Toward a novel approach for assimilation of SAR-based flood observations in a fully coupled hydrological model

洪灾对人类构成重大威胁。除了基础设施和农业遭到破坏外，每年都有数以千计的人因洪水而流离失所。卫星观测越来越多地被用于洪水监测。星载合成孔径雷达(SARS)是一种特别吸引人的观测来源，因为它们能够在任何天气条件下返回地球表面的有意义的信息。随着在轨合成孔径雷达系统的数量和数据量的不断增加，开发在自动化处理中使用这些数据的方法是很有意义的。其中一个被称为数据同化的过程将观测数据与预报模型的输出相结合，为预报模型提供改进的初始条件。这是用来生成天气预报的方法。在洪水预报中，从SAR图像同化水位观测值(WLOS)是一个相对较新的领域。以前的研究已经使用规定的地面几何形状(水深测量)从SAR获得了WLOS。拟议的研究将使用一个完全耦合的水文模型，在该模型中，水深测量随洪水演变，以开发一种从SAR数据中提取和同化WLOS的新方法。长期目标是改善洪水预报。

Flooding poses a significant risk to humanity. Thousands of people are displaced each year due to flooding, inaddition to destruction of infrastructure and agriculture. Satellite observations are increasingly being used forflood monitoring. Satellite-borne synthetic aperture radars (SARs) are a particularly appealing source ofobservations due to their ability to return meaningful information of the earth's surface during all weatherconditions. With increasing numbers of SAR systems in orbit, and growing data volumes, it is of interest todevelop methods to use these data in automated processes.One such process, known as data assimilation, combines observational data with output from a forecast modelto provide an improved initial condition for the forecast model. This is the method used to generate weatherforecasts. For flood forecasting, assimilation of water level observations (WLOs) from SAR imagery is arelatively new area. Previous studies have obtained the WLOs from SAR using a prescribed geometry of theground surface (bathymetry). The proposed research will use a fully coupled hydrological model in which thebathymetry evolves with the flood to develop a novel approach to retrieve and assimilate WLOs from SARdata. The long-term goal is to improve flood forecasts.