Skip to main content
Logo of Goethe university

Weather Forecasts Are Significantly Less Accurate in Poorer Countries

A powerful tropical cyclone swirls across the ocean, creating a swirling vortex of wind and water. The eye of the storm is visible in the center, surrounded by a... Elements of this image provided by NASA.

A study by Goethe University Frankfurt reveals substantial and persistent global disparities in the accuracy of weather forecasts.

A seven-day weather forecast in a high-income country is, on average, more accurate than a next-day forecast in a low-income country. This is the striking finding of a new study by Prof. Manuel Linsenmeier of Goethe University Frankfurt and Prof. Jeffrey Shrader of Columbia University, to be published in Nature Communications. Although weather forecasting has improved worldwide since the 1980s, a substantial gap in forecast quality persists between richer and poorer countries. Differences in how predictable the atmosphere is in different parts of the world play a role, but so do disparities in weather observation networks and forecasting infrastructure.

FRANKFURT. Weather forecasts inform countless everyday decisions – from farmers deciding when to sow crops and energy companies anticipating demand to public authorities preparing for heatwaves. But how useful those forecasts are depends considerably on where people live. Manuel Linsenmeier, Professor of Environmental Economics at Goethe University Frankfurt’s Center for Critical Computational Studies (C3S), and Jeffrey Shrader of Columbia University set out to map differences in forecast accuracy around the world and examine how they relate to countries’ levels of economic development. They analyzed global model forecasts produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) between 1985 and 2020 and compared them with observed weather conditions. Their main focus was short-term temperature forecasts, although they also examined predictions of precipitation and surface pressure.

The results reveal a striking pattern: on average, short-term temperature forecasts are more accurate in higher-income countries. The relationship holds for forecasts from one to seven days ahead. Remarkably, a seven-day forecast in a high-income country is, on average, more accurate than a one-day forecast in a low-income country. “For many people in Germany, checking a fairly reliable weather forecast on their phone whenever they want is something they barely think twice about,” Linsenmeier says. “But our findings show that the information underpinning those forecasts is distributed very unevenly around the world. That matters because weather information is particularly valuable in places where livelihoods depend on weather-sensitive activities or where people are especially vulnerable to extreme weather.”

 

Major Progress, But a Persistent Gap

Overall, weather forecasts have become substantially more accurate worldwide since the mid-1980s, with countries at all income levels benefiting from scientific and technological advances. Yet the gap between richer and poorer countries has narrowed only modestly. Geography accounts for part of the difference: short-term weather forecasting is inherently more difficult in the tropics, where many low-income countries are located. The global distribution of wealth therefore overlaps with the geographic distribution of weather predictability. Linsenmeier and Shrader estimate that geography alone accounts for roughly two-thirds of the worldwide variation in forecast accuracy. Differences in meteorological observation systems also play a role. Low-income countries typically have fewer surface weather stations and radiosondes – instruments carried into the atmosphere by weather balloons – while existing instruments in some countries report observations less frequently. The authors’ statistical analysis suggests that these gaps in observational coverage explain part of the variation in forecast accuracy.

Another factor is the capacity of national meteorological services. Only a small number of countries operate their own global numerical weather-prediction models. Most national weather services rely instead on forecasts produced by a handful of international centers, which they then refine using regional models, local observations and the expertise of their own forecasters. Doing so requires skilled staff, reliable data and substantial computing resources. As a rough measure of this capacity, the authors examined which countries submit official forecasts for their capital cities to the World Meteorological Organization (WMO). Around 76 percent of high-income countries do so, compared with just 19 percent of low-income countries.

“We cannot easily change the fact that weather is harder to predict in some parts of the world than in others,” Linsenmeier says. “But there are concrete ways to improve forecasts: strengthening observation networks, improving models and building the capacity of national meteorological services.” International cooperation, he stresses, is particularly important. “Weather observations and forecasts are a form of global infrastructure,” he adds. “Investing in better observations in underserved regions benefits local communities, but it also improves forecasts across national borders because forecasting models draw on data from around the world.”

 

Weather Forecasts as a Tool for Climate Adaptation

The study forms part of a broader research program led by Linsenmeier examining the social and economic value of weather information. Since summer 2025, he has been Professor of Environmental Economics at Goethe University Frankfurt’s Center for Critical Computational Studies; before joining Goethe University, he conducted research at Princeton University.

A central question in this work is how better weather information translates into tangible benefits for people. In a 2026 paper published in the Proceedings of the National Academy of Sciences (PNAS), Linsenmeier, Shrader, Stephan Thies, Laura Bakkensen and Derek Lemoine showed that more accurate temperature forecasts – particularly on hot days – reduce mortality. The researchers argue that as climate change increases the frequency and severity of extreme heat, further improvements in forecast accuracy will become increasingly important.

Linsenmeier is now exploring these questions in Germany. Working with a German health insurer, he is investigating which groups of people are particularly vulnerable to extreme temperatures and the extent to which weather forecasts and heat warnings issued by Germany’s national meteorological service, the Deutscher Wetterdienst (DWD), can help reduce the health impacts of a warming climate.

Editor: Dr. Dirk Frank, Press Officer/ Deputy Press Spokesperson, PR & Communications Office, Theodor-W.-Adorno-Platz 1, 60323 Frankfurt, Tel.: +49 (0)69/798-13753, [email protected]