Analysis of RASP HLM+ sensitivity to improve national flood risk assessment
A sensitivity analysis of the high level method plus risk assessment of flood and coastal defence for strategic planning in national flood risk assessment.
Documents
If you use assistive technology (such as a screen reader) and need a version of this document in a more accessible format, please email: enquiries@environment-agency.gov.uk. Please tell us what format you need. It will help us if you say what assistive technology you use.
Details
Details This analysis is for Environment Agency staff and consultants who are working with risk assessment of flood and coastal defence. This is for strategic planning methods through research and development and related projects. It’s also for people involved in the National Flood Risk Assessment (NaFRA) projects.
We researched the methods and the sensitivity of the HLM+ to the various input data sets and model criteria. We also measured the sensitivity of outputs, such as the probability of flood and average economic damage. In addition, we focused on providing the relative contribution of different input data sets to the variability in the RASP HLM+ outputs.
We used three pilot sites to include a fluvial, estuary and coastal location. These included the:
- Stour fluvial site in south-east England
- Thamesmead site on the Thames Estuary
- Skegness coastal site in Lincolnshire
You can use this research to:
- improve the NaFRA calculation approach
- improve the methods for estimating uncertainty - to streamline these to focus on the more sensitive input data and model criteria
- prioritise data collection activities in the Environment Agency to focus on those criteria which changes the risk by a significant amount
Methods
We used a perturbation approach to the sensitivity testing. We varied input data within realistic ranges based on knowledge of the associated field measurement techniques.
Conclusions
The sensitivity analysis concluded that:
- we cannot make generic conclusions about the sensitivity of coastal sites from the existing pilot results - further exploration of sensitivity for a coastal site with measured toe level information is recommended
- the coastal element showed that good quality data on toe levels and crest levels is very important in order to do more probabilistic analysis of flood risk
- you should choose pilot sites for any further work on sensitivity and uncertainty analysis based on data availability and quality
- we need to take a rational risk based approach to data quality and model accuracy.
This report was published in March 2007