17 October 2017
The Bellavista heap leach gold mine landslide in 2007
On 21st October 2007 a landslide effectively closed the Bellavista heap leach gold mine in Costa Rica. This landslide is subject to a court case that is currently being heard in the Ontario Superior Court of Justice, with a claim if about $100 million being made against SRK Consulting (US), who undertook engineering and design work, by B2Gold, based in Vancouver.
Cyanide heap leaching involves creating piles of crushed ore, mined from an open pit site in this case, which are then washed with cyanide solution. The leachate is then collected at the foot of the pile, allowing the gold to be extracted. Clearly the risks of pollution are sufficiently high that great care is needed to protect the integrity of the heap. The Bellavista site is located in heavily weathered, humid tropical mountains that are also seismically active. Slope instability is a challenge in such environments, as multiple examples on this blog have shown.
The Tierra Group has a description of the work that they undertook at the site in 2007, which includes this image of the ore heap:
In their description of the project, the Tierra Group state that:
…our Team performed a geologic hazard assessment which ultimately identified existence of a mega-landslide (35,000,000 m3) underlying a waste rock dump and flanking a heap leach pad. A site wide Environmental Audit was prepared for the Secretaria Técnica Nacional Ambiental.
This landslide induced instability in the heap; unconfirmed reports suggest that movement was observed in 2006. The mine was reportedly closed in July 2007. In October 2007 a major landslide occurred in the southeast corner of the heap. The Ecologist has this image of the landslide, taken in January 2009:-
After this major landslide, the mine was closed and has subsequently been remediated. Of course I cannot comment on the liability issues that this case raises (and I cannot comment on the validity of any of the legal claims for obvious reasons), but once again problems of slope instability and mining are clear.
16 October 2017
Phu Cuong commune: a major landslide has killed 18 people in Vietnam
Over the last few days, heavy rainfall has affected the northern and central provinces of Vietnam. The most serious incident appears to be a landslide that struck Phu Cuong commune in Tan Lac district, located in the northern province of Hoa Binh. Latest reports suggest that 18 people were killed in this event. Eleven has a good report about the impact of this major landslide:
The villagers were asleep when a massive amount of soil and rocks, triggered by heavy rainfall, rolled down an adjacent hill at about 1am Thursday, the newspaper reported. At least eighteen people, most family members, were buried alive.
One of the survivors, Bùi Văn Dũng, 26, was still shaken when recalling the disaster, 35 hours after it happened.
“We were sleeping. It was pitch-dark,” he told Dân Việt. “I opened my eyes to a loud ‘bang!’ and found myself lying under the sky, my legs buried in rocks. Everything was shaking.”
Horrified, Dũng grabbed a flashlight to find his wife and daughter in the same situation some 10 metres away. With his screams for help muted by rumbling sounds, he used the wood beams that were lying around to dig up the dirt and rocks that were crushing his wife and daughter.
The family then ran out of the area, at the sight of massive rocks rolling down from the top of the Khanh Waterfall.
The best images of this landslide can be found in this article by the IBTimes:
Whilst this image shows the damage caused by the landslide:-
This image, from VietNamNet Bridge, appears to show the site of the landslide:-
Note the enormous overhang (now disappeared) and the very obvious rockfall deposit in the foreground. A major landslide at this site should come as no surprise.
In total the heavy rainfall has left at least 72 people dead, with further rainfall forecast in the days ahead as Typhoon Khanun passes by.
9 October 2017
Lahars: an analysis of their role in losses from volcanic eruptions
A lahar ( /ˈlɑːhɑːr/) is a type of mudflow or debris flow composed of a slurry of pyroclastic material, rocky debris, and water. The material flows down from a volcano, typically along a river valley.
In a paper just published online (and open access) in the Journal of Applied Volcanology (Brown et al. 2017), Sarah Brown and colleagues from the University of Bristol have looked at the role of lahars, and other phenomena such a pyroclastic density currents, play in causing loss of life in volcanic events. This work analyses a volcanic fatality database that extends from 1500 AD to 2017, and includes 635 separate volcanic events that have resulted in almost 280,000 fatalities. In terms of landslides, they record three types of event:
- Lahars: 72 events, 49,938 fatalities
- Secondary lahars: 41 events, 6,377 fatalities
- Avalanches: 9 events, 3,525 fatalities
In this case, secondary lahars are defined as events that occur after an eruptions (sometimes years later), whilst avalanches are defined as being “inclusive of debris avalanches, sector collapse and landslides: These result from the collapse of unstable edifices due to seismicity, eruption or intense rainfall. Avalanches discharging into lakes or oceans can generate tsunamis; resultant fatalities are classified under tsunami.”
Thus, landslides of one type or another have been responsible for 59,840 fatalities, about 21.5% of the total, and roughly the same number as pyroclastic density currents.
One other aspect of this work is really interesting, which is that the research has examined how far away from the volcano the victims were located when they died, classified by the phenomenon that killed them:-
Lahars have a notably different distance profile from the other major volcanic hazards, with significant numbers of deaths occurring at more than 20 km from the volcano. This is a good illustration of the role that landslides can play in extending the reach of a hazardous phenomenon over a long distance, in a manner similar to the way that tsunamis extend the reach of earthquakes.
Sarah K. Brown, Susanna F. Jenkins, R. Stephen J. Sparks, Henry Odbert and Melanie R. Auker 2017. Volcanic fatalities database: analysis of volcanic threat with distance and victim classification. Journal of Applied Volcanology, 6, 15. https://doi.org/10.1186/s13617-017-0067-4
6 October 2017
Murchison Glacier: more information about the slowly developing rockslope failure that is affecting the hut
Murchison Glacier: more information about the ongoing rockslope failure that is affecting the hut
Earlier this week I posted about the reasoning behind the closure of Murchison Hut, the climbing and skiing refuge located above the Murchison Glacier in Aoraki/Mount Cook National Park in the South Island of New Zealand. To recap, analysis of aerial imagery and satellite images by Dr Pascal Sirguey at the University of Otago and Dr Simon Cox from GNS Science has indicated that the location of the hut has shifted 9 m laterally and 9 m vertically since November 2008, indicating that there is a large-scale rockslope failure occurring beneath the hut. Clearly there is the potential for a major failure event onto Murchison Glacier at some point (without further monitoring this is essentially impossible to predict), such that the hut has had to be closed.
Pascal has very kindly provided some additional images to illustrate the problem, and I post them here with his permission. First, this is a Pleiades high resilution satellite image draped onto a digital terrain model, that shows the location of the hut, the morphology of the slope, and the tension cracks that extend into the snow pack. The scale of the rockslope failure that is developing is clear:-
This image shows the tension cracks that have developed in the snow pack, and the change in location of the Murchison Hut as the rockslope undergoes creep:-
And finally, Pascal has also produced a surface elevation difference model between 2008 and 2017:-
This appears to be a site in which INSAR could very usefully provide insight into the patterns of movement, and any change in rates of creep. Anticipating what will happen next is very hard without this level of analysis; the slope could collapse today or it might still be standing in two decades.
5 October 2017
Hurricane Maria: USGS landslide impact maps from Puerto Rico reveal the extent of the devastation
From across the Atlantic, the response of the White House to the Hurricane Maria disaster in Puerto Rico looks astonishing. With little fanfare, the USGS has now started to release maps of the impacts of landslides triggered by the hurricane, created through the analysis of satellite imagery. USGS staff are past masters at this sort of mapping, which is both challenging and time-consuming. They have released the first tranche of data, for a part of northwest Puerto Rico:-
It is important to note that this not a landslide map, it is a map of the impacts of landslides – i.e. the locations in which a landslide has affected a road, a building or another element of infrastructure. This is illustrated by one of the images that the USGS has released alongside the map:-
There are eight mapped impacts (the red dots) on this image, but far more landslides. Another example of the extent and impact of this event is this image, from the San Antonio Express-News, that shows multiple landslides and extremely high levels of damage in the upland areas of Puerto Rico:-
The Washington Post has published a rainfall map for Hurricane Maria in Puerto Rico:-
This suggests that the mapped area had some of the higher levels of rainfall, but this amount is replicated across much of the rest of the island. In turn, this suggests that similar levels of landslide damage are likely elsewhere.
3 October 2017
Murchison Hut: an interesting landslide problem in Aoraki/Mount Cook National Park
The Murchison Hut is a climbers and skiers hut located on a slope above the Murchison Glacier in Aoraki/Mount Cook National Park in the South Island of New Zealand. An iconic location, it provides 10 bunks for those willing to experience the wilds of this beautiful national park:-
The Alpine Club New Zealand web page for this hut indicates that it is “closed due to geological instability until further notice”. The New Zealand Herald posted a piece last week explaining the background to this decision. They noted that:
“Scientists have revealed the satellite images of mountain movement that have forced the closure of a climbers’ and skiers’ hut in Aoraki/Mount Cook National Park. Murchison Hut was closed last month over fears the mountainside it sits on could suddenly collapse into the glacier about 200m below, killing anyone inside the building. The urgent decision was made by hut manager the Department of Conservation and owner the New Zealand Alpine Club on advice from experts at Otago University and GNS Science.”
The work was undertaken by Dr Pascal Sirguey at the University of Otago and Dr Simon Cox from GNS Science, who compared aerial photographs of the site collected in November 2008 with aerial imagery from November 2015 and satellite imagery from February 2017. The image below plots the location of the hut on each of those occasions:-
The hut was found to have moved laterally towards the valley by about 9 metres and to have moved downwards, also by about 9 metres. Almost 4 metres of this movement has occurred since November 2015. This is of course consistent with the hut being located on a large, progressive landslide, estimated to have a volume of about 500,000 m³. The imagery supports this interpretation as tension cracks have been observed extending up through the seasonal snowpack, indicating that movement is ongoing.
The behaviour of such a landslide is very hard to predict, but there is the potential for an acceleration to failure. This would not be survivable for occupants of the hut, which has as a result been closed.
2 October 2017
Karnali Highway in Nepal: the danger of collecting landslide videos (viewer beware)
The Karnali Highway is the road between Surkhet and Jumla in the Nepal Himalayas. It is a long highway – 232 km in total – through a high mountain area. Despite being a fundamentally important strategic link within the Mid-Western Development area, it is a road with many problems. Wikipedia describes it as follows:-
According to “A Value Chain Analysis of Apple from Jumla”, and the intervention strategy indicates that more than 85 percent of the Karnali highway is still unsafe as of July 2011. Many rural inhabitants along the highway have poor access to markets, healthcare facilities and schools and deal with high transport costs. Inadequate roads make it hard for farmers to transport and market their crops. There is a pressing need to provide a functional road system in the area, made more urgent by current concerns over food prices and shortages, high energy costs and social and health needs Between 60 and 75 percent of children under five are chronically malnourished, and up to 64 percent of the population live in poverty.
The road is commonly described as being the most dangerous highway in Nepal (and there are many dangerous roads in Nepal). However, it should not be forgotten that it has opened up a huge area of Nepal, improving access to education, healthcare and markets for many people. But road users are subject to substantial levels of hazard, and landslides are common, especially in the monsoon.
A video was posted to Youtube last week that illustrates this in a terrifying manner. The video is hard to watch as a man ends up being hit by a piece of flying rock. I urge discretion in viewing it.
The location of this landslide along the road is not entirely clear – the text that accompanies it describes it as follows:-
Recently the Landslide at Dahi Kohla and now at radhauney or Kagney Khola killed a bike riders and blocked for a weeks because of massive and big stones on roads.
The Jumla Nepal blog has a report about this event as well – it describes the landslide as occurring at Bagauney, although again this location is not clear to me. It is also unclear as to whether the person who was hit by the rock was killed – I cannot find any report in the Nepal media about this event.
This is a stark illustration of the dangers of filming landslides. The rockslide rapidly transitions from a relatively small event into a very major slide, presumably because of a larger rockslope collapse upslope. The size of the blocks that are moving – fast – is notable as this still from the video shows:
The impact of this block into other boulders close to the road causes it to fracture, releasing pieces of fly rock. The dangers of these substantial pieces of rock are all too clear.
I often feature landslide videos on this site, and they have given us new insights into the way that slopes fail and move. However, landslides are incredibly dangerous events, and it is never worth risking your life for the sake of a piece of footage.
29 September 2017
Rize, Turkey – an extreme rainfall event yesterday triggered landslides and flash floods
The Rize area of Turkey yesterday suffered an extreme rainfall event. At the moment details are quite sketchy – the Watchers reports that:
At least one person has died after heavy rain caused major flash floods and landslides in the Rize region of northeastern Turkey, on September 28, 2017.
The person who died was identified in reports as a 40-year-old, Songul Tolan. The incident is reported to have happened in the village of Gurpınar, near Cayeli.
Weather stations in the region recorded 150 mm (5.9 inches) of rain within 6 hours.
There are some very dramatic videos online showing the effects, most notably the large landslide captured in this recording:-
This appears to be a large, rapid, potentially damaging landslide, but I can find no further details as yet. I have tried to improve a still from the video:
NASA 24 hour Global Precipitation Measurement data shows a rain cell sitting over the Rize area of Turkey, and a much larger one (with indicative maximum rainfall levels of over 229 mm in places) to the east in Azerbaijan:
Interestingly, there also reports of landslides in Georgia yesterday, close to the western coast of the Black Sea. The rainfall event that I presume was responsible for this is also visible in the GMP image.
Turkey is of course no stranger to highly destructive landslides, and there are several publications, such as this one, that highlight the risks. In my paper on human losses from landslides (Petley 2010) I noted that Turkey suffered 22 fatal landslides between 2004 and 2010, with 85 resultant fatalities (see my blog post about this work). In the period since there have been many more landslides associated with heavy rainfall.
Petley, D.N. 2012. Global patterns of loss of life from landslides. Geology 40 (10), 927-930.
27 September 2017
Heifangtai terrace: a location of spectacular loess flowslides
Loess, a windblown silt deposit, mantles about 6% of the land area of China. As I have noted previously, these thick loess deposits generate some of the most spectacular landslides on earth because of their potential to undergo fluidisation, allowing landslides to transition into highly destructive, long runout flowslides. In a recent paper published in the journal Engineering Geology, Qi et al. (2017) describe loess landslides on the Heifangtai terrace, which is located at the confluence of the Yellow River and the Huangshi River in Gansu Province, northwestern China. This terrace, which is located at 36.0916, 103.3089 if you want to take a look, has a cap of about 26-48 metres of loess. It is both populated (primarily by people relocated from the site of a hydroelectric plant) and irrigated. Qi et al. (2017) describe 17 loess landslides on the margins of the terrace – you can see many of these in the Google Earth image below:-
Qi et al. (2017) highlight in particular a pair of landslides that have developed in the Moshi Gully on the flanks of the terrace between 2004 and 2013. This image shows the site of these two landslides, on the southern edge of the terrace, in December 2004:-
By November 2012 notable instability had started to develop at two sites on the terrace edge. I have marked the headscarps of the two landslides. At this stage the landslides appear to have low levels of mobility:-
By February 2015 one of the two landslides, the one further from the camera, had developed significantly. As Qi et al. (2017) note, this slide had shows retrogression of the headscarp and had started to develop a clear flow-type structure. Note though that the mobility is still low:-
But by September 2015 the landslides have changed dramatically:-
Both slides have failed spectacularly, with the one in the foreground in particular showing very high mobility and extreme retrogression. Qi et al. (2017) report that this landslide occurred in a two stage process in April 2015. The first failure, involving 80,000 m3, occurred on 29th April 2015 at 7:50 am. The second happened about three hours later in three distinct phases. The landslide travelled almost 800 metres.
The most recent image, from May 2017, shows that this landslide has undergone little change, but the other has suffered a further failure event:-
Worryingly, landslide activity is clearly developing on other sections of the terrace edge.
Xing Qi, Qiang Xu and Fangzhou Liu 2017 Analysis of retrogressive loess flowslides in Heifangtai, China. Engineering Geology.
26 September 2017
A still from the Youtube video of the landslides at El Jale in Mexico.
El Jale, Ixtapaluca: a stunning earthquake triggered landslide video from Mexico
On the day of the earthquake in Mexico I noted that there was a high chance that it would have triggered landslides. It was always unlikely that this would be efficient in so doing in the manner of, for example, the Kaikoura earthquake (due to the depth of the rupture), it had the potential to generate slides over a wide area. A video has recently been posted to Youtube that shows landslides being triggered by this event, reportedly El Jale in Ixtapaluca in Mexico City. It is spectacular:
This is a Google Earth image of the Ixtapaluca area:-
The video appears to show the collapse of a quarry wall in a comparatively rural area. The one quarry shown above does not seem to fit the video. A better fit to the video appears to a set of quarries located further to the east:-
These quarries are likely to be mining volcanic materials, which would explain both the instability and the large amounts of dust generated. It is not clear as to whether there was any losses from these landslides. One early report did suggest that a quarry worker was killed in the event, but the location was given as Morales state, so it is unlikely to be this event:
Morelos Gov. Graco Ramirez reported on Twitter that at least 42 people had died in his state south of Mexico City. Gov. Alfredo del Mazo told the Televisa news network that two people died in the State of Mexico, which also borders the capital: a quarry worker who was killed when the quake unleashed a rockslide and another person who was hit by a falling lamppost.
So far this part of the area affected by the earthquake has not yet been fully images by Planet Labs. I’ll take a look when images become available.