Landslides

, Volume 10, Issue 5, pp 599–610

Landslide management in the UK—the problem of managing hazards in a ‘low-risk’ environment

Authors

    • School of Earth and Environmental SciencesUniversity of Portsmouth
  • M. G. Culshaw
    • School of Civil EngineeringUniversity of Birmingham
    • British Geological Survey
  • C. Dashwood
    • British Geological Survey
  • C. V. L. Pennington
    • British Geological Survey
Original Paper

DOI: 10.1007/s10346-012-0346-4

Cite this article as:
Gibson, A.D., Culshaw, M.G., Dashwood, C. et al. Landslides (2013) 10: 599. doi:10.1007/s10346-012-0346-4

Abstract

The UK is a country with limited direct experience of natural disasters. Whilst landslide losses are not negligible and fatalities are rare, accounts are under-reported. Financial losses from landslides are poorly understood but likely to be considerably in excess of £10 million per year. As a result, a strategic management framework has evolved based upon small, low-impact events punctuated by occasional larger events or larger landslides affecting urban areas. We present an overview of the different landslide management mechanisms in the UK and discuss them in context of cases studies to explore their effectiveness. We conclude with three issues that may have implications for landslide management in the UK and other low-risk countries. Firstly, the evidence base by which landslide hazards and risks are measured is insufficient and limitations in existing information need to be better understood. Secondly, existing guidance on strategic and responsive management needs to be assessed for its fitness for purpose. Thirdly, we encourage debate about the importance of near misses.

Keywords

ManagementLand-use planningResponsive planningNear misses

Geohazards and landslide hazards in the UK

In an international context, the UK is relatively unaffected by major disasters. The country is seismically stable, contains no active volcanoes and due to a temperate climate, has few problems with droughts or wildfires. Economically, storms and flooding are the most significant hydro-meteorological hazards. For example, the floods of June–July 2006 that affected areas of England and Wales caused £4 billion of damage comprising £3 billion in insurance claims (Pitt 2008) and about £1 billion in further costs (Chatterton et al. 2010). However, flood events are not regular and only affect vulnerable floodplain and coastal areas. Swelling, and particularly shrinkage, of clay soils are believed to cause the largest ongoing financial losses of any ground condition, predominantly in the highly populated southeast of the country (Culshaw and Harrison 2010) but do not cause injury or loss of life.

Information on the real impact of these hazards and of landslides is difficult to obtain. Despite the presence of over 14,000 landslides in the UK (Foster 2008), there is little information on their impacts, due perhaps to their age, size, isolation or inactivity. Better information exists for large active coastal landslides, such as Black Ven, Sidestrand, Ventnor and Scarborough or for significant events where an emergency response has been required such as Aberfan, Nefyn or the Scottish Debris Flow cluster in 2004.

The central contention of this paper is that even though overall losses due to landsliding in the UK may be low, we may be underestimating the true nature of the problem until they are properly understood and discussed. In the authors’ experience of working with geologists, civil engineers, planners, politicians and the general public, landslides are recognised as a problem, but the threat that they pose to life, property and infrastructure is not always well understood. Many scientists and engineers seem to regard them as a localised problem, occurring only at a limited number of sites or under extreme weather conditions. In the light of this perception, using recently available information on landslide susceptibility and from case studies, to discuss how landslides are managed in the UK.

All recorded deaths due to landsliding in the UK have resulted from rock falls, debris fall or debris flows (Table 1). The largest loss of life caused by a single landslide in the UK occurred on 21 October 1966 when a waste tip formed by coal spoil failed and destroyed a school in the South Wales town of Aberfan. One hundred forty-four people were killed, most of whom were children (Anon. 1967). In recent years, the only other landslide incidents to result in multiple fatalities involved a school party at Lulworth Cove, 21 February 1977, and an incident at Boscombe, Wiltshire, which killed three workmen in 1925.
Table 1

Recorded fatal landslides in the UK categorised by type and by geographical setting

Landslide event

Year

Killed

Type (after Varnes 1978)

Geographical setting

Newbiggin, Northumberland

2010

1

Rock fall

Coast

Whitehaven, Cumbria

2007

1

Debris fall

Coast

Ben Nevis, Lochaber

2006

1

Rock fall

Mountain

Nefyn, Gwynedd

2001

1

Debris flow

Coast

Newquay, Cornwall

1986

1

Rock fall

Coast

Durdle Door, Dorset

1979

1

Rock fall

Coast

Lulworth Cove, Dorset

1977

3

Rock fall

Coast

Swanage Bay, Dorset

1976

1

Rock fall

Coast

Kimmeridge Bay, Dorset

1971

1

Rock fall

Coast

Aberfan, S Wales

1966

144

Debris flow

Hillside/spoil tip

Alum Bay, Isle of Wight

1959

1

Rock fall

Coast

Boscombe, Wiltshire

1925

3

Rock fall

Coast

Loch Ness, Scotland

1877

1

Rock fall

Mountain

Indirect fatalities such as vehicle collisions or dam bursts are not included nor are those fatalities occurring in quarries

Despite the small dataset, the nature of fatal landslides provides important points relevant to our discussion. First, there are very few fatal landslides in the UK. Second, fatal landslides are spatio-temporally isolated and, with obvious exceptions, tend to involve only single individuals. Third, there is no consistent pattern from year to year.

Taking into account the limited number of data and the statistical influence of the Aberfan disaster, there appears to be a trend of increased reporting of fatal landsliding in the UK during and since the 1970s. This is in line with the findings of Petley et al. (2005), a study of worldwide trends. A detailed research is underway to investigate this pattern in the UK, but it is possibly a consequence of increasing utilisation of coastal sections for leisure activities (French 1997). The trend may also reflect a general increase in UK society’s sensitivity to the impacts of natural disasters.

Regardless of the statistical significance, it is difficult to underestimate the wider impact of the Aberfan disaster upon British society. Prior to this event, there was little government interest in mitigating risks posed by hazardous ground conditions—rebuilding the economy and national infrastructure following World War II was the priority, with resources prioritised towards developing the manufacturing sector, and supporting exploitation of energy and mineral resources. The disaster prompted a government-appointed Tribunal of Inquiry into the landslide (Anon 1967). Research into mapping landslides was stimulated in the 1960/1970s after unexpected problems during construction projects highlighted ground movement as a potentially significant problem in other parts of the country (Early and Skempton 1972; Skempton and Weeks 1976; Chandler 1970).

In the mid-1980s, the UK Government’s Department of the Environment (DOE) commissioned a large study to build a national register (database) of landslides in Great Britain (not the UK, therefore omitting Northern Ireland, the Channel Islands and the Isle of Man). This was completed in 1994 by Geomorphological Services Limited in collaboration with Rendel, Palmer and Tritton. The work was a desk study of readily available information sources including the literature, British Geological Survey (BGS) maps, selected local government records and selected consultancy reports. Due to the diverse nature and detail of the information available and of database software then available, a pro-forma-based procedure was undertaken to ensure that key information was captured as consistently as possible (Lee and Moore 1996). This study was effective in providing the first evidence base for the pattern of landsliding and landslide hazards across Great Britain based upon 8,835 landslide records (Jones and Lee 1994). As discussed by those authors and Foster et al. (2008), the study was constrained by the amount of information available, particularly in terms of activity and precise location but did provide sufficient information to inform government on the development of planning guidance (Department of the Environment 1990; Jones 1999). The survey also identified the need for detailed studies commissioned by the government in, for example, South Wales (Jennings and Siddle 1998; Siddle et al. 2000) and the Isle of Wight (Lee and Moore 1991). Following completion of these studies, anticipated long-term funding to develop the national database was not forthcoming and further large-area formal studies ceased.

Since 2000, a new national landslide database has been developed incorporating the DOE Database, post 1994 mapping campaigns and other literature sources. The methodology and design of this inventory is described by Foster et al. (2008) and Pennington et al. (2009). By the end of 2010, this national landslide database contained information on over 15,000 landslides in Great Britain.

Since 2006, the database has also incorporated information reported in the media (Table 2). Although not a complete record of all events, web and media sources provide an efficient and cost-effective way to gain understanding of landslide trends (Petley et al. 2005). The data showed that on average, about 28 landslides are reported in this way in the UK every year. However, not all these landslides will be added into the BGS Database, as some information reported lacks detail and cannot be verified without field evidence, between 10–15 new landslides entered each year.
Table 2

Landslides reported in the media in the UK and considered for inclusion in the database, April 2005-December 2010

Year

Number of landslides

2010

38

2009

38

2008

10

2007

38

2006

32

2005 (from April)

11

A geographical information system (GIS)-based system, GeoSure, has also been developed to assess the principal geological hazards across the country. The model is based upon 1:50,000 scale geological mapping, information from the database, information from the National Geotechnical Database and from expert guidance from regional mapping teams (Foster et al. 2008; Walsby 2007). Output is a series of GIS layers that provide ratings of the susceptibility of any location to a range of geohazards or ground conditions (Walsby 2008). The level of susceptibility is communicated differently to different audiences, either as a number range, a rating A–E or as ‘low’, ‘moderate’ or ‘significant’ (Fig. 1). With output at 1:50,000 scale and combined with locations of population centres and properties, the model can be used to provide a useful estimate of the significance of landsliding nationally. According to the dataset, 350,000 households in the UK, representing 1 % of all housing stock, are in areas considered to have a significant landslide susceptibility.

Land-use planning for landslides in the UK

There is no centralised legally binding mechanism for the management or mitigation of landslides in the UK. In England, issues relating to landslide problems are dealt with under common law and the primary responsibility is with the landowner (Department of the Environment 1996b). The planning system, through the Town and Country Planning Act (1990), regulates the development of land, and its powers are exercised through local authorities. The Housing Acts (various) and Building Regulations Act allow local authorities to exercise control over certain aspects of development. Similarly, the Construction (Design and Management) Regulations (Anon. 2007) are relevant to safe construction. Further considerations of public safety are made in the Coal Mining (Subsidence) Act (1991), Coal Industry Act (1994) and Occupiers Liability Acts (1957, 1984). The elements of these regulations and acts relevant to local authorities regarding landslide management are as follows:
  1. 1.

    Strategic planning: production of a development plan, which, once approved by the central government, is a blueprint for development of the whole area for at least the next 10 years (Alker et al. 2002).

     
  2. 2.

    Development control: provide planning permissions for (re)development that accord with the development plan ensuring that these meet all statutory and regulatory requirements (Alker et al. 2002).

     
  3. 3.

    Building control: ensure that buildings are designed, constructed or altered so as to be structurally safe and robust (Anon. 2004).

     
  4. 4.

    Emergency planning: plans prepared for, and enacted in the event of, crises and disasters (Alexander 2002).

     

The government has advised planning authorities, through Planning Policy Statements (PPS), that the planning system must regulate the development of land in the public interest and take into account whether proposals would affect amenities, building and land which should be protected in the public interest (Office of the Deputy Prime Minister 2005; Communities and Local Government 2007, 2008). Further guidance covering unstable land (including landslides) has been given in Planning Policy Guidance Notes, PPG14 and Annexes (Department of the Environment 1990, 1996b; Department for Transport, Local Government and the Regions 2002) which set out the responsibility of the developer to determine whether the land is suitable for the proposed purpose. PPG14 and its annexes provide guidance on how slope instability should be considered in any planning decision and that, if landsliding is a known issue, ‘a developer’ must provide evidence that any development activity will not exacerbate landslide activity and that any building will be safe. The guidance note related to slope instability is not legally binding. It does state that ‘The stability of the ground in so far as it affects land use is a material consideration which should be taken into account when deciding a planning application.’ However, it goes on to state that ‘Many local planning authorities may not, in any event, have the required expertise available to them. Where relevant expertise is available on issues such as mineral planning, waste disposal, land reclamation, building control, surveying or engineering, the local authority should endeavour to make use of it. The list does not include geological or geotechnical expertise, but details of some sources of information are provided, including the BGS. There is no legal compulsion for a planning authority to understand the extent or nature of landslide hazards within their area of concern and, thus, include them in planning decisions.

Building regulations provide a further control on the impact of slope instability requiring that The building shall be constructed so that ground movement caused by (a) swelling, shrinkage or freezing of the subsoil; or (b) landslip or subsidence (other than subsidence arising from shrinkage), in so far as the risk can be reasonably foreseen, will not impair the stability of any part of the building’ (Anon. 2004). Again, the liability is placed upon the developer, but there is no compulsion to seek out or use information that might indicate landslide hazard in or around the development site. Building regulations and planning policy guidance were reviewed by Brook (1991, 2007). Brook and Marker (2008) warned about the threat to the perceived status of PPG/PPSs, particularly PPG14, which covers many geohazards, as the current planning system is ‘reformed’. These reforms may further diminish the degree to which planning controls can be used as a mechanism to control landslide hazards. In the same paper, they draw attention to the fact that many authorities lack sufficient expertise with which to make detailed landslide assessments with the information they have available.

The delivery mechanism for health and safety legislation is a complex series of regulations that can oblige operators of utilities and infrastructure such as roads, pipelines and railways to protect citizens from any potential harm caused by their operations. These mechanisms can place responsibility on the operators requiring them to prove to the government Health and Safety Executive that their activities do not compromise the safety of others. For instance, industry-specific regulations require that quarry, pipeline and chemical warehouse operators and other such organisations must consider the potential impact of landslides upon their operations and any potential impact upon third parties. Although a complex system, it addresses concerns for specific industries. For instance, operators of chemical warehouses fall under regulations that require the submission of a report for the Health and Safety Executive which demonstrates that a risk analyses has taken into account all relevant ‘off-site accident initiators’ which could include seismic events, flooding, subsidence and landslip. Current guidance in this matter suggests that operators will ‘assign [event] frequencies on the basis of engineering judgement or historical records’ and demonstrate that appropriate action has been taken to reduce risk to an acceptable level (Health and Safety Executive 2012).

In Scotland, The Town and Country Planning (Scotland) Act 1997 is the primary document governing land use and development. Scottish Planning Policy (2010) provides broad guidance to developers and planners under which environmental problems including unstable land should be taken into consideration. This refers to the guide for planners and developers in Great Britain (Department of the Environment 1996a). Specific guidance in Scotland is also given for peat slides affecting wind farm construction (Scottish Executive 2006) and debris flows affecting the road network (Winter et al. 2005).

Individuals are also responsible for not causing damage to their property or that of others by exacerbating landslide hazards. For the most part, this will be managed by the application of planning or building code regulations, enforced by local governments. The Highways Act (applies to England) and the Roads (Scotland) Act both allow the roads authority to require third parties not to act negligently, etc. and gives the powers for them to act to rectify and recharge (see Winter et al. 2005, Section 2.5). Citizens, landowners, operators and others will all come under the umbrella protection offered by the insurance and re-insurance industries. Most insurers provide some sort of geohazard cover as part of buildings/construction insurance policies. Insurers will, of course, assess risk according to their own policies and procedures, but it is useful to consider that not all companies utilise geological data in their appraisals.

None of these mechanisms or supporting documentation provides, or was intended to provide, comprehensive, up-to-date guidance on the following:
  • the information that should be sought to assess landslide hazard;

  • sources of appropriate information;

  • how landslide information should be used and interpreted;

  • how landslide information should be presented;

  • what measures are appropriate in terms of public safety; nor

  • what measures are appropriate for long-term remediation.

The responsibility for accessing and using appropriate information always lies with the developer, operator or their technical advisor. Recently, the international scientific community has produced a series of guidelines for local and national government officials, land-use planners, geo-professionals and project managers on landslide susceptibility, hazard and risk zoning (Fell et al. 2008a, b). These guidelines are similar to those produced in Australia (Australian Geomechanics Society 2007a, b). The guidelines are highly technical, and no attempt is made to relate them to specific planning regulations in individual countries. More accessible guides on landslide management and investigation have been produced, notably McInnes (2000, 2007). Although these are specific to southern Isle of Wight, they provide broad advice that can be applied elsewhere.

Most planning policy guidance predates the era of GIS and advises that individuals or organisations consult geological maps and the Department of the Environment Landslide Database. However, both datasets were difficult to access and virtually unintelligible to anyone without a geoscience background. Much of this information, and other datasets held by government, councils, companies and government institutions is now available online. As it stands, even though there have been significant improvements in geohazard models and the generation of hazard data, much at public expense, there is still no requirement to use or even consider the most modern data within a landslide hazard assessment. To the authors’ knowledge, only the most recently published guidance takes the National Planning Framework for Scotland refers its audience to online flooding maps as the suggested data source.

The insurance of landslide losses in the UK

Losses caused by landslides were incorporated into domestic property insurance policies in the early 1970s (Doornkamp 1995a, b; Wyles 1983). The umbrella term ‘subsidence’ came to be used for natural ground movements whether caused by landsliding, dissolution, swelling and shrinkage of clays or some other geohazard process. Indeed, the vast majority of losses incurred by the UK insurance industry have been caused by shrinkage of clay soils during extended periods of low rainfall. These losses have occurred mainly in the south east of England.

However, claims against subsidence losses, including those caused by landslides remained very low throughout the 1970s and 1980s. Then, following a particularly dry period between 1989 and 1991, the UK insurance industry paid out £1–2 billion. Since then, payments have averaged between £3–400 million per year (Culshaw and Harrison 2010). Less than 5 % of these losses are thought to relate to landsliding. The industry realised that it needed to understand better the risks to which it was exposed from geological hazards. As a result, a digital geohazard information system (GeoHAzard Susceptibility Package) was developed, which came to be used by around 35 % of the industry (Culshaw and Kelk 1994). This system was developed at a time when digital geological information was only just becoming available. Following the digitisation of most of the BGS spatial data in the early 2000s, a much improved geohazard information system (GeoSure) was developed (Walsby 2007; Foster et al. 2008). However, such information systems do not appear to be widespread, even in the more developed world.

Though usually provided by mortgage providers, buildings insurance is not compulsory in the UK; when obtained, it does provide the property owner with a good measure of financial cover in the event of landslide damage occurring despite policy holders having to pay part of the cost (excess) of any claim. However, premiums and excess contributions are often higher for owners in landslide-prone areas, though the value of cover is of course of particular value to owners of protected properties, especially those constructed before the planning guidance and building control standards, described above, were developed.

Landslide risk to infrastructure

It is the responsibility of infrastructure owners (pipeline, power line, road, railway, canal, etc.) and certain landowners to assess the risks to their operation and third parties. However, this is regulated by health and safety legislation, and depending upon the institutional interpretation of guidelines, the level of ‘acceptable’ risk must be set in line with guidance provided by the UK Health and Safety Executive. As a result, infrastructure owners must assess the risk of damage, injury and death resulting from failure of their utility and take steps to reduce it to the acceptable level if the risk is too high. Such an assessment should take into account natural hazards, though of course, in the context of other potential risks, landslides may be of little significance and warrant little detailed study.

Gibson et al. (2005) described research for the major national distributor of natural gas to determine the distribution and potential severity of geohazards across the UK. Of the many geohazards that affect the UK, landsliding and the dissolution of soluble rocks were considered to pose the greatest threat to the transmission network. BGS national hazard datasets for landslide and dissolution hazard were truncated to buffer zones centred upon the pipeline. These data were enhanced by detailed information from the BGS National Landslide Database and Karst Database (for instance, Cooper 2008; Cooper et al.2001). The result of this research was a set of GIS layers that showed the level of susceptibility to landsliding or dissolution of soluble rocks at any location on the pipeline network in the UK. The operator used these to improve their risk assessment methodology, inform discussions on pipeline safety with their safety regulator and improve their surveillance strategies where it is considered that geohazards represent a potential threat to the integrity of the pipeline.

Response to major landslides in the UK

As with the planning system, response to landslide events in the UK is essentially a local matter, with little guidance available on the responsibilities to be taken or procedures to be followed. In the event of a major landslide, initial response is controlled by the civilian emergency services, police, fire service and medical personnel, working with the local authority and others within the framework of the local authority emergency plan (Siddle 1999). A multi-agency control centre may be formed to deal with initial rescue and security efforts. For example, following the Holbeck Hall landslide, police and the coastguard provided security for the first 72 h to prevent access to the slipped area (Clements 1994). Once this had been completed, responsibility for the site was passed to the landowner of the property affected or from where the landslide was sourced. Often, it is the local government body or operator of a road or pipeline that will be responsible for repairing or stabilising the site to the extent that it will no longer pose a threat.

Less significant events, where a landslide may not cause immediate threat to life or property but may cause disruption to roads or railways, such as an embankment collapse, also tend to be dealt with locally. Typically, the landslide will be cleared within a few hours, with no record kept of the nature of disruption or repair. The Highways Agency, for instance, which is responsible for the maintenance of motorways and strategic roads in England, does not routinely record landslide events that affect its network; rather, it records that a repair action has occurred at a certain chainage, with little supporting information on the nature, cause or impact of the failure.

Some aspects of how this system operates in reality for both major and less significant landslides are illustrated by four case studies:
  • Holbeck Hall Hotel, North Yorkshire, England, 1993

The Holbeck Hall Hotel in Scarborough, North Yorkshire, England, was destroyed as a result of a landslide that took place between the night of June 3 and 5, 1993 (West 1994; Lee et al. 1998; Lee 1999; BGS 2008a). The hotel had been built in 1880 close to the coastal cliff top; the coast was known to have been susceptible to coastal instability and erosion including some notable landslides (though not recognised as landsliding at that time). Following an unusually wet period in May (West 1994), the slope underlying part of the hotel failed, possibly in part due to pre-existing failures. The foundation of the hotel was partially undermined and most of the structure of the hotel subsequently collapsed (Fig. 2). The remains of the hotel were demolished. Following a rapid movement, where tens of meters of cliff were lost in a few hours, the main part of the hotel was situated above the rear scarp where movement was relatively slow (sufficiently slow to allow monitoring visits to the site over the next few days by the second author here amongst others). The slow movement at the hotel also meant that all residents could be evacuated and that electricity and gas supplies could be switched off before they caused fire.
https://static-content.springer.com/image/art%3A10.1007%2Fs10346-012-0346-4/MediaObjects/10346_2012_346_Fig1_HTML.gif
Fig. 1

Landslide susceptibility map of Great Britain derived from the BGS GeoSure model. This version of the dataset has been produced for the general public, with susceptibility zones used to depict areas of nil to significant landslide potential

https://static-content.springer.com/image/art%3A10.1007%2Fs10346-012-0346-4/MediaObjects/10346_2012_346_Fig2_HTML.jpg
Fig. 2

The Holbeck Hall Hotel landslide, Scarborough, North Yorkshire. The image shows approximately 60 m of recession that occurred on the morning of June 4, 1993, with the hotel situated partially over the rear scarp

The landslide occurrence led to a series of legal arguments due to a dispute over who was ultimately responsible for the landslide—whether it was the servant landowner (the owner of the hotel), the local authority upon whose land the actual slide took place, or the organisation that had, some years before, been commissioned to investigate the potential for slope instability in the area. The nature of legislation, where a duty of care is owed by the landowner (but only if they know the extent of the problem and where there is no compulsion for them to fully investigate the problem), led to a situation where there was no clear responsibility for the event (Holbeck Hall Hotel Ltd v. Scarborough Borough Council, English High Court, London 2000).
  • St. Dogmaels, Pembrokeshire, South Wales, 1994

On 14 February 1994, following heavy rainfall, landsliding was reported on slopes above the village of St. Dogmaels (also known as Llandudoch), Pembrokeshire, Wales, which lies on the left bank of the Afon (River) Teifi (Fig. 3). Though the landslide appeared to be slow-moving and did not immediately threaten the village, provisions were made to implement the local emergency plan, which involved evacuating the village. This preparation was led by the police with the support of the fire brigade, ambulance service and the military. On advice from engineering geologists, it was decided not to implement the emergency plan, though people from houses located on the landslide were evacuated. Several houses were severely affected. The landslide caused damage to an 11-kV power line and to main water supplies. In the event, most of the properties in the village were unaffected, but residents did suffer some distress.
https://static-content.springer.com/image/art%3A10.1007%2Fs10346-012-0346-4/MediaObjects/10346_2012_346_Fig3_HTML.gif
Fig. 3

Landsliding at St. Dogmaels. (Courtesy of Dyfed Police)

Subsequent investigations showed that the landslide was a reactivation of part of an ancient (12,000 years old), complex retrogressive failure. The landslide was moderately deep-seated and took place in glacio-lacustrine deposits which were up to 60-m thick infilling a deeply incised valley (Fletcher and Siddle 1998). Ground movements of up to 1 m began at the head of the landslide and gradually progressed downhill. Movements continued for almost 8 months until drainage measures were implemented. The ground movement affected the upper part of the ancient landslide, which, itself, extended down to the edge of the main part of the village of St. Dogmaels.

The landslide appeared to be caused by the presence of artesian and sub-artesian groundwater levels in the glacio-lacustrine deposits and periods of higher-than-average rainfall in the months and weeks leading up to the failure (Maddison 2000). Other factors, such as the presence of a waste tip at the head of the landslide, poorly maintained land-drains on the landslide and poorly controlled drainage of water from springs upslope from the landslide, may have contributed to the failure.

Planners in the local authority had no previous knowledge of the landslide. This is not surprising, as it was not shown on the then current geological map which dated from the nineteenth century nor was it described in the scientific literature.

One outcome of the landslide was the commissioning of a geological mapping research project to identify geological factors relevant to planning and development in the catchment of the Afon Teifi (Walters et al. 1997). The study included a detailed landslide survey which identified 385 landslides; almost none of which were recorded previously. The results of the study have enabled planners to provide information, to developers and others seeking planning permission for building and construction, on the potential for landsliding. This must be taken into account in the submitted planning application.
  • Glen Ogle, Stirling, Scotland, 2004

The Glen Ogle debris flows were part of a cluster of landslides that took place in the summer of 2004 in the Highlands of Scotland including the A83 trunk road between Glen Kinglas and to the north of Cairndow and the A9 to the north of Dunkeld (Winter et al. 2006; BGS 2008b).

The Glen Ogle events occurred on slopes formed within the Dalradian (Late Pre-Cambrian/Cambrian) Ben Lui Schist Formation. These are characteristic of a classic glaciated terrain, with flat-topped hills (blanketed in peat) above steep (15–45°) slopes that lead down into flat-bottomed valleys. Near the valley floor, the A85 road traverses the side of one of these valleys, as a narrow two-lane road. Existing streams and channels in the schist are culverted underneath the road.

On 18 August 2004, an intense rainfall event followed a period of lower intensity rain above the upper slopes. This intense rainfall displaced a small peat slide (estimated in the field by lead author shortly after the event as 30 m2). This peat slide set off a cascade within the stream channel, first ripping up weathered bedrock, then mobilising a relict deposit of granular head/landslide material. This dense flow moved rapidly down slope, scouring superficial materials on the way and eventually overwhelming the drainage culvert designed to protect the road (Fig. 4). A second flow, also overwhelming its culvert, then blocked the road south of the first flow, trapping 58 people in 20 vehicles between the two flows. Fifty-seven people were airlifted to safety by military helicopter, and one person was leaving the scene by foot (Winter et al. 2005). Vehicles were retrieved over the next 48 h, after the debris flow had been cleared. The emergency evacuation was a complete success. However, this disruption to the road was significant and lasted for 4 days after the flow.
https://static-content.springer.com/image/art%3A10.1007%2Fs10346-012-0346-4/MediaObjects/10346_2012_346_Fig4_HTML.jpg
Fig. 4

Main debris flow deposit at Glen Ogle, a second flow south of this flow (to the right of image) blocked the roadway, trapping 58 people on the road

A key outcome from the Glen Ogle and related events that summer was the realisation by Transport Scotland that there was little strategic information on the debris flow problems that affected upland Scotland. As a direct result, a study was commissioned, which developed a model of the distribution of potential debris flows in Scotland and the risks to the road network associated with them (Winter et al. 2005). The final results of this study were published by Winter et al. (2008) and will probably form the basis of a programme of remediation and further investigation of the problem.
  • Nefyn, Gwynedd, North Wales, 2001

The fourth example presented involved a fatality and further illustrates the danger in the UK posed by debris on steep slopes. The slide involved a mass of only 100 m3 of material, moving just 12 m down slope. The landslide occurred at Nefyn, Gwynedd, on the north coast of Wales. The coastal slopes in this area comprise a complex sequence of glacial till and pro-glacial outwash deposits. Over the past 16,000 years these deposits have been affected by deep-seated rotational landslides, coastal erosion, periglaciation and weathering, resulting in a coastal cliff 60 m in height, with a distinctive stepped profile and a series of sub-horizontal benches separated by low cliffs. The horizontal benches have provided flat ground where debris, comprising weathered soil from the cliff and rotten vegetation, had accumulated in thicknesses of up to 8 m. The days preceding the landslide had been characterised by very high rainfall, which had left the debris accumulation saturated. On 2 January 2001, a small landslide in the upper cliff moved down the slope onto the debris accumulation, causing this to fail by undrained loading. Two separate debris lobes moved down the slope, engulfing a number of parked vehicles, pushing two of them over a low sea cliff, resulting in a fatality to an occupant of one of the cars and serious injury to another (Fig. 5).
https://static-content.springer.com/image/art%3A10.1007%2Fs10346-012-0346-4/MediaObjects/10346_2012_346_Fig5_HTML.jpg
Fig. 5

Landslide at Nefyn. Note the dark mudslide at the top of the slope (centre bottom of image) that could have triggered a similar debris flow above a residential property. (Courtesy of North Wales Police)

The emergency response to the landslide was controlled by local police and fire officers who made the area safe and allowed access to investigation crews and specialist personnel. Following the event, a survey was commissioned to carry out an investigation into the cause of the landslide and the surrounding area. The result was a hazard map that provided guidance on future development in the vicinity (Gibson and Humpage 2002; Gibson et al. 2002; Jenkins et al. 2007). The report recommended that no development be allowed on, or below, the debris-covered slopes; this recommendation has been followed, and no new development has been allowed. However, there is no mechanism to prevent alterations, improvements or repair to existing property in the area. Since the 2001 landslide, a number of nearby properties have been damaged by small landslides and have been rebuilt. There is no legal mechanism to prevent this so long as there is no risk posed to third parties by the actions of the property owners.

Responsive management of landslides in the UK

Although there are only a few examples presented here, they have been selected to illustrate the response to landslides that had a direct impact on people. When a landslide event occurs in the UK that requires emergency action, the response, by the emergency services, military, local government officials and general public is exemplary. It appears, from the case studies, that incidents are dealt with promptly in a well-structured and well-considered manner to ensure that public safety is maintained. In the case of the Nefyn landslide, this was confirmed by the Coroner’s Court set up to inquire into the death that occurred. Each of the events presented here was followed by some form of investigation or review of the cause and impact of the landslide and led to follow-on actions, usually in the form of a public inquiry followed by a change in local planning policy, or in the case of the cluster of flows in the Scottish Highlands in 2004, a national-scale study to fully investigate and assess the problem.

Financial losses due to landslides in the UK

A brief examination of recent events presented by this paper shows that there have been, on average, around 28 landslide events in Britain each year reported by the media since 2005 (Table 2). As recording of new events and further research into past events continues, this picture may become clearer. Several of these events caused significant disruption to important transport links and had financial impacts of several million pounds.

Little information has been found about the total economic costs of landslides in the UK. For the Holbeck Hall landslide, economic costs, including emergency response, engineering works and insurance claims have been estimated as being in excess of £3.5 million (£2 million insurance claim and £1.5 million for the emergency protection scheme needed to protect the slope from further landslides [Byles 1994; Clements 1994; Isle of Wight Centre for the Coastal Environment 2006; Forster et al. 2006]). For the St. Dogmaels, landslide costs were around £2.5 million (£1.9 million for the remediation scheme [Anon. 1999] and at least £0.5 million for the demolished or damaged houses). However, costs are not available for the 2004 Scottish debris flows (including the Glen Ogle debris flows) and the Nefyn landslide of 2001. Costs for smaller landslides are very variable according to the assets they affect. For example, the Pennan landslide in Aberdeenshire in 2007 cost around £600,000 for stabilisation alone (BBC 2009); the Cooper’s Hill landslide in Gloucestershire cost £1.24 million for stabilisation (BBC 2011).

These estimates of cost relate mostly to the direct cost of remediation. Little information is available on the indirect costs such as the cost of disruption to traffic and to the local economy. These figures do not take into account a further category of landslides: those well-known complexes that have a continuing impact upon specific locations. These include the large areas of landsliding that affect the coastal towns of Ventnor, Isle of Wight (for example, Hutchinson and Bromhead 2002; Lee and Moore 2007; McInnes 2007), Lyme Regis, West Dorset (for example, Clark et al. 2000; Sellwood et al. 2000), Scarborough, North Yorkshire (for example, Lee and Clark 2000) and the World Heritage Site of Ironbridge, Shropshire, (for example, Culshaw 1973a, b; Carson and Fisher 1991). The continuing costs of some of these major UK landslide complexes are available. For example, the total cost of the Ventnor Undercliff landslide on the Isle of Wight has averaged £1.4 million per year over the last 20 years in terms of structural damage, insurance costs, engineering measures and monitoring (McInnes et al.2000). However, Lee and Moore (2007) estimated annual losses at Ventnor at £4.64 million per year.

Since 1990, investigation and engineering works at Lyme Regis have cost over £30 million (including works not directly involved with slope stabilisation or coast protection), with a further £23 million requested for future developments (Lyme Regis Environment Improvements Team personal communication 2011; Anon. 2010). Landslide research, investigation, mitigation and remediation have been going on at some of these locations for more than 50 years. For example, in the Ironbridge Gorge, Henkel and Skempton (1954) discussed the causes of the Jackfield landslide that ultimately contributed to the closure of a railway line and the placing of previously buried services on the ground surface where they need regular maintenance. Carson and Fisher (1991) suggested that the landslide may have been active for over 200 years. In addition, a government review of the cost of coastal erosion in the UK found that the average cost was £126 million per year, a significant part of which (not identified by the study) actually resulted from landsliding (Office of Science and Technology 2003). However, such schemes are usually integrated with environmental improvement schemes that have design briefs far outside that of simple landslide remediation.

The figures above cover major landslides, long-term landslide complexes and small disruptive events. Further small losses are experienced in relation to houses. We estimate that 350,000 houses in the UK are in areas where there is a significant landslide susceptibility. In December 2010, the average purchase price of a house in the UK was approximately £162,435 (Halifax 2010). Using these figures, we estimate that, if these figures are correct and that all properties in the significant landslide susceptibility are equally vulnerable, there is potentially, £57 billion worth of housing stock at risk of landslide damage in the UK. Hypothetically, if just 1 % of those properties received 1 % damage in any given year, the total losses would be in the region of £5.7 million per year.

Until detailed studies are completed, we must rely on these broad indices from major landslides and potential housing stock vulnerability to understand the economic impact of landslides in the UK. At present, specific data on insured losses to housing is not easily available. This is partly because losses caused by a range of geohazards (swell-shrink, dissolution, compressible soils and landslides) are generally grouped together under the category subsidence (Culshaw and Harrison 2010). It is also difficult to compare the losses from landslides with those resulting from other natural hazards such as flooding. Losses from flooding are around £500 million per year (Dailey et al. 2009), while losses from geological hazards (subsidence) are of a similar magnitude (around £350 million per year (Culshaw and Harrison 2010)). However, this figure does not separate out the contribution from landsliding alone.

The importance of ‘near misses’

The case studies presented here were chosen also to illustrate a further point. Although each of these events was damaging and one fatal, they could be considered to be ‘near misses’—the consequences of all of these events could have been much more significant.

The Holbeck Hall landslide could have occurred more rapidly, limiting the possibility of escape for the hotel residents. Equally, the landslide could have led to fire or could have resulted in collapse of the building on residents as they tried to escape. Had the landslide occurred during the day, more people may have been on the beach in front of the hotel. The St. Dogmaels landslide could have continued downhill to engulf more densely populated parts of the village. The Glen Ogle debris flow could have resulted in the deaths of some of the persons rescued. The timing of the events, where one debris flow blocked the road and a second trapped the parked vehicles was unusual. The fact that no one was actually struck by either debris flow and that neither debris flow travelled any distance down the roadway into the vehicles was fortunate.

The event at Nefyn, led to the death of a resident of a nearby village and serious injury to her partner. However, the occupants of other vehicles were able to escape injury. It was also fortunate that a separate mudslide (bottom centre in Fig. 5) stopped part way down the cliff. If it had travelled further down slope, it could have triggered more movement within the perched debris accumulation. This could have triggered a similar debris flow to the fatal slide. In this instance, the debris flow would have impacted on a part-wooden building used as a restaurant and could have injured or trapped customers inside at the time.

Between them, the authors of this paper have investigated or reported on many landslides which could be considered to be near misses, and a great many more case studies could have been presented in this paper. However, at some point in the near future (years or decades), there will be a more tragic event that involves multiple fatalities. The international experience strongly suggest that the potential for successful rescue from a landslide depends very much upon the nature of the failure as much as human factors such as the speed of response and medical facilities. Although the emergency response to each of the events described here was excellent—timely, structured and well-executed—there has been an amount of good fortune that the slides were of such a nature that rescue was possible and human losses, low.

Discussion and conclusions

The authors wished to explore some experiences of responsive and strategic planning for landslides in the UK and reflect upon how effective these may be. We hoped to investigate whether a relatively safe natural environment has led to any important omissions in our approach to landslide management.

It is difficult to escape the fact that, in the UK, in a ‘normal’ year, an average landslide will cause little long-lasting damage and will be adequately managed by local action. The case histories given here indicate that the emergency services react promptly and effectively to any event. Similarly, for small landslides that obstruct communication links or impact on properties, local authorities, infrastructure and utility operators respond efficiently. In other words, landslide management in the UK is working well at a responsive level.

Strategic planning for landslides also appears to be working effectively. Guidance for new development should, on the whole, reduce the potential for damage. Planning guidance for geohazards including landslides exists (though its application may be inconsistent), as do appropriate building codes.

However, there are significant weaknesses, possibly resulting from a position of perceived safety that we consider to have potentially serious social and economic implications:
  1. 1.

    The evidence base, by which landslide hazards and risks are measured, is insufficient. There is considerable uncertainty as to the true impact of landslides, and thus, our evaluation of the effectiveness of landslide management is based upon poor data. Geohazard databases developed in the late 1990s provided a valuable overview of hazards in the UK, but lack of long-term investment seriously curtailed their use by planners and developers. New geohazard information systems will rectify this, but fully populating them and making them available will take several years. We have produced a number of estimates based upon available information and suggest that costs are at least in the region of tens of millions of pounds every year. However, until detailed studies are complete, there is a fundamental lack of understanding of the true amount landslides cost, where these are accrued and by whom.

     
  2. 2.

    Despite the guidance given to government and local authorities, it is of a broad nature. Unless authorities have invested over a long period of time to build and maintain an expertise, they may have insufficient information or expertise with which to make informed decisions, even with access to state of the art information.

     
  3. 3.

    There have been sufficient cases, a few of which are illustrated here, to demonstrate that there have been significant near misses where serious loss of life or property has been narrowly avoided. Understandably, planning guidance, insurance premiums and health and safety policy is based upon the best recorded, most representative events. As mentioned above, in the UK, these are small, relatively low-impact events (often unreported). It is our contention that, particularly in countries with perceived low-risk environments, it is important to take proper account of near misses in any calculation of risk or mitigation strategy. These events have the potential for huge impact yet are often poorly understood and poorly reported. We hope, in this paper, to encourage discussion on this matter and promote the publication of case studies to expand the evidence base, within the context of realistic risks of what might have happened as well as what has.

     

Acknowledgments

This paper is published with the permission of the Executive Director of the British Geological Survey (NERC). The authors would also like to extend their thanks to the two anonymous reviewers for their candid views, helpful comments and factual contributions to the paper.

Copyright information

© Springer-Verlag 2012