Investigation into Feasibility of Data Assimilation Approach for Flood Level Estimation Using Temporal-Spatial State Space Model

This study presents flood level estimation using state space model. Traditional water-level sensors for large-scale river can detect water-level rising, though they have limitations in detecting urban flooding depth. Therefore, we propose a flood estimation method, which integrate observed data and flood analysis simulation with minimizing the number of water-level sensors. The estimation method of this research consists of three procedures: flood simulation using a continuity equation and motion equations, compensation using geographic characteristics and data assimilation using temporal-spatial state space model. We apply state space model to actual flooding data of Typhoon in 2017. The analysis result shows that estimated values agree with the observed values. This study is an approach to know the flood level in the urban area with the limited observation data for detecting the real-time flooding of small rivers and waterways, and the flooding process of living areas like the street.