Humanitarian Crisis Detection With Nightlight

Early detection saves lives but many humanitarian crises begin in places that are hard to reach or monitor. This project explores the use of nighttime lights (NTL) satellite data to enhance the detection of humanitarian crises. It is a collaboration between the International Committee of the Red Cross (ICRC), experts in qualitative and computational conflict research at CSS, and remote sensing and machine learning scholars at EPFL. The project is funded by the Engineering for Humanitarian Action (EHA) initiative.

This project aims to identify how and under what conditions NTL can be used to measure different aspects of humanitarian crises. In contrast to other satellite data, NTL reflects human activity and can therefore capture the impact of armed conflicts on populations. The outcome of our research will be a tool designed to detect humanitarian crises in conflict contexts by using NTL data. Additionally, we will produce a report for practitioners, outlining the tool's applications and limitations. The aim is to make the tool usable by ICRC and other humanitarian actors, enhancing their ability to respond to crises in a timely manner.

ETH PIs: Dr. Corinne Bara

ETH project team: Dr. Valerie Sticher and Dr. Jenniina Kotajoki

Partners: Prof. Dr. Devis Tuia (PI, EPFL), Filip Dorm (EPFL), Dr. Emanuele Dalsasso (EPFL), Dr. Thao Ton-That Whelan (ICRC)