California is sinking. In the Central Valley, the most productive agricultural region in the US, some areas drop an inch or two per year. Telephone poles slump, roads crack, canals fail. In time, all that sinking adds up. A recent state survey found one patch of farmland off I-5 near the town of Arbuckle had fallen 2 feet in nine years.

The culprit: overdrafted aquifers. The process speeds up during periods of drought, when rivers run dry and farmers scramble to find other ways to water their fields. “It’s a tragedy of the commons situation,” says Alex Johnson, fund director for the Freshwater Trust, a conservation nonprofit. “Everyone is incentivized to pull as much [water] as they can as fast as they can.” Usually, that means less water for future droughts; the rare wet years in between fail to replenish California’s subterranean emergency fund.

That’s poised to change under new state rules. A 2014 state law for the first time laid out guidelines for stabilizing vulnerable groundwater basins. It largely left the task of meeting those goals to local groups, who were asked to develop plans to reduce their use to a sustainable level by 2040. In the most parched areas of the state, those plans are due next year. The new rules come with a potential sweetener: the ability to sell groundwater to neighbors who can make better use of it. Given the option of selling, a farmer might, for example, switch to a less water-intensive crop—say, from alfalfa to strawberries. It’s a local cap-and-trade regime for groundwater, basically.

Sounds good, in theory. But the question is how local farm groups are supposed to get those markets started and trade a resource that, until now, has been fiercely contested but barely even tracked.

In Solano County, near Sacramento, Johnson is working on what he says could be a model for parched ag regions around the state. It involves throwing a lot of tech at the problem. Last month, working with IBM and a company called SweetSense, Johnson’s team began deploying simple, solar-powered sensors, originally developed to monitor creaky groundwater pumps in East Africa. The sensors will be used to detect how much water is flowing in real-time. Since cell service is poor, the data will be sent via pint-sized satellites known as cubesats, courtesy of a startup called Swarm. Farmers will use that data to trade their water on (what else?) a blockchain platform.

Water on the blockchain? Your eye-roll is forgiven. But local water trading, in one form or another, is likely to soon become a force in California agriculture, says Ellen Hanak, director of water policy at the Public Policy Institute of California. In a study published in February, Hanak found that to comply with the 2014 law, farmers in the San Joaquin Valley south of Sacramento could be forced to leave 15 percent of their fields fallow each year. That translates to about $3.5 billion in lost crops. Local water trading, she says, could reduce the losses by nearly half, simply by encouraging a better distribution of water.

So far, there hasn’t been much progress toward setting up those markets. While surface water trading is common in California, governed by a warren of statutes and water rights, there’s no similar framework for groundwater trading. And it has unique complications. To start, water users need to form local groups and agree on who gets what allocation to start, which could be a years-long process. Then come the geological headaches. The geography of underground aquifers is mysterious; what looks like a distinct basin might, in fact, be linked to another or divvied into smaller parts. If a market isn’t designed to take those complexities into account, UC Berkeley researchers say, water trading could inadvertently suck part of an aquifer dry. But the biggest problem may be accounting: a reliable way to tell how much water is pumped, and by whom, to keep the water traders honest.

In Solano County, Johnson is hoping to prove the new tech could help water users better understand their water use and come to a consensus. The idea, he says, is to build a system that’s cheap (hence the sensors developed for rural East Africa) and trustworthy. They’re testing it in a place with relatively low stakes, he notes. Thanks to its proximity to the Sacramento-San Joaquin Delta, the region’s groundwater situation isn’t particularly dire. But the region still needs to come up with a groundwater plan that incorporates a potentially drier future. “What we’re trying to do is make sure the plan is flexible,” says Chris Lee, an environmental specialist in the Solano County Water Agency. “Where we can get away without installing $4,000 water gauges, using something like satellite imagery or these sensors, we’ll absolutely do that.”

It’s on the point of trust where blockchain is helpful, says Johnson. “Water users are a suspicious bunch. They don’t love government interference and suspect other water users are going over their allocations,” he says. That makes water and blockchain a good fit, he argues: It allows a group of people who don’t necessarily trust each other to make deals, without the need for third-party oversight. The plan, when the blockchain pilot starts this summer, is to allow only landowners to see their water use data. The deals themselves will be conducted via smart contracts—computer programs that execute when certain conditions, such as pumping levels, are met.

It’s unclear whether farmers will find the whiz-bang of blockchain and sensors useful in practice. But Johnson says that’s the point of testing in Solano before going to areas where groundwater cuts will be more divisive and drastic. As Hanak points out, changes—and likely, new technology—are coming, but it’ll likely be a lengthy pivot. “It’ll all depend on how comfortable people are working together,” she says. “We’re in the early days of this.”