Triangulating satellite and administrative data can provide low-cost alternatives to primary data collection, but there are some limits. Mathematica used satellite data from the Google Earth Engine platform to estimate area under production and cultivation by season in the Delta Activity areas and to review and demonstrate changes in farming practices over time, which was triangulated with administrative data. As use of satellite data becomes more precise, including with higher resolution and development of accurate methods to identify crops, some changes can be measured without relying on local data systems. However, satellite and administrative data alone cannot answer all questions, including how farmers make decisions and why subgroups do or do not benefit from an intervention. In the final evaluation, Mathematica corroborated the remote data with interviews with farmers and cooperative leaders who provided insights on how intensification is happening, and who is doing it, revealing farmers faced constraints cultivating the same plot during multiple seasons. To cultivate rice in the adjacent hot dry season and rainy season, farmers must plant early, harvest their hot dry season, and quickly prepare the fields for the rainy season, which pushes farmers to choose between the two seasons. Therefore, despite the potential of satellite imagery, evaluations should continue to triangulate with qualitative and quantitative data for more holistic understandings of outcomes.
Lesson Learned