Fluvial v Pluvial
Severe storms, intense rainfall and other climate change phenomena will increase the likelihood of floods. Hydro-numerical models and 3D-geospatial data are important for improving flood-simulation models. They enable better forecasting and help in advance prevention of floods. Flood simulation models for fluvial (river) flooding are advanced and established. However, pluvial (surface water) flooding, which occurs when natural and manmade drainage systems lack the capacity to run off water, happens quickly, is hard to predict and usually has a high impact on urban areas. In the UK summer floods of 2007, around 70% of damage was ascribed to pluvial flooding. Prediction and prevention would be helped by 3D urban models that included information on ground coverage (type of soil, sealed surfaces) and drainage systems, because this type of fundamental and valuable data makes it possible to improve simulation models. Flooding happens naturally and cannot always be prevented or predicted in advance. What can be done is to reduce the likelihood of its occurrence and minimise impact. Public awareness of flood risk and emergency management must also be raised. Before it can take effective evasive action the general public needs to be aware of a risk of flooding yet only 60% of affected residents in England and Wales claim to know they live in a flood-risk area. In his PhD thesis Householder Responses to Flood Risk , (University of Middlesex, 2007), ­
T. Harris showed that the figures can be even worse. Only 39% of those previously flooded had taken any action to prepare for floods, falling to 6% of those with no prior experience of floods.

Seeing is Believing
New means of communication are required to bridge the big gap between knowing about flood risk in general and acquiring the insight that it might happen ‘in my backyard’. Traditional 2D flood-risk maps give a general overview of affected areas. However, it would seem that most people have difficulty in interpreting flood forecasts and local flood warnings. In particular, estimating the impact of water height requires considerable visualisation skills, apparently beyond the capabilities of the non-expert. Visual representations are much more effective or, as somebody quoted in an article by Basic, Handmer and Cartwright (see MODSIM 2005 proceedings) stated, “Heights don’t say much to me. I don’t know the difference between the normal and raised river height… using a visual aid helps me to understand what is happening”.

VEPS
3D flood-maps of urban areas, available via internet, will tremendously increase flood-risk awareness. People will be able directly to observe the impact of a certain water height for their own house and neighbourhood, and will no longer need to attempt the difficult task of interpreting abstract numbers. In the VEPs (Virtual Environmental Planning) project ( www.veps3d.org ), we developed a first prototype for interactive 3D flood visualisation, see UDMS 2007 proceedings. The VEPs project has received international recognition as an innovative initiative, winning the International Digital Earth 3D-visualisation Grand Challenge Award. Three-dimensional urban models and an appropriate data interchange format play a vital role in creating such 3D flood-maps. As CityGML is on its way to becoming an OGC standard data interchange format for 3D-urban models, this would be a good starting point for integrating flood data into the development of a CityGML Application Domain Extension for dynamic 3D flood information. For the future, integrating 3D urban models and flood information in urban areas will be essential for improving simulation models to help prevent and reduce flood risk and increase awareness of these.

Prof. Dr Volker Coors, University of Applied Sciences, Schellingstr. 24, D-70174 Stuttgart, email: Volker.Coors@hft-stuttgart.de