Satellite Analysis for Measuring Human Progress

Exploring deep learning and satellite imagery to measure poverty, vulnerability, and global progress.

In a rapidly globalizing world, the pursuit of accurate, comprehensive data on welfare and development has become imperative, especially to make progress on alleviating poverty, social vulnerability and climate change. Satellite imagery offers a nice standardized way to explore this idea.

In my final year at Bowdoin, I wrote a thesis on deep learning techniques for measuring poverty and vulnerability, focusing on Campinas, Brazil as a case study. This forms the basis for my continuing work on understanding how the rich, diverse and open‑source set of satellite imagery, including multispectral imagery, can help humans tackle global issues including climate change, poverty and social vulnerability.