Session summary
This GeoField 2026 open session brings together four presentations on how Earth observation can improve agricultural, environmental, and policy-relevant evidence, while also introducing new measurement, validation, and implementation challenges. The presentations include research and operational examples from Mali, India, Senegal, Zimbabwe, and multi-country agricultural datasets.
Moderator: Aaron Eubank
Geospatial Impact Evaluation of Ag Intensification in Mali
Dilini Abeygunawardane presents a geospatial impact evaluation of an agricultural intensification project in Mali. The study examines whether an intensification package for sorghum and millet generated land-saving effects or rebound effects in the Sikasso region. Using Landsat-derived measures of woody cover and cropland, field validation data, matching, and spatial difference-in-differences, the analysis shows how Earth observation can help assess land use change beyond conventional agricultural outcomes.
Cloudy With a Chance of Significance: Specification Search and the Illusion of Exogeneity in Remotely Sensed Weather Data
Jeff Michler presents research on the use of remotely sensed rainfall data in economics and agricultural research. The study compares multiple rainfall products, metrics, dependent variables, and model specifications across multi-country household datasets. The presentation highlights how different choices of rainfall product or specification can change results, underscoring the need for clearer theoretical justification, robustness checks, and caution when using rainfall as a causal predictor or instrumental variable.
Integrating Earth Observation for Assessing Agricultural Impacts in High-Biodiversity Landscapes of India
Aditya Sharma presents a framework for identifying environmentally stressed agricultural regions in Assam, India. The study combines indicators related to biodiversity, climate variability, soil and water stress, socioeconomic dependence, population pressure, and emissions to create a composite stress index. By overlaying this index with cropland data, the analysis identifies agricultural areas where environmental vulnerability and production pressures intersect, supporting future planning, field validation, and biodiversity-sensitive agricultural policy.
From EO Workflows to Operational Impact Assessment: Implementing Sen4CAP Approaches through the ESTELLA Project in Senegal and Zimbabwe
Lorenzo de Simone presents FAO’s work to integrate Earth observation into operational agricultural monitoring systems in Senegal and Zimbabwe. Through the ESTELLA project, FAO supports national partners in improving georeferenced survey data, crop type mapping, crop area estimation, drought monitoring, farmer registries, and seasonal agricultural baselines. The presentation emphasizes that Earth observation is most useful when embedded in national systems that are routinely maintained, budgeted, and used for decision-making.
Together, the session highlights both the promise and difficulty of using Earth observation in agricultural decision-making. EO data can reveal land use responses to intensification, improve crop area estimation, identify stressed agricultural landscapes, and support operational monitoring systems. At the same time, the presentations show that results depend on measurement choices, validation data, institutional capacity, data governance, and careful interpretation.