Remote Sensing Analysis Of Forest Disturbance In Southeast Asia For Recent Decades

Forest, has been hailed as "The lung of the earth", is the place most likely to absorb carbon dioxide in the air. It’s also an important component of terrestrial ecosystems, maintaining the material and energy balance of the earth-atmosphere system, the stability of terrestrial ecosystem. Moreover, the forest plays an extremely important role in human survival and development. However, the forest has been subjected to the endless deforestation and eroded in recent decades. Satellite remote sensing provides a powerful mean for monitoring the forest disturbance of the entire Southeast Asia, which characterized by a short observation period, a large range of spatial information, achieving rapid quantitative analysis.Based on MODIS MOD13A2enhanced vegetation index and MOD11A2land surface temperature data in2002-2012, the method of instantaneous index (DIinst) and non-instantaneous disturbance index (DInon) is used to monitor forest fires and deforestation in Southeast Asia in recent decade respectively. The algorithm was evaluated by comparing with the existing burned area products and the FORMA data. The forest disturbance area was extracted from2003to2012in the Southeast Asia, and the spatial and temporal patterns were analyzed. The conclusions are as follows:Ⅰ. The method of coupling of land surface temperature and enhanced vegetation index data can reduce the requirements of absolutely clear sky effectively, and the ability to ensure that when fire to produce the maximum land surface temperature, the corresponding vegetation index value do not appear before the fireⅡ. Compare the DI monitoring method in this paper with MODIS burned area products, the proposed algorithm can better detect the disturbed areas of forest fires in Southeast Asia, and has a high degree of spatial consisitency with remote sensing image.Ⅲ. The results of this paper are compared with the same year FORMA organize data show that the disturbance boundary spatial pattern of the two products extracted is similar. However, the proposed algorithm can better detect forest disturbance district by comparing with remote sensing image.Ⅳ. In this research, results show that the forest disturbance areas are mainly concentrated in Kalimantan, Sumatra, New Guinea and Burma, where containing high carbon storage.