A Research Of Forest Fire Early Warning And Detection Based On MODIS

The occurrence of forest fires has contingency, the place of that is generally untraversed and remoteprimeval forest and half primeval forest, the influence factors of that are many, and is difficult to prevent.This research is based on the existing science and technology, that is American free MODIS data are usedfor the early warning of forest fire before the disaster and dynamic monitoring after the disaster, which issimply opening the subject for discussion and expecting more and more scholars to investigate this subjectto realize the disaster prevention and reduction effectively.Associated with the practical situation of Yunnan province, a model of both the forecast of fire dangerand dynamic monitoring of forest fire is built by applying MODIS data to obtain the parameters closelyrelated to forest fire, such as temperature, humidity, the types and quantities of combustible and watercontent and the feasibility of that is verified by the actual MODIS datum of forest fire of this area in recentyears. The main research contents include:(1) The pre-processing of MODIS dataThe pre-processing of MODIS data part mainly include: removing cloud, geometry rectification,atmospheric and radiometric correction, filtering stripe noise and eliminating ’double-fold eyelids’.(2) Building and verifying the forecast model of forest fire danger based on MODIS dataThe high efficiency, high quality and practical forecast methods of fire danger are selected byinvestigating the existing forecast models of fire danger and analyzing the advantages and disadvantagesand the ability to adapt of that in southwest area. After repeated demonstrations and experiments, Americanfire potential index(FPI) is selected to be main study method of this paper, the advantages of that are: one isthe parameters needed in the models are easy to obtain, the other is that can describe the real situation offorest fire of Yunnan province accurately. The research contents of FPI are that obtain the real-timemeteorological data and the dynamic data of dead and live combustible from MODIS data and build FPImodel using obtained data.(3) The construction and verification of forest fire monitoring modelThe forest fire monitoring model is constructed based on FPI method by analyzing the existing themonitoring models of grassland and forest, finding out the advantages and disadvantages of each, figuringout the monitoring methods and the selective methods of parameters and the data of former years is used toverify the feasibility of which and improve which.Role of remote sensing technology in disaster prevention and mitigation work is more and moreobvious, especially for forest fire prevention contribution is very prominent, and the research is more andmore deep. Many remote sensing data, the advantage of the application of MODIS data in this area is very obvious, it is not only cheap, easy to access data. The most important is the time, spatial resolution is verysuitable for monitoring forest fire early warning and disaster. Especially for forest fire larger, MODIS canbe very convenient access to related information real-time, accurate, to disaster relief, human resourceallocation, post earthquake loss estimation for outstanding contributionsThis study shows that MODIS has a large research space and high research value in early warning andmonitoring of forest fire. However, there are also lots of deficiencies in the study on MODIS, the existingresearch area is very narrow, the study on that is less and not deep enough. In order to serve our country’sand even global forest fire monitoring and early warning work well, lots of outstanding achievements areurgently needed.