10 September 2021
Landslides from the 7 September 2021 M=7.1 Guerrero earthquake in Mexico
In the evening of 7 September 2021 (local time) a M=7.1 earthquake struck close to the city of Acapulco in Mexico. This event was shallow, leading to potentially high levels of damage in the worst affected areas. Wikipedia is calling this the 2021 Guerrero earthquake, but time will tell as to whether this is adopted as the formal name for the event.
At least three people were killed and 20 more were injured, a comparatively light toll for an event of this magnitude. Interestingly, one of the major impacts appears to have been landslides, although fortunately there are no known fatalities from these events.
Some images are emerging of landslides triggered by this event. The best set I have found so far is in the Daily Sabah, which has a gallery of images of the impacts of the earthquake, several of which show landslides:-
Given the topography and the scale of the earthquake, it seems likely that larger landslides will have occurred in the more remote mountainous areas of the epicentral zone. As yet I have not seen any images of such events, but it is likely that reconnaissance and fact-finding visits will be organised in the coming days. I would welcome any information that loyal readers can provide.
9 September 2021
Paimio: a very unusual landslide in Finland
Last week a very unusual and interesting landslide occurred close to the town of Paimio in Finland. The image below, published on ts.fi, provides a wonderful overview of the landslide:-
Once again I don’t think I’ve seen one quite like this before. The source area appears to be an aggregate storage area from a nearby quarry. The aggregate pile has has clearly failed – there is a scar and some tension cracks visible. The material in the fields appears to be clay, common in formerly glaciated areas. The failure of the aggregate pile has propagated through the clay substrate over a distance of perhaps 100 metres, forming a set of compressive ridges. This is unusual.
It is interesting to ponder what might have happened here. My working hypothesis is that this might be a bearing capacity failure in the clay below the aggregate pile – essentially the clay did not have the shear strength to withstand the stresses imposed by the heap. Its failure caused the pile to collapse.
The nearest comparative event I can remember of the Hatfield Colliery failure close to my now home in Sheffield, in England. In this case colliery spoil was piled on gravels that subsequently failed. However, in the case of the landslide at Paimio the failure has propagated a relatively longer distance from the source, presumable reflecting the properties of the clay.
Fortunately no-one was killed or injured in the landslide at Paimio, and only one building was damaged.
The location of the landslide is, I believe, 60.434, 22.610. There is a Google Earth image of the site from 2018:-
Interestingly this image does not show the large pile of aggregate that was involved in the failure, so this might be a comparatively recent addition.
Many thanks to loyal reader Tomi for highlighting this one to me.
8 September 2021
The Pretty Rocks landslide in Denali National Park in Alaska
Last year I posted about landslides in Denali National Park in Alaska, including a very interesting case study, the Pretty Rocks Landslide. This failure, located at 63.537, -149.815 if you wish to take a look, is heavily disrupting the single road through the national park. Time magazine has an online article about this landslide that is worth a look. Sitting behind that is considerable more detail on the Denali National Park website, including a set of images of the landslide:
The Denali National Park article notes that technically this is a rock glacier, and that it is moving quite rapidly. The rate of movement of the landslide has increased in recent years in response to the warming climate, with highttps://www.nps.gov/dena/learn/nature/pretty-rocks.htmh rates being noted in particular in periods of summer rainfall. The website describes the change in pattern of movement of the landslide:-
The landslide at Pretty Rocks has been active since at least the 1960s, but has evolved from a manageable maintenance concern into a much more extensive maintenance challenge. Before 2014, the landslide only caused small cracks in the road surface and required maintenance every 2-3 years.
Between 2016 and 2017, the full width of a 100-yard (90-m) section of road slumped up to 0.2” (0.5 cm) per day, steepening the road gradient and limiting sight lines for drivers approaching the sharp turn
. but not warranting long-term closures.
However, by 2018 the slumping increased to almost half an inch per day, and then to three and a half inches per day by August 2020. Early August rains in 2021 appear to have triggered the rate to increase significantly, with much of the landslide currently moving downhill at over ten inches per day.
NPS road crews maintained a safe drivable surface for most of the summer of 2021 by importing as much as 100 dump-truck loads of gravel per week to fill the slump. On August 24, park managers recognized that this solution was no longer tenable or safe and enacted a road closure to the east of Pretty Rocks. As the area becomes more unstable, traditional road maintenance methods have become inadequate. The park’s ability to adapt to future conditions will require transitioning to more expensive and novel construction methods to maintain road-based access.
Denali National Park has produced a wonderful time lapse video of the landslide movement that is incredibly instructive:
Note the continuous movement of the slide, wbich extends right to the toe of the slope. The dumping of gravel to maintain the road is clearly evident, as is the pitfall of this approach. The gravel is steadily loading the head of the landslide, increasing the driving force. Resistance is not increasing at the same rate, so the tipping is likely to be making the problems with the slope slightly worse.
The National Park are right to cease this approach. Unfortunately, as the National Park notes, alternative ways to maintain the road are likely to be costly and difficult.
7 September 2021
Jeori: a large highway landslide in Shimla district, Himachal Pradesh
Yesterday (6 September 2021) a large landslide blocked National Highway 5 at Jeori in Shimla district, Himachal Pradesh in northern India. The landslide, which occurred at about 9 am, did not cause any casualties. This landslide is particularly interesting as it was caught on video from at least four different angles. The first was collected by someone located on the road itself, fortunately a safe distance away:
— Vipin Kumar (@chauhanVKgeo) September 6, 2021
Meanwhile, another video was captured from a greater distance away, which provides less detail but gives a better appreciation of the process:-
Earlier today, a massive landslide occurred at Jeori in Himachal Pradesh's Shimla. The Shimla-Kinnaur national highway has since been closed. No loss of life reported yet. #Landslide #HimachalLandslide
— The Weather Channel India (@weatherindia) September 6, 2021
And a third video was captured from the other end of the road:-
— Rajinder S Nagarkoti (@nagarkoti) September 6, 2021
And a further video was captured further along the road:-
Today's Landslide video Near Jeori , Shimla . NH 5 blocked (Shimla-Kinnaur)
— Weatherman Shubham (@shubhamtorres09) September 6, 2021
I am not sure I’ve seen a landslide captured from four different perspectives before, so this is a cool event.
Photographs of the site in the aftermath of the landslide site, extracted from one of the videos, suggest that the road was on an artificial bench with an unsupported cut slope above:-
Note the small pile of debris on the road before the main collapse occurred. This indicates that there was some precursory deformation, which presumably provided warning of an impending collapse.
The landslide appears to have occurred in dry weather, but we are in the tail end of the monsoon, which may have played a role in progressively weakening the hillslope.
Works are under way to reopen the road. As usual, these operations carry considerable risk.
6 September 2021
Catoca mine in Angola – using satellite imagery to understand recent events
According to Wikipedia, Catoca mine in Angola is the fourth largest diamond mine in the world. Located at -9.399, 20.301, it consists of a big open cast pit and a very large tailings storage area.
Last month, CRREBaC released information about a set of very significant pollution events in the rivers of Angola and Congo, linked to mining in Angola. Reports have included the pollution of hundreds of kilometres of the river system, deaths of fish and hippos and, in some cases, suggestions of up to twelve fatalities. Detailed information to evidence these huge losses is lacking, but there is little doubt that serious pollution occurred.
Whilst this blog is about landslides, I have very often written (and indeed campaigned) about the scandalously poor management of tailings in the mining industry, and I have highlighted several major tailings dam failures. My interest in the events in Angola and Congo results from concerns that the events might have been one or more tailings dam collapses.
One of the candidate sites for the events in Angola is the facility at Catoca, and indeed the operators have reportedly admitted that a release did occur, but only of water and sand.
This is a satellite image of the Catoca facility collected by Planet Labs on 21 July 2021:-
The pit is in the bottom right hand corner, the tailings facility in the huge orange area on the left side of the image. The tailings dam is located just above the centre of the image, orientated NNW-SSE. There is no evidence of any problem in this image.
The first sign of change appears in the image of 24 July 2021. The change is subtle but significant:-
I have put a black circle around some clear pollution on the downstream side of the tailings dam. Note that the tailings dam was intact (and remains so). A day later the problems had become very much worse:-
By 25 July 2021 a large plume of pollution had appeared on the downstream side of the tailings dam (highlighted with a circle again). The had entered the watercourse, which was now showing very clear signs of pollution (highlighted with an arrow).
The situation today, a few weeks later, is interesting:-
The plume has developed considerably , although it has not enlarged significantly (black circle). The polluted watercourse is very clear. There is clear evidence of works around the plume (new roads for example) and on the upstream side of the dam (white circle).
From the images above the pollution might not at first sight look serious, but satellite images downstream might tell a different story. This image was collected on 25 July 2021 at about 8:56 UT, downstream of Catoca at -9.127, 20.346:-
The river is flowing from the south to the north. In the lower part of the image the river is bright orange with pollution. In the northern part of the image the river is unpolluted. The marker, at -9.127, 20.346 is the approximate front of the pollution moving downstream from Catoca. Note the profound change in the water as the pollution front moves through.
This is even more profoundly illustrated in the image below. The small tributary from Catoca flows in from the southwest. It meets the main channel, flowing from south to north. Note the huge change in water quality from this point onwards:-
This location is -9.309, 20.362.
In a statement last month, the company admitted that there had been a “rupture in the pipeline that works as a spillway.”
The satellite images are consistent with this as the source of the pollution, although the ecological damage looks to be more severe than some have suggested.
However, there is an anomaly. The CRREBaC report indicates that substantial pollution was first seen in the rivers of Angola from 15 July 2021:
On the basis of the Sentinel images published by Visio Terra (Equipe Sentinel Vision, EVT919, 2021) and our preliminary investigations from riparian communities, this pollution has been observed since 15 July 2021 from the source in the Angola part of the basin and would have taken 15 days to reach the city of Tshikapa, and 21 days for the city of Ilebo in the Democratic Republic of the Congo (DRC).
The pollution observed on 15 July 2021 does not seem to have come from Catoca according to the images, which suggests a second serious event occurred in a different mine in the area.
Planet Team (2021). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
3 September 2021
Landslides from the 14 August 2021 earthquake in Haiti
On 14 August 2021 a M=7.2 earthquake struck Haiti to the west of the capital Port-au-Prince. Whilst this event affected an area that is less densely populated than was the case for the 2010 event, it still killed over 2,000 people.
The earthquake affected an area that is mountainous at a time of year in which rainfall is common, so landslides were inevitable. In a post soon after the earthquake, NASA highlighted that substantial numbers of landslides had indeed happened.
The image below shows the landscape around Pic Macaya National Park, which was close to the fault that ruptured to trigger the earthquake in Haiti. The landscape is densely vegetated and mountainous:-
Some landslides are visible, particularly in the central south of the area and to the east. However, the image below shows the same area after the earthquake:
The image shows a very dramatic increase in the number of landslides, particularly in the catchments in the centre of the image. The satellite image below provides more detail of some of these landslides:-
There are hundreds, possibly thousands, of landslides in the images, with densities approaching those seen in areas of the Wenchuan earthquake in China, although over a smaller area. Most of the images appear to be shallow disrupted slides, as is usually the case in such events.
The area was affected by a large rainstorm two days after the earthquake, when Tropical Depression Grace crossed Haiti on 16 August 2021. It is likely that the landslides in the images is a consequence of the combination of the earthquake and mobilisation of slope materials and landslide debris in the storm.
It is not clear how many people were killed by landslides in the 2021 Haiti earthquake, but the numbers are likely to have been substantial.
Reference and acknowledgement
Planet Team (2021). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
1 September 2021
George County: a drone video of the site of a fatal landslide triggered by Hurricane Ida
The landfall of Hurricane Ida across the southern USA earlier this week caused high levels of rainfall, inducing flooding and landslides. In terms of mass movements, one of the most significant events occurred in the Benndale community in George County in Mississippi, where a substantial failure destroyed a section of MS26. The landslide had substantial consequences – it killed two people and injured a further 10, three of them critically.
The SunHerald has a detailed article about the event, and has published a drone video of the aftermath. The site is quite interesting, as this overhead view shows:
The view below, also from the drone video, provides an understanding of the form of the landslide:
“It is a slide, which means the ground under the roadway and embankment was super-saturated and we can tell right now that’s what caused the slide,” said Kelly Castleberry, district engineer for the Mississippi Department of Transportation. “The ground liquified and it spread several hundred feet to the south.”
The road appears to be on embankment, which has failed.
13 August 2021
Nalda: a valley-blocking landslide in Himachal Pradesh, India
This morning (Friday 13 August 2021) a major landslide occurred near to the village of Nalda, in Lahaul and Spiti district in Himachal Pradesh, northern India, blocking the Chenab River. The landslide event itself was captured in a video that has been posted to Youtube:
The landslide has left a large barrier across the Chenab River close to Nalda. The image below, from Tribune India, provides an overview of the situation:
This is not likely to remain breached for long given the topography of the land and the form of the barrier – indeed some reports suggest that some flow has been re-established. But even a few hours of impounded water has the potential to generate a substantial local flood, so there is a substantial hazard for villages downstream.
Note the very clear planar structures in the landslide scar. The failure is clearly defined by two existing joints or faults, meeting with a c.90 degree angle. Note also the landslide scar immediately to the left of the landslide scar. That must also have been a substantial failure event.
This landslide comes on the back of a significant landslide in India on Wednesday, when a slope above a road in India collapsed, burying several vehicles including a bus. Recovery operations continue at the site, but at the time of writing 15 bodies have been recovered. There are fears that there may be substantially higher levels of loss. The Indian Express reports that Chief Minister Jai Ram Thakur has indicated that up to 60 people were feared trapped under the debris at the time of the accident. Rescue operations are being hampered by ongoing instability on the slope.
5 August 2021
Understanding the recent monsoon landslides in Maharashtra, India
Last month, heavy rainfall in the Maharashtra area of India triggered a number of deadly landslides. The most serious of these occurred at Taliye village, where at least 53 people, and possibly as many as 84 people, were killed. However, this was not the only substantial landslide; there were at least another six fatal landslides, which took an additional 50 lives.
My friends at South Asia Network on Dams, Rivers and People (sandrp.in) have done a heroic job in collating information about these landslides, their locations and impacts. I have used their information to produce a quick Google Earth map of the landslide locations:
I have included two significant non-fatal landslides as well. It is interesting to compare the landslides to a map of the rainfall. The map below is a 3 day rainfall total for the period ending 24 July 2021 (the Taliye landslide occurred on 22 July 2021), from the NASA Global Precipitation Measurement mission:
The cluster of high rainfall totals exactly where the landslides were triggered is clear. The map below (from Researchgate) shows the topography of Maharashtra:
The map shows that there is higher elevation ridge running approximately parallel to the coastline. It is clear that the high rainfall totals reflect this ridge (presumably due to an orographic effect), and of course it also provides the topography for the landslides. The main landslide cluster is to the southeast of Mumbai (Bombay).
But why were there so many landslides? The answer may well lie, at least in part, in degradation of the environment. Removal of natural vegetation, cutting of slopes, quarrying and road building are taking a huge till across the upland areas of India (and indeed across the hilly parts of much of South Asia). As usual, in the aftermath of monsoon landslide disasters there is discussion about warning systems and suchlike (which in general have a very patchy success rate in rural settings in less developed countries). Although they can and do have a role, this is to miss the salient point, which is that we need to work on preventing increases in susceptibility as a priority, mainly through better management of slopes.
Climate change is going to exacerbate these problems at pace, as will continued damage to slopes. If we are to reduce these disasters, we need manage the slopes properly.
Many thanks to my friends at the South Asia Network on Dams, Rivers and People, and in particular Parineeta Deshpande-Dandekar, for help with the data in this post.
4 August 2021
A rockfall video from the Couloir du Gypaète in France
With thanks to those who spotted this on Youtube, a very surprising rockfall video was posted online last week by Rémi Bourdelle, who was flying above the Couloir du Gypaète in France when a rockfall developed. The video provides an utterly unique perspective on rockfall processes. Take a look:-
My first reaction to this is of course wow! The video shows boulders bouncing down the gulley and then through the trees below. As noted previously the force of these rocks is extraordinary.
But the most surprising element of this video is the trajectory of the rock seen at the start of the video. Fortunately, this boulder (which appears to have been fragmenting) left a trail of dust that marks its path through the air:
This is the sort of trajectory that is normally associated with fly rock from blasting rather than a rockfall. It is quite extraordinary; I have not seen this previously (have any readers?).
Presumably, the boulder struck an inclined surface, whilst travelling at a very high velocity, causing a ricochet that created this ballistic trajectory.
The other obviously possibility would be ejection during fracturing, but this seems less likely to me.
I’m not sure if this is a freak, or something that is quite common. Any views?