New tool helps forecast olive harvests under climate change
The Universidad Politécnica de Madrid participates in recent smart farming research which combines artificial intelligence, satellite imagery, and climate data to predict olive and oil production in Mediterranean olive groves.
08.07.2026
A research team from the Universidad Politécnica de Madrid (UPM), in collaboration with the company AgrowingData and the ICAI School of Engineering at Comillas Pontifical University, developed a new methodology capable of improving early predictions of olive and oil yields in Mediterranean olive groves affected by drought and high temperatures. The study integrates Sentinel-2 satellite imagery, climatic variables, and soil properties to analyze how olive groves respond to environmental conditions amidst increasing climate variability.
The research was conducted in olive groves in the province of Córdoba, one of Spain's primary olive oil-producing regions. To this end, the team analyzed over 1,100 agricultural plots, combining remote sensing data with information on temperature, precipitation, and soil characteristics.
Location of the study area (Córdoba province, Spain) and spatial distribution of the olive grove geometries included in the dataset. Source: https://www.mdpi.com/2073-4395/16/7/722
A key advancement of the study is the use of "thermal time” or Growing Degree Days (GDD)—a metric that tracks the crop's physiological development based on accumulated heat rather than relying solely on calendar dates. "This approach makes it easier to compare widely different agricultural seasons and identify more robust production-related patterns," explains Ana María Tarquis, a UPM researcher who participated in the study.
The research findings, recently published in the scientific journal *Agronomy*, reveal that specific rainfall periods and vegetation development patterns observed via satellite are closely linked to future olive and oil yields. Furthermore, the study demonstrates that soil properties determine an olive grove's ability to withstand water and heat stress.
According to the authors, such tools could prove particularly valuable for cooperatives, farmers, and agricultural managers, enabling more accurate yield forecasting and optimized agronomic and commercial decision-making. The work also highlights the potential of combining artificial intelligence, Earth observation, and data science to advance toward agricultural models that are more sustainable, efficient, and adapted to changing climatic conditions.
The research is part of an industrial PhD program conducted jointly by AgrowingData and the Centro de Estudios e Investigación para la Gestión de Riesgos Agrarios y Medioambientales (CEIGRAM-UPM). This program focuses on applying artificial intelligence, data analysis, and satellite imagery to enhance the resilience of Mediterranean agricultural systems in the face of climate change.
Rosa Gutiérrez-Cabrera, Javier Borondo, Ana María Tarquis. Climate-Smart Framework for Olive Yield Estimation: Integrating Soil Properties, Thermal Time, and Remote Sensing NDVI Time Series. Agronomy 2026, 16(7), 722.