Warming is not drying the world evenly. It is moving the water.

Water is moving, not just disappearing

Warming is often described as drying the world out. Three preprints published this week suggest a more useful description: it moves water around. A drought attribution study, a mountain snowfall projection and a reforestation experiment all reach the same shape of result, that some places get more water, others less, and the winners and losers are decided by circulation and evaporation rather than by a simple change in rainfall. For anyone planning water supply, hydropower or land use, the question is not whether there will be less water, but where it will go. m:…5474 m:…8858 m:…3419

A drought that heat made longer

The drought case is the Western Mediterranean drought of 2021 to 2024. In a preprint, anthropogenic warming prolonged and intensified it by increasing evaporative demand, the atmosphere pulling more moisture out of soil and plants, rather than through a shortfall of rain alone. In Catalonia, extreme-drought duration rose from 38 days under 1950s climate conditions to 61 days in the present-day climate. Under a scenario of two degrees of additional warming, the duration reaches 94 days and the affected area reaches 98.6 percent. The mechanism matters for planning: reservoirs and crops can be sized for a rainfall deficit, but a drought driven by hotter air evaporating water faster behaves differently and lasts longer. m:…5474

Snow is moving north, not disappearing

Mountain snow shows the same redistribution. A preprint using the CMIP6 multi-model ensemble, checked against NCEP/NCAR and ERA5 reanalysis, reproduces historical extreme snowfall climatology with an R-squared above 0.89 and error below 6.5 percent, which gives its projections more weight than a model that cannot reproduce the past. Its future picture is split: extreme snowfall increases poleward of 60 degrees north and falls in mid- to low-latitude mountain ranges. Changes in synoptic-scale circulation, the large weather patterns that steer storms, are the dominant mechanism behind the split, not a uniform warming effect. Ski resorts, hydropower operators and flood managers in different latitudes should therefore expect opposite problems from the same warming. m:…8858

Planting trees moves rain somewhere else

Reforestation does the same thing in a different way. Large-scale tree planting changes rainfall not mainly by recycling local moisture, but through large-scale circulation adjustments. In the preprint's coupled Earth system model experiments, reforestation weakens the Atlantic overturning circulation, which shifts the tropical rain belt southward, and uncoupled models and offline moisture-tracking diagnostics miss these responses entirely. That means a reforestation plan judged only on the carbon its trees store, or on the rain it returns locally, can miss the remote effect: rain falling somewhere else, or not falling at all. Where trees are planted, and at what scale, decides which of the competing circulation mechanisms wins. m:…3419

The pattern

The three studies do not agree on a single number, and they should not be read as one forecast. They agree on the shape of the change: warming and land use move water between regions through evaporation and circulation, and the local outcome can run in either direction. Plans built on the assumption that a place simply gets drier or wetter will misjudge both the risk and the opportunity. The useful question for water planners is which mechanism dominates in their basin, evaporation, storm-track shifts, or a remote circulation change, because each one points to a different adaptation. m:…5474 m:…8858 m:…3419