CHAPTER 6: FLOOD MAPPING
Methodology
6.88 This chapter deals with the conversion of the model results into flood outlines. For many watercourses flood levels from the hydraulic model are then extrapolated and compared with local topography to define the floodplain. Where there are complex overland flow paths or where the floodplain is unlikely to be represented well by this extrapolation; 2-dimensional floodplain flow modelling using JFLOW has been undertaken. These schemes were modelled using the weir flow values from the hydraulic models, input into JFLOW, to derive the flood extents over the scheme reach. The flood outlines are provided in the Database and in the Appendices of the site specific reports and an example can be seen in Figure 6-1 below.
Figure 6-1: Example Flood Outlines Fort Augustus
NEXTMapDTM
6.89 The NEXTMap Digital Terrain Model of England, Scotland & Wales was completed by Intermap Technologies and is currently the most accurate countrywide DTM available. It has a nominal vertical accuracy of 1.0 metre ( RSME) and horizontal nominal accuracy of 2.5 m in areas of minimal vegetation and terrain slope of less than twenty degrees. Readings are available at 5 m postings. The NEXTMap was supplied for this study by Scottish Executive.
6.90 The NEXTMap topographic data was collected using Intermap's Star 3i IFSAR (Interferometric Synthetic Aperture Radar) using a Learjet flying at an altitude of 28,000 ft. This is a similar methodology to the orbiting radar imaging satellites but with an increased accuracy and by using two radar antennae, a significant reduction in time for data collection is achieved.
6.91 NEXTMap is provided in several different datasets:
- A Digital Surface Model ( DSM), which uses the first return values for and minimal filtering of data that represents the surface features of the topography including buildings and trees.
- A Digital Terrain Model ( DTM), which represents the 'bare earth' of the topography and has the surface features filtered out. As a result of the filtering, the elevation values have a reduced accuracy.
- Orthorectified Images ( ORI) - Photographs taken from the sensor of the ground and georeferenced.
6.92 For the purposes of this study the NEXTMapDTM was the most suitable dataset to use for extending survey cross sections across floodplains and for flood mapping. The DTM's accuracy varies and it is less accurate in areas of high profile, highly urbanised or wooded valleys, partly as a result of filtering techniques.
NEXTMap and Survey Levels
6.93 For the majority of the modelled schemes, threshold levels were measured as part of the survey to assist both with the flood extents and calculation of economic damages.
6.94 The surveyed river cross sections used to create the river models were geo-referenced to allow the extension of sections using the NEXTMapDTM. For a number of scheme sites the elevation values of the last surveyed point differed considerably to that of the underlying DTM (This is illustrated by Figure 6-2). A number of reasons could be accounted for this including steep river banks, tree canopies and vegetated areas or the proximity of buildings.
6.95 In general we found that the accuracy of the DTM in urban and wooded areas was poor but an increased accuracy was obtained in more rural open areas. Where detailed surveys were available it was still difficult to find areas consistently within the error margin quoted even in relatively open areas. The NEXTMapDTM did, however, provide a good indication of the general landform in the vast majority of locations.
6.96 This difference between the survey data and the underlying DTM varied for each scheme with the maximum difference being approximately 4 m. Where this occurred on a high steep bank it had little effect on the flood extents. Part of the survey work involved the collection of threshold levels on properties within the floodplain. These threshold levels were used where necessary to modify and validate the DTM, particularly in flatter areas. In several schemes enough threshold and spot levels were taken during survey work to allow the production of a more accurate DTM, which could be stamped onto the underlying NEXTMapDTM.
Figure 6-2: Typical HEC- RAS output showing the difference between the NEXTMapDTM and survey cross sections
Delineation of Flood Extents
6.97 Floodplain extents have been derived using modelled levels and refined using JFLOW for some locations. Where applicable river cross sections were extended using the DTM and the HEC- RAS levels projected. Contours were generated using the NEXTMapDTM to allow this to be performed with greater accuracy. For schemes where the overland flow path was not obvious JFLOW was used in conjunction with the DTM to determine the likely flow paths. Where available, historic flood outlines have been referenced as part of the flood extent generation.
Description of flood outlines
6.98 For all schemes in detail, flood outlines were determined for the 100yr (1% AEP), 200yr (0.5% AEP), 1000yr (0.1% AEP) and the Standard of Protection events. These outlines were derived for both the scheme with defences in place and the scheme with no defences (i.e. before the scheme was constructed). As a number of schemes have a standard of protection lower than 100 years the database also shows the areas that benefit from the scheme and the associated return period and probability.