Real-time fluvial flood mapping for impact-based flood early warning in Denmark

Estimating flood (inundation) extent for river systems is a key element for impact-based flood forecasting and warning. Hydrodynamic Inundation mapping is widely used for planning for local flood mitigation measures and climate adaptation strategies. An interesting but challenging option is to use these models in real-time to provide locally relevant and impact-based flood warning. The challenge is that these methods can be computationally demanding and therefore may not be able to provide timely forecasts and effective warnings for early action. One approach used in practical applications, is to simulate pre-defined flood scenarios that are calculated, ahead of time, and then used as a look-up table by flood forecasters. However, this approach may not necessarily capture the actual flood dynamics and requires time-consuming manual interpretation.

The Danish Meteorological Institute (DMI) has recently been appointed as the national authority for flood forecasting for Denmark, and is tasked with developing and implementing a flood forecasting and early warning. The initial focus is on informing decision-making for local and national emergency services. In this study, we explore approximate, but computationally efficient, flood mapping for flood early warning. As a starting point, we have formulated a static flood mapping approach, based on an extension of the deterministic 8-node approach, and established an approximate hydrodynamic model, using LISFLOOD-FP for the Vejle River in Denmark. Adopting these simpler approaches recognizes that for flood warning the most relevant information is the identification of the areas at risk during an extreme flood event rather than the precise extent and magnitude of flooding. The township of Vejle, located near the mouth of Vejle River in a deep lowland glacial valley, is subject to frequent flooding from the coast as well as fluvial flooding from heavy rainfall and cloudbursts. Severe flooding from extreme rainfall occurred in the Vejle River during both February and March 2019. To evaluate the impact of the approximations used, we have compared the resulting flood map with drone observations of flood extent, photographs and satellite data during the flood in March 2019. These evaluations will guide DMI in developing operational flood mapping for flood early warning and emergency actions across the whole of Denmark.