This blog provides a commentary on landslide events occurring worldwide, including the landslides themselves, latest research, and conferences and meetings. The blog is written on a personal basis by Dave Petley, who is the Wilson Professor of Hazard and Risk in the Department of Geography at Durham University in the United Kingdom.

This blog is a personal project that does not seek to represent Durham University.

Wednesday, 31 December 2008

10 New Year Resolutions for Safer Slopes

Given the time of year, I thought I would try to put together ten resolutions that could lead to a decrease in the impact of landslides. They are in no particular order and apply at a range of scales. I would welcome any comments.
  1. Before building infrastructure, we should always map the slopes to check for potential or actual landslides. This is not a complex or difficult thing to do, but it can save a huge amount in the future. The costs of a landslide on a road, in a city or on a railway can be really serious;
  2. Reverse deforestation. There is no doubt that chopping down forests is bad news for slope stability. Increasing levels of forest loss are a huge threat in many less developed countries, and will continue to be so as population increases. Reversing deforestation is not just a task for less developed countries - there is a need to grow more, and increasingly sustainable, forests in the developed nations too.
  3. Determine how to save people trapped by landslides. In the recent landslide in Kuala Lumpur, Malaysia it took more than two weeks to find the remains of one of the victims, despite the massive mobilisation of resources. This highlights how poorly we know how and where to look for victims. There is an urgent need to research this properly.
  4. Maintain slopes properly. Far too often an assumption is made that once slope protection has been emplaced then nothing more is needed. One grows tired of hearing how a slope has failed because a simple drain was blocked. Frequent, low level maintenance can make all the difference.
  5. Mine waste and fly ash can be very dangerous if they are allowed to fail. A disconcerting trend this year has been the number of large flowslides in mine wastes and in power station ash. We have known how dangerous these materials are for over 40 years. They can be safely managed - there should be no excuse for allowing these types of disaster to occur.
  6. Landslides are a major issue during earthquakes in mountain areas. Both the Wenchuan (Sichuan) and Balochistan (Pakistan) earthquakes have highlighted the impact that landslides have when earthquakes strike, and also the impediment that they represent to rescue and reconstruction. Preparation for an earthquake in a mountain range must include preparation for landslides too.
  7. We can't ignore climate change, which will have a substantial impact on slope stability, especially in high mountain areas. I am increasingly vexed by the rantings of the small group of people who insist on denying the evidence for anthropogenic (human) climate change. In the real science community there is no essentially doubt that climate change is real, and indeed I have noted a growing sense of quiet pessimism. In permafrost areas the rate of rockfall activity is now notably increasing, and are likely to get much worse.
  8. The biggest landslide impacts occur in less developed countries. The lack of investment in safe slopes in developing countries is frankly appalling. There is an urgent need to undertake proper research in less developed countries to understand the mechanics of the slopes properly, to identify hazardous locations and to develop sustainable ways to support and or drain them.
  9. People are the key. In all environments in which the toll from landslides has been reduced this has been achieved by creating a cadre of professional people with the skills to identify and mitigate unstable slopes, and the means to so do. Information exchange is critically important. If we wish to reduce the impact of slope failures in less developed countries then we will need to do the same thing there. It is simply naive to believe that we can achieve the same thing with community-based disaster reduction strategies or without the use of appropriate engineering.
  10. Urban growth is a major threat. The rapid growth of cities, in particular in hot and humid countries, is placing an increasing pressure on slopes that will lead to an increase in accidents unless due care is taken.
I still maintain that of all the natural hazards, landslides are the most amenable to mitigation. That thousands of people are losing their lives each year in avoidable slope accidents is a tragedy. We can and should do better!

Happy New Year to you all.


Tuesday, 30 December 2008

Losses in catastrophes in 2008 - initial statistics from Swiss Re

Tropical Cyclone Nargis montage from the Brisbane Times

Swiss Re Sigma has just released its initial statistics on losses in catastrophes in 2008. Of course it is important to understand that reinsurance look at catastrophes in a particular way, meaning that the statistics are weighted in particular to low frequency - high magnitude events, rather than iterative processes, and to losses causing a high level of financial loss. There is no problem with this of course, but it is important to keep it in mind.

So what do the statistics show? Unfortunately the picture is pretty grim. Swiss Re estimate that these large loss events cost over 238,000 lives, most notably as a result of tropical cyclone Nargis in Burma (Myanmar), which led to 138,400 fatalities, and the Wenchuan (Sichuan) earthquake, which killed 87,400 people. In terms of economic losses, Swiss Re estimate that catastrophes cost about US$225 billion in 2008, of which about $50 billion was insured. The greatest costs were from the Wenchuan earthquake ($85 billion) and Hurricane Ike (USA) ($40 billion).

All in all 2008 has been the second worse year for insured losses on record (2005 was the worse because of Hurricane Katrina). I guess that this is unwelcome given the global financial malaise.

I have to say though that for me perhaps the interesting aspect is the following graph within the report (download a hard copy of the report here):

So let's take a quick look at the recent trends. First, the palest blue line is the total insured losses. There is a huge amount of scatter from year to year, but the general trend is clearly upwards. However, the other two losses show that this increase arises primarily from the weather related perils (the darker blue line). Over the last decade there is no obvious trend in losses from man-made disasters, with a single obvious peak in 2001 - no prizes for guessing the cause of that. Note also that very large-scale earthquake losses to the insurance industry are comparatively rare - most earthquakes occur in areas that have very low levels of insurance cover. Thus the 2005 Kashmir earthquake and the 2008 Wenchuan earthquake hardly appear on the above graph.

The trends in losses interests me greatly. I wanted to check that there is a net increasing trend in losses from weather-related perils. A quick, back of the envelope graph shows that there is indeed a clear upward trend:

I fitted a fairly simple exponential function above, although it may not necessarily be the best choice. There is a very clear upward trend in insured losses from natural perils. This rise in economic losses from natural perils is generally ascribed to increasing vulnerability - i.e. we have more economic assets in the "firing line". However, note that insured losses from man-made perils are not following the same trend. Presumably, this is also the case for man-made perils, but in this case the improvement in management and control is keeping track.

I will return to these data in the future.

Wednesday, 24 December 2008

Ash flowslide at Knoxville, Tennessee

A few months ago Shaanxi Province in China suffered a dreadful flowslide when the dam holding back mine wastes collapsed, releasing an avalanche of material onto the town below. About 260 people were killed. On 22nd December, what appears to be a similar failure occurred at Knoxville in Tennessee, USA, when the retaining wall holding back coal ash from the Tennessee Valley Authority Kingston Steam Plant in Harriman power station collapsed, releasing an estimate 2 million cubic metres of waste, which then flowed down slope (image from Knoxnews).

Fortunately, the area below was only sparsely inhabited, so in this case no-one has been killed, but about a dozen homes have been rendered uninhabitable and, as the picture above shows, a train was also hit (image from Knoxnews).

Microsoft Virtual Earth has this rather nice black and white image of the site, which I have annotated to show the salient features (click on the image for a better view in a new window):

The image below shows the crown of the failed area (see annotation) - comaprison with the above image confirms which part of the storage pond has failed (image from Knoxnews):

However, as the image below shows, the failure affected a very large part of this lower set of ponds (image from Knoxnews):

My current interpretation of what has happened is shown below, which is an annotated zoom-in of the MicrosoftVirtual Earth image above:

Interestingly, the mass of mobile ash also appears in places to have caused the ground to fail without flowing. This image shows the access road to the site - note how the material has moved forward over the road without fluidising (image from Knoxnews):

That such a failure should occur is extraordinary, given that the danger of flowslides has been known for over 40 years. The only fortunate aspect of this is that no-one was killed, which seems to be a matter of luck given the volume and mobility of this landslide. However, coal ash is an unpleasant material (which is why it is stored in ponds like this), sometimes containing lad, arsenic and mercury amongst other heavy metals, although in very low concentrations. The major issue will probably be dealing with the sludge before it enters the watercourses.

Accidents like this should not be allowed to happen - they are utterly avoidable. I hope that a review is underway to ensure that there is no repeat.

Tuesday, 23 December 2008

Brazil landslides

Over the last month Brazil has suffered terribly from heavy rainfall that has induced extensive flooding and landslides. Over130 people have been killed to date.

This week, Minas Gerais state has been heavily affected. The Latin American Herald Tribune has today released this image of some of the damage. It shows how small, shallow soil slips are causing extensive damage to houses.

The Burgess Shale - its all down to landslides! The Burgess Shale is probably the World's most famous assemblage of fossils, made famous in particular by Stephen Jay Gould's book Wonderful Life. In a nutshell, the shale contains an extraordinarily well-preserved collection of Cambrian marine fossils, now generally considered to be 505 million years old. The morphology of the animals is generally somewhat extraordinary:

Walcott Quarry, the site at which the Burgess Shale is exposed, is located in British Columbia in Canada. Whilst studies over the last century have, understandably, focused on the fauna of the deposit, in recent years there has been a growing level of interest in trying to understand why this assemblage of fossils has survived in such a remarkable state of preservation. Earlier this year, Gabbott et al. (2008) published a paper that examined the mudstone in which the fossils are located in order to try to understand better the conditions in which the fossils were able to survive. Interestingly, a key factor appears to the occurrence of "mud-rich slurries" that buried the fossils rapidly.

Preservation of intact marine fossils requires that a number of conditions are met. First, of course, the organism needs to be buried in order for it to be preserved. This burial cannot be too dynamic or the organism will be broken up, but must be rapid to avoid decomposition. Once buried, the conditions must prevent decomposition. Finally, the material must survive intact the long geological time period to the future - this will probably include burial to considerable depths, where there are both high pressures and high temperatures - the Burgess Shale is thought to have reached depths of 10 km below the sea bed, in which temperatures would have been in the order of 250-280 degrees C! In many ways it is amazing that so many fossils survive.

Gabbett et al. suggest that in the case of the Burgess Shale, burial occurred in a series of fluid debris flows that came down from the slope, which is thought to have been about 160 m tall. These mud slurries buried the fauna almost instantaneously, smothering them without causing them traumatic damage, although some movement during the landslide is likely. Burial was sufficiently deep (tens of centimetres) that they did not float free as gases formed during decay accumulated and were also below the surface layers of sediment, in which large amount of bacteria are active.

Gabbett et al. conclude that it was this rapid but non-traumatic burial that allowed such remarkable levels of preservation of the fossils, rather than chemical processes that inhibited decay. An interesting case in which landslides have proven to be a good thing (apart from the organisms themselves of course).

S.E. Gabbott, J. Zalasiewicz, D. Collins (2008). Sedimentation of the Phyllopod Bed within the Cambrian Burgess Shale Formation of British Columbia Journal of the Geological Society, 165 (1), 307-318 DOI: 10.1144/0016-76492007-023

Monday, 22 December 2008

Landslide on State Route 87, Arizona

Arizona Geology, the Blog of the Arizona State Geologist Lee Allison, has a new post highlighting the reactivation of an ongoing slope failure on Highway 87, which links Phoenix with the town of Payson. Movement on this slope was first noted in March 2008, whereupon emergency works were initiated to try to reduce the immediate hazard. Nonetheless, the road was closed for six days at that time.

The failure is quite unusual in terms of its morphology, as the image below shows (image from the Arizona Department of Transport - click on the image for a larger view on their web site):

The slide is clearly rather large (look how far back from the road the cracks entend), and fairly deep-seated and complex, given the heave in the road (image from the same web site):

After the initial slide in March the Arizona Geological Survey published a reconnaissance report about the slide, which can be accessed from a dedicated website here. This report notes that the slide is part of a much larger, ancient complex of landslides extending over a kilometre down slope (see image below from

The slide, which first occurred in a period of heavy rainfall, is located in wet, sandy sediments with clay rich layers (classic landslide materials). The emergency repairs seem to have consisted on some regrading of the slope together with some reinforcement of the road area, supplemented with a state-of-the-art monitoring system. It appears that the latter has successfully detected the recent movements, which have occurred in another period of heavy rain. Larger scale mitigation works are planned for the Spring.

Thursday, 18 December 2008

Spatial patterns of deaths from natural hazards in the US

There is a very interesting paper in press in the International Journal of Health Geographics on the spatial patterns of mortality (deaths) from Natural Hazards in the United States. The paper, entitled "Spatial Patterns of Natural Hazards Mortality in the United States" by Borden and Cutter is in pre-print form but can be downloaded as a PDF here. First up, lets be clear that the authors are reputable - Susan Cutter in particular has made a massive contribution to our understanding of the social impacts of natural hazards and, unlike many working in this field, she has managed to keep her materials sensible, balanced and approachable.

The paper looks at the spatial pattern of mortality from hazards using an improved dataset orginally based upon the Spatial Hazard Event and Loss Database (SHELDUS), which is available at, covering the period 1970-2004. Much of the paper relates to the reliability of the data and the quality of the database, but there are also some interesting outcomes in terms of the spatial distribution and also the cause of death.

First, the spatial distribution is perhaps not what one would expect (Fig. 1). Intuitively one might say that the highest level of mortailty would be on the west coast given the earthquake hazard, but in fact the map shows no clear spatial pattern, with pockets of high mortality occurring across the whole of the country. There is a slighlty higher level of mortality in the midwest than in the east, reflecting the pattern if temperature extremes perhaps.

Figure 1: The map of the spatial distribution of mortality across the USA for the period 1970-2004. This map is Fig. 3 from Borden and Cutter (2009) (Biomed Central Ltd).

The reason for this is that there is a vast range of causes of death in natural hazards in the USA, with temperature-related deaths (either from heat or cold) dominating.

So what of landslides? Well, the combined datasets indicate 170 landslide fatalities out of a total of 19,958, which is 0.85% of the total. In comparison, geophysical hazards (presumably earthquakes and volcanoes) provide 302 deaths (1.5% of the total).

One of the things that the study highlights, although does not really discuss explicitly, the key importance of time in such a study, bearing in mind the low frequency but high magnitude nature of natural hazards. For instance, the time period does not include the occurrence of Hurricane Mitch in 2005 - I suspect that the map might look a little different if the >1800 fatalities in this event were included. Similarly, the map does not include a very large earthquake in California or on the Cascadia subduction zone. Therefore, great care must be taken in the use of such data to evaluate risk rather than impact.

The paper finishes by acknowledging that much work is needed on this type of analysis. The paper presented here is a very useful step along the way.

Borden, K.A. and Cutter, S.L. 2008 in press. Spatial Patterns of Natural Hazards Mortality in the United States. International Journal of Health Geographics, 7:64. doi:10.1186/1476-072X-7-64 .

Tuesday, 16 December 2008

Dave's landslide blog - one year on

Yesterday we threw a small party to celebrate the first birthday of this Blog, which is today. As you can imagine, it was well-attended by A-list celebrities, some of whom have a surprising interest in landslides. Did you know for example that Britney Spears has a deep fascination with flow dynamics (although rumours that the lyrics of her song were originally "my debris flow is killing me" may well be exaggerated). Robbie Williams has gone one step further, having completed a PhD in low cost landslide mitigation in mountainous environments. News that his song "Angel" was actually written in appreciation of gabian wall design ("...and through it all, it offers me protection") should be treated with a pinch of salt. Reports that he is considering a reunion with Take That in order to record a cover version of Transmission Vamp's "Landslide of Love" are something that we can all get excited about.

On a more serious note, I started the blog as an experiment to see what would happen, primarily with the aim of trying to break the barriers between academics and the wider community. I think it has worked - certainly I have had a lot of fun doing it and I have learnt a surprising amount - although it has not all gone to plan. Unfortunately I didn't put a web counter on the page until April, so my data for views of the blog only runs from there. I guess it is probably true to say that the site didn't get a huge amount of traffic before that though. So here are the stats:

Number of posts since 16th December 2008: 173
Number of views since 18th April 2008: 63,228
Number of visitors since 18th April 2008: 32,980.

The graph of page downloads since 18th April is shown below (click on the graph for a better view). You will see that the number of visitors varies hugely according to whether there is a large, high profile landslide in the news.

The highlight remains the post of 12th May 2008, which I put up at 07:27 UT, 29 minutes after the Wenchuan Earthquake occurred. This post said "it is reasonable to assume that this earthquake will have triggered large numbers of landslides as this is a very landslide-prone area...if the initial reports on this earthquake are correct then its impact could be fearsome."

Sadly all too true, although I guess I hadn't really envisaged just how bad it would actually be.

Finally, thanks to everyone who has tipped me off about videos, images, news stories, etc. I couldn't do this without you.

Best wishes,


Sunday, 14 December 2008

Landslide sessions at EGU in 2009

Around this time of the year I always start to turn my attention to the European Geociences Union meeting in April. This is a huge get-together of earth scientists from around the world (last year >8,500 people attended). One of the largest divisions is "Natural Hazards", and the landslide section, of which I am the secretary, is the biggest single part of NH. Next year (2009) the meeting will be held in Vienna on 19th-24th April and it looks to be the best yet - we have seventeen proposed sessions for which we are inviting abstracts. These are as follows:

NH4.1/GM6.3: Landslides, ground-failures and mass movements induced by earthquakes and volcanic activity
Convener: V. Del Gaudio | Co-Conveners: R. W. Jibson , D. Keefer , J. Wasowski

NH4.2/HS11.7: Hydrological processes in landslide research: analysis and quantification
Convener: T.A. Bogaard | Co-Conveners: L. Borgatti , F. Lindenmaier , J.-P. Malet , E. Zehe

NH4.3: Landslides Triggered by Rainfall Events
Convener: K.-t. Chang | Co-Convener: M. Borga

HS11.1/NH4.4: Rainfall triggered landslides and debris flows and their effect on erosion and sediment yield in river catchments
Convener: J. Bathurst | Co-Conveners: G. Crosta , P. Frattini

GM6.2/NH4.5: Processes and rates of rock slope erosion: weathering, detachment, and transport
Convener: JR Moore | Co-Conveners: M. Krautblacher , S. Loew

NH4.6: Hydrological, hydraulic and mechanical effects of plants for slope stability
Convener: F. Florineth | Co-Conveners: F. Graf , H. P. Rauch , F. Rey

NH4.7/HS2.7: Natural and anthropogenic hazards related to water reservoirs
Convener: J. Rohn | Co-Conveners: L. King , T. Scholten

NH4.8: Large slope instabilities: from dating, triggering, monitoring and evolution modelling to hazard assessment
Convener: G. Crosta | Co-Conveners: L. H. Blikra , M. Jaboyedoff , O. Korup

NH4.9: Landslides monitoring and characterization using high resolution DEM, LIDAR and other DEM techniques
Convener: M. Jaboyedoff | Co-Conveners: R. Couture , M.-H. Derron , C. Froese

NH4.10: Impacts of climate change and land-use change on landslides
Convener: J. Wasowski | Co-Conveners: T. Dijkstra , N. Dixon , M. Winter

NH4.11: Time and intensity prediction in landslide hazard assessment
Convener: F. Catani | Co-Conveners: J.-P. Malet , J. L. Zezere

NH4.12: Remote sensing and geophysical techniques for investigating unstable slopes
Convener: J. Wasowski | Co-Conveners: V. Del Gaudio , H.-B. Havenith

NH4.13: Terrain Instability Analysis and Mass Movement Prevention
Convener: X. Meng | Co-Conveners: J. Ma , D. Wang

NH4.14/HS11.6: Landslide Forecasting
Convener: F. Guzzetti | Co-Conveners: G. G.R. Iovine , M. Parise

NH4.15: Landslide risk assessment methods and strategies
Convener: P. Reichenbach | Co-Conveners: A. Guenther , F. Guzzetti

NH4.16: Documentation and monitoring of landslides and debris flows for mathematical modelling and design of mitigation measures
Convener: L. FRANZI | Co-Conveners: M. Arattano , F. Tagliavini

NH4.17: Rockfalls - Analysis, Simulation and Protection
Convener: A. Volkwein | Co-Conveners: F. Berger , L. Dorren

This covers a vast range of landslide topics. Details are available here. Abstracts need to be submitted by 13th January 2009. Please do come along and join in!

Thursday, 11 December 2008

The Vaiont (Vajont) landslide of 1963

For some 12 or so years I have maintained a set of notes on the amazing Vajont (sometimes spelt Vaiont) landslide of N. Italy. This is the most deadly landslide in Europe in recorded history. For a while I have been meaning to move the notes over to here - today I have finally got around to it, so here they are:

The Vajont reservoir disaster is a classic example of the consequences of the failure of engineers and geologists to understand the nature of the problem that they were trying to deal with. During the filling of the reservoir a block of approximately 270 million cubic metres detached from one wall and slid into the lake at velocities of up to 30 m/sec (approx. 110 km/h). As a result a wave over topped the dam by 250 m and swept onto the valley below, with the loss of about 2500 lives. Remarkably the dam remained unbroken by the flood.

Location and background
Vajont is located in the south-eastern part of the Dolomite Region of the Italian Alps, about 100 km north of Venice. It was built as a part of the on-going, post-war development of Italy in order to provide HEP for the rapidly-expanding northern cities of Milan, Turin and Modena. Whilst a proposal to site a dam at this location was made in the 1920's, excavation of the site began in 1956 and the dam was completed in 1960. The completed doubly curved arch dam was, at 265.5 metres above the valley floor, the worlds highest thin arch dam. The chord of the dam was 160 m, and the volume of impounded water was 115 million cubic metres.

The dam was built across the Vajont Valley, a deep, narrow gorge. The geological setting of the valley was fully understood. In this area, the mountains tend to be characterised by massive, near-vertical cliffs formed in the Jurassic Dogger formation and underlying Triassic formations. The local valleys tend to be associated with outcrops of the weaker formations, particularly the Upper and Lower Cretaceous and Tertiary units, which contain more clays and are more thinly bedded. Thus the generalised geological structure is of a syncline cut by the valley. The syncline is based in middle Jurassic limestone, overlain with successive layers of upper Jurassic limestone with clay and Cretaceous limestones.

Order of events
The order of events should be examined in conjunction with this diagram (click on it for a better view in a new window):

1. Prior to the Completion of the Dam
It appears that during the construction of the dam the chief engineer was concerned about the stability of the left bank of the dam, and a number of reports were compiled on this during 1958 and 1959, which identified a possible prehistoric slide on the right bank. Whilst there was considerable discussion of the stability of the valley walls in view of the inclined synclinal form of the strata and the possibility of old slides in this area, it was concluded that deep-seated landslides were extremely unlikely as (see Muller 1964 for a review of this):

  • areas of weakness were not identified in the three test borings;
  • it was assumed that any shear plane would have a 'chairlike' form that would exert a 'braking effect';
  • seismic analyses had suggest that the banks consisted of very firm in-situ rock with a high modulus of elasticity.
Smaller slides in the looser surface layers were considered to be likely, although volumes and velocities of movement were expected to be low.

2. During the First Filling of the Reservoir
Filling was initiated in February 1960, before final completion of the dam (which occurred in September 1960). By March 1960 the level of the reservoir had reached 130 m above the level of the river, when the first small detachment occurred. Continued filling of the reservoir occurred whilst monitoring of the movements in the banks was undertaken. In October 1960, when the depth of the reservoir had reached 170 metres, a rapid increase in the rate of displacement to approximately 3.5 cm/day was observed. At the same time a huge joint of 2 km length opened up, defining an area about 1700 m long and 1000 m wide, suggesting that a very large landslide had been mobilised. This is the crack:

On 4th November, with the depth of the reservoir at 180 m, a large failure occurred when 700,000 cubic metres of material slid into the lake in about ten minutes. As a result the level of the reservoir was gently dropped back to 135 m. At this point movement reduced to close to 1 mm/day. This is the 1960 failure:

It was realised by the designers of the dam that the large mass of the left bank was inherently unstable. However Muller (1964) stated that:

'It appeared hopeless to arrest the slide artificially, because all means that would have had to be applied were beyond human bounds. It was also impossible to either seal the surface of the area, to shift the weight or to cement the rock by means of injections. On the other hand the possibility of accelerating the sliding movement in order to let the entire mass to slide down all at once had to be excluded. The danger arising for the formation upstream of the slide by an uncontrollable level of the storage lake would have been too great.'

Thus it was decided that an attempt could be made to gain control of the sliding mass by varying the level of water in the reservoir whilst controlling the joint water thrust within the rock mass by means of drainage tunnels. It was realised that this could lead to the blockage of that section of the reservoir by the landslide mass. However the volume of water in the unblocked (upstream) section would still be sufficient to allow the generation of electricity. Hence a bypass tunnel was constructed on the opposite (right) bank such that if the reservoir was divided into two sections the level of the lake could still be controlled.

It was assumed that by elevating the level of the reservoir in a careful manner movement of the large landslide mass could be initiated. The rate of movement could be controlled by altering the level of the lake. It was realised that a final sudden movement might occur, and it was calculated that, so long as the movement did not exceed a rate that would lead to filling of the reservoir by the landslide in ten minutes or less, over-topping of the dam would be avoided.

3. First Draw-Down of the Reservoir
Creep had been initiated by the initial filling of the reservoir. As the level was subsequently drawn down, rates of movement decreased from a maximum of about 8 cm/day to 3 mm/day at a level of 185 m and less than 1 mm/day at 135 m. By this time the main landslide mass had moved an average of about 1 m.

4. Second Filling of the Reservoir
From the beginning of October 1961 through to early February 1962 the water level was raised to 185 m, followed by a phase of slow impoundment such that in November 1962 the level had reached 235 m. During the early part of this phase velocities did not substantially increase, but by the end of the phase velocities had increased to 1.2 cm/day.

5. Second Draw-Down of the Reservoir
In November 1962 a second lowering of the level was slowly undertaken, with the water depth decreasing to 185 m after four months. Initially displacements remained high but in December they began to reduce and, by early April when the water height had reached 185 m, the rate was effectively zero. The experiences gained from the second phase of filling and the subsequent draw-down confirmed to the engineers that control of the landslide was possible by altering the level of the reservoir. In consequence a third filling of the reservoir was undertaken.

6. Third Filling of the Reservoir
Between April and May 1963 the reservoir level was rapidly increased to 231 m. Slight increases in velocity were noted, but rates never exceeded 0.3 cm/day. During June the level was increased to 237 m and the rate of displacement increased to 0.4 cm/day. In mid July the level reached 240 m and some of the control points indicated small increases in displacement to 0.5 cm/day. The level was maintained through to mid-August, but during this time velocities increased to 0.8 cm/day. In the latter part of August the level was increased once more such that by early September the depth of water was 245 m. In some parts of the slide velocities increased to as much as 3.5 cm/day.

7. Third Drawing Down of the Reservoir

In late September the water level was slowly dropped to bring the rates of creep back under control. By 9th October a depth of 235 m was reached. However velocities of movement continued to slowly increase, and rates of up to 20 cm/day were recorded.

8. Catastrophic Failure
At 22:38 GMT on 9th October 1963 catastrophic failure of the landslide occurred on the slope shown below.

Before failure

After failure

The entire mass slid approximately 500 m northwards at up to 30 m/sec. The mass completely blocked the gorge to a depth of up to 400m , and it travelled up to 140 m up the opposite bank. Movement of the landslide mass ceased after a maximum of 45 sec. At the time the reservoir contained 115 million cubic metres of water. A wave of water was pushed up the opposite bank and destroyed the village of Casso, 260 m above lake level before over-topping the dam by up to 245 m. The water, estimated to have had a volume of about 30 million cubic metres, then fell more than 500 m onto the villages of Longarone, Pirago, Villanova, Rivalta and Fae, totally decimating them. A total 2500 lives were lost. The image below shows the location of these villages after the flood. The valley floor, on which the villages were located, has been wiped clean by the water. The flood wave came down the Vajont valley, which can be seen in the upper right of the image:

However the dam was not destroyed and is still standing today. The by-pass tunnel is used for the generation of HEP.

Causes of the landslide
Since the catastrophic failure, a huge range of work has been undertaken on the causes of the failure. Initially the was a large amount of speculation about the location of the sliding surface, but more recent studies have confirmed that it was located in thin (5 - 15 cm) clay layers in the limestone. It is claimed by some that as such it represents a reactivation of an old landslide (Hendron and Patten, 1985; Pasuto and Soldati, 1991), whilst others claim that it was a first-time movement (Skempton, 1966; Petley, 1996). It is likely that increasing the level of the reservoir drove up pore pressures in the clay layers, reducing the effective normal strength and hence the shear resistance. Resistance to movement was created by the chair-like form of the shear surface. Dropping the level of the reservoir induced hydraulic pressures that increased the stresses as water in the jointed limestone tried to drain. It has been estimated that the total thrust from this effect was 2 - 4 million tonnes (!?) (Muller, 1964). Failure occurred in a brittle manner, inducing catastrophic loss of strength. The speed of movement is probably the result of frictional heating of the pore water in the clay layers (Voight and Faust, 1982, 1992).

References - my papers on Vajont
Petley, D.N. 1996. 'The mechanics and landforms of deep-seated landslides'. Brooks, S., and Anderson, M (eds). Advances in Hillslope Processes, John Wiley, Chichester.

Kilburn, C.J. and Petley, D.N. 2003. Forecasting giant, catastrophic slope collapse: lessons from Vajont, Northern Italy. Geomorphology 54, 1-2, 21-32.

Petley, D.N. and Petley, D.J. 2006. On the initiation of large rockslides: perspectives from a new analysis of the Vaiont movement record. Evans, S.G., Scasrascia Mugnozza, G., Strom, A., and Hermanns, R.L. (eds) Massive Rock Slope Failure. Kluwer, Rotterdam (NATO Science Series, Earth and Environmental Sciences 49), 77-84.

Petley, D.N., 2006. The Vajont (Vaiont) Landslide. Geo-strata, March-April 2006.
References - other
Hendron, A.J., and Patten, F.D, 1985, The Vaiont Slide. US Corps of Engineers Technical Report GL-85-8.

Jaegar, C., 1980, Rock mechanics and Engineering. Cambridge University Press, 523 pages.

Kiersch, G.A., 1964. 'Vaiont reservoir disaster'. Civil Engineering, 34, 32-39.

Müller, L., 1964, The rock slide in the Vaiont valley. Felsmechanik und Ingenieur-geologie, 2, 148-212.

Pasuto, M. and Soldati, A. 1990. 'Some cases of deep-seated gravitational deformations in the area of Cortina d'Ampezzo (Dolomites)', The Proceedings of the European Short Course on Applied Geomorphology, 2, 91-104.

Skempton, A.W. 1966. 'Bedding-plane slip, residual strength and the Vaiont landslide', Geotechnique, 16, 82-84.

Voight, B. and Faust, C., 1982, Frictional heat and strength loss in some rapid landslides Geotechnique, 32, 43-54.

Voight, B. and Faust, C. 1992. 'Frictional heat and strength loss in some rapid landslides: error correction and affirmation of mechanism for the Vaiont landslide', Geotechnique, 42, 641-643.

Tuesday, 9 December 2008

Saving landslide victims

In the last few days there have been two landslides that have led to searches in the hope of finding buried victims. The higher profile slide was this one that occurred in the Bukit Antarabangsa suburb of Kuala Lumpur (Malaysia) on Saturday (pictures from The Star):

This landslide, which has dominated the news in Malaysia over the last few days, killed four people and injured a further 15. One person is reported to still be missing - searches are continuing for her.

The other slide happened in Papua New Guinea when a slope above a camp at a goldmine collapsed, burying ten people. The Australian Government dispatched a search and rescue team, but no victims were recovered alive.

These two events set me thinking about how we go about recovering people who have been affected by landslides. As far as I know there has yet to be a proper scientific study of how and why people survive landslides and of where we should look for victims. This is a crucial shortcoming that urgently needs to be corrected. The key must be the type of landslide that has occurred. So lets start with our basic landslide classification (this diagram, from the Geonet web site, is pretty much the best that is available - click on the image for a better view):

So let's look at these one by one:
1. Falls
All falls basically involve free movement through the air, meaning that objects at the toe of the slope are likely to be buried. Therefore, if the likely victims were located at the toe of the slope then the victims are likely to be under the rubble. This is what we usually see with rockfalls on roads for example:

Image from

However, away from the toe of the slope movement is often parallel to the ground surface, meaning that objects in the way get pushed outwards. Hence, vehicles away from the toe of the slope often get pushed off the road rather than buried:

In many cases, on roads this results in vehicles ending up in the river. For buildings this means that structures close to the toe of the slope will be buried, whereas those further away will be pushed outwards and usually collapsed.

2. Topples
Topples on the other hand rarely have any substantial horizontal component to their movement. Therefore, in a toppling failure the victims are likely to be buried under the rubble. This was illustrated by the recent Manshiet Nasser failure in Cairo:

Here, almost all of the buildings were buried beneath huge boulders. Unsurprisingly, few people were recovered alive. However, in the case of topples buildings may at least partially survive and there are often gaps between the boulders, which means that there is a real possibility of recovering victims if search operations are undertaken quickly and efficiently.

3. Slides
In the case of slides the key issues are:
  1. The location of the victims at the start of movement;
  2. The rate of movement;
  3. The distance moved; and
  4. The degree to which the slide breaks up.
There is a fundamental difference between the impact of the event if the victim is on the slide body rather than in its path. If the victim is on the top of the slide then the survival rate is probably quite high, even for long and reasonably fast events. In fact the Malaysia slide shows this quite well - structures on the surface are recognisable at the end of the event:

Of course there is usually a huge amount of internal movement of the landslide body, so buildings will often collapse. Thus, there is a real danger that people will be killed or injured in buildings, as above. Even in very fast and turbulent slides people sometimes survive on the surface. Below is the Hattian Bala landslide that was triggered by the 2005 earthquake in Kashmir. Two women who were cutting grass near to the crown (top) of the slide walked off unhurt at the bottom:

At the toe of the slide, objects in the way are hit and may be buried. Interestingly, in general it seems that in most cases structures are pushed outwards by the slide rather than being incorporated within it. Therefore, victims are likely to be found in the arc at or near to the toe rather than where the building originally was located:

Inevitably the level of damage to structures, and the survivability in general, is dependent upon the rate of movement and the distance that the landslide (and the victim) travels. Faster landslides moving longer distances have a far higher chance of killing the victim.

4. Spreads
Spreads are generally low velocity and therefore unlikely to kill, although buildings can and do collapse. The exception is in earthquakes, when spreads can be quite rapid and very large. Collapse of structures is the greatest hazard.

5. Flows
As with slides, flows depend on the location of the victim and the rate of movement. In this case, people on the surface when movement starts are far less likely to survive as the turbulence of the movement means that they can easily be drawn into the slide. Structures are likely to collapse completely. A mistake that is sometimes made is to look for buildings in their original location, whereas in fact they are likely to have been displaced a considerable distance. The best advice to rescuers is to look for debris as an indicator of the location of survivors

The survival of structures in the way of a debris flow depends on the speed and size of the flow, and on the properties of the material that forms it. Unprotected people stand very little chance is the flow is anything other than small, but well-built structures are surprisingly resilient:

Image from USGS

However, larger flows and those with a high proportion of soil and rock, are very destructive:

Here the key is took for the location of debris from structures that might contain survivors. For weaker buildings this is unlikely to be in the location of the original structure. There is some evidence that the flow pushes debris ahead of it, at least for a short distance, so looking around the toe of the slide might be a good option, as was the case for the second La Conchita landslide:

Image from the New York Times

The danger to rescuers
Landslides often occur in stages. Therefore, there is a very real danger to rescue staff that they might be affected by another event. This story, from Taiwan last week, is a graphic illustration of this issue:

"The body of a worker who was buried in dirt from landslides at a landfill construction site in the central city of Taichung for more than 10 hours was recovered late Sunday after many twists and turns, city officials said. Wu Cheng-wei showed no vital signs when his body was pulled out of dirt at 11 p.m., and he was pronounced death on arrival (DOA) after being rushed to nearby Taichung Veterans General Hospital. Wu was the first to be buried in a series of landslides at the suburban site where a landfill retaining wall reinforcement project was underway, but he was the last to be recovered from the dirt because subsequent landslides claimed another victim and injured two municipal Fire Department staff members. Wu was buried when the first landslide hit at around 9 a.m. Sunday. The city government's Fire Department dispatched a rescue team immediately after it received a distress call. While Fire Department staff and volunteers were busy searching for Wu, there was a second landslide which buried Fu Chia-sen, a volunteer firefighter. Fu's colleagues then rushed to mount a new search mission. At around 1 p.m., two Fire Department staff members, identified as Lin Hung-yi and Chen Chieh-chi, spotted part of Fu's clothing and tried to pull him out. However, another landslide hit, causing the pair to sink into the soft dirt. Lin Hung-yi was rescued first and was sent to the nearby Taichung Veterans General Hospital, where it was found that he had sustained only slight injuries. Chen Chieh-chi, who was buried in dirt from his chest down, was not rescued until around 5 p.m. He complained of acute pain in his legs and was sent to hospital for examination and treatment of his injuries. Fu Chia-sen was also pulled out of the dirt at 5: 35 p.m., but he showed no vital signs. He was pronounced DOA after being sent to the hospital."

I guess my take-home messages are:
  1. We really need to look at this issue properly!
  2. The type of landslide determines how, where and why people might survive. Therefore, getting someone who knows about landslides, and about landslide rescues, involved at an early stage is critical;
  3. There is also a need to protect the rescuers against further landslides. A landslide specialist can help in this;
  4. If in doubt, look for the location of debris from buildings. This is most likely to tell you where the victims are.
I would welcome any comments or thoughts on this. Has anyone any experience to share?

Friday, 5 December 2008

November 2008 fatal landslide maps

So here are the scores on the doors for fatal landslides in 2008. First the statistics:
Number of fatal landslides: 48
Number of fatalities: 396.

This is the map of the fatal landslides themselves (click on the map for a better view):

As is normal for this time of year the spatial dsitribution is quite scattered, with slides occurring in SE. Asia, S. Asia and C./S. America. This year there has been a particular focus in Colombia and in Brazil - in fact the map does not really show that there were multiple landslides in a small area (Santa Catarina) in S. Brazil that have resulted in >100 fatalities.

This is the map for the year to date:

Note the line of landslides running from NNE to SSW in China. Does it remind you of the aftershock map for the Wenchuan earthquake?

Wednesday, 3 December 2008

Another amazing Brazilian landslide video

Thanks to Christoph for the heads-up on this one. Following on from the two last week, another amazing landslide video has emerged from Brazil. This can be seen at this site:

Click on the video on the web page to play it. The ambulancemen were very fortunate. The rate of movement and the violence of the landslide is quite remarkable.

Tuesday, 2 December 2008

Burning landslides revisited...

Back in August I posted on the strange phenomenon of burning landslides. This is an interesting and surprisingly common phenomenon in which a slope failure exposes materials in the soil or rock that oxidises to generate large amounts of heat. At times this effect can be strong enough to actually ignite - in Dorset in England for example organic matter in the Kimmeridge Clay (which is the main source rock for North Sea Oil) has often ignited after landslides.

In the latest edition of the journal Geology, a paper by Robert Mariner and his colleagues at the USGS at Menlo Park in California report upon a forest fire that occurred about 40 km northeast of Santa Barbara. The fire was quickly extinguished, but subsequent investigations showed the seat of the blaze was a 7 hectare landslide, from which hot gas was being emitted via a series of fumaroles. The team measured the temperatures in some of the fumaroles and also in a borehole drilled near to one of them. The measured temperatures were as high as 262 C in a fumarole and 307 C in the borehole. The ignition temperature for dry grass is 150-200 C, so it is not hard to work out where the fire started!

The team report suggest that pyrite in the shale bedrock was exposed by the slide and oxidised rapidly, generating heat. The temperatures were high enough to ignite carbonaceous matter in the shale, which then burnt to create the fumaroles, which in turn started the forest fire. This view is supported by the isotopic composition of the gases emitted by the fumaroles.

Interestingly, the paper notes that the team are "surprised that the phenomenon is rare because siltstone-shale sequences commonly have pyrite and organic matter and are involved in landslides", suggesting that maybe this happens more often than we realise.

Overall this is an excellent paper on a slightly esoteric but very interesting topic.

Robert H. Mariner, Scott A. Minor, Allen P. King, James R. Boles, Karl S. Kellogg, William C. Evans, Gary A. Landis, Andrew G. Hunt, and Christy B. Till, 2008. A landslide in Tertiary marine shale with superheated fumaroles, Coast Ranges, California. Geology, 36 (12) 959-62.