This is a discussion piece on Fraser R (2024) An updated reference table for extreme flood hydrology methods at UK dams. Dams and Reservoirs 34(4): 163–166, https://doi.org/10.1680/jdare.2024.13.
This short Note is a response to the invitation made after the paper by Fraser (2024) to contribute to the discussion of the topic covered by that paper. The timelines of rainfall and rainfall–streamflow models presented and discussed by Fraser (2024) give the complex history of adjustments to FSR methods and models since 1975. This Note is restricted to the rainfall–streamflow modelling component of currently recommended design flood estimation procedures. It has been suggested (Faulkner and Benn, 2019) that process-based models using high-resolution digital terrain data may be better suited for reservoir flood estimation than the UH approach currently recommended. However, it is possible that where Faulkner and Benn (2019) refer to ‘limitations’ of the UH concept they mean the FSR/FEH/ReFH UH specifically (and they would be right); they may not have considered the IHACRES UH approach, which overcomes many of the problems associated with the FSR/FEH/ReFH UH methods, as has been well documented.
The British Hydrological Society (BHS) Occasional Paper 15 (OP15) ‘Unit Hydrographs and United Kingdom Hydrology 1990 to 2020: IHACRES Rainfall-Streamflow Modelling’ (Littlewood, 2021) includes a section (Section 4, pp. 37–41) on ‘Engineering Hydrology and IHACRES’. Section 4 of the BHS OP15 discusses some of the pros and cons of the IHACRES UH compared with the FSR/ReFH UH. To include this discussion here would be repetitive, but BHS OP15 is downloadable from the BHS website (Littlewood, 2021).
Table 1 (similar to table 13 in BHS OP15) summarises several differences between FSR/FEH and IHACRES UHs and indicates that the IHACRES UH overcomes several problems.
Comparing FSR/FEH and IHACRES unit hydrographs (after table 13 in Littlewood, 2021)
| Type of unit hydrograph | ||
|---|---|---|
| Feature | FSR/FEH | IHACRES UH |
| What flow is modelled? | Direct flow (non-measurable) | Streamflow (measurable) |
| High flows modelled? | Yes | Yes |
| Low flows modelled? | No | Yes |
| Event-based modelling? | Yes | Yes |
| Continuous streamflow simulation? | No | Yes |
| Prior hyetograph separation required? | Yes | Noa |
| Prior hydrograph separation required? | Yes | No |
| Automated hydrograph separation provided? | No | Yesb |
| Smoothing of UHs identified from hydrometric data required? | Yes | No |
| UH parameters transferable to ungauged catchments? | Yes | Yes |
| Type of unit hydrograph | ||
|---|---|---|
| Feature | FSR/FEH | IHACRES UH |
| What flow is modelled? | Direct flow | Streamflow |
| High flows modelled? | Yes | Yes |
| Low flows modelled? | No | Yes |
| Event-based modelling? | Yes | Yes |
| Continuous streamflow simulation? | No | Yes |
| Prior hyetograph separation required? | Yes | No |
| Prior hydrograph separation required? | Yes | No |
| Automated hydrograph separation provided? | No | Yes |
| Smoothing of UHs identified from hydrometric data required? | Yes | No |
| UH parameters transferable to ungauged catchments? | Yes | Yes |
Note: Italic bold typeface indicates a relatively good feature
Automatically by a parameter grid-search method within model calibration
Dominant quick- and slow-response components of streamflow
Apart from discussion in Section 4 of BHS OP15 (and earlier papers cited therein), and by Faulkner and Benn (2019), there does not appear to have been public discussion of possible ways forward. It is suggested that the options are:
- (a)
do nothing, that is continue to use the FEH/ReFH UH–based model (for ever?), possibly with further minor adjustments
or, with a vision towards replacing the FEH/ReFH approach with something better, assess the potential for assisting with design flood estimation by using
- (b)
the IHACRES UH-based rainfall–streamflow modelling approach or
- (c)
a process-based rainfall–streamflow model with a digital terrain model or
- (d)
any other reasonable proposed approach.
It would make sense to consider (b), (c) and (d) together rather than exclusively.
A response to this discussion can be found at Fraser R (2025) Response to discussion: An updated reference table for extreme flood hydrology methods at UK dams. Dams and Reservoirs, https://doi.org/10.1680/jdare.25.00096.
