Machine Learning and Remote Sensing for Impact Evaluation

This project uses satellite data to improve geospatial impact evaluation. The Scarf Fellow will work with research team members to:

Use large foundation models to improve measurement of economic and demographic changes using publicly available satellite imagery.
Use remote sensing data to adjust for otherwise unobservable confounding variables in environmental impact evaluations.
Use remote sensing data to evaluate the impacts of forest carbon offset projects on carbon sequestration and quantify the reliability of carbon credits.

Requisite Skills and Qualifications:

Students are required to have advanced R programming skills and have experience or have taken classes in at least one of the following areas:

Causal inference
Machine learning
Remote Sensing
Geographic Information Systems

We are looking for students who are interested in contributing to longer-term projects and potentially working with the SPIRES lab beyond the duration of the Scarf fellowship, either through a Tobin Fellowship or through continued work as a research assistant. In this position you will work with Prof. Sanford, postdoctoral scholars, PhD students, and other undergraduate RAs and can expect to develop skills across the above areas.