Based On The MODIS Reservoir Monitoring Algorithm And Application Research

Traditional hydrological model for hydrological forecasting, there will always bea certain degree of prediction error. In order to improve the forecasting accuracy, it isnecessary to cut down the prediction error. With the development of spatialinformation technology, using remote sensing data to monitor hydrological becomesone of the important application areas.In this paper, from remote sensing theory, first, homomorphism filtering, waveletdecomposition and fusion method and multi-scale Retinex analysis, aimed at mist filterthe MODIS image. By the experimental results of the study compared the threemethods to determine which method is more suitable the mist filter out of the MODISsatellite images. Second, based on remote sensing theory, this paper detailed analyzedspectral characters of MODIS data in various types of surface and water bodies in eachband to find the logical relationship between the water body and non-water body.Water body extraction model have already been constructed, based on the previousconstruction model ideas. In validation of the model, we found that there were certainproblems in the extraction accuracy, such as easy to extract some non-water part, andcan not extract the part of water strictly. By re-sampling of the target sample, theanalysis of the different characteristics of the three MODIS image, the model has beenproposed to be improved, thereby obtaining a new model for water extraction. It hasbeen proven that the improved model can be better separation of water andnon-aqueous body portion and has better extraction accuracy. Finally, the improvedmodel is applied to actual reservoir water monitoring. Extraction of water body ofMODIS images Guxianhu reservoir in2011three different time phase has been made. we can clearly see the changes in the water level of the reservoir area and realizereal-time monitor of Reservoir water resources.In this study, the monitor of reservoir water resources based on MODIS data has acertain practical significance and application value flood control and disasterabatement, hydrological data collection, planning and construction of waterconservancy and post-disaster reconstruction.