AGROMINDS - Historic and projected global agro-meteorologic indicators for major summer crops

Farmers have always watched the sky, knowing that a few weeks of heat or a delay in rain can turn a bountiful harvest into a failure. For scientists trying to predict how climate change will affect the food we eat, the challenge is similar but far more complex. They must track not just the weather, but how that weather interacts with the specific life cycle of a crop. A plant does not grow on a fixed calendar; it grows based on the heat it accumulates. In a warm spring, a crop might mature quickly, while in a cool year, it takes longer. Traditional climate data often averages conditions over fixed months, which can miss the critical moments when a plant is most vulnerable. To understand the future of agriculture, researchers need a way to measure the weather exactly as the crop experiences it, from the moment the seed is planted until the harvest is gathered.

A team of researchers at the International Institute for Applied Systems Analysis has created a new global dataset called AGROMINDS to solve this problem. Instead of looking at weather in generic monthly blocks, they built a system that follows the actual growing season of four major summer crops: maize, soybean, rice, and spring wheat. The dataset combines daily weather records with detailed information about when farmers plant and when crops are ready to harvest. By using computer models to estimate how quickly crops develop based on daily temperatures, the team can define the exact start and end of the growing season for every patch of farmland on Earth. This approach allows them to calculate a wide range of climate indicators—such as total rainfall, average temperature, and the number of days with extreme heat—specifically for the time the crop is actually in the ground.

The researchers used this method to create a comprehensive record of climate conditions from 1971 to the year 2100. They covered both historical weather, based on reanalysis of past observations, and future projections based on five different climate models and four different scenarios of how human emissions might change. The data is organized into a grid that covers the entire globe, focusing on areas where these crops are currently grown. For each year and location, the system calculates how much heat the crop accumulated, how much water it received, and how many days it faced stress from drought or scorching temperatures. Crucially, the system breaks the growing season down into smaller, biologically meaningful stages, such as the time when the plant is establishing its roots, when it is flowering, and when it is filling its grains. This level of detail helps scientists see exactly when a crop is most at risk.

When the team applied this dataset to look at the future, the results painted a clear picture of a warming world. In simulations for the end of the century, the average temperature during the growing season rises consistently across all regions and all climate models. The increase is particularly sharp under scenarios where greenhouse gas emissions remain high. Along with the heat, the number of days with dangerously high temperatures also increases, meaning crops will face more frequent heatwaves during their most sensitive growth stages. While rainfall patterns become more variable, with some areas getting wetter and others drier, the overall demand for water by the plants increases because hotter air dries out the soil faster. The dataset shows that in many places, the combination of higher temperatures and increased water demand will create more days where the atmosphere pulls more moisture from the soil than the rain can replace.

This new tool offers a more realistic way to study the relationship between climate and agriculture than ever before. By aligning the weather data with the actual biological rhythm of the crop, the researchers have provided a resource that can help scientists build better models to predict crop yields. It allows for a clearer understanding of how climate shocks, such as a sudden heatwave during flowering, might impact food production in different parts of the world. The dataset is designed to be flexible, allowing other researchers to adjust the parameters for different crops or regions, ensuring that the analysis remains relevant as farming practices and climate conditions evolve. Ultimately, AGROMINDS provides a detailed map of the climate challenges facing global agriculture, offering a foundation for developing strategies to protect food security in a changing world.