Urban flood risk mapping

The sources and pathways of flooding can be particularly complex in the urban environment (Fig. 18.12). Flood risk pathways in an urban system can comprise raised defences, surface flow pathways and the below-ground drainage system. How flood flows in these systems interact in terms of conveyance and storage and the consequences of the flooding are key challenges.

A study reported by Hankin et al. (2008) conducted as part of Making Space for Water (see Chapter 16) reviewed flood risk modelling and mapping approaches for multiple sources of flooding. To produce accurate models and maps requires large amounts of data collection and flood maps generated for non-fluvial and coastal sources on a wide scale using current flow modelling and data acquisition techniques are likely to be less accurate than the existing models for fluvial and coastal flooding. Table 18.7 gives an overview of possible approaches for sources of flooding other than rivers and coasts based on the findings of the report.

Fig. 18.12 Sources and pathways of flooding in the urban environment.

Table 18.7 Methodologies investigated by the Making Space for Water (see Chapter 16) Flooding from Other Sources Scoping Study (Hankin et al., 2008) for the feasibility of wide-scale potential flood hazard mapping

Flooding process

Methods reviewed for mapping

Comments

Direct runoff, integrated urban drainage

Sewerage and drainage system flooding from pipe capacity exceedance Sewerage and drainage system flooding from 'other causes' (blockage and collapse)

Restricted outlets from drainage systems due to high flood levels in the receiving watercourse

Blanket rainfall and 2D overland routing

Different levels of integration of sewer and overland routing

Different levels of integration of sewer and overland routing

As above, with blockage scenarios or probability matrices

Control the levels in receiving watercourses from outlets of hydraulic sewer model

Demonstrated for direct runoff with simple assumptions about drainage exceedance in terms of rainfall return period

Would require collation of a large amount of asset data and co-operation between data owners. Not currently feasible for a national product As above

Not feasible for national probabilistic mapping in the near future, although further investigation by the DTI SAM project (http://www.dti-sam. co.uk/) Would also require a process for combined probability analysis and national access to data sets as above

Table 18.7 Cont'd

Flooding process

Methods reviewed for mapping

Comments

Surcharge from small (ordinary) and 'lost' watercourses

Heavy, long duration rainfall leading to ponding on e.g. roads or fields

Rising groundwater levels in major aquifers

Groundwater rebound owing to rising water table and failed or ceased pumping

Groundwater flooding through alluvial aquifers

Canal breach, over-topping or leakage

Reservoir dam break (for reservoirs covered and not covered by the Reservoir Act 1975)

Block escape or limit capacity of floodwater in the 2D routing model

Topographic screening

Name search

Borehole flood frequency analysis

Buffer around historical records

Geostatistical methods

Theoretical modelling using the Theis equation

Generic groundwater modelling

Simple risk assessment with 2D overland routing Risk ranking method for prioritisation of reservoirs in need of risk mapping; 2D overland routing where inertial effects not dominant (as for canal embankments)

High costs. Approach can be included in an integrated urban model if a detailed survey is undertaken Yes, with remote sensed digital terrain data, but only for screening purposes. Requires further development of technique with larger rural dataset No, this approach did not give strong spatial correlation with historic flooding Yes, at a screening level, and subject to spatial coverage and record length of borehole level time series Yes, this would give detail of the specific locality of flooding for flood warning purposes Yes, but at a very coarse level. The interpolation of BFIHOST (see Chapter 13) is useful where the geology varies significantly, but for areas on the chalk, it is uniformly high Screening level for known problem areas. This would require collation of abstraction records and borehole data Further research is needed into how the approach can be used to extend the National Flood Map in areas likely to be affected Possible with costs of additional survey data requirements Subject of current scoping study to specify a national method for reservoir inundation modelling (RIM)

2D, two-dimensional.

The combined risk from fluvial and sewer flooding can be a combination of overland flow routing and the urban drainage infrastructure. While the most detailed studies can afford to gather the data and invest the time needed to build complex urban drainage models, in many places there is also a need for more rapid assessment tools. Some suitable approaches were reviewed and tested as part of a study in 2008 (Balmforth et al, 2006) in response to the Pitt Review into flooding of the summer of 2007 in the UK. The study assessed five different methods of varying complexity. This investigated how well the methods replicated known flooding for location and spatial extent. The methods tested were:

• topographic index analysis;

• 2D overland routing of a spatially uniform rainfall event;

• decoupled hydraulic sewer model and one-dimensional overland routing;

• decoupled hydraulic sewer model and 2D overland routing;

• coupled hydraulic sewer model and 2D overland routing.

A national map of surface water flood risk was produced in 2008 to obtain a rapid assessment of areas naturally susceptible to surface water flooding Design rainfall profiles for a 6-h duration, 1 per cent AEP storm were created using the rainfall depth duration frequency (DDF) models given in Vol. 2 of the Flood Estimation Handbook (see Chapter 9) on a 5 km x 5 km grid over the whole of the UK.

The design rainfall for each grid tile was then applied to a 2D hydraulic model to route the water over a digital terrain model (DTM). The resulting flood depths were used as an indicator of risk by applying thresholds at 0.1m, 0.3 m and 1.0 m depth. The resulting map, although not incorporating detailed information about the operation of sewer network, has been found to be consistent with recoded flooding incidents in many places and provides a high-level screening tool for more detailed investigations; Fig. 18.13 shows an example for a city centre area where roads, railway lines and some natural topographic drainage lines at risk of flooding are visible.

0 0

Post a comment