Many of the world’s most agriculturally productive regions sit on top of fossil aquifers — underground water reserves that accumulated over geological time scales and that recharge from current precipitation only very slowly, if at all. These aquifers have been pumped, principally for irrigation, at rates that exceed recharge for most of the past half-century. The High Plains Aquifer in the central United States, the North China Plain Aquifer, the Indus Basin Aquifer, and several large regional aquifers in the Middle East are all in measurable, and in some cases severe, decline.
What aquifer depletion actually looks like
The local symptoms are familiar: wells must be drilled deeper to reach the water table, pumping costs rise, well yields decline, and in some regions the land surface itself subsides as the water that previously supported the soil structure is removed. Farmers respond by switching to less water-intensive crops where possible, by abandoning marginal land, and by investing in drip and precision irrigation systems that reduce per-hectare water use.
None of these adaptations changes the underlying arithmetic. The aquifers are being drawn down toward exhaustion. In the North China Plain, the rate of decline has slowed in some sub-regions where governmental controls on planting have been imposed. In most other major depleted aquifers, the rate of decline has not.
Surface water has its own version of the problem
Major snow-fed river systems — the Colorado, the Indus, the rivers fed by Himalayan glaciers, the rivers of the European Alps — are seeing earlier and smaller spring runoff peaks as snowpacks shrink and melt earlier in the season. The consequence is that the same total annual flow arrives at a different time of year than it did historically, with reduced flow during the late-summer period when irrigation demand is highest. Reservoir capacity built for the historical hydrograph is sized incorrectly for the new one.
The Colorado River Basin’s allocation system, established in the 1922 compact, was based on a flow estimate that turned out to have been measured during an unusually wet period. The river has produced less water than the compact promised in most years since, and the ongoing renegotiations between the basin states are an early case study in how water-allocation institutions adapt — or fail to adapt — when the underlying resource shrinks.
The agricultural and political question
Approximately seventy percent of global freshwater withdrawals go to agriculture. The leverage on water demand is therefore principally in agricultural decisions: which crops are grown where, what irrigation methods are used, what subsidies and water-pricing structures the state imposes. These decisions are politically difficult in nearly every jurisdiction because they affect rural livelihoods, food prices, and historical water-rights claims that often predate any of the relevant regulators.
The water question is, more than most environmental questions, fundamentally a political-economy question dressed in hydrological clothing. The physical resource is straightforwardly measurable. The institutions for allocating its decline are, in most regions, not yet built.