Weather forecasts are significantly less accurate in poorer countries, analysis finds

This analysis, published in Nature Communications, uncovers significant disparities in the accuracy of weather forecasts across countries of varying economic development. Researchers Manuel Linsenmeier from Goethe University Frankfurt and Jeffrey Shrader from Columbia University examined global model forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF) from 1985 to 2020 and compared them to actual weather observations.

Their study focused on short-term temperature predictions, but also looked at precipitation and surface pressure forecasts. The key finding is that higher-income countries consistently enjoy more accurate temperature forecasts, even when the forecast interval ranges from one to seven days. A seven-day forecast in a high-income nation is more reliable than a one-day forecast in a low-income country.

This disparity highlights the uneven distribution of weather forecast information globally, which is crucial for communities where livelihoods are directly affected by weather conditions or where people are more vulnerable to extreme weather events. While overall global weather forecasting has improved since the mid-1980s due to advancements in science and technology, the gap between the most developed and least developed nations has not narrowed substantially.

Geographic factors, particularly the challenges of forecasting in tropical regions where many low-income countries are situated, contribute to the variance. Low-income countries also typically have fewer weather monitoring stations and observation systems, and their existing instruments may not report data as frequently as in wealthier nations.

The capacity of national meteorological services plays a significant role, with only a limited number of countries operating their own global weather prediction models. Most rely on forecasts from a few international centers, requiring skilled personnel, reliable data, and substantial computing resources. The study emphasizes that while the physical unpredictability of weather in certain regions is unavoidable, improvements can be made by enhancing observation networks, advancing weather models, and building the capabilities of national meteorological services.

International collaboration is crucial, as weather observations and forecasts constitute a form of global infrastructure. Investing in better observations in underserved regions not only benefits those local communities but also enhances forecasts worldwide, given that these models rely on data from around the globe. This research is part of a broader initiative by Linsenmeier to explore the social and economic value of weather information, including its impact on public health, especially in the context of climate change-induced extreme heat.