Identification Of Ecosystem Functional Types In The West Of Songnen Plain Based On MODIS Data

With current environmental changes, the functions of terrestrial ecosystems fromregional to global scales have partly degraded. People have realized the importance ofthe management and protection of the heterogeneity in the ecosystem composition,structure and function. However, to observe and quantify the pattern of ecosystemfunction and its change is challenging, the difficulties in the management andprotection are obvious. In recent years, ecologists emphasize the importance ofincorporating the analysis and monitoring of carbon gains, nutrient cycling and waterdynamics in regional conservation programs. Therefore, the characterization ofecosystem functional types and the construction of ecosystem functional pattern arecritical for understanding and preserving the integrity of biological communities,ecosystem services and climate. Studies on how to use remote sensing technology andsatellite-derived information to identify and analyze ecosystem function diversity andregional pattern have been increasingly concerned.The west of Songnen Plain locates in transition zone from semi-humid climate tosemi-arid climate with many types of ecological landscape. Under the pressure offrequent natural hazards, population growth and excessive land exploitation, theecological environments that have been fragile become more vulnerable and unstable.It is one of the typical regions in Northeast China sensitive to global changes.Based on MODIS data products for the period of2003-2012, five indices thatcould reflect vegetation ecological functions, namely NDWI (Normalized DifferenceWater Index), LST (Land Surface Temperature), EVI (Enhanced Vegetation Index),PSN (Net Photosynthesis) and Albedo, and three variables derived from the abovefive indexes, namely the average value in growing season (April to October), themaximum value in growing season, and the difference between the maximum andminimum values in growing season are selected and computated. In total,15variablesare proposed as the integrative indicators of ecosystem functional types in the studyarea. According to the classification indicators, Fuzzy C-Means clustering method isutilized to identify the ecosystem functional types in the west of Songnen Plain. Thenthe study area is divided into9EFTs, and the characteristics of each EFTs arerevealed based on the classification results. Finally, the Correspondence Analysis (CA)on EFTs and Land Cover Types are performed, and the unity and heterogeneity between them analyzed. Because the identification of the EFTs combines water,energy and carbon balance in the west of Songnen Plain, the result is of constructivesignificance to manage and protect different ecosystem functional types in the studyarea.