Dynamic Monitoring Of Poyang Lake’s Area Based On Landsat Data And Its Relationship With Water Level

Lake is one of the most important parts of the terrestrial ecosystem,and the changes in lake water’s area have a major impact on environmental changes and human production activities. Poyang Lake is ranked first in China’s freshwater lake in area, and is an important wetland with international influence. Affected by the bottom topography and the local climatic factors, Poyang lake formed the natural landscape of "like a river when in low water level and like a lake when in high water level ",which also makes the flood and drought disasters occur frequently in this region.Therefor, it is of great significance to carry out dynamic monitoring in Poyang Lake.In recent years, the technology of remote sensing is extensively used in water body identification of large area and flood monitoring by its wide monitoring range, high-speed and synchronization, and it realize the real-time monitoring of flood and drought disaster. This paper choose the Landsat satellite data of 2000-2014,combining with the observation data of water level in the same period of 2009-2013 to investigate the changes of the Poyang Lake in this fifteen years, the main contents are shown as follows:(1) Satellite remote sensing images’ selection and data preprocessing. In order to reach the purpose of water body extraction, this paper used Landsat data, the image preprocessing includes radiation correction, projection conversion and strip repair for Landsat 7 ETM+ images because of the presence of the banding phenomenon. For the water level data used in this paper, this paper has carried on the conversion of height datums, the datum of water level data were converted to the Wusong Datum.(2) The extraction of water body from the remote sensing images of 123 phases during 2000-2014. According to the spectral characteristics of water, this paper contrasted four kinds of water extraction method, include water index method of NDWI and MNDWI, multi band spectral relation method, AWEI- automatic water extraction index method. Through the accuracy verification, this paper selects the AWEI-automatic water extraction index to extract the water body information of Poyang Lake because of its higher accuracy, and calculated the area of the water bodyin Poyang Lake extracted from each image.(3)Analysis of the changes of Poyang Lake’s area and the relationship between the Lake’s area and water level. According to the water extraction results of the Poyang Lake from 123 phase cloudless images in 2000-2014, this paper studied the monthly and inter annual variation trend of the lake’s area. Based on the number of water flooded, the spatial distribution of the flood frequency in Poyang Lake area was analyzed. Combined with the water level data of 38 days from 2009-2013, the representative water level station was determined according to the correlation between the water level value and the area of the lake. And the linear, logarithmic, exponential,and quadratic polynomials models of representative water level station were established by regression analysis. According to the water level, the relationship model was established piecewisely. After precision verification, the subsection model in Wucheng water station was determined as the optimal model of lake’s area and water level. So, this model can be used to predict the lake’s area in flood season by the water level observation, and remedy shortcomings which visible remote sensing couldn’t monitor the Poyang Lake’s flooding situation when it’s rain.