Abstract
As decentralization of power, local governments are pressed to draw up their own master plan. In this paper, we try to support their endeavor by presenting systems for fundamental data collection and for evaluation of alternatives on fire prevention planning. First, we develop an algorithm to convert “land cover classification map” to “attributed mesh map” and do 20 case simulation, applying fire spreading simulation system previously developed by authors to maps generated by the algorithm mentioned above. Secondly, we construct regression model based on these simulation results, in which the ratio of wooden building, wind velocity and number of regions blocked by firebreak are selected as independent variables, and we get a coefficient which determinant is 0.85. We define rank of regional vulnerability to fire spreading based on this regression estimate calculation. Finally, we show a mapping system of these regional data using chromatic difference of red-color and demonstrate to output such colored map applying the system to K-city. We also demonstrate to output reevaluation map of fire prevention alternative in this city.
