Remote Sensing Analysis for Documenting Human Rights Violations in Zones of Armed Conflict: A Systematic Review of Empirical Research

Raphaela Edler^1*Georg Glasze¹Katrin Kinzelbach²Blake Byron Walker¹

• ¹Institute of Geography, Department of Geography and Geosciences, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Bavaria, Germany
• ²Institute of Political Science, Chair of Human Rights Politics, Friedrich-Alexander University Erlangen-Nuremberg, Erlangen, Germany

This paper investigates the use and potential role of satellite remote sensing (RS) data in documenting conflict incidents and related human rights violations. In particular, we scrutinized the methods applied by researchers and human rights practitioners as well as the characteristics of those human rights violations that were documented on the basis of analysis of remote sensing data. To this end, we reviewed 901 articles, out of which we chose 48 articles that comprised 51 empirical studies for an in-depth review analysis. The articles included in our literature corpus were published between 2006 and 2023. Our review of the studies focused on the following aspects: research organisations, geography of the conflict, spatial resolution of RS data, methods, validation, reported challenges. Despite the salience of human rights violations in recent history and the adoption of digital data-driven investigations in this context, we find that an increased focus on human rights is not detectable in RS-based research. Employing a critical remote sensing perspective, we result that a few conflicts dominate the attention of RS analyses, while other conflicts remain unobserved. As most studies (still) rely on visual interpretation, which requires very high-resolution data, we detect widespread dependencies on commercial providers. A perceptible rise in the use of machine learning approaches for documenting conflicts goes hand in hand with a similar increase in the involvement of academic researchers. The latter's involvement changes validation processes to the extent that these processes shift towards relying exclusively on remote data instead of local sources. Our findings highlight the untapped potential for interdisciplinary research and emphasize the need for more collaborative work, conjoining remote sensing and human rights expertise.