4 January 2023
Joshimath: an Indian town with a serious landslide problem
Joshimath is a town built in a beautiful setting on the side of a mountain in Chamoli District in Uttarakhand in northern India. Located in the same district as the 2021 Chamoli landslide and debris flow, this is an area with many slope stability issues. The town itself, situated at 30.553, 79.563, looks like this on Google Earth:-
A rapid inspection of the site suggests a host of underlying landslide concerns with this slope. The large scoop missing from below the town in the centre right of the image is particularly intriguing for example. History tells us that slope problems are often greatly exacerbated by urbanisation, including disruption to the drainage, the uncontrolled discharge of water and undercutting of slopes.
Over the last few days, a number of media outlets in India have run stories about the scale of the landslide issues that are occurring at Joshimath, and the rate at which they are developing. For example, PKB News has quite a detailed story about the impacts of the slope failure. Their story notes the scale of the problem:
In Joshimath of Uttarakhand, people are in panic as cracks have appeared in around 559 houses. Due to the landslide, not only the houses, but also the roads along with the hotels and hospitals present there have been cracked.
The Times of India has a set of photographs of the catastrophic damage that is occurring to buildings in Josimath, whilst the PKB News story is illustrated with this image:-
Interestingly, the Times of India reports that an investigation in 1976 reported that the town is built upon an ancient landslide, and that the long term prognosis was poor.
Clearly there is an urgent need to investigate the hazards at Joshimath, and to develop a plan to manage the hazard. Reports indicate that a survey of damaged properties is under way, but this needs to be a part of a major larger investigation. Unfortunately, managing a landslide on this scale is likely to be very expensive.
It is of course worth noting as well that this is an area of significant seismic hazard, so the potential behaviour of these slopes during a major shake should also be factored in.
23 December 2022
Losses from debris flows in Brazil
It is well-established that Brazil has a substantial problem with landslides, most notably in the urban areas in the southeast of the country. An Open Access paper just published in Landslides (Cabral et al. 2022) provides a detailed analysis of one type of landslide in the country, debris flows. The paper looks at recorded events in the period between 1920 and 2021 inclusive.
In total, the authors have documented 45 debris flow events over this period in Brazil, causing 5,771 fatalities and US$5,5 billion in economic losses. The most serious events occurred on 11 January 2011 in the vicinity of Rio de Janeiro, when heavy rainfall triggered three separate debris flows that killed a total of 892 people. But very large events have occurred at other times. For example, on 18 March 1967 a debris flow at Caraguatatuba, Sao Paolo killed 436 people, and on 19 February 1988 a debris flow in Rio de Janeiro killed 289 people.
In 2010 a debris flow triggered within a deactivated landfill area in Niterói killed 48 people and left a further 221 missing. There is a contemporary report on this event in the San Diego Union Tribune, which includes this image:-
Perhaps the most interesting aspect of this study is the analysis of events and losses through time. (Cabral et al. 2022) report the following:
There has not been a significant increase [in the number of events] in recent decades, with the frequency averaging at about seven per decade since the 1960s. Prior to 1960s, the documentation of debris-flow events is scarcer, which can be associated to lower levels of urbanization and, consequently, a lower societal impact. The number of fatalities along the years has also been steady since the 1960s, with the decades of 1960s and 2010s standing out as those with the highest number of fatalities, particularly in the years of 1967 and 2011. These two years are characterized by the high-magnitude debris-flow events in Caraguatatuba and Serra das Araras (1967), and Teresópolis, Petrópolis and Nova Friburgo (2011).
On the other hand, economic losses have accelerated dramatically, probably related primarily to increased urbanisation and economic activity, meaning that more assets with a higher value were in the path of debris flows.
Cabral et al. (2022) conclude the abstract with the following, which is hard to deny:
…these results highlight the urgent need for investments in disaster prevention and preparedness programs.
Cabral, V., Reis, F., Veloso, V. et al. 2022. The consequences of debris flows in Brazil: a historical analysis based on recorded events in the last 100 years. Landslides. https://doi.org/10.1007/s10346-022-01984-7
21 December 2022
The Urni landslide: a recurring slope failure in India
On Monday 19 December 2022 a significant slope failure occurred at the Urni Landslide in Kinnar, Himachal Pradesh in northern India. This event was notable for the fact that it was caught on camera by users of National Highway 5, which passes close to the landslide site:-
Fortunately there was no loss of life at this site, but the road was closed for repairs.
The landslide is located at 31.519, 78.129 on the banks of the Sutlej River in very steep terrain. It is captured beautifully in Google Earth:-
The archive images of the site on Google Earth provide a classic case study of the development of a landslide through time – take a look. The slope is now deeply problematic – note the very the diversion of the road across to the other bank of the river for a short section to try to avoid the landslide toe/
The Urni landslide was written up in an open access paper (Kumar et al. 2019) that was published in the journal Geoscience Frontiers. The aim of the paper was to model a major failure that might dam the river. The authors noted that the landslide has developed in gneissic rocks on steep slopes, and that there are numerous cracks upslope of the existing failure, indicating a high probability of future events. This came to pass this week, at least in part. The failure was initially a structurally-controlled wedge failure, but has developed into a much more complex landslide.
Movement is primarily triggered by heavy rainfall in the monsoonal climate, but the precarious nature of the topography inevitably means that collapses can occur at any time.
Of interest at this site is the ongoing works to construct the Karcham Wangtoo Hydro Power Project – some elements of this are visible in the image above. It has long been asserted that these works have led to an increase in the occurrence of landslides in this area. For example, the South Asia Network on Dams, Rivers and People has an article from 2014 documenting some of the impacts. Construction started in 2005.
Of course, of immediate concern is the potential risk to the houses upslope of the Urni landslide. In October 2022 the crown of the landslide was perilously close to the houses. A major failure could exacerbate this situation.
16 December 2022
A major, deadly landslide in Genting Highlands, Malaysia
On this, the 15th anniversary of the Landslide Blog, there is tragic news of a major landslide in the Genting Highlands of Malaysia. The situation remains dynamic as I write, but it is clear that the landslide inundated a campsite at a location known as Father’s Organic Farm. News reports indicate that there were 81 people registered to stay at the site, 30 of whom were children, plus of course a number of staff. In total 53 people have been reported to have been rescued. As I write, there are 16 confirmed fatalities, with a further 17 believed to be missing. These numbers will change in the coming hours.
There are some videos on Twitter collected via drones showing the configuration of the landslide. These videos in particular are revealing:-
KEMASKINI KEJADIAN TANAH RUNTUH SEHINGGA JAM 1041 HRS ;-
Jumlah keseluruhan (94) orang, (60) dijumpai selamat, (9) meninggal dunia, (25) masih belum dijumpai. pic.twitter.com/gxcvotFtIz
— BERNAMA TV 🇲🇾 (@BernamaTV) December 16, 2022
Based on these videos I believe that the site of the landslide is at 3.423, 101.757. This is my interpretation of the location on Google Earth:-
The landslide occurred at 2:42 am, which sadly will have amplified the human cost of the disaster – to be in tents that are struck by a high mobility landslide in the dark is true horror.
There is already considerable speculation in Malaysia about the likely causes of this tragedy. I suspect that there might be a hint in the archive of Google Earth images, although I can only hypothesise at this point. This image, from 2002, is extremely interesting. As above, I have marked the site of the landslide:-
It appears to me that the slopes both above and below the road have been engineered. The videos indicate that the slope above the road remains intact, but the slope below the road has failed. The 2002 Google Earth image would seem to indicate that the slope was terraced. I wonder if this was a fill slope, and that the fill has failed to generate the landslide. This would explain the location, the dynamics and the impact of the landslide.
The second element that is interesting is from a 2018 Google Earth image, although the quality is not great:-
In this image there appears to be bare soil at or close to the toe of the postulated fill slope. It is not clear to me as to what was happening at this time – in subsequent images this area has revegetated, but this needs investigation.
JKR Malaysia Slope Engineering branch has responsibility for analysing these types of events, and a team is reported to be on site today. I am sure we will see a full interpretation of the causes of this landslide in due course. Meanwhile, the recovery work will continue, but sadly the likelihood of further successful rescues seems low.
Thanks to Riza and Liew SS for their kind emails pointing to information about the landslide at Genting Highlands.
Fifteen years of the Landslide Blog
Today marks the 15th anniversary of my first post to the Landslide Blog. I find it hard to believe that it has been so long. My first post was on 16 December 2007, when I posted about the La Honda landslide in California (see image below), based on a field visit with Nick Rosser, who was working with me as a post-doctoral researcher at Durham University (Nick is now a professor at Durham). Today’s post is number 2,470 on this blog, which works out at about 164 per year. The blog started out on blogger but after a few years moved over to the AGU Blogosphere. I have been hugely honored to have been hosted by the AGU, who have been immensely generous to me.
The site still receives around 1,000 individual visits per day on average, although this fluctuates greatly. It has evolved with time – my posts tend to be a little longer now, and I benefit greatly from the support of many people around the world who provide information about landslides in their domain. I’m not always able to respond as much as I would like, but thank you to you all.
Over the years I have focused on some specific landslide events – in 2008 for example I wrote a great deal about the Wenchuan Earthquake, in 2010 about the Attabad landslide. I have developed some recurring themes – tailings dam failures, the rapidly evolving ability of satellites to allow rapid assessment of failures, landslides associated with major dams, and the horror of mining landslides, especially in Myanmar, for example.
Along the way I have only really had two bad experiences, both of which have been associated with senior and well-respected landslide scientists. Whilst those interactions were shocking and disturbing in many ways, the rest of my interactions have been a huge pleasure and a privilege.
Over the 15 years of this blog my day job has changed hugely. At the start I was a Professor at Durham University, and in the later stages I was also Dean of Research there. In 2014 I moved to become Pro-Vice-Chancellor (Research and Enterprise) at UEA, then in 2016 I became Vice-President (Research and Innovation) at the University of Sheffield. In September 2022 I moved again, to become Vice-Chancellor at the University of Hull, which is my role today. The blog has been a constant through this period, and has allowed m to remain rooted to my discipline.
Even with my current somewhat busy schedule I plan to maintain the blog, even as many others that started at the same time have fallen away. I write it very early in the morning, at the start of my day, as a way to get my brain in gear. I hugely appreciate spending time thinking about my research focus, even though I rarely have time for research itself. Of course my pieces on individual events are only ever an initial take; the definitive accounts result from detailed studies by others, and must always be seen as being more important.
I write about things that interest me, making the topics somewhat eclectic. I’m constantly amazed and delighted that others find them to be of interest. I will continue to do so if and when I can.
So, finally, thanks to you all for the last 15 years, and thanks in particular to the AGU, to LarryO’, to Planet Labs and to many others for your kind help.
15 December 2022
The occurrence of landslides in Pakistan
The combination of topography, rainfall, seismicity and people means that Pakistan is heavily affected by landslides. However, until now I have not been aware of a systematic analysis of landslides at a national level in a way that has been performed for many other countries, a major gap in our understanding of this hazard. Pleasingly, a paper (Shabbir et al. 2022) has just been published in the journal Environmental Science and Pollution Research that seeks to fill in this gap. Even better, it is open access.
The researchers have compiled a dataset of landslides across Pakistan from 2003 to 2019 inclusive. The dataset includes all reported landslides, not just those that have caused fatalities (as in the case of my own research). In total, 1,089 landslides have been identified, of which 180 led to loss of life. In total, 1,072 people were killed by landslides over this period. Rainfall triggered 63% of the recorded events, although interestingly the seasonal effects that are seen in other parts of South Asia (for example Nepal) is less strong in Pakistan.
The map below shows the distribution of recorded landslides:-
There is huge richness in this map. The most obvious conclusion is the immense focus of landslides on the northeast of the country, in the high mountain areas. Pakistan is particularly interesting as the Karakoram Highway means that many people live beyond the frontal range of the high mountains. This is reflected in the landslide distribution, which extends a long way to the north. But even here there are two distinct clusters, one located in the area of Muzaffarabad and associated districts, and the other in the far northeast, around Baltistan. The first cluster is at least in part in the area affected by the dreadful 2005 Kashmir Earthquake, which triggered many landslides (most of which are poorly constrained, and so are not captured in this dataset), whilst the other is the area in which the 2010 Attabad landslide occurred.
This is a really valuable study, and I commend the authors for putting it together. I would really like to see further analysis, in particular of the fatal landslides, which are likely to form a more consistent dataset. Understanding these patterns in time and space is incredibly valuable as a step towards managing the risk.
Shabbir, W., Omer, T. & Pilz, J. 2022. The impact of environmental change on landslides, fatal landslides, and their triggers in Pakistan (2003–2019). Environmental Science and Pollution Research. https://doi.org/10.1007/s11356-022-24291-z
14 December 2022
Palos Verdes Estates: a coastal cliff landslide captured on video
On Friday 9 December 2022 a landslide occurred on a coastal cliff at Palos Verdes Estates in California. Thanks to loyal readers Ernie Roumelis and Ken for highlighting this one – it is particularly interesting because it was caught on video from a helicopter. A Fox 11 helicopter was in the area, and succeeded in capturing the runout of the failure:-
The Daily Breeze has a couple of images of the aftermath of the landslide, including this good image of the source area:-
The image suggests that the failure was a structurally controlled wedge failure that turned into a dry debris avalanche. It is fortunate that the landslide occurred when the beach was largely unoccupied. Meanwhile CNN has a good image of the debris and the (fortunately unoccupied) pick-up truck that was struck by the landslide:-
The location of the landslide at Palos Verdes Estates is 33.804, -118.394. This is a section of coast, in common with many other locations in California, is no stranger to landslides. It is interesting that the event occurred in dry weather, ahead of the major rainstorm that affected this region over the weekend. Coastal cliff failures often seem to occur in dry weather (this is an area of active research across multiple groups at the moment). It should also be noted that artificial irrigation is common in this area, although I have no evidence of the likely cause in this case.
For those who are interested in the underlying geology, and the associated landslides, Ernie has highlighted a really interesting map produced by the California Geological Survey. Almost every cliff along this coast is likely to have experienced landslides of one form or another over time.
California remains remarkably prone to landslides of multiple types, both on the coast and inland, resulting from the combination of geology, climate, seismicity and the actions of people.
13 December 2022
The Gaguney landslide in India: evolution over time
Yesterday I wrote about the work that Save The Hills has been doing to highlight the plight of people of Pathing in northern India who are being affected by the Gaguney landslide. This landslide had been problematic for some time, but in November of this year the situation became dramatically worse, with a new area of instability threatening many more houses.
A combination of Google Earth and Planet Labs imagery provides an interesting perspective on the development of this failure. There is a good Google Earth image from 2011 – I have place the marker at the crown of the landslide at this time:-
At this point the landslide was quite well developed, but the impact on the village was quite low. The image below shows the landslide in late 2020:-
In the intervening nine years the landslide had clearly developed considerably, but again the impact downslope was quite limited. However, from this point the situation rapidly deteriorated. This is the landslide in late 2021:-
The surface area of the landslide had expanded greatly and the debris area was large. The track into Pathing was now being affected and the village itself was being impacted.
We now switch to the lower resolution but still valuable Planet Labs PlanetScope imagery. This image is from 15 October 2022, with the marker at the same point is in the images above:-
The landslide had dramatically increased in size. The headscarp had retrogressed by over 200 m, bringing a huge increase of mass into the active system. A much larger area of debris is evident downslope, and the impact on the village had also increased substantially. Perhaps most worryingly, a second area of active failure had developed to the south and east of the main landslide.
The landslide subsequently developed quickly. On 27 November 2022 the increased level of activity was starkly evident:-
The main slide had further increased in size, whilst the secondary failure had also developed considerably, and there was evidence of debris impacting much more of the village. Note the substantial part of the access track that was now covered by landslide debris.
Perhaps the most worrying sign is the large volume of displaced mass that is now in the headscarp area, directly above the smaller landslide that threatens the main part of the village. This material needs to be examined in more detail – the PlanetScope imagery has a resolution that is too low for proper analysis. The threat to the village needs to be properly assessed.
Planet Team (2022). Planet Application Program Interface: In Space for Life on Earth. San Francisco, CA. https://www.planet.com/
12 December 2022
Save The Hills: raising awareness on impact of landslides on rural communities
Long time readers of this blog will be aware of my admiration of a wonderful NGO, Save The Hills, which is a group of concerned citizens who seek to raise awareness of landslides in the Darjeeling-Sikkim Himalaya. I have been privileged to be in regular contact for many years with the founder, Praful Rao, and many other landslide experts have worked with the group too. They run a wonderful blog about landslides in this region. This is grass roots activism as its very finest.
As part of their efforts to raise awareness of landslides, Save The Hills has produced two short but powerful documentaries about the impact of landslides in a remote village, Pathing. The first video was made in November 2022, highlighting the way in which the large Gaguney Landslide is steadily and systematically destroying the loves of this rural community. The video combines testimony from the local people with stunning drone footage:-
Whilst this landslide has been active for two decades or more, in recent months it has become more active, spalling rocks almost continually. The debris has inundated land and blocked roads, as well as representing a major hazard for the local residents. The still from the video below shows the scale of the problem at Pathing:-
Tragically, Save The Hills visited the site a week later to find that the situation had deteriorated desperately. The footage was collected at the end of November, when the landslide had become much more active. The government ordered evacuation of the village.
The issue appears to be that the crown of the landslide has expanded, putting many more houses at risk. The second video provides a graphic and powerful account of the scale of the problem:-
One key element of this second video is that it highlights the desperate plight of people being forced from their homes. Relocation is rarely simple, trading one problem for many others.
This image shows the scale of the new landslide source area, where the rock mass is clearly undergoing extensive raveling:-
Save The Hills also has a blog post online that highlights the challenges at Pathing, illustrated by multiple photographs. Pathing is at 27.295, 88.392, and is well captured on Google Earth.
Pathing is of course not alone in this problem, which is replicated in many other locations across the Himalayas. In Nepal for example, Khabar featured another example on Saturday, this time the village of Maya in Pachaljharana Rural Municipality-8 in Kalikot district.
2 December 2022
BR-376: a deadly highway landslide in Brazil
On Monday 28 November 2022 a large landslide occurred on the BR-376 highway in western Brazil. The image below captures will the aftermath of the landslide:-
The image below, from the state news agency of Parana, shows the general set up of the slope that failed. This image was collected after much of the debris had been removed from the slope:-
There is a good gallery of images on the MetSul website. Reports indicate that there were two failures. The first occurred at about 3:30 pm on Monday, blocking one side of the BR-376. Four hours later a second, larger, failure occurred whilst traffic was queueing to pass the spot. This is the reason why so many vehicles were caught up in the failure.
At the time of writing two fatalities had been confirmed, whilst rescue operations continue. Early fears were that there might have been as many as 30 fatalities, but thankfully this now seems unlikely.
The site was known to have stability issues, and reports indicate that work had been ongoing. The failure was triggered by heavy rainfall – MetSul reports about 200 mm on the day of the failure.
I believe that the site of the failure is -25.874, -48.937. This is the Google Street View image of the location, taken in March 2022:-
It is notable that works were ongoing when this image was captured, although these appear to have been focused on a site a little further along the road. The images in the aftermath of the landslide appear to show a temporary building (with a red roof), which is not present in the images from earlier this year. Is this an indication of ongoing engineering works?
The rescue and recovery operations at this site are very complex. Much of the debris and the vehicles are located off the road, below a large retaining wall. Perhaps more importantly, there is a large displaced but stalled mass on the slope that is potentially unstable. This is sufficiently large to pose a very substantial hazard to those in the debris field.