to do: conserve harder, recycle more, and negotiate a share while a deal was still on the table. Camarillo poured its effort into a plant, and then, too late, into a courtroom. What it never did was get in front of the problem the model was only describing.

V. The Legal Framework

A few load-bearing doctrines make the rest legible. The first is the word the whole Pleasant Valley case turns on — safe yield, the older water-rights concept of how much a basin can give up year over year, which the court here used to size the pie. It is a cousin of, but not identical to, the sustainable yield that the Sustainable Groundwater Management Act asks agencies to manage toward; sustainable yield is capped by the worst of six “undesirable results,” seawater intrusion and water-quality degradation among them, and so is often the more conservative number.¹⁰ A related distinction — “native” yield (natural recharge) versus “total” yield (which can fold in imported, returned, or recycled flows) — is exactly the seam Camarillo tried to pry open, arguing the court improperly mixed legally distinct sources of water. The argument lost on the evidence, but the conceptual fight over what counts as supply is the one every adjudication eventually has.¹⁰

The second doctrine governs the division to come. California ranks groundwater rights: overlying landowners (the farmers) are senior; appropriators who move water elsewhere (a city’s system) are junior. Among overliers, the rights are “correlative” — shared — but a 2021 appellate decision in the Antelope Valley cases clarified that they need not be shared equally, only equitably, and that “dormant” rights belonging to owners who aren’t actually pumping can be subordinated to nothing.⁹ The practical rule that emerges rewards current beneficial use and punishes paper claims — and it is administered through a physical solution, the equitable, watermaster-run allocation that ends the rights trial, which the coalition’s complaint expressly seeks.⁴

The third doctrine answers a suspicion the case’s history invites. Three different judges handled this matter before it reached its merits — the entire Ventura County bench, then Judge Anderle, then Judge Geck — which can look like maneuvering. It is the opposite. California’s comprehensive-adjudication statute disqualifies every judge of a county that overlies the basin, assigns the case centrally through the Judicial Council, and — critically — bars the peremptory challenge that lets litigants shop for a friendlier judge elsewhere.¹¹ The Ventura bench was removed by operation of law; Judge Anderle was later disqualified for cause, on a motion pointing out that he owned property over the basin. Every change was a neutrality rule working as designed, in a basin so small that even the judges own a piece of the ground. There was no forum shopping here because the statute made it impossible.

VI. The Basin Next Door

The most instructive contrast is not in Antelope Valley or Santa Maria, the usual exemplars, but one subbasin over. The Las Posas Valley basin adjoins Oxnard and Pleasant Valley and is managed by the same agency, Fox Canyon. It was every bit as overdrafted, and it was adjudicated on the same clock — and it chose the other path. Rather than litigate the yield to exhaustion, the Las Posas parties negotiated a stipulated settlement and a physical solution; some 87 percent of the basin’s extractor-parties signed on to the allocation, Fox Canyon was installed as watermaster, and on March 5, 2026, the Second District Court of Appeal affirmed the whole framework, dormant-rights subordination and all.¹² Las Posas cooperated its way to a durable, appellate-blessed judgment. Pleasant Valley, next door, is still fighting — its safe yield set over one party’s objection, that party’s appeal rebuffed by the Supreme Court, and the actual division of water still unwritten.

The older comparators tell the same story over longer arcs. Antelope Valley spent more than fifteen years and some seventy thousand landowners reaching a settlement and a physical solution; Santa Maria was resolved on appeal in 2012 in favor of the public suppliers who negotiated.¹³ Each ended where Las Posas ended and where Pleasant Valley has not: in a negotiated, enforceable allocation that the parties built together rather than a number a judge imposed after a trial nobody fully won. The agencies that treated overdraft as a problem to solve jointly bought themselves certainty. The ones that treated it as a case to win bought themselves delay — which can look like a victory to a party that would rather keep pumping, except that a basin does not pause while the lawyers argue. Every year the case runs is another year of overdraft, the salt creeping further inland; the reckoning still comes, only later, costlier, and with less water left to divide.

Conclusion

Set the two rulings down side by side, and the through-line is unmistakable. One watershed, one threat — the slow chemistry of salt against fresh — and two agencies that could see it coming and waited anyway. United Water fought a fish listing; Camarillo fought a safe-yield number; both lost on the same ground and in the same way, contesting the evidence of a problem rather than the problem itself. And both now owe the very bills they went to court to postpone — a fish-passage retrofit on one side, a constrained desalter on the other.

Which is the real lesson, and it has little to do with models or expert witnesses. The watershed’s predicament — too little fresh water, too much salt, a basin and a river that can no longer give everyone what they long took for granted — has been visible since the wells first turned brackish two generations ago. The only question each agency faced was whether to get in front of it or wait for the law to force the reckoning. United Water waited, and the Endangered Species Act forced it. Camarillo built its desalter but never got in front of the fight over the water that feeds it, and the adjudication forced it. By the time the problem reached a courtroom, the only move left was to dispute that it was real — and you cannot litigate your way out of hydrology. The basin next door is the proof: Las Posas got ahead of its overdraft and negotiated the hard number while it still had a hand to play, and it is done, living with a deal it helped write, while its neighbors remain in court, bound by figures imposed on them. A straight oar looks bent where it enters the water. The agencies that thrive are not the ones insisting at trial that it is straight; they are the ones who dealt with the bend long before anyone reached for a judge.

Sources