03. Juni 2026
Current crop models systematically underestimate the water requirements for crop irrigation—with potentially serious consequences for global food security. This is shown by a new study from the Leibniz Centre for Agricultural Landscape Research (ZALF), published in the journal Nature Water. The researchers identify three key weaknesses in modeling that lead to inaccurate forecasts.
Crop models that calculate water requirements for climate change impact assessments often use standardized automatic irrigation schemes. These assume that fields are irrigated whenever the soil falls below a certain moisture level. “These models treat water as a resource that is directly available at the field—similar to rain,” explains Dr. Ehsan Eyshi Rezaei, lead author of the study. “But in reality, water is a lateral resource: it flows in rivers, is pumped from aquifers, or stored in reservoirs. Its availability depends on how much water other users in the same watershed are withdrawing.”
Example: A model calculates that all fields in a river basin can be sufficiently irrigated. But in practice, farmers upstream withdraw so much water that there is no longer enough left downstream. The result: The models overestimate the buffering effect of irrigation, especially in water-limited regions.
Temperature differences: Why plants need more water than assumed
Irrigated crops are often 3 to 6 degrees cooler than the surrounding air due to evaporative cooling. This effect is particularly relevant during heat waves when the evaporation rate is high. “Most models use air temperature to calculate water requirements,” says Rezaei. “But plants experience a cooler canopy temperature, which increases water requirements.” Two effects exacerbate this problem:
- Increased evaporation demand: The warmer the air, the more water must evaporate to maintain the cooling effect.
- Extended growing season: Cooler crop canopy temperatures slow plant development, which extends the irrigation period.
Efficiency Gaps: Why Irrigation Is Often Inefficient in Practice
Crop models usually treat irrigation efficiency as a fixed value. But efficiency depends on the climate, as in a warmer atmosphere with higher vapor pressure deficit, more of the applied water evaporates or drifts away before reaching the roots, so less arrives at the crop than the models assume. In high-income regions such as Europe or the U.S., where irrigation is well managed, these climate-driven physical losses dominate. In resource-limited regions, a second problem dominates - practical obstacles that lead to inefficient irrigation:
- Logistical challenges: Shared water pumps or limited infrastructure.
- Water rights: Prioritization of certain crops (e.g., vegetables instead of staple foods).
- Costs: High operating costs lead farmers to allocate water based on the value of the crop.
“Models assume optimal conditions, but in practice water is distributed unevenly”, says Rezaei. “The different gaps pull in opposite directions, but together they mean we are underestimating either the water that will be needed or the regions where adaptation will fail.”
Consequences for food security
The authors stress that this is not an argument against irrigation, but a call for more realistic assessment. The tools to do better already exist: crop models have been successfully coupled with energy-balance and hydrological models. Yet these models are rarely used in the large-scale assessments that guide food-security policy and irrigation investment.
The study shows that current projections of water demand are fraught with significant uncertainties. “If we rely on these models, we risk irrigation systems being undersized,” warns Rezaei. Regions already suffering from water scarcity today, such as parts of India or the U.S., are particularly affected.
The researchers call for:
- Better data: Measurements of actual irrigation practices and water availability.
- More realistic models: Coupling climate models with hydrological data to account for lateral water flows.
- Local adaptation: Incorporation of social and economic factors into planning.
Project partners:
- Leibniz Centre for Agricultural Landscape Research (ZALF), Müncheberg
- University of Potsdam, Institute of Biochemistry and Biology
- Global Change Research Institute CAS, Brno (Czech Republic)
Further information:
DOI: https://doi.org/10.1038/s44221-026-00642-9
Note on the text:
This is a summary of the original text generated using artificial intelligence:
Eyshi Rezaei, E., Nendel, C. (2025). Gaps in irrigation representation constrain projections of water security under climate change. Nature Water.
https://doi.org/10.1038/s44221-026-00642-9
The text was carefully reviewed and revised in accordance with ZALF’s AI guidelines.
