Spatio-Temporal Modeling of Violent Conflict and Fatality in Nigeria: A Point Process Modeling with SPDE Approach

Abstract For many decades, Nigeria has been plagued by a consistently high rate of violent events, resulting in countless fatalities and the displacement of citizens. This study aimed to model the spatio-temporal patterns of these events and the associated fatality to gain insight into the chain of events and provide a basis for swift and strategic intervention. To this end, a Cox point process model through the stochastic partial differential equation was adopted, taking into account the location randomness exhibited by violent events occurrence. The data analyzed was derived from the Uppsala Conflict Data Program – Georeferenced Event Dataset (UCDP-GED) version 22. The results revealed that violent events are particularly prevalent in the country’s northeast region, with a probability of 0.42 of at least one death occurring per violent event. These findings suggest a need for urgent intervention through informed policymaking, impeding the influx of illegal arms and ammunition in porous borders, and strategically tackling poverty.