We must remember first that, while we might expect more flooding as our climate changes, it is not a simple matter to predict the extent of flooding. It’s not just a winter issue. There may be a higher frequency of sudden, intense summer storms, causing urban flooding, as well as a greater risk of sustained winter rain causing river basin flooding. The urban summer flooding is largely an issue for collection systems, while the winter rains give problems for managing river systems.
However, in both cases modelling of the system can help in risk assessment and in pointing to the key areas where additional capacity or control is best provided to reduce risk. And we are of course talking about reducing risk – not eliminating it.
The essence of modelling is being able to look at situations before they occur in reality. So modelling is an ideal approach to the examination of extreme conditions such as flooding. However, if the condition is off the scale of experience, what faith can you place in the model? How can you trust the model and its calibration?
The answer lies in the way the model works. If the model is based solely on observations of the inputs and outputs of the system under known conditions, and the results of extreme conditions are based simply on extrapolation of known conditions, there is no theoretical basis to validate the results. However, if the model represents the detailed hydrodynamic and hydrological processes that are at work throughout the system, then the results have a solid basis. Many of our models work at this detailed level, and our users therefore apply them confidently to the What If situations of evaluating and mitigating flood risks.
To help this process, work is ongoing here at HR Wallingford to work out procedures for calculating flood risk maps, to include the likelihood of flooding, the depth of flooding and the extent of damage that would be caused. It’s an important area for the future, and we’re very keen that what we come up with should be practical and not too onerous – it would be counterproductive to over-complicate the process. So in our InfoWorks modelling system we have pulled together a good user interface for engineering analysis, tools for efficient model building and inbuilt data management to make sure the procedures are accountable.
But it’s not all about planning. Models can also play a vital role in operations. We have a product called FloodWorks that runs models in the control room to forecast flooding incidents and to help manage under flood alert and flood conditions. Driven by live operational data and meteorological forecasts, the outputs provide real-time decision support to the control room staff as they manage a flood risk situation or an actual flood.
Finally, I hope that nothing I’ve said implies that considering flood risks is a one-off activity. A thorough approach to flood risk evaluation and flood management requires a long term approach. It’s never too late to look at flood risk, but it’s also true that it’s never too early to start planning.