基于高时相遥感影像的洪水演进模型参数优化方法

种子点蔓延算法是洪涝演进过程模拟仿真的主要方法之一,其中,洪水演进模型参数由于受到不同地理环境影响而发生改变,具有一定的不确定性,进而导致洪涝模拟结果可靠性不高。随着高分4号卫星的发射,其具有高时间分辨率特性的影像将成为洪涝模拟、灾情评估的第一手资料。本文在模拟仿真过程中,接入高时效卫星影像并提取淹没范围和洪水体积,通过与模拟仿真产生的淹没范围和洪水体积进行比对,迭代调整洪水演进模型参数μ值、速度V,进而提高模拟结果的准确性。选取云南华坪山洪为典型案例,验证了本文方法的可行性与可靠性,为快速灾情评估和预防决策服务提供更为科学的依据。

A Parameter Optimization Method of Flood Simulation Based on High Temporal Resolution Remote Sensing Images

Seeds spread algorithm is the primary method of flood simulation. The parameters of flood simulation model are uncertain due to the different and complicated geographical environments, which leads to unreliable flood simulation results and products. High temporal resolution remote sensing images will be more important data source for flood simulation and assessment with the launch of GF-4 satellite satellites. The method of parameter optimization is to constantly adjust the value of V and μ by comparing the flood inundation area and volume obtaining by flood simulation to the flood inundation area and volume extracting from high temporal resolution remote sensing images to gain more reliable flood simulation results. As a typical case, the simulation of Huaping flood in Yunnan province suggests that this method can significantly improve the accuracy and reliability of flooding information including flood submerging range and flood submerging depth, which provide more scientific basis for quick assessment and prevention decision services.