Flood Depth Calculation Model Of Wetland Based On The "3S" Technology

Flood disaster is one of the worst natural disasters, according to statistics, currently the world because of the economic losses caused by floods, natural disasters accounted for more than 30% of total losses, and particularly severe flooding in China. Therefore, the monitoring is to strengthen the flood prevention, an important element of disaster reduction, with the conventional method to obtain flood information, you need to spend a lot of manpower, material and financial resources, but also time consuming and often difficult because of flooding from the ground on the acquisition. Flood monitoring need to consider the timeliness of the information, and its access to information and can not be a one-time, but dynamic, real time, which is more difficult to get a big ground. Thus, a more rapid and effective method is to obtain flood hazard information.In this paper, used normal of the ETM+ and flood Radarsat as data base, in the geometric correction, image enhancement and other remote sensing image preprocessing, respectively, to extract pre-disaster and disaster in the water bodies, use of superposition methods and rapid extraction of the flood range of services for rapid disaster assessment; Based on the image and the measured reflectance values of the correlation between water depth,ETM+in the normal combination of the measured water depth, respectively, to establish single-factor regression, multi-factor regression, BP neural network, support vector machines Depth inversion model, and comparative analysis of four models; in order to quickly and accurately calculate the depth of flood distribution, Based on RS and GIS technology, extraction of submerged area combined with the DEM, the establishment of a suitable wide range, can be approximated as a static body of water depth measurement model.The results show that the inversion of the normal water depth of four models, SVM model was the best. Water depth by remote sensing inversion is a nonlinear process, SVM is a nonlinear dynamic system with strong nonlinear mapping ability and generalization ability, can establish the depth inversion model can be a good solution to the modeling of small sample data, with some practical value; Flood depth calculation model is suitable for wide range, can be quickly and accurately calculate the flood depth points, provides an efficient and practical use of remote sensing information on flood depth algorithm.