19 November 2020
Guinobatan: a major lahar in the Philippines triggered by Super Typhoon Goni
Super typhoon Goni (known locally as Rolly) brought extensive heavy rainfall to the Philippines over the weekend of 31 October and 1 November 2020. Widespread damage occurred, including landslides in many places. Very significant damage occurred in the town of Guinobatan in Albay, caused by a lahar from the flanks of Mount Mayon, an active stratovolcano.
The best image that I can find of the lahar itself appears in a video interview of the governor of Albay province by ABS-CBN News:-
Reports indicate that the lahar struck around about 150 houses in and around Guinobatan, with seven fatalities. There is now considerable controversy as to whether quarrying on the slopes of Mount Mayon may have been responsible for the lahar.
Planet Labs have captured a good image of the aftermath of the lahar at Guinobatan:
The image shows that the lahar passed to the east of the village of Maninila and then through the isolated houses at Tamdarora before striking the northeast margin of Guinobatan (in the southwest of the image above). it is clear that many properties were destroyed in the lahar.
The imagery does not provide strong evidence to support or refute the premise that quarrying may have been an issue. However, Mount Mayon has a long history of deadly lahars. Wikipedia has a good description of these events, which includes a major event on 30 November 2006, in which at least 1.266 people were killed. Whilst most of the damage occurred in Padang, on the outskitrts of Legazpi City, large areas of Guinobatan were also destroyed. Another major lahar occurred in October 1766, again triggered by a typhoon, and again in 1825, when 1,500 people died.
Reference and acknowledgement
Planet Team (2020). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
18 November 2020
Argaga: the major rockfall on La Gomera on 14 November 2020
Videos were widely shared on Twitter over the weekend showing remarkable footage of a major rock slope collapse at Argaga on the Spanish island of La Gomera on 14 November 2020. This is a large, near vertical collapse event, with the debris falling onto a road and into the sea. Fortunately no-one was killed.
There are actually two videos of this event, both capturing slightly different elements. Together they provide a very useful documentation of the failure. The first starts about a minute before the collapse. When the video starts there is some dust close to the rock slope. Over time, isolated (though not trivial) falls occur across the slope:-
The person collecting this video is clearly concerned about the potential for a displacement wave (sensibly), and so evacuates when the collapse occurs.
The second catches less of the precursory activity, although the main collapse is well constrained:
But this one captures the aftermath of the landslide, including the displacement wave that is generated by the collapse, as well as the very impressive cloud of dust.
In this case failure has been controlled by near vertical joints in the volcanic rocks. Collapse would have been controlled by the development of new fractures that linked the existing joints together to create the release surface. As this proceeded the rock mass crept and deformed at an accelerating rate. It is this deformation that pops off the precursory rockfalls. These precursors are of course invaluable as they often warn people in the vicinity that failure is developing, and in this case it also meant that people started to video the slope.
El Pais reports that rockfalls are common on this stretch of cliff. The owners of this vehicle might consider themselves somewhat fortunate:
Quickslide 1: Hurricane Iota has made landfall in Central America
Hurricane Iota crossed the coast in Central America yesterday, bringing exceptional rainfall and very high winds. We await to see the level of damage that it has caused, but landslides are inevitable.
Quickslide 2: Investigations of the peat landslides in the Republic of Ireland
17 November 2020
Mount Eagle: another peat slide in Ireland
Amongst the focus on the major peat slide in Meenbog in the Republic of Ireland over the last few days, it slipped my attention that another event occurred, this time at Mount Eagle in County Kerry. The Farmers Journal has an article that provides the detail. The landslide has occurred on a Sitka spruce plantation in a peat bog area of County Kerry on land owned by Coillte, the state owned forestry company.
The best imagery that I have seen has been posted to Twitter by Save Kerry (@SaveKerry):
15/11/2020 Bog-slide in Mt Eagle County Kerry. Planting sitka spruce on peat is never a good idea 👇👇@EPAIreland @InlandFisherIE @jerosullivanRK @radiokerry @FSC_Watch @npwsBioData @countykerry @agriculture_ie @pippa_hackett #specialprotectionarea pic.twitter.com/cvzF1BnIWG
— Save Kerry (@SaveKerry) November 15, 2020
On this occasion the peat slide appears to have started in the middle of the Sitka spruce plantation. It has displaced a large amount of bog, and as observed previously movement is shallow, consisting of rafts some of which include upright trees. The image below is a screenshot showing these rafts:
As the above image shows, the deformed land is highly chaotic, without the ordered structure seen in the Meenbog slide. I suspect that this may indicate that the this is not an area that has been affected by retrogression of the landslide. I would like to see an image that captures the whole of the source area of the Mount Eagle peat slide.
Downslope, the Mount Eagle peat slide has transitioned into a channelised flow with a long runout, in common with the many other failures of this type that I have described on this site. There are concerns for drinking water quality and for the ecological impacts downstream.
The trigger for these landslides is undoubtedly the recent heavy rainfall. The spate of events that we have seen this year so far reflects the periods of extremely wet weather that have occurred over the British Isles in 2020. But the precise mechanism of the initiation of the failures is not clear. I think that there is a particular question to be asked about the role that the spruce plantations play in initiating the slides (this may be related to the way that the trees create rafts of material and/or the ways that they change the drainage of the bogs), as well as the impact of wind farm construction in triggering instability.
As I have noted previously, global heating will mean that these prolonged spells of rainfall, and shorter duration high intensity rainfall events, will increase, so we might expect to see more failures of this type. Blanket bogs are an important habitat in the UK and in Ireland, and of course they are a major carbon store. Restoration is a very slow and difficult process. So, in my view, there is an urgent need to undertake research to understand these landslides better, and to ensure that the land is managed properly. A combination of detailed field monitoring and novel lab testing should yield some fascinating results.
16 November 2020
The Meenbog peat slide in the Republic of Ireland
There was a great deal of interest over the weekend in a peat slide that was caught on video at Meenbog in the Republic of Ireland. The landslide occurred on 13 November 2020. This is the video on Youtube:
The video is remarkable, showing rafts of bog with intact and upright trees sliding along a low angle slope. The peat blog landslide is large – as is so often the case with landslides in peat, the landslide has runout a very long distance, all the way to the major watercourse. There is an excellent drone video of the landslide track, also on Youtube:
The image below, a still from the video above, captures the upper part of the landslide:-
The upper portion of the Meenbog peat landslide is an extensional zone, almost certainly caused by retrogression (uphill expansion) of the slide after the initial failure. It is possible that failure started towards the downslope end of this bowl. It is unlikely to be a coincidence that a ditch crosses the slope at this point. Downslope from the landslide becomes a channelised flow.
The landslide has occurred in an area in which construction is underway for the Meenbog wind farm, a project that has been controversial. According to the Irish News, “The wind farm is owned by global retail giant Amazon, which aims to have the site operational by 2022“.
I have written about peat landslides previously, most notably in relation to the broadly similar Dawn of Hope peat landslide earlier this year. Peat is a very strange geotechnical material, with very low density, low compressive strength but some strength in tension due to the fibrous nature of the material (which is how it can generate the rafts seen on low angled slopes). It generally has an exceptionally high water content. Once disturbed it has a very low shear strength, allowing long runout landslides to develop. Especial care is needed to avoid instability in areas of blanket peat coverage.
The trees involved are Sitka spruce. According to the USDA Forest Service, these trees have roots systems that are “shallow and platelike with long lateral roots with few branchings. On deep well-drained soils the root system may reach depths of 6.5 feet (2 m), especially on alluvial soils. Root grafting often occurs between roots of the same tree and adjacent trees”. This explains how trees can be involved in the rafts of the peat slide.
The threat of peat slides from wind farms is well-established – indeed in 2003 a major peat slide was triggered by the construction of a wind farm in Derrybrien. As the Galway Daily notes:
“The Irish State was found to be in breach of environmental safeguards during the construction phase of the project. On top of the €5 million lump sum which the Department of Housing, Planning and Local Government has confirmed as paid over, Ireland still faces daily fines of €15,000 until the situation is rectified.”
I find it hard to understand how this can happen again.
Quickslide 1: Hurricane Iota is heading for Central America
Quickslide 2: The anticipated landslide in West Dorset occurred on 15 November
13 November 2020
Quang Nam Province: A video that shows why the aftermath of a landslide can be very dangerous
Further heavy rainfall is triggering landslides in Vietnam. Unfortunately Typhoon Vamco is likely to make landfall over the weekend, bringing further rainfall into areas that are already saturated. Substantial problems are likely.
In the meantime, the ongoing landslides, triggered in initially by the succession of recent typhoons, are having a major impact. This is clearly illustrated by a frightening video of a small but nonetheless deadly landslide from Quang Nam Province, which has been posted to Youtube:-
The video shows a group crossing the site of a recent landslide. Fortunately someone shouts a warning and the group seeks to escape the landslide, as this still shows:
Apparently two people were caught in the landslide in Quang Nam Province – I suspect that this will have been the individuals on the far side of the failure when it initiated. There is some variation in the reports of the consequences, but it seems likely that one person was killed and another suffered a broken leg.
The video illustrates two key points. The first is that failures often repeatedly occur at the same site, so moving into the danger zone is exceptionally perilous. This is a particular issue for those undertaking rescues. The second is that even small failures can be exceptionally violent, especially when static liquefaction is invoked.
Quickslide 1: Gutemala requests international assistance
As Guatemala tries to recover from the recent floods and landslides, it is formally requesting assistance from the international community. Given the current chaos at the top of many international governments, I wonder whether it will receive the help it needs.
Quickslide 2: a large incipient landslide in Dorset
ITV reports that a large crack has opened up on the coastal path near to West Bay in Dorset. This is undoubtedly the development of a new failure on a section of coast in which landslides dominate the geomorphology.
11 November 2020
The Queja landslide: Planet Labs imagery
The search for the victims of the Queja landslide appears now to be ceasing, with the site of the disaster being declared a “camposanto”, or a resting place for the dead, and the location will be declared as being uninhabitable. Whilst understandable, one should not underestimate the impact on the remaining population, whose future will now be very perilous. AP has a nice syndicated article about the vulnerability of Guatemalan communities to future events. This will not be the last catastrophic landslide in Guatemala.
Meanwhile, the skies have finally cleared, meaning that optical satellite imagery has become available for the site. Planet Labs captured a good image of the location using their 3 m resolution PlanetScope instrument yesterday:-
Note the scale bar , which gives an idea of the size of this landslide. On first inspection this image supports my sense that there are in effect two landslide sources here, both of which had an element of rotation before transitioning to a more flow style of movement. The larger to the two sources (source 1), to the north and west has generated a flow that has spread laterally, inundating a larger area. The other source (source 2) has spread rather less, but has had a longer runout. It is this part of the landslide that appears to have caused the greatest damage. As I noted in my first post on this failure, there is a stand of trees between the two sources.
I suspect some interesting work could be done on the timing of the two failures as there is some evidence of material from source 1 in the west side of the flow that is mostly from source 2.
Quickslide 1: a deadly landslide at a resort in Malaysia
Two people were killed in a hot springs resort landslide in Tambun, Malaysia. The failure was triggered by heavy rainfall.
Quickslide 2: The tenth anniversary of the Pike River Mine accident
Next Thursday (19th November) will be the tenth anniversary of the Pike River Mine disaster in New Zealand. Otago Daily Times has a good article about events on that tragic day. Attempts to re-enter the mine continue, but it is not possible to recover the victims.
Planet Team (2020). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
10 November 2020
The Queja landslide at San Cristobal Verapaz: drone footage of the aftermath
Over the last 24 hours further information has emerged about the Queja landslide at San Cristobal Verapaz in Guatemala. The search for the victims continues, but with little success, primarily because the debris is an estimated 15 metres deep. There seems to be a growing recognition that recovery will be impossible in the circumstances. Reuters has a report of the impact of the landslide on local Mayan families, for whom the Queja landslide is a catastrophe at every level.
Meanwhile, improved images and some high quality drone footage have been posted about the landslide. Drone footage of both the source area and the debris track and deposit were posted to Twitter:
This footage has caused me to slightly change my interpretation of the event from yesterday. As I noted, it does appear that there are two main portions to the landslide, but the main impact on the communities seems to have come from the right hand, rather than the left hand, portion. The image below, posted by Yahoo News, shows this more mobile portion of the landslide:-
These images have made it possible to track down the source area of the Queja landslide on Google Earth, which is in the vicinity of the coordinates I posted yesterday (15.391, -90.632). The image below highlights the approximate source area of the landslide and the main damage track. Note the distinctive house with a brown and grey roof on the edge of the landslide on the image above – this can also be seen in the Google Earth image below:-
Quickslide 1: Landslide induced pipeline ruptures in Ecuador prove to be extremely expensive
Reuters reports that Ecuador expects its crude oil production to miss its expected annual totals in part because of the impact of a pipeline rupture induced by a landslide:
Ecuador declared force majeure on oil exports in April after the state-run SOTE pipeline and privately held OCP pipeline burst due to a landslide in the Amazon region, forcing a temporary cut to crude production.
Quickslide 2: A rainfall induced landslide in Napier, New Zealand
Torrential rainfall in Napier, New Zealand yesterday triggered a landslide that destroyed a house. The occupants were lucky to survive.
9 November 2020
San Cristobal Verapaz: the deadly landslide in Guatemala triggered by Hurricane Eta
Over the weekend more information has emerged about the large, deadly landslide that struck San Cristobal Verapaz in Guatemala last week. The landslide, in an area that is remote and steep, destroyed a number of houses in the village of Queja. Unfortunately the area continues to be affected by heavy rainfall, and further landslides are occurring, disrupting the rescue and putting further lives at risk.
In terms of understanding this landslide, the two best images that I have found to date are below. The first, published on CP24, provides a good oversight of the source area of the landslide:-
On first inspection this appears to be a complex landslide, possibly occurring in two parts (note the stand of trees in the centre left of the landslide scar). In both elements there is possible a degree of rotational failure, but the upper portion appears to have transitioned into a slide or flow that has passed to the left of the trees (I’m using left in this case as shown in the image).
The second image is the one below, published by France 24:-
This is the lower portion of the landslide; I suspect from other images that this has derived from the upper failure shown in the first image. Here the landslide appears to be a rockslide or possibly even a flow. Note the amount of standing water on the surface, but also the large boulders. The damage to the structures is clear; any building in the direct path would have been destroyed.
The number of people killed at San Cristobal Verapaz is not yet clear – to date only three victims have been recovered. However, on Friday President Alejandro Giammattei indicated up to 150 people might have been buried in the landslide. One woman has reportedly lost 22 family members in the disaster:-
“All her family is gone, she’s the only survivor. Her dad, mom, siblings, aunts and uncles, grandparents, they’re all gone. Twenty-two family members and it’s just her alive”. It is hard to imagine this level of loss.
A search for the location of Queja in San Cristobal Verapaz yields a location of: 15.391, -90.632 on the Geonames database. At present the satellite imagery is too cloudy to verify this as the correct site.
6 November 2020
Hurricane Eta: dozens killed by landslides in Central America
In the last few days a major tropical storm, Hurricane Eta, has made landfall in Central America. In many ways this story has been lost in the maelstrom around the US election (and in the UK the imposition of a new lock down), but the impact has been severe. So far it appears that the greatest losses have occurred in Guatemala, where at least 50 people are reported to have been killed in landslides. Earlier news reports suggest that at least 37 of these fatalities have occurred in three landslides. The picture is quite confused, but the best report I have found so far is provided by CBS7:-
“On Thursday afternoon, Guatemalan President Alejandro Giammattei said a water-soaked mountainside in the central part of the country had slid down onto the town of San Cristobal Verapaz, burying homes and leaving at least 25 dead.
“Two other slides in Huehuetenango had killed at least 12 more, he said. The president initially said more than 50 people had died in slides, but the individual incidents he cited did not reach that total. Later, David de León, spokesman for the national disaster agency, said there were reports of 50 people missing in the Verapaz slide, but government rescue teams had not reached the site.”
News reports in Guatemala suggest that the losses at San Cristobal Verapaz may be much higher. This image apparently shows the landslide:
These news reports indicate that up to 75 houses might have been buried and that the loss of life might be about 100 people.
Unfortunately, the losses from landslides extend further. In Panama, five people, including two children, were killed in a landslide that buried two homes in Chiriqui province near to the border with Costa Rica. In Costa Rica itself two people were killed in a landslide. In Nicaragua, two gold miners were killed when they were buried by a collapsed slope. In Honduras, there are reports of five landslide fatalities, including a child who was killed when her house was buried in San Pedro Sula.
It is very possible that the losses will prove to be rather higher in due course. I’m reminded of the terrible impact of Hurricane Mitch in 1998. Fortunately, at the moment this event does not seem to be on a par – Mitch killed 11,000 people. This included a terrible lahar from Casita volcano in Nicaragua, which killed about 2,500 people. At the moment this event seems to be far less severe, but it is a reminder of the terrible vulnerability in Central America to these extreme rainfall events.
Quickslide 1: The rising impact of tropical cyclones in a warming world
Nature World News has a good article about the increasing impact of tropical cyclones as global heating continues to develop. Spoiler alert – seven out of 10 of the strongest storms have occurred since 2006.
Quickslide 2: The rapidly growing cost of making coal tips safe in South Wales
According to the BBC, Chris Bryant MP and Labour colleagues have written to the UK Chancellor Rishi Sunak about the costs of mitigating coal tips in south Wales. The letter reveals unpublished details from a review by the Welsh Government. It indicates that the cost of making the tips safe might exceed £500 million.
4 November 2020
The February 2018 Mangapoike landslide, New Zealand: an intriguing failure mechanism?
On a morning in which the media is full of a landslide that fail to materialise, I can find solace in the real thing. On 24 February 2018 an 8 million cubic metre landslide dammed the Mangapoike River in a remote upland area of eastern North Island in New Zealand. I blogged about this landslide at the time – it is a rather beautiful example. The height of landslide dam was managed through blasting, although the landslide remains intact, leaving interesting questions about the cause of the failure. This is a Google Earth image of the landslide:-
An interesting paper has been published in the journal Landslides about the Mangapoike landslide (McGovern et al. 2020). This is provides a detailed description and analysis of the feature, which was a wedge failure in weak rock sandstones and mudstones, with the slide surface being defined by a smoothed, northwest-dipping bedding plane. The elegant diagram below, which is an exemplar of how to use graphics to present landslide information, shows the source area of the slide:
The landslide mass itself partially disaggregated into blocks, which of course meant that the river was more effectively dammed.
The really interesting aspect of this paper though is its consideration of the causes and triggers of the Mangapoike landslide. The underlying causes are clear (and are well-explained), but what triggered the failure? McGovern et al. (2020) demonstrate clearly that there was no unusual rainfall event prior to the failure. Indeed, they are remarkably definitive about this:
“…it seems implausible that rainfall and elevated groundwater were important triggering factors”.
Instead McGovern et al. (2020) propose that fluid overpressure might have played a role. As they put it:
“The Makaretu Sandstone is a known hydrocarbon reservoir formation within the Tolaga Group, with overlying mudstone units acting [as] seals”.
There is geodetic evidence that prior to the landslide the region underwent approximately 15 mm of vertical movement, which may indicate that there was a fluid overpressurisation event occurring. The authors postulate that gas migrated vertically though the sandstones but was trapped by the mudstones at the base of the landslide, reducing the normal effective stress and triggering the failure.
This mechanism is a hypothesis, and indeed the authors state that:
“Whether fluid overpressure was a contributing factor to the Mangapoike landslide is equivocal.”
There have been previous suggestions that gas overpressurisation might be a factor in certain landslide failures. The interesting thing about this one is that the data is so good that it invites detailed modelling to understand the potential mechanism, and its efficacy, better.
McGovern, S., Brook, M.S. & Cave, M. 2020. Geomorphology and triggering mechanism of a river-damming block slide: February 2018 Mangapoike landslide, New Zealand. Landslides (2020). https://doi.org/10.1007/s10346-020-01572-7