Likelihood selector (%): Depth selector (cm): Description:
1%AEP event estimate for Carlisle
Disclaimer: this is a visualisation research tool. Flood forecast extents are purely for testing purposes.
Likelihood of
depth exceeding cm
';

Definitions:

AEP and return period

AEP is the annual exceedence probability. This value expresses the probability (as a percentage) that a flood event will exceed a given magnitude. The reciprocal (100 divided by AEP) gives the average amount of time expected to contain 1 event of the stated magnitude.

Percentile

Likelihoods can be expressed as percentage values. Here an expression such as "80% chance that the 1%AEP event will be larger than that shown..." means the study that estimated the size of the 1%AEP event found that 80% (or 8 out of 10) of the acceptable computer simulation results showed a flood larger than the flood shown on the map.

Likelihood of exceedence

This webpage shows that flood extent forecasting can never be exact. This is because flood forecasting is based on computer estimates of what might happen during a real flood. One way to communicate the range of possibilities for what might happen is to specify the chance that a flood will be bigger than the one shown on the map. For example a likelihood of exceedance of 20% means that the computer simulation estimates that the 1%AEP event has a 20% (or 1 in 5) chance of being bigger than the one shown on the map.

About

This web-based tool was built by David Leedal (Lancaster Environment Centre, Lancaster University) as part of the work carried out by the Flood Risk Management Research Consortium 2. The concept is to produce a means of communicating probabilistic flood risk modelling results in a user friendly and interactive way.

The original modelling study was carried out in LISFLOOD-FP by Jeff Neal (School of Geographical Sciences, Bristol University), with multivariate input flows calculated by Caroline Keef (JBA Consulting).

The flood inundation study estimated the probabilistic range of inundation resulting from the statistical properties of the 1%AEP inflows to Carlisle from the Eden, Caldew and Petteril. The results from the study were a set of 16 probability grids providing the estimated probability of exceedence for each of 16 depth values between 0 and 150cm. The grids cover an area of 4 x 6km at 10 x 10m resolution. The 16 grids were used to produce georeferenced overlay images of the exceedence extent at each percentile. This tool converts these to Google Map overlays and allows the user to select them easily using slider interfaces.

The data from the 16 grids is also loaded into a 3D JavaScript array (using JSON) which allows the user to inspect the probability of exceeding all the depth values at a specific point. This is done by converting the Google Coordinates into the appropriate x,y reference into the array and extracting the entries in the z dimension. These are then plotted using the Google Charts API.

The wiki and forum tabs are there to demonstrate that as well as statistical (aleatory) uncertainty, as dealt with and visualised by the model study and this tool, there is always a degree of knowledge-based (epistemic) uncertainty. This can be managed in more qualitative ways using discussion, investigation, debate, and opinion. The wiki and forum provide a space for this.

This work has a GNU General Public License v3.