Conflict Hotspot Trajectories in Somalia from 2015-2023: Descriptive Spatial Diagnostics to Inform Measles Program Monitoring

Background Conflict in Somalia is geographically uneven, with persistent concentration in specific areas that challenge humanitarian access and routine health service delivery. Program monitoring and situational awareness often rely on national summaries that mask subnational patterns. Descriptive spatial diagnostics provide a transparent approach to documenting where and to what extent conflict clusters evolve over time, supporting planning without requiring causal assumptions. Methods A descriptive spatial analysis was conducted for 2015-2023 using geocoded events from the Armed Conflict Location and Event Data (ACLED) dataset, aggregated to GADM v3.6 administrative level-2 (district) polygons. Violent incident types retained were Battles, Explosions/Remote violence, and Violence against civilians. Global spatial autocorrelation was assessed annually with Moran’s I using queen contiguity. Local clustering was identified with the Local Getis-Ord Gi* statistic; districts with z-scores > 1.96 were classified as hot spots and z < −1.96 as cold spots. Annual hotspot prevalence was defined as the percentage of districts classified as hotspots. Results Conflict showed significant spatial clustering each year (2015: Moran’s I = 0.27, z = 4.61, p < 0.001; 2023: Moran’s I = 0.33, z = 5.50, p < 0.001). The share of districts classified as hotspots increased from 5.4% in 2015 to a peak of 10.8% in 2019 and was 10.8% in 2023. Multi-year hotspot maps illustrate persistent concentration in specific areas rather than dispersed patterns . Conclusions Descriptive spatial diagnostics offer a reproducible approach to detecting and tracking conflict clustering in Somalia, revealing persistent geographic concentration from 2015-2023. These findings support sub-national monitoring and planning without relying on causal assumptions. All code and data sources are documented in Supplement A to facilitate transparency and reuse.