PREDICTING SLOPE INSTABILITY

Background

The coastal environment of the WHS is a combination of steep slopes, fractured rocks, high rainfall and active marine erosion. It is hardly surprising therefore that as a region it is characterised by a wide range of naturally active mass movements (landslides). In a recent study of these failures, McDonnell (2000) mapped three different factors that exert major controls on where failures occur.

  • Moisture – as indicated by the presence of groundwater seepage zones or ‘percolines’.
  • Slope steepness.
  • Geological structure – obtained by feeding rock properties, fracture patterns and densities into a numerical modelling package for slope stability analysis – UDEC.

From this analysis she identified that the area of greatest hazard lies in The Amphitheatre and around the headland of Lacada point. However, the majority of the cliffs in the immediate vicinity of the Giant’s Causeway also into the moderate or higher hazard classes. Within this broad zone, two areas of particular hazard were identified: Aird Snout and the section of cliff between the current gateway on the lower path and The Organ.


Triggering Slope failures

Whilst it is arguably possible to identify locations where there is a heightened risk of slope failure it is not possible to predict with the same precision when an individual fall, slide or flow will occur. This is because timing is a threshold phenomenon that depends on a complex balance between the strength of the slope material and the gravitational stress to which it is subject. When stress exceeds strength, this threshold is breached and failure occurs.

The difficulty is that there are many ways by which this threshold can be approached and ultimately breached. In the long-term, for example, the strength (competence) of a material will be gradually reduced by weathering. In some materials this may involve complete alteration through a combination of physical and chemical weathering. Whereas, a well-jointed, but otherwise resistant rock such as basalt may loose competence through selective weathering and opening out of the joint system, leaving the joint-bounded block relatively untouched. Conversely, the stress on a slope can be increased through, for example, undercutting and over-steepening it or by loading the top of it in some way.


Failures in unconsolidated material

It is rare, however, for gradual change alone to cause slope failure. This is because as strength and stress converge they become susceptible to short-term fluctuations in local environmental conditions. The most important of these is the reduction in the cohesion of weathered and unconsolidated materials brought about by increased wetness and the build-up of pore water pressure.

Eventually a point is reached where the material begins to deform under its own weight (plus that of the contained water) and fails through plastic deformation as a landslide. If there is a convenient failure plain within the material, for example, where weathered soil and regolith overly unweathered, cohesive material, it is likely that the material will slide downslope without rotating as a translational failure or landslide. Where there is no natural near-surface failure plane, and where the weakened material is of considerable thickness, the forces acting on the slope can resolve themselves into a curved, internal failure plain along which the material fails as a deep-seated rotational failure or landslide. In the most extreme case, slope material can become so wet that a positive pore water pressure is created, water begins to flow out of the material on to the surface, it looses all cohesion, takes on the properties of a fluid and moves downslope as a mudflow. In most cases, however, as the mudflow moves downslope water drains out and as a degree of cohesion returns it ceases to flow and continues to move through plastic deformation as a slide. Because of this most ‘mudflows’, especially if they are of any length, are more accurately referred to as ‘flow/slides’.

predicting1

Palaeosols (left at the outcrop) and till layers (right at the outcrop, overlaying palaeosol) are very prone to translational slides, mudflows and creeping. These shallow failures (some creeping can be seen on the right of the outcrop) are generally of limited significance, but can disturb human activity by directly affecting infrastructure such as roads or buildings.


Failures in unweathered material – rock

In unweathered, but well-jointed rock, failure is most likely to occur through the toppling or collapse of cliffed sections of slope to form a rock fall. This is because, although cohesion is not reduced at the micro-scale of individual grains and pores responsible for flows and slides in, for example, weathered materials, there can be a reduction in cohesion at a macro-scale as the friction along joints is reduced by a water film and/or localised weathering along joint planes.

Where well-jointed, competent rocks are underlain by less cohesive materials (for example, where basalt columns overly the weathered Inter-Basaltic Bed at the Causeway) the combination of loading and any reduction in the cohesion/strength of the lower material can produce a large-scale rotational failure involving both strata. This is especially the case if the open joints of the overburden also facilitate moisture flow into the underlying material. It was a failure of this type at Lacada Point in 1994 that lead ultimately to the closure of the lower cliff path (Figure 4.3). At a smaller scale, but a much higher frequency, the release of individual blocks from a jointed cliff face may be triggered by a localised increase in stress associated with, for example, overnight freezing and early morning thawing. There is anecdotal evidence that this is a common feature of the cliffs at the Causeway. Although it must be emphasised that freeze/thaw is only the final trigger mechanism and is underpinned by a long ‘preparation period involving wetting and drying, chemical weathering and the gradual opening out of joints near the unconstrained cliff face. There is also an argument that the regular, localised release of individual blocks could act as a ‘safety mechanism’ that prevents the accumulation of debris on the cliff face and the less frequent occurrence of major rock falls.

predicting2

These well jointed columnar basalts are very prone to toppling due to water and material infilling joints behind the columns and undercutting of their base by translational slides within the debris mantle below. Strong wind and rain storms are often the final trigger for failure.


Factors that trigger slope failures

In light of the above observations it can be seen that there are a number of factors that can combine to trigger slope failure.

  • A long-term decrease in material strength through thorough weathering of slope materials and a loss of cohesion.
  • Selective weathering of slope materials along joint planes.
  • Lubrication and loss of friction along joint planes by water ingress and flow. Freeze/thaw within already weakened rock close to its threshold of failure.
  • Undercutting and over-steepening of a slope, either by natural processes such as basal wave attack or human intervention to create, for example footpaths around the site.
  • External loading of slope materials through, for example, construction, vehicle traffic and debris accumulation.
  • Internal loading of slope materials through an increase in moisture content.
  • A reduction in strength (cohesion) through a prolonged and/or short-lived increase in moisture content.

Of all of the above factors it is arguably the last that is, in general terms, the most significant at the Causeway WHS. It is also true that the factors controlling moisture content are complex. However, at sites that are not fed from below by groundwater seepage, moisture content of slope materials is in general terms dictated by:

  • Rainfall duration.
  • The total amount of rain within a rainfall event.
  • Rainfall frequency, through its influence on antecedent moisture conditions, i.e. how wet the material is when it starts to rain and how much more rain is needed to saturate it.
  • Rainfall intensity. Up to a point, intense rainfall can more quickly saturate slope materials. However, if intensity exceeds the infiltration capacity of the material it can be lost as surface runoff, before the underlying soil becomes fully saturated.

On the basis of these observations, the ideal conditions to promote saturation of unconsolidated slope materials are a period of long duration, low intensity rainfall that comes on top of a prolonged period of wetness. Slope failures are most likely to be triggered if, within this long duration rain event there is a period of intense rainfall that can infiltrate rapidly enough to generate a positive pore water pressure.


An exception to the rule: Summer slope failures

The possible exception to the above rule concerns the failure of well-jointed, cohesive rocks such as the basalt columns of the Causeway Coast cliffs. There is, again, anecdotal evidence to suggest that many rockfalls are triggered by isolated periods of intense rainfall during otherwise dry periods. This is possibly because there is little constraint on the rapid infiltration of rainwater along joints that that had previously dried out and widened. Such a rapid uptake of moisture could drastically compromise the strength of the basalt through reduction of friction along joint planes. This is especially the case at the base of columnar strata, where infiltrated moisture may pond above a less permeable underlying layer comprised either of the entablature of the preceding flow or a weathered, inter-basaltic layer. Because of this, ‘dry weather storms’ may be a significant factor in the triggering of toppling failures – especially if the cliff has been previously undercut during, for example, footpath construction. However, as with freeze/thaw, any sudden failure is likely to represent the climax of a longer-term, cumulative decrease in local material strength and/or a gradual increase in stress.

predicting3

Translational flow – slide that occurred in August 2008. The summer of 2008 was a very wet season, which lead to numerous failures across the site.

up


Home  | People  | www.qub.ac.uk | www.uva.es | www.nationaltrust.org.uk |