Development of a fire prediction model at the urban planning stage: Ordinary least squares regression analysis of the area of urban land use and fire damage data in South Korea

Fire risk assessment at the urban planning stage is crucial for developing fire prevention strategies and designing mitigation measures. Many studies have been conducted to support fire risk assessments in cities. However, the fire prediction models developed in those studies have been based on factors that continuously change in the urban development process. Thus, their models cannot be used in fire risk assessments or fire prevention planning at the urban planning stage. To overcome this limitation, the correlations between fire damage and land use area, as determined at the urban planning stage, were identified in this study and two fire prediction models were proposed based on the results of the analyses. To develop these models, the land use area and fire damage data for 230 districts in South Korea were collected from 2010 to 2019. The fire prediction models were proposed based on correlation analysis and ordinary least squares regression analysis. The two fire prediction models that were finally derived showed accuracies of 69% and 71%, respectively. The findings of this study can be used as basic data to support fire service planning when city plans and urban development plans are established.