Optimizing Targeted SAR Acquisitions for Flood Extent Assimilation to Improve Inundation Forecasts

In the absence of accurate flood inundation model forecasts, floods can have extremely expensive and often fatal consequences. The lack of sufficiently precise global elevation data and inflow uncertainties propagated from precipitation forecasts leads to inherently erroneous flood inundation model outputs, frequently impeding their operational application. Recent studies suggest that the assimilation of independent flood observations can be used to diagnose the inherent uncertainty in hydraulic modelling. Synthetic Aperture Radar (SAR) sensors on-board an increasing number of Earth observation satellites, have enhanced the possibility of monitoring flood dynamics from space, due to their all-weather/all-day imaging capabilities. SAR-based flood extents can either be used to retrieve floodplain water level observations for flood data assimilation or be assimilated directly. Flood extent assimilation can also be used to characterize and diagnose localized model errors. In this study, numerical experiments are used to simulate multiple spatio-temporal SAR acquisition scenarios, to identify the optimum measurement design for targeted satellite acquisition, to best facilitate flood estimation. A particle filter based flood extent assimilation framework was developed using the hydraulic model LISFLOOD-FP, and implemented for the 2011 flood event in the Clarence Catchment, Australia. A real world scenario was emulated for the open loop model ensemble, with the consideration of uncertainties in inflows, initial conditions, parameters (channel roughness and geometry), and topography. The impact of assimilating flood extent at hydrodynamically uniform reaches, with different combinations of first visit and revisit times have been investigated. Results indicate that the optimum timing and frequency of targeted SAR acquisitions differs with respect to reach hydraulic characteristics and the spatiotemporal position in the catchment relative to the flood wave. Findings from the study will allow diagnosing model deficiencies as well as enable the optimal utilization of SAR imagery to overcome these localized uncertainties, such that the accuracy of inundation forecasts can be maximized.