Author(s)

Rizwan Ali^*, Mobushir Riaz Khan, Hannan Mehmood

Institute of Geo-Information and Earth Observations, PMAS Arid Agriculture University, Rawalpindi, Pakistan.

Violence comes first by which human lives are being destroyed. Violence poses serious challenges to law enforcement agencies and policy makers as well as it threatens the writ of the Government. Violence condition of a country can be visualized through an accurate violence risk map. This paper presents the methodology to develop an incidence of violence (IOV) risk map using GIS. IOV data for year 2010 was collected from Pakistan Institute for Peace Studies (PIPS) that have 3104 records and cover 16 categories of attack types. IOV data was further geocoded by using online Google geocoding service. IOV risk maps were developed through two available methods: kernel density and getis-ord-gi* by giving single parameter “frequency”, and one new indexing method by giving two additional parameters as well “severity” and “probability”. For each district, IOV frequency and IOV severity values were calculated from PIPS data whereas IOV probability value was derived from the Benazir Income Support Program—Poverty Scorecard Survey (BISP—PSS). A value ranging from 1 to 5 was assigned to each of three parameters against each district and then all three parameters were multiplied with each other to generate IOV risk index that had values ranging from 1 to 125. IOV risk index was further classified through natural breaks into three categories: low risk (1 - 40), moderate risk (41 - 70) and high risk (71 - 125). For validation purpose, spatial overlay analysis was conducted between year 2011 IOV data (classified through natural breaks) and year 2010 IOV risk maps (developed through kernel density, getis-ord-gi* and indexing method). The indexing method has proved as a reliable method to develop an IOV risk map with an accuracy of 93 percent then getis-ord-gi* and kernel density having accuracy of 58 percent and 89 percent respectively. Furthermore, indexing method predicted IOV risk areas more efficiently in terms of spatial distribution. Indexing method highlighted Khuzdar, Zhob, Upper Dir, Khyber Agency, Orakzai Agency, Peshawar and Karachi Districts under high-risk category where actions are needed from the law enforcement agencies and stakeholders to minimize the violent incidents. This study has showed that GIS has the incredible capabilities that facilitate us in capturing, analyzing, and visualizing the IOV data.

GIS, Incidence of Violence, Risk Map, Getis-Ord-Gi*, Kernel Density, Indexing, PIPS, Pakistan

Ali, R. , Khan, M. and Mehmood, H. (2017) Incidence of Violence Risk Mapping Using GIS: A Case Study of Pakistan. Journal of Geographic Information System, 9, 623-636. doi: 10.4236/jgis.2017.96039.