Cookies

We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Department of Geography

Departmental Research Projects

Landslide Assessment and Flood Erosion Risk from the North Yorkshire Floods

A research project of the Department of Geography.

Background

The recent well-publicised floods in North Yorkshire (19 June 2005) have again focussed attention on the devastating impact of flash floods in the UK. This adds to a catalogue of recent events including catastrophic hillslope failures in Western Ireland and Shetland in September 2003; the Boscastle disaster in August 2004; devastating landslips and debris flows in Central Scotland in August 2004; and an unprecedented number of shallow landslides in Northern Cumbria in January 2005. The North Yorkshire floods were considered so severe that they prompted an emergency debate in Parliament (29.06.05) with MPs demanding ‘We must learn the lessons from events like this’. The area most severely affected was the upper river Rye in the North York Moors National Park (150 km2) where there was widespread valley floor flooding and multiple landslides, which added to the flood devastation. Our preliminary survey has identified over 100 landslide scars in this area and possibly the largest peat landslide in England.

Local variations in: rainfall intensity, topography, surficial geology and runoff have combined to influence the pattern and magnitude of slope failure and landslide-derived debris has impacted on the extent of flooding and channel erosion. Patterns of erosion, and the debris carried by the flood are critical components in determining loss of valuable upland habitat and farmland; destruction of farming infrastructure; loss of livestock; elevation of flood levels; structural damage to river structures (e.g. bridges, culverts, etc.); ecological impacts along the main water course; and deposition of sediment in homes and properties downstream. Assessments of sediment impacts are very rarely undertaken because the focus is generally on the hydrology of the flood and evidence of erosion and sedimentation is rapidly lost in the landscape. However, understanding the spatial controls on landsliding and the significance of landslide-derived flood debris in accentuating valley floor flood inundation patterns are arguably the biggest challenge in the hazard assessment of upland flooding. Experience suggests that much of the key evidence of how and why landslides occurred is lost very quickly as subsequent rainfall erodes the scar surfaces and deposits. Floodplain high water marks and flood debris are rapidly masked by vegetation or disturbed by clean up operations making reconstructions of flood levels and flood inundation patterns very difficult. Hence when events occur it is particularly important that impacts are assessed rapidly and accurately. Improving the efficiency and accuracy of data acquisition is fundamental to understanding flash flooding, as critical evidence from flood events is transient and is rapidly masked by other landscape processes in the first few months following a flood.

Aims

The objectives of this work are to:

  1. Assess the impact of the flood in terms of the sediment contribution from peat mass movements and shallow mineral landslides; and their link with channel erosion and sedimentation at the catchment scale
  2. Quantify the amount of sediment removed by detailed morphological survey of a sample of landslide scars using aerial photography, modelling of LIDAR data and ground survey techniques. We aim to deliver a new methodology for calculating shallow landslide volumes in upland terrain using 3D remote sensing
  3. Map the extent of valley flooding and compare this to Environment Agency flood extent predictions and recent 2D flood diffusion wave models. Use this with sediment delivery data to assess the impact of sediment volume on flood extent
  4. Correlate erosion and sedimentation patterns with rainfall intensity measured by rainfall RADAR4

It is important that this is done with urgency for two main reasons:

  1. before definition of the flood features is lost (this particularly applies to deposits)
  2. so we can report back quickly on our key findings and our analysis can progress in parallel with the Environment Agency in their hydrological analysis thus ensuring maximum synergy between the two groups

Our work is distinctive, yet would compliment the hydrological investigations by the EA by demonstrating the interactions between sediment delivery and flooding in the upper catchment.

Methods

Our project will progress through five key stages:

  1. Acquisition of new high resolution imagery for mapping slope failures and site conditions (with visible, near infrared and thermal imagery), setting out ground control and field survey of key landslide sites and geomorphic features
  2. Detailed quantitative morphometric modelling of landslide volumes and patterns of erosion using height data modelled from photography and LiDAR calibration with field survey results
  3. Mapping flood inundation patterns from remote sensing data and comparison with the EA flood extent predictions; examination of residual patterns of difference
  4. Comparison of landslide sediment delivery data with flood extent maps to assess the impact of sediment volume on flood extent
  5. Correlation of erosion and sedimentation patterns with rainfall intensity measured by rainfall RADAR

The proposed project relies on rapid acquisition of remotely sensed data and detailed recording and mapping of the field evidence. We have already secured low-resolution digital aerial photography from the NERC Airborne Remote Sensing Facility (as part of their urgency procedures) in the immediate aftermath of the flooding. The existing data however, is only low resolution digital photographic imagery and is insufficient to provide quantitative analysis of the landslides hence the need for detailed ground survey and further airborne remotely sensed imagery including LiDAR. The NERC ARSF imagery is able to provide photography, LiDAR and thermal imagery with sufficient geometric quality to make the processing and analysis at key sites in the Ryedale catchment a realistic data processing exercise. The results will provide an unparalleled opportunity to compare and contrast novel remote sensing methods with more established assessment techniques which are time consuming and expensive. Storm tracking will be carried out using Met Office (NIMROD) rainfall radar data from the British Atmospheric Data Centre (BADC).

Staff

From the Department of Geography