The Targeting of Civilian Buildings in Internal Armed Conflict: Uncovering and Explaining Spatial Patterns Using Satellite Imagery

Armed conflicts often cause extensive damage to civilian buildings hosting essential services, cultural property, and residential areas but do so to vastly varying degrees. While ample scholarly attention has been dedicated to violence against civilians during armed conflicts, the nuances of damage inflicted on civilian buildings are far less understood. This dissertation addresses this gap by proposing a project on the question of why civilian buildings are being targeted in armed conflicts. In doing so, it makes three key contributions. First, it will introduce a comprehensive database on conflict-related building damages by harnessing open-source satellite imagery and advanced methods from remote sensing and machine learning. The database will be interoperable with existing conflict event data, allowing for integration into conflict research. Second, this integration is made possible by conceptually defining civilian buildings, differentiating harm to civilians from damage to civilian buildings, and introducing conflict damage events as a new subtype of conflict events. Third, theoretical expectations about actor-specific patterns are established, focusing on the strategic damaging of civilian buildings by armed groups in internal armed conflicts. The proposed arguments highlight factors such as an armed group's interwovenness with the civilian population, strategic logics for one-sided violence, and the spatial separability of civilian supporters of the armed groups. Through the integration of these three key contributions, this proposal advances the understanding of conflict-related damages to civilian buildings while paving the way for future research endeavors. The results of this project have the potential to improve the protection of civilians, their homes, and the infrastructure they rely on.