The Water Information Extraction And Land Cover Classification Of Nansi Lake Based On SPOT5 Remote Sensing Image

Nansi Lake is one of the largest freshwater lake in northern part of China and is an important water source in Shandong Province. It is also the important channel and storage reservoir of the east route of South-to-North Water Diversion Project. In recent years, increasing interfere activities cause the water resources shortage, reduction of lake area, frequent drought and floods, the destruction of biological diversity and wetland ecosystems were seriously destroyed. As people gradually realized the value of lake resources, the protection of ecological environment in Nansi Lake is imperative. Thus, the water environment and resources management in Nansi Lake is of vital important. Using remote sensing to study the environment of Nansi Lake can provide scientific foundation for the protection of the lake.At present, the distribution of the water body and land cover of Nansi lake are rarely reported. Few research classified and analyzed the land cover of the Lake using Landsat TM image data, but the accuracy is still to be improved. In this study, Two scenes of SPOT5 imageries (middle of June, 2006) were interpreted affiliation with global positioning system data to classify the land cover of Nansi Lake, using ERDAS Imagine and Arc View. The change of water bodies, the function to prevent floods and the land cover of the lake are studied and analyzed. The dynamic change of water bodies is quantitatively analyzed. The land cover type, area, distribution and functions are discussed, the area of enclosure aquaculture in Nansi Lake is accurately calculated. This study provides the background data of the lake ecosystems and biological diversity. Multi-level and multi-temporal information and scientific analysis are provided for the research and management of Nansi Lake, and provide scientific information for the dynamic monitoring of the ecological environment and the protection of biodiversity in the lake. Main contents and conclusions of the study are as follows: (1) The two scenes of SPOT5 satellite multi-spectral images of Nansi Lake are interpreted affiliation with global positioning system data to classify the land cover of Nansi Lake, using ERDAS Imagine and Arc View. First, the images are preprocessed including geometric correction, image mosaic and subset. As the fishpond and the open water-body were difficult to distinguish, unsupervised classification and manual interpreting were used as the classification method. The accuracy of interpretation classification was 83.3% after checking.(2) The land cover classification types were mapped and the area of each type was calculated. The overall area of Nansi Lake is 1206.9 km~2, only 45.54% belong to open water. Others lands are covered with wetland vegetation, fishpond, farmland and forests, with proportions of 21.06%, 15.60%, 11.23%, 5.66%, respectively. The Normalized Difference Vegetation Index (NDVI) is also mapped and the vegetation distribution is analyzed.(3) Nansi Lake water information is extracted through decision tree classification using ENVI software. Based on the texture feature of the image, the reclaimed lands from Nansi lake are classified. The distribution and characteristics are studied.(4) the capacity of storing floods are reduced by the water area diminishing. The results showed that conversion of water area to other land cover type would reduce the lake storage capacity. In order to protect the ecological environment of Nansi Lake, the land reclamation should be prohibited and enclosure should be controlled. Then the wetlands vegetation will be gradually restored.