Abstract:

Taking East Dongting Lake National Nature Reserve as the research region, this paper was purposed to explore the correlation between the wetland ecosystem health state in East Dongting Lake and the water level of Chenglingji Hydrometric station. By analyzing the spectral features of remotely sensed images such as the normalized differential vegetation index, band reflectance, the brightness and wetness after ENVI Tasseled Cap Transform, the model of decision tree for classification was built. And the distribution area of different features in East Dongting Lake including water, mudbank, reed, poplar, shrubland and field were extracted out, based on CBERS remotely sensed images of 12 phases. Then, as the wetland ecosystem has productivity, stability and ability of selfadjusting, the wetland ecosystem health state evaluation model was established by integrating productivity, structure, elasticity with function, in which the productivity element reflects the primary vegetation productivity of wetland, the structure element reflects the structural integrity and landscape diversity of the wetland system, the elasticity element reflects the stability and selfregulation of the wetland, and the function element reflects the water storage capacity of wetland. On the basis of the theory of ecosystem health and the characteristics of East Dongting Lake wetland, the average normalized differential vegetation index was chosen as the index of wetland productivity, the wetland structure element was measured by landscape diversity index, landscape shape index and patch density index,and the average elasticity was chosen as the index of wetland elasticity, as well as the water area was chosen as the index of wetland function. So the evaluation indexes system of wetland ecosystem health state was established. By combining remote sensing and GIS technology, special information of each wetland ecosystem health evaluation index was extracted from different data, and the standardization value of evaluation indexes were obtained through standardized analysis. Then, based on the standardization value of each evaluation index, the wetland ecosystem health state evaluation model was used to analyze the ecosystem health state of 10 phases from 2000 to 2006 in East Dongting Lake wetland. 〖JP2〗Subsequently, the water level data of Chenglingji Hydrometric station was integrated into the study, and the correlation between the wetland ecosystem health state in East Dongting Lake and the water level was discussed. The results show that the wetland productivity was negatively correlated with water level, the correlation between the wetland structure and water level was indistinctive, and the wetland elasticity was positively correlated with water level, as well as the wetland function. By integrating productivity, structure, elasticity and function of wetland, we came to the conclusion that the relationship between wetland ecosystem health state of East Dongting Lake and water level of Chenglingji Hydrometric station tended to be quadratic polynomial, and the East Dongting Lake wetland ecosystem health state would be the best when Chenglingji water level was about 30 meters deep. Finally, the data of East Dongting Lake wetland of two phases in 2007 was used to inspect the correlation between wetland ecosystem health state and water level, and the inspection result showed that the correlation was considerably accurate