Research in brief
- The cause of rock fall hazards at coastal cliffs cannot only be attributed solely to environmental drivers, such as rain or marine erosion.
- Rock falls evolve over time and respond to a range of preparatory and triggering factors.
- Hard rock cliff faces that appear solid and stable are likely to be experiencing an ongoing reduction in rock mass strength which may in time result in failure.
- A rock fall model that is based on physical data and is inclusive of rock mechanics that determine the collapse of cliffs would improve understanding of how rock falls occur along coastlines.
- Precursors may exist for large rock falls, a sequence of smaller rock falls for example, and these provide a warning before the main rock fall event takes place.
The coastal cliffs of the North York Moors National Park in Yorkshire overlook one of the most breathtaking stretches of the British coastline. Varying in height from 25 to nearly 100 metres, the cliffs on this coast are rugged and steeped in a rich industrial heritage. Yet these cliffs have been transforming over thousands of years and today continue to change constantly.
Since 2002 a team of researchers led by Dr Nick Rosser based at Durham University’s Department of Geography and IHRR, have been monitoring this coast to build an improved understanding of the controls on the behaviour of cliff erosion, rock fall, and coastline retreat.
Perhaps surprisingly, the timing of rock fall from the cliffs cannot be easily attributed to simple environmental drivers alone, such as rain, wind and marine erosion. What appear to be contributing to the cause of rock falls at the cliffs are more complex processes that remain poorly understood.
Coastlines have received more attention recently due to the predicted threat of sea level rise and changes in storminess induced by climate change. Most attention is given to cliffs made up of softer rocks that recede more quickly, but many coastal areas are formed of hard rocks, and more research is needed to understand how these change over time.
Where rates of change appear extremely slow, making them challenging to monitor, long-term data is essential for building a proper understanding of retreat rates and controlling processes. This may be one reason why the causes of rock falls from cliffs remain difficult to pinpoint in many instances.
Understanding these hazards is important for risk mitigation, especially for coastal communities, where property values and the sustainability of infrastructure are closely linked to erosion risk. Despite knowledge gained from decades of scientific research of how coastlines change, there are as of yet no methods available that predict accurately when and where rock falls may occur.
These cliffs have been transforming over thousands of years and today continue to change constantly.
The cliffs of the coast of the North York Moors National Park are predominantly formed of hard rocks from the Jurassic and Cretaceous periods. They appear stable, with the exception of abrasion caused by incoming tides and waves. But closer inspection reveals that the cliffs are not only eroding at the toe, but across the entire cliff face, although this is easy to miss when observing them by eye alone.
Changes in coastal cliffs are difficult to spot because they are sometimes so slow or subtle that they go undetected. Using aerial photography for surveying cliffs is limited in practice, particularly on steep rock faces that are obscured when viewed from above.
In order to understand these unseen changes taking place in the coastal cliffs that can lead to collapse, new techniques are needed that allow researchers to monitor coastal cliffs more accurately and over long time periods. Fortunately, there is an advanced technology available that helps geoscientists to monitor the cliffs in ways never before possible. This technique is known as Terrestrial Laser Scanning (TLS).
A terrestrial laser scanner is equivalent to a 3D camera but instead of capturing a coloured pixel for every position upon the cliff, as with a digital camera, it uses a laser to measure the shape of the cliff to build a 3D point cloud of the surface. Comparing this data month to month helps researchers to monitor the underlying changes in rock cliffs that are otherwise invisible to the eye.
‘It’s accurate’, says Rosser, ‘it’s fast and convenient because you don’t have to go near the cliff, and we can capture extensive sections of the coast within one survey. The system captures data at very high-resolution, measuring the cliff surface every 3–5 cm across the face’.
Rosser’s research team travel to the same part of the coastline in the North York Moors National Park each month to capture the shape of the cliffs. The laser scanning system is fast enough to allow researchers to scan 3 km of the coastline in only one visit.
They have been doing this from the foreshore during low tide for 11 years in total, and now have the longest monitoring dataset of coastal cliff erosion. This makes their research site the most intensively monitored coastal cliffs in the world.
The data collected using laser scanning of the cliff was in some ways surprising in comparison to previous research on hard rock coastal cliffs. Erosion at the cliff toe is normally used to explain deterioration of the cliff via rock fall, but based on research by Rosser and his team, there is clearly more happening at the cliff face than previously thought.
While the cliff face may appear solid and stable, it is ‘constantly evolving’, says Rosser, ‘you can’t necessarily relate erosion or subsequent retreat to any simple or single environmental condition using conventional approaches’.
Rosser adds that some aspects of cliff behaviour are predictable. For example, sequences of rock fall are observed via a progression from small to medium to large rock falls. At present, coastal erosion models that only consider how wave action drives erosion may be ‘a bit misleading’, he says.
Instead, Rosser and his team suggest that monitoring data, which is inclusive of the rock mechanics that determine collapse, and that operates over the timescales of change observed at the coast, is essential to building an accurate model of how rock falls occur along coastal cliffs.
Much of the research on coastal cliff hazards such as rock falls is on cliff retreat – the movement of the coastline – and not erosion itself. Whilst erosion at the toe is important, it is not the only contributing factor to the development of the whole cliff via rock fall. This may be especially true when it comes to the larger rock failures when a larger part of the cliff suddenly gives way.
A rock fall recorded at one of the field sites on the North York Moors coast caused the cliff to step back 13 metres, generating a local tsunami. Rock falls on this scale are not well represented in current models because of the degree of complexity involved and the lack of monitoring data that includes such infrequent events. The timing and triggers of this rock fall and similar failures remain tricky to identify.
An important finding from this research has identified that ‘the development of the rock fall is primarily driven by external processes, such as wind and rain, but beyond this, internal controls, such as the fracturing and cracking of the rock itself, control when the rock fall occurs, its size and nature’, explains Siobhan Whadcoat, who is modelling rock falls identified in research for her PhD in the Department of Geography. Evidence to support this idea comes from sequences of smaller rock falls which occur before larger events; these smaller events can be precursors to a larger collapse later.
There is currently debate over environmental changes that may affect coastlines in the future, particularly those that can potentially be accelerated by the impacts of climate change. By definition, the previous focus only on the external mechanisms that affect coastal rock cliffs overlooks the internal processes that cause cliff rocks to break down over time.
Similar to landslides observed elsewhere, which sometimes occur without an immediately obvious trigger, coastal cliff rock falls are highly unpredictable, making them challenging to model conventionally. A rock fall may, for example, occur according to the gradual fracturing of the cliff rock, via a mechanism termed ‘progressive deformation’, where microcracking takes place within the cliff rock over time.
Research in rock mechanics is useful for understanding these processes, and might apply to coastal cliff failure, showing the influence of structure, strength, size, and tendency to fracture.
Waves continue to influence the cliffs during low tide: breaking waves on the foreshore, but also from further offshore, both transfer energy to the sea bed, which propagates as seismic waves directly to the cliff. Such effects are typically at their most intense during high spring tides where high water allows larger waves to reach the toe of the cliffs where they break.
These processes cause the cliff to vibrate, a process that holds some influence on occurrence of observed cliff rock falls. This observation is one of the main findings from research by Dr Emma Norman, who recently completed her PhD on coastal cliffs erosion on the North York Moors coast. ‘A small transfer of energy, say, from ocean waves, could be enough to trigger a rock fall, but the magnitude of the energy delivered by the waves doesn’t appear to directly relate to the magnitude of the rock fall that happens at that point in time’, says Norman.
During storms, as large waves hit the cliff face, there may be a cumulative effect that gradually or progressively weakens the cliff rock. In addition, if a small rock fall occurs, it can weaken the area around it by removing support, initiating a sequence of rock falls to occur over time. This would explain why a cliff along the coast that appears perfectly stable one moment collapses suddenly without warning the next, without an obvious trigger.
‘When a fragment of rock falls away there’s a change in the stress distribution through the cliff rock which then may cause further fracture, and, eventually, failure’, says Norman.
Processes affecting the rock cliff may include strong wind action, salt crystal growth, heavy rain or, during a storm, large waves crashing into the cliff, transferring large amounts of energy that can cause the cliff to shake. During large storms, local residents who live near the cliff have claimed that they could ‘feel’ such ground motions.
While it is clear that waves transfer energy to the cliff face, it is unclear whether they alone are causing rock falls to occur. Rock falls are complex, but there may be indicators that scientists can look for in order to predict when they occur.
In their research paper published in the journal Geology, Rosser’s team reported on sequences of rock falls that occurred at the coastal cliffs of the North York Moors. ‘Rock falls cluster in space, but they cluster in time as well’, says Rosser. The study identified precursors to the largest rock falls. ‘Even though the monitoring is high-resolution, the data is very noisy’, says Rosser, ‘there is rarely a perfectly clear signal’.
This is because there are many different things happening at the cliff face including weathering, but also variations in the rock, such as its strength, along with the shape and structure of the cliff itself.
Identifying risk indicators for rock falls before the ‘big one’ is not as straightforward as it would seem, because the cliff is undergoing so many changes at once at different scales.
However, Rosser and his research team have managed to identify patterns in the ‘noisy’ cliff.
Whilst smaller rock falls at the cliff face may appear to occur randomly, successive terrestrial laser scan data demonstrate ‘clustering’ of rock fall events in space and time. Small rock falls are always much more frequent than large rock falls.
Previously Rosser and his team recorded over 500,000 rock falls, but less than 100 of those made any noticeable difference to the cliff line observable by the human eye, and the changes were too small to be picked up by aerial photography or in maps.
‘Imagine the cliff as equivalent to a big game of Jenga, you pull a chunk out of the bottom, then after a certain amount of time the bit of the Jenga block directly above it might fall, because it’s weakened’, says Rosser.
Most of the failures the team monitored with TLS were shallow in depth, but would they have occurred regardless of any external environmental drivers, such as wave erosion? With more data the research team will be able to confirm whether small sections of the cliff falling one after the other is indeed an emerging pattern of rock falls.
Further investigation is needed. ‘We’re trying to look at those spatial relationships and also how fast that happens through time’, says Rosser.
Siobhan Whadcoat is developing a model that will use the long-term data set on the North York Moors coastal cliffs to simulate these patterns for the first time. This model incorporates failures around joints in the rocks, how failures interact, and how rock falls develop over time.
Whadcoat’s PhD research aims to account for the internal processes that cliff erosion models have not represented well in the past. She will also investigate the ways rock falls cluster together, which seem to hold clues as to how other rock slopes may also behave. ‘Overarching failure patterns could also apply to rock fall failures in non-coastal environments, in areas such as Yosemite in California’, says Whadcoat.
In researching the mechanisms that control rock falls, Whadcoat hopes to uncover what triggers events to occur suddenly, and to develop over long periods of time.
This is important because some of these unique characteristics of rock cliff behaviour are most likely universal, and worth knowing for any community that lives near a coastline or in a mountainous region where adaptation or mitigation of rock fall hazards is of paramount importance.
Brain, M.J., Rosser, N.J., Norman, E.C. & Petley, D.N. (2014) Are microseismic ground displacements a significant geomorphic agent? Geomorphology, 207, pp. 161–173.
Norman, E.C., Rosser, N.J., Brain, M.J., Petley, D.N. & Lim, M. (2013) Coastal cliff-top ground motions as proxies for environmental processes. Journal of Geophysical Research – Oceans, 118 (12) pp. 6807–6823.
Rosser, N.J., Brain, M.J., Petley, D.N., Lim, M. & Norman, E.C. (2013) Coastline retreat via progressive failure of rocky coastal cliffs. Geology, 41, pp. 939–942.
For further information about this research from the Coastal Behaviour and Rates of Activity (COBRA) project visit: http://www.community.dur.ac.uk/cobra/. The research is supported by Cleveland Potash Ltd. Contact Dr Nick Rosser: firstname.lastname@example.org