Combining satellite imagery and machine learning to predict poverty

TLDR

Nighttime satellite lighting is a coarse proxy for economic wealth, revealing that many developing countries have sparse illumination. The authors fuse publicly available nighttime lighting maps with high‑resolution daytime satellite imagery and apply machine‑learning techniques to generate accurate household consumption and asset estimates. The combined imagery yields accurate estimates of household consumption and assets in low‑income settings. Jean et al., Science, this issue p.

Abstract

Measuring consumption and wealth remotely Nighttime lighting is a rough proxy for economic wealth, and nighttime maps of the world show that many developing countries are sparsely illuminated. Jean et al. combined nighttime maps with high-resolution daytime satellite images (see the Perspective by Blumenstock). With a bit of machine-learning wizardry, the combined images can be converted into accurate estimates of household consumption and assets, both of which are hard to measure in poorer countries. Furthermore, the night- and day-time data are publicly available and nonproprietary. Science , this issue p. 790 ; see also p. 753