9 February 2015
Simulating submarine landslides
Submarine landslides are deeply interesting – these are largest slides on earth in terms of both runout and volume. Recently, my twitter feed has been dominated by interest in a section through a Japanese landslide deposit that was tweeted to me:
Zane Jobe has a very nice blog post explaining that deposit. Submarine landslides typically generate these very complex deposits, primarily because of their complex dynamics. However, the processes involved in submarine landslides are rather hard to measure given the inaccessibility of the setting.
But, a very nice video has been published on Youtube by Joris Eggenhuisen showing an experiment to simulate the processes involved in submarine landslides. This is from the Eurotank Flume at Utrecht University:
For information, I have copied the supporting text from the Youtube page here:
This video shows a real-time recording of a scaled-down experiment of a subaqueous massflow, made with an under-water camera. The width of view is 0.5m, the depth of view is ~2.5m (considerable foreshortening due to low camera angle). The experiment demonstrates the sequence of events during mass transport events, which occur on the modern seafloor, and are recognized in geological deposits: A trigger (in this case an earthquake) destabilised the sediments near the seafloor. The down-slope pull of gravity then exceeded the strength of the seafloor sediment, and the sediment started to break-up, and slide down the slope, all the while transforming into a jumble of blocks. Near the bottom of the slope, where gradients decreased, the movement came to a halt. The change of light halfway through the video is associated with a time lapse of 24 hours during which the experiment was left untouched. Another earthquake was generated, which destabilised another patch of sediments on the upper slope. Some blocks could be seen to accelerate to much higher speeds due to a mechanism called “hydroplaning” in massflow literature. The deformation of the flowing sediment blocks completely transformed part of the massflow into a debrisflow. The massflow deposit of the previous day’s event acted as a buttress at the base of the slope, that was partly overriden by the 2nd flow. In the geological record, such a combination of deposits created by a number of consecutive massflow events coming from the same source area is called a Mass Transport Complex.
The video is really interesting, especially the way that the second slide develops a higher velocity because it is moving down the shear surface formed by the first failure, although its runout is short because of the block imposed by the deposit from the first slide. Noe the less it does over-ride the first landslide debris. And note also the very complex scar and deposit combination that is set up by even this small, simple landslide system.
4 February 2015
Harbury Tunnel landslide
Yesterday, Network Rail put out a press release providing a great deal more information about the status of the Harbury Tunnel landslide in Central England. From a landslides perspective the key information is that they had been undertaking stabilisation work on the slope after the landslides on 14th February 2014. They are now estimating that 350,000 tonnes of material will need to be removed, such that the railway line will remain closed for some weeks.
Perhaps the most helpful aspect is the images that they released of the site. This one shows an overview of the failed slope:
Whilst this annotated image shows the failed slope in more detail:
The openness and the clarity of information provided by Network Rail is very admirable. The Harbury Tunnel landslide is complex – the big (4 m) drop of material at the top of the slope appears to support a rotational mechanism, but the very complex deformation in the area of the gravel drains is more difficult to understand and requires proper mapping. The top image suggests that the slope is bulging in the toe area. Weather conditions are currently cold and mostly dry, and fortunately there is no heavy rainfall in the forecast for the coming days. Once the slope stops moving then it might be possible to start to carefully remove material from the crest, which will improve the stability of the site, allowing proper mitigation.
An important question is centred on how this failure occurred during the stabilisation works. Network Rail and its consultants have both very skilled engineers and robust design and construction practices. It will be interesting to find out in due course what unanticipated factor led to this slope failure. At present no structures at the top of the slope have been affected, and the tunnel itself looks to be outside the boundaries of the slide. From those perspectives the Harbury Tunnel landslide could have been far worse.
3 February 2015
Yesterday afternoon a really interesting landslide occurred at Vestfold, to the south of Oslo in Norway. This is the site of an elevated two lane in each direction highway – the major road that runs southwest from the capital. The landslide, which occurred on the edge of a small river, destabilised one of the pillars supporting the motorway, causing massive deformation of the road deck. Fortunately no-one was hurt. The event is described nicely in this report in English. These two images, from this forum, show the landslide and the deformation of the bridge deck:
This looks like a classic Norwegian quick clay landslide to me, but that is just speculation. Such landslides are often triggered by a comparatively minor disturbance that then triggers massive instability – the Rissa landslide being the most famous example. And interestingly, the images suggest that there was work going on under the bridge, because right in the middle of the landslide there is a bulldozer:
Unconfirmed posts on the forum suggest that this machine was doing some work reprofiling the field without the knowledge of the highways agency. Unfortunately this is going to take a long time to repair.
2 February 2015
The Harbury cutting and tunnel landslide
An unsually large railway slope landslide occurred on Saturday afternoon at the site of the Harbury cutting and tunnel, on the train line between Leamington Spa and Banbury in Warwickshire, Central England. The railway line is blocked. The Network Rail press release about the incident described it as being “extremely significant” and estimated that a volume of about 350,000 cubic metres of debris will need to be removed from the site to reopen the line. The images of the landslide are impressive:
The press release also suggests that the landslide is still moving and that the tunnel itself has been affected, so this may be quite complicated to put right.
In the UK there is a large community of railway enthusiasts. One of their websites has a discussion from four years ago about the Harbury cutting and tunnel site, which includes this image:
The tunnel is in the background. But perhaps most interesting is this historic image of the Harbury Tunnel and cutting being dug in April 1884:
“The problem of slippages in the very deep cutting of Harbury resulted in the GWR taking radical action. The dark areas adjacent to the line indicate what is left of the original cutting. As can be seen the GWR widened the cutting to create a more gentle slope adjacent to the line. The track is clearly standard gauge yet they are formed using the Broad Gauge method of construction of using longitudinal timbers under the rails with ties at six to eight feet apart. April 1884″
This site appears to have suffered some instability during the winter storms of February 2014, and news reports at the time suggested that the line was closed for a while. The cutting itself was constructed in 1847 and was for a while the deepest man-made cutting in the world, dug of course entirely manually.
1 February 2015
Retaining wall failure
I don’t know the background to this spectacular retaining wall failure, which was posted on Youtube on 28th January:
If you watch closely you can see the acceleration of the landslide in the period leading up to the collapse – right down to individual grains of soil. An additional interesting aspect is the mobility of the landslide – the large object (the truck) was much more mobile than most of the retaining wall.
I wonder what the retaining wall was founded upon? (Other than the obvious “not enough” of course).
29 January 2015
A very interesting slope problem appears to have developed at the site of the new Pakyong Airport in India. The airport project is described well in Wikipedia:
Pakyong Airport is an airport under construction near Gangtok, the state capital of Sikkim, India. The airport, spread over 400 ha (990 acres), is located at Pakyong village about 13 km (8.1 mi) south of Gangtok. It is the first Greenfield airport to be constructed in the Northeastern Region of India and the first airport in the state of Sikkim…The land for the airport was carved from the mountainside using massive geotechnical ‘cut and fill’ engineering works. These state-of-the-art geogrid soil reinforcement and slope stabilisation techniques were employed as traditional retaining structures and embankments were ruled out as being unfeasible. Irish geotechnical company Maccaferri executed the project that envisaged a 550 m (1,800 ft) wide, 1.7 km (1.1 mi) long corridor on which the runway and airport buildings are to be constructed. The company, which completed the project with partners Mott MacDonald and Punj Lloyd, won the ‘International Project of the Year’ award at the Ground Engineering Awards 2012 for its work in constructing 70 m (230 ft) high reinforced soil walls and slopes at the site. The project is among the tallest reinforced soil structures in the world.
The site is impressive – this is a recent Google Earth Image of the location:
Perhaps the most important aspect in terms of this blog is the very extensive slope cutting that has been required to create the bench for the airport site. Unfortunately, since the airport construction project started there have been repeated indications of slope instability on the upslope side of the airport site, although this is often reported in the media as subsidence. An example occurred this week (via the Tibet Earthquake Twitter feed):
The location of the monastery is shown as “temple” in the above image. Meanwhile, last week there were reports of landslides on other parts of the slope:
A key question must be what is occurring on this slope to cause these instability issues, and what are the implications for the airport site? Whilst the link between the airport slope works and the instability is of course not proven (and of course instability occurs naturally on slopes in this area), there must be real cause for concern.
27 January 2015
Tsarap landslide images
The Tsarap landslide crisis in northern India continues, although it is good to see that there has been an appropriate level of response to the hazard from the authorities. This is an interesting contrast to the initial indifference shown by the Pakistani authorities to the Attabad landslide five years ago. Over the weekend, efforts started to airlift people trapped to the north of the landslide – in total 300 people were trapped when the Chadar Trek was suspended. The same report also notes that the expert group formed by NDMA to advise on the management of the landslide dam will visit the site this week, so it will be interesting to hear what they have to say.
Meanwhile, also over the weekend, two sets of satellite images were published of the landslide site. The ever-wonderful NASA Earth Observatory published a beautiful Landsat 8 OLI image of the area, collected on 18th January, that very clearly shows the slide:
This is without doubt the best image of the landslide so far, showing both the source area and the slide itself, together with the impounded water. I remain concerned that the material forming the dam is potentially erodible and that the dam crest is short, both of which suggest that there is the potential for a rapid breach event when the water overtops the barrier. There remains some urgency to get a monitoring system in place for this landslide, and then to start the detailed planning for what will happen when snow melt starts to fill the lake more rapidly.
23 January 2015
This is probably one of the best landslide videos so far. It was posted on Youtube yesterday, with the title: “Terrible Landslide in the Mountains of Dagestan!“
It is a classic and very spectacular example of an earthflow. The people trying to rescue the car were fortunate, and indeed I suspect that with a winch the vehicle would be recovered. The rapid transition from slow creep to fast flow, occurring at the point of the screenshot below, remarkable.
22 January 2015
The Tsarap River Landslide in Zanskar
A new set of much better images of the landslide on the Tsarap River in the Zanskar Valley have been posted on Facebook by Jaskrit Bawa. The origin of the images appears to be the DC Office of LAHDC in Kargil. The most revealing image is this one, taken from a helicopter upstream of the barrier on the Tsarap River:
The landslide barrier appears to be tall but to have quite a narrow crest. Whilst there are many boulders in the debris, from these images there appears to be a great deal of fine-grained material as well. As an aside, the geological structure on the valley above seems to be quite unstable, with slope parallel discontinuities (bedding plans) that can act as detachment surfaces.
A downstream view is also quite interesting:
This image both confirms the fine-grained nature of the landslide materials but also suggests that there may be some seepage through the landslide mass occurring, given that there appears to be water in the Tsarap River channel. That water is not clean, which may suggest that there is some internal erosion occurring.
The National Disaster Management Agency is reported to be forming an expert group to help manage the hazard, which is welcome and wise. This is based on a very sensible assessment of the situation by the local authorities:
“The lake has been created around 90 km from the Padam area of Zanskar and beyond 43 km no one can go by foot. After consulting all local engineers, including Army engineers, we were not able to do anything,” said Muhammad Sadiq Sheikh, Deputy Commissioner, Kargil.
It is hard to know how serious the hazard is given the available information. On the face of it the size of the barrier, the length of the lake, the narrow dam crest and the possibly weak materials would suggest that it is potentially quite risky. But that will depend on the location of people and assets downstream and the nature of the materials that form the dam. The first step must be to put a monitoring system on the dam – in principle this is not difficult, although getting the information out of the valley will be a real challenge – in high mountain areas communications can be quite problematic.
21 January 2015
Zanskar Valley landslide
Colin Stark at Columbia University has used Landsat 8 images from 1st December 2014 and 18th January 2015 to identify the location of the landslide in the Zanskar Valley, and has kindly sent them to me. This is the January 2015 image, showing the landslide scar and deposit:
The landslide scar deposit can be seen blocking the river in the centre of the image, the source of the landslide is on the southern valley wall. Colin has draped this image on a digital elevation model to generate a perspective view:
The landslide source and deposit are indicated on the above image. From this Colin has a location: 33.29N, 77.286E, which is on the Tsarap River. The calculated surface area of the source zone is about 140,000 square metres, and of the deposit is about 90,000 square metres. If we assume an average deposit thickness of 30 m (and note that is a big assumption), we get a ballpark figure of about 2.7 million cubic metres. This is quite a large landslide. So this is the slope that failed, as shown in Google Earth imagery from 290th June 2014: