A Model Of Water Body Extraction Based On EOS/MODIS Data & Its Application On Flood Disaster Monitoring

Flood prevention and disasters decrease information, including rainwater, flood and drought, water capacity and water quality, hydrology engineering and so on, is greatly related to spatial distribution. It is necessary and important for spatial information technology to be applied in the field of flood prevention and disasters decrease. However, MODerate-resolution Imaging Spectroradiometer (MODIS) is not widely used in this field. It is a new optical remote sensing instrument that combines images with spectra in the world. MODIS is of many special advantages, such as moderate spatial resolution, higher time resolution and free data reception, which determines it is valuable in real time earth-observation and disasters-monitoring. This study just aims at using spatial information to set up a model of water body identification (or water body extraction) based on MODIS data, and then apply this model in flood disasters-monitoring. According to the foundation of remote sensing and spectrum theory, the general extracting principal of remote sensing information is introduced, the characteristics of water body on MODIS images are analyzed, and a model of water body extraction is set up in this paper. Concretely, after collecting and analyzing a large amount of swatch from MODIS images, the characteristics of water body and other contents are found out. Based upon this, the fundamental model of water body extraction is set up. Following the model, a lot of authenticating examples are listed, and the bug of this model is pointed out, that is, water body and city cannot be distinguished through this model. To solve this problem, the water body and city swatches are colleted once more, then the improved model is built and evidence suggests that this model is feasible and useful. At last, this model is successfully applied in the dynamic monitoring of flood disasters of Poyang Lake. This study is provided with great application value, for the flood disaster monitoring based on MODIS data is the basis of layout and construction of hydraulic engineering and hydropower engineering and the basis of decision-making of flood prevention and disasters decrease.