Bread, pasta and breakfast cereals sit far from the fields that produce them. But a new study published in the journal Earth's Future makes the supply chain considerably shorter: researchers have established that the extent of severe water scarcity across the world's major wheat-growing regions can explain around 74 per cent of annual fluctuations in global wheat prices between 2000 and 2021.

The international study, which includes researchers from Aarhus University, Mendel University, Rothamsted Research, the University of Cambridge, Duke University and more than a dozen other institutions, introduces a new metric called Severe Water Scarcity (SWS). Rather than measuring drought at the local or national level, SWS tracks the proportion of global cropland experiencing water stress during the growth stages when crops are most vulnerable. The distinction matters: it is not a single failed harvest that moves markets, but drought hitting several major production regions at the same time.

Historically, roughly five per cent of the world's wheat-growing area has experienced severe water scarcity in any given year. In particularly dry years (2000, 2010, 2012 and 2020), that figure rose above 15 per cent. Each of those years produced notable price spikes that were felt through supply chains globally.

Wheat showed a markedly stronger relationship to water scarcity than either maize or rice, for which researchers found no comparable price signal. That specificity has direct implications for the baking industry, which is more exposed to wheat commodity volatility than almost any other food manufacturing sector.

Professor Jørgen E. Olesen, Head of the Department of Agroecology at Aarhus University and a co-author of the study, notes that the evidence now moves beyond general statements about climate and agriculture:

"We already know that climate change affects agricultural production. What is new here is that we can document a clear link between widespread drought across the globe and the prices faced by consumers and food markets."

What the climate models project

Using climate projections, the research team modelled how SWS is likely to develop as temperatures rise. At approximately two degrees of global warming, the model estimates an average wheat price of around USD 273 per tonne. At three degrees, that rises to approximately USD 364 per tonne (roughly three times the inflation-adjusted global wheat price recorded in 2010).

The authors are careful to note that drought is not the only driver of wheat prices. Energy costs, fertiliser prices, grain stocks, trade policy and geopolitical conflict all contribute. What the study does is isolate and quantify the climate-driven component, which had previously proved difficult to measure. That is the methodological advance here: not a new argument about climate risk, but a demonstrated, data-backed link between compound drought events and commodity pricing.

Food security framing

The study's broader recommendation is that food security assessments should not limit themselves to gradual shifts in average crop yields. Short-term but severe and geographically widespread drought events (the kind projected to increase in frequency as warming continues) can cause price shocks that travel far beyond the regions where rainfall failed.

For the baking sector, that framing is useful. Ingredient procurement teams and supply chain planners already factor in year-to-year wheat quality variation. What this research adds is a quantitative basis for treating coordinated, multi-region drought as a systemic risk category rather than an exceptional event.

"Wheat is one of the world's most important food crops, which means climate-related production problems can quickly have global consequences," says Olesen. "This study shows that we need to view drought as an international challenge rather than a series of isolated local events."

The full paper is published in Earth's Future. Collaborating institutions include the Global Change Research Institute CAS, the International Institute for Applied Systems Analysis (IIASA), the Leibniz Centre for Agricultural Landscape Research, the University of Potsdam, Universidad Politécnica de Madrid, Rothamsted Research, the University of Cambridge and Aarhus University, among others.