Probabilistic ReFH Modelling Function

Probabilistic ReFH Modelling Function

Running the ReFH uncertainty analysis is a two-stage process within the Flood Modeller interface. The first consists of the generation of a number of hydrographs (defaulting to 33) which can then be ranked in order of peak flow, time to peak or flow volume in order to determine exceedence.
From this, the modeller may then select a smaller subset, for example based on required exceedence percentiles, for which to run a full hydrodynamic simulation.
Uncertainty is therefore estimated by relating a percentage exceedence to model output.

Step-by-step guide

1. Load the model into Flood Modeller
2. Select Hydrology (tab) > Probabilistic ReFH
3. From the 'Parameters' tab, select the ReFH boundaries (at least one) in the model at which to apply the distributions.
   
4. The distribution parameters (fse in the case of the log-normal distributions; max/min factors in the case of uniformly distributed parameters) may also be amended on this tab - recommended by the advanced user only.
5. Select the 'Hydrology runs' tab and choose the base event file - this will define the model run parameters to be used to run the hydraulic simulations. This will ultimately be augmented with the sampled ReFH boundary data when the hydraulic simulations are run.
   
6. On this tab, click 'Run hydrology' - this generates the sampled hydrological boundaries for the selected boundary units. This uses the combination of distributions of each parameters dictated by the OLH method.
7. In order to view the output of each hydrological boundary for each sample, click the 'View detail' button.
8. The 'Hydraulic Runs' tab will initially rank the output from the hydrological boundaries, defaulting to by peak flow at the first listed boundary. The boundary can be changed, as can the parameter, to either total volume or time to peak. On making a selection, a re-ranking will automatically take place. An exceedence probability is then assigned to each sample based on this ranking.
   
9. The user should then select the required samples from which to run a full hydraulic simulation, e.g. based on required exceedences from the 'Include in run' column and then click 'Run' to run the selected hydraulic simulations.
   
10. The run summary then appears on the 'Results' tab, detailing where to find the results - csv files of maximum water levels are automatically produced, and further analysis may be performed on the 1D results as usual, e.g. via the user interface or TabularCSV postprocessor.
    
11. The 'Advanced Parameters' tab contains combinations of the individual distributions for each parameter used by the OLH method to produce the hydrological boundary samples, and gives the opportunity to alter or load a new OLH distribution. This should only be changed by the advanced user; the combinations chosen should conform to an orthogonal Latin Hypercube in order to preserve an equally probable subspace.

References

1. Revitalisation of the FSR/FEH rainfall runoff method , Joint DEFRA/Environment Agency Flood and Coastal Erosion Risk Management R&D Programme R&D Technical Report FD1913/TR (2005)
2. McKay, M.D., Beckman, R.J. and Conover, W.J. (1979). A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code, Technometrics (American Statistical Association) 21 (2): pp239-245.