Detection and Prediction Systems of Peat-Forest Fires in Central Kalimantan
Fire detection systems and fire prediction systems were developed in peatland of Central Kalimantan, Indonesia. Through the fire detection systems, the fire occurrence information with more than 1 km2 coverage is obtained by 4 pilot villages at an average of 13–16 h. As a result, in order to send messages to stakeholders more smoothly, we needed to estimate traffic congestion time of the short message system (SMS) of real traffic in Indonesia. Next, in order to confirm all record of ten hotspot data (July 2009) and two current firing hotspot data (September 2012) detected by the improved algorithm, we introduced the unmanned aerial vehicle (UAV) into the systems, so that all hotspot data were confirmed to be burnt or burning area by UAV photographs and also the wireless sensor network (WSN) was confirmed to be useful to fire detection and prediction through the artificial field-fire experiment. Thirdly, fire spread prediction time in the fire prediction systems is about 4 h for 2 km area from the pilot villages when applying simplified fire-extension model. When we consider the interval of satellite image acquisition, predicted fire spread coverage error becomes within 50 % if the velocity of hotspot center is less than 2 m/min. In order to improve the simplified fire-extension model, we have to examine and establish the velocity of the movement of hotspot center determined by wind velocity, soil moisture and vegetation, and verify the precision of the simplified fire-extension model through the time-series hotspot data analysis and on-site inspection. As a result, we will be able to estimate the time-series CO2 emission due to fires based on the time-series hotspot data. Finally, the systems developed in the project was validated thorough the fire practical operation in the peat land, so that it was found that we have to discuss and coordinate about fire communication systems with local government authorities for their realization. In addition, it is necessary to review the possibility of making accumulated data open for researchers and practitioners to share observation data among researchers and practitioners even if the project is finished.
KeywordsHotspot satellite data Unmanned aerial vehicle Fire detection and prediction Wireless sensor network
This work was supported by SATREPS (Science and Technology Research Partnership for Sustainable Development) project entitled as “Wild fire and carbon management in peat-forest in Indonesia” founded by JST (Japan Science and Technology Agency) and JICA (Japan International Cooperation Agency).
- Honma T, Kaku K, Usup A. and Hidayat A (2014) Fire detection and fire prediction group activities in JST-JICA project: current status and planning. In: Proceedings of international symposium on wild fire and carbon management in Peat-Forest in Indonesia 2013, pp 25–31Google Scholar