The rural limits of conflict monitoring using nighttime lights

To reduce the suffering caused by armed conflicts, humanitarian organizations are exploring the use of remote conflict monitoring, including through satellite data, to complement human efforts in detecting crises early. Nighttime light (NTL) is a promising data source for such initiatives, as it can capture changes in human activities that may not have physical manifestations detectable by other satellite data. We assess a key requirement for using NTL data for conflict monitoring by examining how well baseline NTL—defined as minimal light emissions from an area—covers populations in conflict-affected areas. Combining NTL, population, settlement, and conflict data, we find that nearly a quarter of the global population, and almost a third of those affected by violence, live in areas without detectable NTL. In many conflict-affected regions, the population density at which NTL is reliably detected is much higher than in other regions due to the mutually reinforcing relationship between poverty, weak states, and armed conflict. Additionally, even within conflict countries NTL coverage is biased toward urban settlements and hardly reaches peripheral areas, indicating that the availability of data often considered “objective” shares a bias with news-derived conflict data. Our findings suggest that NTL should serve as a complementary, rather than a standalone, data source for conflict monitoring.