Spatiotemporal Evolution of Conflict and Risk in South Asia Based on Multi-Source Data

Abstract

South Asia shares a large territorial border with China, and is the latter's land "position" to enter the Indian Ocean and an important part of the "21st Century Maritime Silk Road." Based on the conflict data in South Asia collected from the Armed Conflict Location and Event Database, the Global Terrorism Database, and the Uppsala Conflict Database, the spatial and temporal changes and intensity distribution of conflicts during 2011-2020 were explored using geostatistical methods such as nuclear density, spatial autocorrelation, and spatial hot spot detection and analysis. The results revealed the following: (1) From 2011 to 2020, the number of conflicts in South Asia increased, but the number of casualties decreased. From the perspective of spatial distribution, regional conflict events are concentrated in the "outer ring of South Asia," and the cluster centers form a "trapezoid" pattern. The Kashmir-Sindh line is a trapezoid at the top bottom, and the Dhaka-Chettisgar-Tamilnad-Colombo line is a trapezoid at the bottom. The conflicts in South Asia exhibit a tendency of similar aggregation, and their distribution is affected by the surrounding region. The secondary administrative regions with positive spatial autocorrelation accounted for about 1/5 of the total, and the aggregate effect of war events was the most significant, andthat of armed conflict events was weak. The distribution of H-H clustering regions for each event type was inconsistent, but L-L clustering regions were mainly distributed in the center of the continent. (2) The distribution range of conflict intensity first increased and then decreased, reaching the maximum in 2017. In Pakistan, there were many areas of conflict intensity distribution, which did not change significantly at the end of the period. In India, the range of conflict intensity distribution first increased and then decreased. The intensity hot spots first spread to Baluchistan, and then shifted to the northeast and southeast, forming two hotspots centered on Kashmir and Chhattisgarh in 2020, with no cold spots. At present, the activities of non-state actors are more fluid and involve more areas, so it is difficult to form new hotspots. Kashmir will remain a hotspot of conflict, and the concentration trend of Chhattisgarh will decline. Baluchistan should be watched closely. (3) The regional conflicts in South Asia are affected by political parties, and power, identity, natural, and other factors. The source of the conflicts is the differences in national identity and religious belief, and they are promoted by state actors, ethnic separatist organizations, religious extremist organizations, and other actors. State factors play a leading role, while the religious and ethnic factors are intractable diseases. The current study has the following limitations. First, this study utilized a large amount of data, with more loyal manual participation in data fusion, cleaning and other work, and there is room for further improvement of accuracy. Second, the driving mechanism of South Asia has not been deeply discussed. The causes of conflicts in South Asia are complicated, and quantitative detection is the next research direction.

Keywords

• regional conflict
• multi-source data
• spatial autocorrelation
• spatial hotspot analysis
• south asia