How do we measure and predict the floodwater footprint when ground conditions are changing so rapidly?

With so many innovations in flood alleviation measures that have been developed and implemented in recent years, how can we be sure that they are all working in sympathy with each other? Surveying rivers for the last 28 years and measuring post flood levels has provided me with practical experience of how different river catchments react to high rainfall. I have seen destruction in housing areas, caravans in trees, raw sewage in living rooms, sanitary products strewn across floodplains and people who have lost their homes and businesses. It's grim. I have seen flood protection schemes and products work in many areas but in some it has shifted the problem elsewhere. An example of this could be a simple flood barrier across a driveway that deflects floodwater away from a property but causes flooding to someone else downstream. The national government agencies concerned with alleviating flooding do a marvelous job of joining up main rivers and tributaries in a hydraulic model to evaluate downstream effect of flood protection schemes but there is a problem with this. The density of the data is often too coarse or not current enough to highlight some of the critical geospatial features of river channels and floodplains that can cause flooding; and if they are not measured and included in the model it will cause a misrepresentation in the final model results. So how can we improve the calibration of catchment models so that they can better predict flows and flood extents whilst taking into account some of the more subtle flood defence methods such as property level protection, property flood barriers, upland debris dams, locally excavated river channel spillways and improved field drainage? This is open to discussion and I would welcome peoples views of which I have a few of my own!