31 December 2020
An update on the Gjerdrum landslide in Norway
Overnight, the emergency operations at the site of the Gjerdrum landslide in Norway have continued. The number of people unaccounted for has reduced to ten now, but I note a change in the tone of the reporting within Norway, with an increased emphasis on this event being a tragedy. The likelihood that the missing people are away from the site and out of contact is reducing, but hopefully it is not yet zero. Unfortunately, the Norwegian police have said that they are reasonably sure that there are people in the landslide area. The weather (in particular the cold, but also the snowy conditions) and the instability of the site are hindering the rescue effort.
A total of eight additional homes have now collapsed into the landslide, illustrating the continued instability of the area. Some of the buildings were multiple occupation units; in total it is reported that 31 homes have been lost. About 1,000 people have now been evacuated.
During the night, a dog and handler were lowered onto the landslide site from a Sea King helicopter. This was an extremely brave act, intended to search the partially collapsed houses that might have a space in which someone had survived. Unfortunately the search proved to be fruitless.
The one bright spot is that a dog, a dalmatian named Zajka, was rescued yesterday. This dog belongs to a family that successfully evacuated the site. The dog is injured but is expected to survive, and will be reunited with its family.
Support at the site is being provided by NVE, who have a statement online about the landslide, and the Norwegian Geotechnical Institute (NGI). The combined knowledge of these two organisations is tremendous. We can be confident that the operations are in the best possible hands.
I will try to provide updates today as they become available. In the meantime, many thanks to all those who have commented and helped. As usual, these events are best understood as a collective effort.
Update – 31 December 17:00 UT
The headline is that 10 people remain missing; it seems increasingly likely that they were present when the landslide occurred. A search team, including a dog, was airlifted onto the landslide this afternoon, but were unable to recover any survivors or locate any victims. The police are using mobile phone tracking, as well as drones, to try to locate those who are missing.
Meanwhile, new ground cracks have developed away from the existing landslide. As a result 46 more people have been evacuated, a road has been closed and the slope is being monitored.
The identities of those missing have not been formally released, but Norwegian media have speculated that it includes a three person family including a small child; a teenage girl and one of her parents; and a woman in her 50s and her son in his 20s.
30 December 2020
Gjerdrum: a quick clay landslide disaster in Norway this morning
Overnight, a large quick clay landslide has struck the village of Ask in the municipality of Gjerdurm in Norway. News reports from this landslide are a little confused as I write this – unsurprisingly – but it is clear that the slide has swept away a number of houses. Reports suggest that up to nine people might have been injured. About 700 people have been evacuated so far. It is reported that 14 buildings have been lost to date. Update: as the afternoon ends in Norway reports indicate that 12 people are not accounted for, some of whom are children. Unfortunately, the rescue services have not been able to enter the area affected by the landslide. This is the correct decision – this slope is likely to still be highly unstable. A further three houses have been lost during the day.
The Norwegian Police have described this event as a disaster. The site of the landslide appears to be the area shown in the Google Earth image below – the location is 60.065, 11.036:-
Reports indicate that the alert was sounded at about 4 am local time. The landslide remains unstable, and some retrogression of the head scarp is still occurring. It is not clear how far downslope the landslide extends, but the images suggest that this is a very substantial failure.
Gjerdrum has suffered quick clay (sensitive clay) landslides before. For example, Google Books has a case study of a significant event in 2012.
This is the NVE quick clay hazard map for the area:
The area that has failed is described as being Low hazard
but a loosening area that is within the source zone for landslides (many thanks to those who have helped me to understand this). However, it is immediately adjacent to an area described as medium hazard.
VG TV has a video of a house collapsing into the landslide bowl. It demonstrates that this is a very large failure (unfortunately I cannot embed this video):-https://www.vgtv.no/video/210593/vgs-reporter-om-hus-i-jordskredsomraade-staar-delvis-i-loese-lufta
The Norwegian Police have now released a map showing the area affected by the Gjerdrum landslide (in red). The large area with green hatching is the zone that has been evacuated today.
29 December 2020
Taku River: another very large landslide in British Columbia, Canada
Loyal reader Hig has alerted me to two postings on Facebook by Daryl Keith Tait, who has identified another very large and very dramatic landslide in British Columbia, Canada. His two Facebook posts include a set of images of the landslide source and debris and several videos of the aftermath. This is one of the images:-
This appears to be a large rock slope failure, with a substantial mass perhaps detaching from high on the slope. If so, this mass has then triggered the failure of face of the slope below. The mass was in free-fall for a considerable distance – on impact with the valley floor it has instantaneously fragmented to create an avalanche with a long run out and considerable dispersion. There is also a large volume of scree at the foot of the slope, suggesting multiple smaller failures after the main collapse.
The videos also seem to show that the landslide deposit has lobate structures at the margins, which seem to form during late stage sliding across frozen surfaces, and hummocky mounds within the landslide mass.
Hig pointed out to me that the Facebook postings give some indication of the timing of the event. On 24 December 2020 at 19:50 UTC a M=2.9 earthquake was recorded very close to this location. It is very likely that this is the record of the landslide itself – the magnitude is about right for a rock slope failure of this scale, and of course a free-falling mass is going to have generated a really substantial release of energy.
Secondly, downstream (and across the border into Alaska) on the Taku River there is a USGS gauging station. This shows a really interesting spike in water temperature and in turbidity late on 24 December and early on 25 December. This is the graph of water temperature for example:-
It is reasonable to hypothesise that these anomalies may be the signal from the impact of the landslide on the river.
At this time of year it is very difficult to obtain satellite imagery of this area, and of course the deposit may become covered by snow. This is another site that we will probably need to wait until the Spring before we can obtain a better understanding.
Thanks to loyal reader Hig for posting this one out to me, and of course to Darryl Keith Tate for both finding it and for posting it to Facebook.
26 December 2020
The Carmen Copper Mine landslide in the Philippines – the slope evolution before failure
It is now clear that the ongoing post-landslide operation at the site of the Carmen Copper Mine landslide in the Philippines has transitioned from a rescue to a recovery operation. The loss of the workers feels both tragic and unnecessary; questions need to be asked as to how such a tragedy can occur in a large, apparently well-managed mine.
Meanwhile, I have been looking at the archive of Planet Labs images of the site. This is not an easy place in which to obtain satellite imagery due to the amount of moisture in the atmosphere, but I have found two good PlanetScope images that are quite revealing.
This image of the site of the Carmen Copper Mine landslide was collected on 14 October 2020:-
Whilst this image of the site of the Carmen Copper Mine landslide was collected on 17 December 2020, a few days before the failure:
The slope that failed last week is at the north end of the mine. I have created a slider that will allow a comparison of the two images, although I am not able to embed this in WordPress.
The image from October shows a slope that retained the mining benches in at least a part, and at this resolution (about 3 m per pixel) there are few obvious signs of instability. However, by December the slope had evolved dramatically. The benches had disappeared, there are signs of a scarp forming on the west side of the slope, there is a large amount of scree in the botton part of the slope and the upper portion appears to be deformed.
Thus, on the face of it I would suggest that the slope had deformed considerably, which makes the tragedy even harder to understand.
I would very much like to see some InSAR data for the site of the Carmen Copper Mine landslide over the last six months.
Reference and acknowledgement
Planet Team (2020). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
24 December 2020
The deadly landslide at the Carmen Copper Mine in the Philippines
On 21 December 2020, in the aftermath of Tropical Storm Krovanh (also known as Tropical Storm Vicky), a very large landslide occurred at the Carmen Copper Mine, also known as the Toledo copper mine, at Don Andres Soriano (Lutopan), Toledo City, Cebu in the Philippines. This landslide is known to have killed four miners, whilst six more remain missing, with no prospect of survivors.
This landslide has raised barely a ripple of interest beyond the Philippines, which is very surprising because there are two remarkable videos of the failure and its aftermath. The first can be seen below. The main phase of movement starts at about 2 minutes 40 seconds if you are impatient, although the first part of the video shows both the creep of the landslide towards failure and the precursory failures from the mass as the mass deforms:
Watch to the end of the video – there is a surprise in store at about 3 minutes 20 seconds. At the very end you can see the aftermath of this secondary event.
As well as the main failure, this video shows the terrifying dynamic displacement wave that was created when the highly mobile front of the landslide impacted the lake. It is this displacement wave that caused the fatalities. The videos implied that the motion of the slope had been anticipated. Perhaps the size and magnitude of the wave was a surprise?
Reports today suggest that new cracks have been observed around the edge of the mine, causing the authorities to evacuate 400 families for the area. This has happened on the day before Christmas.
The best image that I have found of the aftermath of this landslide is this one, tweeted by News5 in the Philippines:
It seems to me that the main failure occurred in the upper half of the slope, with loading and entrainment causing the lower part to become involved. This means that most the material had a large fall height, which will explain the energy when it entered the lake. The slopes around the failure have been left in a very steep condition.
This is yet another major, fatal landslide in a large mine. This has occurred at the same time as a major spike in iron ore prices, driven by a major fatal landslide at the Vale Córrego do Feijão mine in Brazil last week, which killed an employee.
Once again I note that the occurrence of landslides associated with large-scale mining is unacceptably high.
18 December 2020
Arequipa, Peru: landslides from a M=5.6 earthquake on 16 December 2020
The city of Arequipa in Peru was struck by a M=5.6 earthquake at 17:48 UT on 16 December 2020. This was a comparatively deep earthquake (87 km according to the USGS) in an area with a history of significant seismicity. It is not expected that an earthquake of this magnitude at such a depth would cause large amounts of structural damage (or indeed extensive landslides), and indeed there are no reports that I can find of loss of life.
This earthquake is however interesting because it appears to have generated significant landslide activity in the mountains around the city. These were caught on a video that has been posted to Youtube, which is quite dramatic:-
The video of course is showing the clouds of dust generated by landslides (most probably rockfalls) in this extremely dry environment. We have seen this effect previously where earthquakes have occurred in very arid environments, most notably in the 2010 M=7.2 earthquake in the Sierra Cucapah, Mexico, which was also caught on video. In Mexico we studied the landslides via remote sensing and wrote up the results (Barlow et al. 2014). Whilst the videos of the event were very dramatic, and rockfalls were widespread, the volume of material moved by the failures was quite small.
It is difficult to know whether this will be the case here:-
There are some images online of rockfalls triggered by the earthquake, especially on roads, but I have not been able to track down any images of larger failures. These may of course be located in remote areas, if they have occurred.
Barlow, J., Barisin, I., Rosser, N., Petley, D., Densmore, A. and Wright, T. 2015. Seismically-induced mass movements and volumetric fluxes resulting from the 2010 Mw = 7.2 earthquake in the Sierra Cucapah, Mexico, Geomorphology, 230, 138-145. http://dx.doi.org/10.1016/j.geomorph.2014.11.012.
16 December 2020
Bute Inlet: a very long runout proglacial landslide in Canada
News has emerged over the last few days of a recent very large landslide close to Bute Inlet in British Columbia in Canada. This is a really big one – I think the runout is in the order of 13 km, based on a back of the envelope calculation from Google Earth. This landslide has been reported as a variety of phenomenon including a glacial lake outburst flood (GLOF) and a tsunami, and as loyal reader Hig points out below there is an element of these in the chain of events. The location is approximately 50.975, -124.609.
A good starting point is a set of videos posted to Facebook by 49 North Helicopters, who I think discovered the landslide. These show the track of the lower part of the landslide, which was confined within the valley below the glacier. You should be able to access them via the embedded tweets below:-
— Brent Ward (@GeoBrentatlarge) December 14, 2020
Brent Ward from the Department of Earth Sciences at Simon Fraser University has been tweeting extensively about this landslide, and he has also appeared on the TV News to discuss the slide. His interpretation of the events is I think spot on, and my description here is based on this.
The full extent of the landslide can be seen on the Google Earth image, obviously captured before the failure, below. I have annotated the location of the original landslide, the position of Elliot Lake, the track of the landslide and the location of the outwash deposits:-
Planet Labs has imagery from 2 December 2020 that captures the aftermath of the landslide – clearly the event occurred before this date. This is the source area of the landslide and the remains of Elliot Lake:-
Note the scale bar in the bottom right corner of the image – this is a very large failure. The interpretation is that a large failure occurred on the rock slope to the west of the front of the glacier. The very large mass entered the lake, driving a huge displacement wave and entraining both the water and lake sediment to form a catastrophic debris flow (or a hypersaturated flow?) down the valley. The track of the flow is captured very nicely in the 49 North Helicopters videos. This is a still from one of them:-
At the mouth of the valley there is extensive deposition of sediment:-
The flow then travelled westwards down the Southgate River into Bute Inlet, where a large amount of floating timber was observed.
This is a classic compound hazard chain – landslide, displacement wave, debris flow, debris flood and possibly even a submarine density flow in Bute Inlet.
Detailed analysis of this event may need to wait until the Spring, but good data should be available from the regional seismic network, which should yield data on the timing, volume and velocity of the event. It is very interesting that this large failure occurred very close to the snout of a retreating glacier – there are parallels to the Barry Arm landslide in Alaska.
15 December 2020
A high resolution Planet Labs SkySat image of the landslide at Gamsberg mine in South Africa. Image copyright Planet Labs, used with permission.
The large landslide at Gamsberg mine: high resolution satellite images
A month ago, on 17 November 2020, a large slope failure occurred at the South Pit of Gamsberg mine in South Africa. This failure, which appears not to have been anticipated, killed two people. The body of one of the victims had not been recovered as of a week ago. News reports on 7 December indicated that mining operations remained suspended.
I thoroughly recommend that you read a comment to my original post about this failure – it provides some interesting insight.
My friends at Planet Labs have captured a high resolution satellite image of the site, dated 3rd December, using their wonderful SkySat instrument:-
It is worth calibrating this image against the view from the ground that I included in my original post – used together one can get a good impression of what transpired at Gamsberg mine:-
It is clear that the landslide occurred on a slope across which the access ramp crossed. The failure appears to have occurred above the access ramp, although the debris has reached the pit floor, and has covered most of it. The image below is an enlargement of the Planet Labs image:-
Note that the stratigraphy is roughly preserved in the landslide deposit, with the deeply weathered (reddish) rocks being closest to the landslide scar. This might suggest that it was a rockslide with failure starting low on the slope? It appears that there has been a smaller secondary failure to the northeast of the main collapse. The deposits are quite distinct. I am not sure whether part of the mine is obscured by thin cloud or by dust. If the latter then some collapses may still be occurring.
There is not much evidence in the images of heavy equipment working on the landslide on 3rd December, but if collapses were continuing then this might be understandable. Of course a satellite image is just a snapshot, so there might be activity that was undetected.
The size of the bite out of the boundary of the mine is quite striking. The long term questions at Gamsberg mine will be how such an apparently unanticipated failure could occur so early in the mining operations, and how the mine can be operated safely in the future. In the meantime the economic costs of this failure will be accumulating rapidly.
Reference and acknowledgement
Planet Team (2020). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
Thanks to Rob Simmon for collecting and providing the image.
14 December 2020
The Eureka Valley Landslide in Death Valley National Park
Way back in the mists of time, when this blog was fresh-faced, I wrote a brief piece about a very intriguing landslide in Death Valley National Park. This feature, known as the Eureka Valley Landslide, is the beautifully-preserved preserved remains of an ancient long runout failure. It is very clear on Google Earth:-
I have placed the marker at the toe of the main lobe of the landslide, which is located at 37.076, -117.652. Note however that there is another lobe downslope of this point with a slightly different orientation – this is interpreted as being an outwash lobe. The runout distance for the main lobe is about 3.5 km.
This landslide has just been described in a new paper published in the journal Lithosphere (Shaller et al. 2020). Many thanks to the authors for highlighting it to me. The authors have found that the landslide started with 5 million cubic metre detachment up in the steep terrain. The authors tentatively suggest that the landslide occurred about 100,000 years ago.
The most interesting aspect of this study is the detailed sedimentology undertaken by the authors, which has allowed them to construct an interpretation of the likely movement mechanisms. In the early part of its movement, through the canyon system in the mountains, they suggest a “cataclasis mechanism” involving rock fragmentation and energy recycling. However, once out of the rugged terrain they suggest that the mechanism transitioned to a sliding mechanism enabled by liquefaction of the alluvium over which the landslide was moving. The image below provides a full overview of the landslide:-
In broad terms this feels a bit like the change in mechanism that is seem in the Mount Dixon rock avalanche video, where the landslide transitioned from an avalanche phase into sliding, in that case over ice.
There is a huge amount of rich detail in the paper – it is worth a read.
Shaller, P.J., Doroudian, M. and Hart, M.W. 2020. The Eureka Valley Landslide: Evidence of a Dual Failure Mechanism for a Long-Runout Landslide. Lithosphere 2020 (1), 1–26. doi: https://doi.org/10.2113/2020/8860819
10 December 2020
Mindu in Tibet: detecting precursors of an imminent landslide
With the possible exception of some landslides triggered by earthquakes, large slope failures generally develop strain (movement) prior to failure. The failure process involves the progressive deformation of the slope – a shear surface may form in the base of the landslide, tension cracks form, lateral scarps develop, etc. A classic case is the Barry Arm landslide in Alaska, where movement on the slope has led to the development of very large tension cracks. There is strong evidence that over time the rate of movement increases as failure approaches. This increase in deformation is the basis of various methods of prediction of the time of failure (with success in some circumstances).
One challenge of course is to use this knowledge to identify and monitor slopes that might be undergoing failure. The holy grail is to have a remote monitoring system that collects data at a regional or national scale and then identifies slopes that are actively deforming. InSAR provides one potential basis for this, and national scale deformation maps are now available, but identifying correctly slopes that might be dangerous requires more work and a better understanding.
Another approach, applicable for individual landslides at the moment, is to use imaged correlation approaches from optical satellite imagery. In this approach, pairs of images are compared. Perhaps surprisingly, image processing can allows deformations on the scale of 3 to 10% of a pixel to be detected. The Sentinel-2 satellites have an image resolution of about 10 metres, so deformation of less than a metre can be measured.
In an open access paper in the journal Natural Hazards and Earth System Sciences, Yang et al. (2020) have used image processing of Sentinel-2 imagery to examine the movement through time of a developing failure near to the town of Mindu in Tibet. This is a large area of slope deformation – this is the Google earth image collected in 2011 for example:-
The location of this landslide is 30.582, 98.925. Whilst there is little in the way of the landslide, clearly a major rock slope failure at this site near to Mindu would potentially block the river, creating a substantial hazard downstream.
The image processing by Yang et al. (2020) has demonstrated that this landslide is actively deforming. Perhaps most interestingly, between November 2015 and November 2018 the slope showed less than 2 metres of movement. However, between 2018 and 209 the slope moved over 6 metres. The research team were able to then look at a larger number of images in this more rapid movement period, finding that in the rainy season (summer and autumn) the movement rate accelerated.
This study demonstrates that processing of optical satellite imagery can allow high quality monitoring of dangerous slopes to be undertaken. It is another step along the way towards the goal of high quality early warning systems for slopes in high mountain areas.