OPW launches flooding web-portal to help engineers plan for future
24 February 2015
Author: Oliver Nicholson, chartered engineer, Office of Public Works Hydrology and Coastal Section
In June 2014, the Office of Public Works (OPW) launched the Flood Studies Update (FSU) Web Portal (see Figure 1). This web-based applications portal is a key output of the FSU Research Programme that was initiated, managed and funded by the OPW. The programme is a substantial update of the Flood Studies Report (NERC, 1975), which provided methodologies for flood estimation in Ireland and the UK. The rainfall and flood-estimation methodologies that derive from the FSU Research Programme are now implemented in the FSU Web Portal, which may be accessed at this web address.
The portal facilitates the estimation of flood flows, extreme rainfall depths and other hydrological variables at approximately 134,000 river locations/nodes in Ireland using a series of algorithms and location/node-specific parameters. It also provides users with a map-based search tool or Site Orientation Map (see Figure 2) to find their location of interest. It then guides them through each step of a calculation process, resulting in the derivation of final design rainfall or flow estimates for that location.
Some of the other main features of the portal include: Design Rainfall & Flood Estimation Modules; on-screen Stepwise Guidance; FSU Guidance Handbook; Flood Estimation Reports print/export facility; Flood Estimation database with data and software download facility; Users Forum and other common website features such as links to documents and related third-party websites, news page, and feedback facility.
Modules within FSU Web Portal
The design rainfall and flood estimation tools have been arranged into modules within the FSU Web Portal and the following paragraphs provide a brief explanation of the methodologies that are used in each of the most important modules.
- Rainfall Depth-Duration-Frequency
From the FSU research a Depth-Duration-Frequency (DDF) module was developed, which allows for the estimation of point rainfall frequencies for a range of durations at any location in Ireland. Users can calculate rainfall depths for a range of frequencies (from two to 500 years) and durations (from 15 minutes to 25 days). It can also be used to estimate rainfall frequencies if the rainfall depth and duration are known. Rainfall DDF also allows these values to be calculated on a catchment-wide basis.
- Flood Frequencies – estimation of the index flood at ungauged locations
At gauged locations, QMED is the median of the observed annual maximum flow series and is equivalent to a flood with a return period (T) of two years. In order to estimate a design flood of a specified return period (QT), two items of information are required: QMED and the relevant flood growth factor (XT) for that return period. The design flood is calculated from the following equation:
QT = QMED . XT
The new FSU methodologies use a 7-variable equation to derive an initial estimate of QMED at ungauged locations (which assumes there is no urbanisation in the catchment). The variables used in this equation are referred to as Physical Catchment Descriptors (PCDs). The PCDs describe the physical characteristics of a catchment in numerical form. Table 1 provides a summary description of the PCDs used in the FSU methodologies.
|PCD name||PCD Description||Units|
|SAAR||Standard period (1961-‘90) average annual rainfall||mm|
|BFIsoil||Baseflow Index derived from soils data||from 0 – 1|
|FARL||Flood Attenuation from Reservoirs and Lakes||from 0 – 1|
|DRAIND||Ratio of length of river network to catchment area||km/km2|
|S1085||Slope of the main stream between the 10 and 85 percentiles of the main stream length||m/km|
|ARTDRAIN2||Proportion of the river network that is included in drainage schemes||from 0 – 1|
|URBEXT||Proportion of the catchment area that is urbanised (per Corine Landcover 2000 dataset)||from 0 – 1|
|ALLUV||Proportion of the catchment area that is covered by alluvial deposits||from 0 – 1|
|FOREST||Proportion of the catchment area that is covered by forest (per Corine Landcover 2000 and FIPS datasets)||from 0 – 1|
Table 1: Physical Catchment Descriptors used in the FSU methodologies
The 7-variable equation used in the FSU methodologies is as follows:
In order to improve the initial estimate of QMEDrural,PCD from the 7-variable equation, an adjustment is applied based on how the equation performs at a ‘pivotal site’. A pivotal site is defined as the gauging station that is considered most relevant to a particular flood estimation problem at the subject site. Ideally, a pivotal site will lie a short distance upstream or downstream from the subject site at which the flood estimation is required, but it can also be geographically close or hydrologically similar to the subject site.
The general procedure is to infer an adjustment factor to the QMEDrural,PCD estimate by examining the performance of the equation at the pivotal site. This pivotal site adjustment factor, AdjFac, is calculated at the pivotal site as follows (below right):
The adjustment is then made at the subject site by multiplying the estimate of QMEDrural, PCD by the pivotal site adjustment factor (AdjFac). A final adjustment to the estimate to account for urbanisation is made to complete the calculation.
- Flood Frequencies – estimation of flood-growth factors (XT) at ungauged locations
In the FSU methodologies, it is possible to estimate site-specific flood-growth factors. This is in contrast to the FSR methodology which provided a single set of growth factors for all catchments in Ireland regardless of shape, size or catchment characteristics.
In the FSU approach, a pooling group of gauging stations is created that comprises the corresponding catchments which are hydrologically similar to that of the subject catchment. The number of stations in the pooling group is decided using the 5T rule – that is, the group should comprise a set of gauging stations with a total record, in terms of station-years, of at least five times the target return period.
For example, to estimate the 100-year return period flood the pooling group must possess approximately 500 years of annual maxima data. The L-moment statistics of the gauging stations in the pooling group are derived and the weighted average of all these L-moment ratios are used to select the most suitable statistical distribution from a plot of standard curves. The chosen distribution is then used to derive the site specific flood growth factors at the subject site.
- Hydrograph Width – derivation of design hydrographs at ungauged locations
The aspect where the FSU approach departs most from methods used previously in Ireland is in the area of deriving design-flood hydrographs. The rainfall-runoff method used in the FSR has been abandoned in favour of a ‘parametric approach’ to deriving design hydrographs. No rainfall inputs are used in the FSU approach; instead, the method is based solely on the analysis of recorded hydrograph data from flow gauging stations in Ireland.
This analysis found that hydrograph shapes at ungauged locations can be constructed using a combination of two curves. A Gamma curve describes the rising limb of the hydrograph as far as its point of inflection on the recession limb. From this point of inflection, the rest of the hydrograph is defined by an exponential decay curve.
Three different hydrograph shape parameters are used to derive these curves, and hence the design flood hydrograph:
- n is the Gamma curve shape parameter that describes the shape of the rising limb of the hydrograph;
- Tr is the translation parameter for the Gamma curve. The time origin (t=0) of the model is at the peak of the hydrograph. Thus, t = -Tr defines the time at which the hydrograph commences; and
- C is the recession parameter of the exponential decay curve which replaces the Gamma model beyond its point of inflection on the recession limb.
These parameters are estimated using PCDs in the form of regression equations, and are then adjusted with reference to a hydrograph pivotal site.
Intended users of the FSU methodologies
The methodologies described above are fully implemented in the FSU Web Portal which is intended to be used by the OPW, local authorities, third-level institutions and private-sector practitioners and will be continually applied and referred to by professionals working in the area of flood risk assessment and management in Ireland. It is expected to be used mostly for commercial purposes in the design of engineered structures and other flood risk management activities but also by third-level Institutions and other training organisations for educational and research purposes.
The OPW will undertake day-to-day management of the portal, together with periodic updates of the associated methodologies and improvements. The OPW is very interested in receiving feedback from users (via email to: firstname.lastname@example.org) and will use this feedback to identify future improvements to the FSU Web Portal.
In November and December 2014, the OPW provided introductory training lectures on the FSU methodologies and web portal. These lectures were held in Athlone, Cork, Galway, Kilkenny and Dublin. The presentations from these lectures are now available on the Documents page of the FSU Web Portal at the following web address: (see FSU Web Portal Training Documentation).
The final lecture of the series in Dublin was hosted by Engineers Ireland’s Civil Division. This was video recorded and may now be viewed on the Engineers Ireland website via the Engineers TV Archive web page at this web address.http://www.engineersjournal.ie/2015/02/24/opw-flood-studies-web-portal/http://www.engineersjournal.ie/wp-content/uploads/2015/02/New-Picture4.bmphttp://www.engineersjournal.ie/wp-content/uploads/2015/02/New-Picture4.bmpCivilflooding,OPW,software