Abstract
Oil spills cause environmental pollution with a serious threat to local communities and sustainable development. Accidental oil spills can be modelled as a stochastic process where each oil spill event is described by its spatial locations and incidence-time and hence allow for space-time cluster analysis. Spacetime cluster analysis can detect space-time pattern distribution of oil spills which can be useful for implementing preventive measures and evidence-based decision making. This study aims to detect the space-time clusters of accidental oil spills in Rivers state, Nigeria through the Space-time Scan Statistic. The Space-time Scan Statistic was applied under the permutation model to the oil spill data (each for sabotage and operational oil spills) collected at Local Government Area (LGA)-level during the period from 2011 to 2019. The results show that the sabotage oil spill clusters have covered most of the LGAs in the southern part of the state at the start of the study period and then in 2018-2019, it moved to the west covering a single LGA. The operational oil spill clusters covered two neighboring LGAs in the south. In addition, the temporal cluster of sabotage oil spills was seen in 2019 and operational oil spills in 2011-2012. The sabotage oil spills show an increasing trend with the maximum in 2019 while the operational oil spills show a decreasing trend with the minimum in 2019. These findings assist in more effective decision-making for combating the environmental problems and controlling the future spill incidence in the cluster-regions.