Item

Abstract

Nearly-pristine rivers are relatively undisturbed by flood control and flood protection measures. These systems, which are frequently affected by floods, are typically inhabited by low-income communities that develop their own flood management strategies. At these sites, there is usually little data available about flood extent to accurately estimate hazard exposure and ultimately risk.Here, we design and implement a bi-dimensional hydrodynamic flood-model for estimating inundation zones at a nearly-pristine coastal watershed in central Chile that has been repeatedly affected by floods. The hydrodynamic numerical model incorporates high-resolution digital elevation data and bathymetry as well as up to date soil use information. In order to accurately determine all flood-prone areas in nearby communities and to identify worst-case flood scenarios, we test a combination of historic and climate-change predicted extreme discharges as upstream boundary condition and use measured and predicted sea level heights as boundary condition along the coast and river mouth.In addition to the traditional methods used for validating a flood model (i.e. measured water levels and discharge, or historical aerial photographs) and in the absence of in-situ flood extent measurements, the model is validated against novel social cartographies: flood maps developed by the community that incorporate their own experience with floods.This investigation is part of our research project “Climate change and the sociohydrology of floods”, that aims at improving flood risk quantification following a sociohydrologic approach that incorporates historical and predicted climate change trends as well as time-changes in community preparedness and vulnerability in flood-prone areas.