Urban scale has become gigantic in the process of dazzling development, leading to the concentration of population again due to the distribution of a wide range of facilities. However, the risk of damage always lurks behind the convenience of the city. In particular, fire, which is one of artificial disasters, has occurred the most in urban areas, along with traffic accidents and its risk is increasing day by day as urban scale becomes larger and its function increases. As of 2014, 42,134 fire cases occurred in South Korea, causing 2,181 casualties and the property damage of approximately 400 billion won. In this regard, Jinju-si, Gyeongsangnam-do, which was selected as the target area in this study, was rated as grade IV(4) in the dangerous grade of fire damage, according to the area safety index, published by Ministry of Public Safety and Security, which was recorded a higher fire rating in Gyeongnam province. Therefore, various studies should be conducted to prevent the increase in casualties and property damage, caused by fire.
In this regard, this study analyzed the damage of fire occurrence from 2007 to 2014 with the target of Jinju-si, located in the western part of Gyeongsangnam-do. It analyzed the temporal distribution patterns of fire by dividing time into year, month, season, time-zone, and day and night. It also drew the clustering of fire by analyzing spatial autocorrelation.
It calculated Moran''s I index and used Getis-Ord Gi* to assess the clustering of fire occurrence spots by analyzing spatial autocorrelation in the urban space.
Moran''s I index value for the fire occurrence of the entire area in Jinju-si was 0.104, showing clustered distribution, while Moran''s I index value in Haenjeong-dong, where fire was concentrated, was 0.144, which was higher than the value of the entire areas in the clustering. To analyze local clustering, it attempted to draw hot spots by using the Getis-Ord Gi*, the results demonstrate that hot spots were seen in the eastern industrial areas, where the distribution of central commercial areas and industrial facilities was predominant with the concentration of detached houses and sales facilities. It analyzed the spatial distribution according to years and months by converting the fire occurrence points into spot data through geo-coding. As a result, it came to draw hot spots, concentrated in Haengjeong-dong, and calculated fire risk according to facilities by collecting the data of facility conditions in related areas. As for the fire risk assessment method, the rules of the Society of Fire Protection Engineers(SEPE) in the USA were applied to the absolute evaluation. Among four grades, four facilities; business facilities, medical facilities, entertainment facilities, and car facilities were drawn as the fires of A(anticipated) grade that represents the highest fire occurrence rates. Based on this method, it assessed the risk levels of fire occurrence by constructing the risk assessment matrix for the casualties and property damage. The Matrix was constructed by the horizontal axis of the matrix, composed of the fire occurrence frequency, and the vertical axis of the matrix, composed of casualties and property damage respectively to assess risk levels. Finally, it applied natural classification in order to distinguish fire risk into five grades. The highest grade, grade V(5) did not appear in Jinju-si, but the second highest grade, grade IV(4) appeared, which was divided into two; business facilities and industrial facilities. In addition, apartment houses, educational facilities, sales facilities, complex facilities, medical facilities, entertainment facilities, hazardous materials manufacturing facilities, and automotive facilities were classified as Grade III(3), detached houses, accommodation and living service facilities were classified as grade II(2) and other houses were classified as grade I(1), the lowest dangerous grade.
Thus, it is expected that this study will be usefully utilized to perform systematic management of the urban fire and reduce damage size by evaluating the characteristics of fire occurrence and presenting the risk map of fire occurrence according to facility buildings and the spatial distribution patterns of fire occurrence spots after analyzing the temporal change of fire, which occurs in urban areas.