Natural Flood Storage and Extreme Flood Events Final Report
5 CASE STUDY 1 - WHITE CART WATER
5.1 Background
White Cart Water, a tributary of the Clyde, drains a catchment area of approximately 250km 2. The catchment does contain some undeveloped floodplain areas upstream of the south Glasgow-Paisley-East Kilbride conurbation. A number of the tributaries of the White Cart are already regulated by headwater reservoirs. The built-up area covers about 25% of the catchment, though grassland is the major land cover at about 55% (improved or rough). The shallow fast flowing river is prone to flooding, with 20 significant floods since 1908.
Glasgow City Council is actively pursuing a scheme to alleviate flooding caused by the White Cart. The current preferred option involves the construction of three flood storage areas with high embankments (with the crests 9-15m above the river bed level) upstream of the main urban area, together with some new flood defences in southern Glasgow, at an estimated cost of around £30 million. The proposed storage areas are significant engineered features and therefore do not constitute natural flood attenuation.
5.2 Modelling
The first step in the modelling of the White Cart was to estimate peak flow magnitudes for flooding at the downstream risk location, which was chosen to be Overlee Gauging Station (GS) (Table 5-1).
Table 5-1: White Cart Water - Flood flows and return periods
Location | Peak flow (m 3s -1) | Return period (years) | Notes |
|---|
Overlee GS
(NGR NS 579 575) | 104
184
214 | 5
100
200 | Flows derived from White Cart Millennium scheme study |
An outline of the routing model available for the White Cart catchment is shown in Figure 5-1. The model is a variable parameter Muskingum-Cunge model where routing parameters are determined from cross section properties. This was calibrated such that the peak flows agreed (at least to a close approximation) with the values given in Table 5-1 for the corresponding return period.
The downstream hydrographs for the larger events were then compared with the peak flow for the 'threshold' event, and the volume that would have to be stored was calculated (Table 5-2). These figures can be considered as minimum required volumes of flood storage (strictly speaking assumed to exist immediately upstream of the risk location).
Figure 5-1: White Cart Water routing model
Table 5-2: White Cart Water-Storage volumes derived from hydrograph analysis
| Return period of event (yrs) |
|---|
Location | Incident event | Target peak flow (or volume) | Volume (million m 3) |
|---|
Overlee GS
(NGR NS 579 575) | 100 | 5 (flow) | 1.0 |
The routing model and JFLOW flood inundation model were run to create flood outlines for the specified return periods. The area of inundation was calculated and plotted as a function of distance upstream from the risk location. We have then calculated the differences between the areas of the larger (200 year) and smaller (100 year) events to represent the 'natural' area in which water could be held back to mitigate the larger event. These results are shown in Figure 5-2. Individual tributaries of the White Cart have been shown at the appropriate locations. The steep sections of the graph at around 1km and 4km upstream of the flood risk location may offer good potential to hold water back during the larger event within the 'natural' flood extent.
Figure 5-2: White Cart Water - Distance-area curves for natural flood extents
We have made a simple assumption that the volume of storage needed at the downstream flood risk location can be divided by the modelled flooded area to give a notional average storage depth (Table 5-3). We have calculated the required average depth using both the total extent of the 200 year flood (i.e. the flood event we have used in this study as the extent of the 'natural' floodplain), and also the marginal extent between the 100 year and 200 year outlines.
Table 5-3: White Cart Water - Notional average depth of natural flooding
Event return period reduction (years) | Volume
(million m 3) | Available extent | Available area
(km 2) | Average depth
(m) |
|---|
100 to 5 | 1.0 | Within 200 year extent | 2.0 | 0.5 |
100 to 5 | 1.0 | Between 100 and 200 year extent | 0.1 | 10.0 |
Based on these figures and the distribution of floodplain extent (as a function of distance upstream from the flood risk location) we have then calculated the average depth of water that would be required on the floodplain to achieve the required volume of storage (Figure 5-3).
In general, storage is most effective when it is located immediately upstream of the flood risk location. If the average storage depth falls to a practically-attainable value within a few kilometers of the downstream flood risk location, then there may be scope to use land with the natural 200 year extent to provide the required storage. This could be the case for the White Cart as the average storage depth falls to 2m within 2km of the downstream risk location.
Figure 5-3: White Cart Water - Distance-depth curves for natural flood extents
5.3 Environmental assessment
The White Cart catchment contains a number of sites designated as being of conservation and/or historical importance. A map showing the Sites of Special Scientific Interest (SSSIs) and Scheduled Ancient Monuments (SAMs) within the White Cart catchment is given in Figure 6-4. Few of these sites (such as the Cart & Kittoch Valleys SSSI and Polnoon Castle) are located within or near to the modelled 200 year flood outline, which has been used in this project to define the 'natural' floodplain. However, if a more extreme flood outline was modelled then further sites may be required to be considered.
Figure 5-4: White Cart Water - SSSIs and Scheduled Ancient Monuments
The White Cart floodplain also contains many man-made assets that could be affected by enhanced inundation. For example, Scottish Power have provided the project with a GIS dataset showing the locations of all their electricity sub-stations within the White Cart catchment. Figure 5-5 indicates that a number of these assets are located within the 200 year flood outline and would require a more detailed assessment, together with consultation, should any proposed scheme to enhance natural floodplain attenuation be taken forward.
Figure 5-5: White Cart Water - Electricity assets
5.4 Agricultural economic assessment
5.4.1 MDSF-based
The first methodology, based on the MDSF technique derived from the England and Wales conditions, provides an estimate of the flood damage to four land cover classes, namely: arable (non-cereals)/horticulture (including potatoes, brassicas, carrots, sugarbeet, salad crops), arable (cereals), intensive grass and extensive grass.
The estimated total cost of the 5 year, 100 year and 200 year events using the MDSF methodology are shown in Table 5-4. The 100 year and 200 year totals are not significantly different as the modelled total inundated area for these land cover classes was not very different for the two events.
The results indicate that the damage to the arable (non-cereals) and horticulture land caused by the two extreme floods completely controls the overall economic cost. This is due to the very high value of this land cover in comparison to all the other land covers, including arable (cereals).
Table 5-4: White Cart Water - Economic cost of flooding on agricultural land, based on MDSF
Flood return period | Cost
(£) |
|---|
5 year | 2,500 |
100 year | 43,760 |
200 year | 46,690 |
To provide a concise summary, the results are also presented as a function of distance upstream in Figure 5-6. Within the White Cart catchment the much higher agricultural flood damage costs, associated with arable and horticultural land cover classes, exist in the area >6km upstream from Overlee gauging station. Any proposed 'natural' floodplain attenuation scheme should therefore target the areas nearer to the flood risk location.
Figure 5-6: White Cart Water - Distance-cost curve for the 200 year natural flood extent
5.4.2 Single flood compensation payment based
A simple estimation of the potential cost of inundating large areas of agricultural land on the floodplain was used that assumed one single value for all agricultural and rural land cover classes (excluding classes 1, 2 and 3 - water, bare rock/land and built-up) and therefore, by implication, all grades of land quality. Recent publications from England have suggested that a typical annual payment of £300/ha would provide a landowner with a reasonable amount of compensation for allowing his/her land to be flooded.
Table 5-5 provides a summary of the overall costs of permitting the land to be flooded using the single payment for all land cover classes (excluding land cover classes 1, 2 and 3).
Table 5-5: White Cart Water - Annual compensation costs, based on single payment
Flood return period | Area inundated
(km 2) | Cost (@ £300/ha)
(£) |
|---|
5 year | 0.5 | 15,300 |
100 year | 1.6 | 48,900 |
200 year | 1.7 | 51,300 |
The total catchment figures are typically 50% less than those derived from the MDSF methods, again indicating the large impact that the high value arable and horticultural crops have on the MDSF figures. However, these figures provide some indication of the magnitude of the longer term total annual compensation payments that might be required if the full 'natural' floodplain was to be actively managed for flood attenuation purposes.