The development of forest fire monitoring and warning system for agroforestry areas in Uttaradit Province, Thailand

Abstract

© Published under licence by IOP Publishing Ltd. Forest fire occurs regularly in the northern agroforestry of Thailand started in February to April and it causes economic and environmental losses. This study aimed to develop an automatic spatial data analysis system for monitoring and warning and a system for tracking and warning forest fires in agroforestry areas. Near real-time NDVI in dry season was analyzed to estimate the amount of biomass using a relationship between the NDVI and surveyed dataset. The biomass, hotspot, DEM, slope, aspect and relative humidity data were analyzed by Multi-criteria analysis (MCA) for forest fire risk area. The results are presented on web and mobile platform for forest fire warning. The estimated biomass of March was 7.5 to 9.1 tons/hectare, which corresponded with the results obtained from the field survey. The fire risk factors analysis including hotspot density, relative humidity, and landscape were used to produce the five risk levels. Each risk level area from one to five was 11.84 km2, 25.68 km2, 42.28 km2, 30.85 km2 and 20.65 km2, respectively. The forest fire monitoring and warning systems were designed and developed in open-source platform, which can be utilized for forest, fire monitor and warning for local administration as well as farmers in the area.