Sequential data assimilation for real-time probabilistic flood inundation mapping

Abstract. Real-time probabilistic flood inundation mapping is crucial for risk warning and decision-making during the emergency period before an upcoming event. Considering high uncertainties involved in modeling of a nonlinear complex event, providing deterministic map can be erroneous misleading reliable timely decision-making. The conventional hazard maps provided different return periods cannot also represent actual dynamics flooding rivers. Therefore, real-time framework that forecasts areas onset paramount importance. Sequential data assimilation (DA) techniques are well known operation physical models while accounting existing uncertainties. In this study, we present DA hydrodynamic where multiple gauge observations integrated into LISFLOOD-FP model to improve its performance. This study utilizes ensemble Kalman filter (EnKF) multivariate fashion dual estimation state variables parameters correlations among point source taken account. First, synthetic experiment designed assess performance proposed approach; then method used simulate Hurricane Harvey 2017. Our results indicate accuracy reliability by 5 %–7 %, it provides basis sequential updating mapping.