Research On The Remote Sensed Monitoring In Forest Fire Based On MODIS

Forest resources are one of the most important natural resources for human beings. It maintains the dynamic balance of nature and improves the living environment of mankind. Forest fires, refers to the loss of human control, free spreading in woodland, forest, the forest ecosystem and the human bring certain harm and loss of forest fire behavior. The sudden strong, destructive and difficult to dispose of forest fire is difficult. In recent years, with the effect of climate warming and human factor, forest fires have become one of the most serious forestry disasters. It has brought huge losses to the lives and property of the state and people. Monitoring forest fire timely, accurately and effectively has important significance for reducing disaster loss, protecting forest resources and maintaining ecological balance.The remote sensing satellite has a wide range of coverage, high temporal resolution, and has special ability and potential in forest fire monitoring. The MODIS sensor is considered as the requirement of fire monitoring in design, and it is suitable for forest fire information extraction. It has a broad application prospect. Forest in Liangshan Province, Sichuan Province, mainly covered in mountainous areas, the climate is special, the terrain is changeable. This paper takes Liangshan Prefecture of Sichuan Province as the research area, combined with the actual situation of the research area and the ground fire data, in order to study the method of extracting the forest fire point information.The main process of identifying the forest fire points of the forest fire is:(1) Summarize the existing fire point extraction algorithm, select three methods to build the algorithm system. Three algorithms are: channel synthesis, absolute fire point recognition and threshold model.(2) Using the algorithm in the algorithm system for the research area designated information for fire point identification and extraction. According to the characteristics of MODIS data of each channel, using synthetic image band combination means directly to fire point identification, through absolute fire recognition, threshold model algorithm further extraction fire point, study area the results of fire recognition. Comparison of the results of the ground information is given to verify the results of the fire point, and the parameters of the modified algorithm are improved, and the accuracy of the fire spot identification is improved.(3) The algorithm system of the correction factor is used to extract the fire point information from April to May 2005 of the research area, and Validation algorithm availability. At the same time, combined with the existing fire information optimize the algorithm coefficient.Based on the actual fire data, the paper mainly focuses on the parameters of the fire point extraction algorithm, and the optimization of the fire point information of forest fire in Liangshan is achieved more accurate identification and extraction. The improved algorithm for remote sensing of forest fires in Liangshan Prefecture to provide better support, provide better services to real-time monitoring and early warning of forest fires in the study area, and improve the level of disaster prevention and mitigation. Further illustrate the use of remote sensing work, especially the work of MODIS remote sensing data can play a more important role in forest fire monitoring.