The Spatio-temporal Patterns Of Ecsystem Disturbances Based On MODIS Data

Most ecosystems in the world have been influenced at varied degrees by different natural and human disturbances,which have significant impacts on global carbon cycle,climate change and natural resources management.The acquisition and application of disturbance data plays an important role in how to maintain the normal function of ecosystem,track climate change response and formulate ecological conservation planning.In comparison with the traditional methods that usually derive ecosystem disturbance data from the on-ground measurements,the technology of satellite remote sensing has the advantages of being able to cover a larger scale,having a higher temporal resolution and a high efficiency of execution.This study applied the Disturbance Index model based on the MODIS data products of MYD11A2 land surface temperature and MYD13A2 enhanced vegetation index during 2005-2016.The study area consisted of Yunnan.Guangxi,Guizhou,Chongqing,Sichuan,and parts of Qinghai and Tibet in western China.In addition to the overall ecosystem disturbance data,this study also extracted the disturbance data separately for forests and grassland-shrubs by using the MODIS landcover classification data.The results were verified by using disturbance information from other sources,including the MODIS fired land data,on-ground fires spots,fire occurrences reported by various media,Temperature Vegetation Dryness Index and Google Earth imagery.The spatial-temporal patterns of ecological disturbances,including the overall ecosystem disturbance,and forests disturbance and grassland-shrub disturbance,were analyzed.Finally,this study modeled the disturbance intensity based on the disturbance frequency and influence scope,and assessed temporal-spatial patterns of the disturbance intensity.The primary results and conclusions are as following:(1)The ecosystem disturbances data for each year from 2007 to 2016 were extracted by coupling the Enhanced Vegetation Index(EVI)and Land Surface Temperature(LST)in the MODIS datasets.The Disturbance Index used in this study is more sensitive than the Land Surface Temperature or Vegetation Index alone,especially for the fire disturbances.(2)The forest disturbances and grassland-shrub disturbances were obtained by using MODIS landcover classification product as masks.According to news reports,Google Earth imagery,Temperature Vegetation Drought Index and meteorological data,the effectiveness of the Disturbance Index to extract forest disturbances and grassland-shrub disturbances was verified.The results showed that the Disturbance Index was sensitive to forest disturbance and grassland-shrub disturbance events.The disturbance data were highly consistent with the results of Google Earth imagery,Temperature Vegetation Drought Index and meteorological data.(3)The forest and grassland-shrub disturbances almost represented the entire disturbances occurred in the study area.The area of forest and grassland-shrub disturbances accounted for about 90%of the total disturbance area.In addition.ecosystem disturbances were concentrated in the area of above 1,000 m sea level,and the proportion of disturbance area in the regions of above 3.500 m has increased greatly since 2011.(4)In study area,the disturbance intensity was higher in the west and the north,and lower in the east and the south.The areas with the strongest forest disturbances were mainly located in the western Sichuan Plateau,and fire was the major type of disturbance.The areas with the strongest grassland-shrub disturbances were the agricultural and pastoral regions in Qinghai-Tibet Plateau,and meteorological drought was the major type of disturbance.