Plan for an Ensemble Data Assimilation System for Forest Fire Prediction

JI-SUN Kang* and Minsu Joh
Disaster Management HPC Technology Research Center, Korea Institute of Science and Technology Information, Daejeon, Korea

About 70 % of area is occupied by mountainous terrain in Korea and 43% of trees are coniferous trees which have been known as flammable woods. In addition, we recently have very dry spring and summer while the temperature tends to increase continuously according to the 30-year climatology. Thus, it is necessary to prepare forest fire prediction system to save people’s lives and properties from possible fire. We plan to develop a forest fire prediction system as well as an ensemble data assimilation system, as one important component of a decision support system for natural disaster management. We intend to use high-resolution regional NWP model (such as WRF) coupled with fire spread model which allows feedback of wind, humidity and heat flux variables between NWP and fire models. We would like to discuss a strategy of such high temporal/spatial resolution of NWP simulation. Moreover, we plan to assimilate meteorological variables first. We expect that the improved wind field will effectively improve the fire spread prediction. Then, we can consider assimilation of satellite image data which detect fire area. We will present our plan and progress of this work at the workshop.

*Preference: Poster