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.

Sunday, 31 January 2010

Latest news from the Attabad landslide in Pakistan

The latest news from the site of the Attabad landslide in Hunza, Pakistan, as relayed by the Pamir Times, is somewhat mixed. There is now a very helpful video of the landslide site on Youtube, which gives a proper view of the slide and its deposit from the scarp area:



The scarp area appears to be highly fractured and unstable.

Meanwhile, this report states that the lake continues to rise as about 2.8 feet (85 cm) per day, which is pretty high. The lake is now 11 km long. Two houses were submerged on Friday, another three will be lost imminently. The helicopter service to the areas upstream has been suspended for two days due to the weather. Interestingly, the article also reports the impact statistics of the landslide itself:
  • 19 fatalities plus 6 people missing
  • Seven people injured
  • 54 houses destroyed
  • 60 houses damaged
  • 1652 people displaced
  • 1.3 km of the Karakoram highway blocked.
On the plus side, the work to create a drainage channel is now well underway. The Pamir Times has printed this image of the site:

Four excavators are working at the site.

I have two observations to make:
  1. The along valley thickness of the dam, as shown on the video, should mean that it the chances of failure by piping or slope instability of the dam itself is not high;
  2. The images all show a dam that is made up of mostly quite fine-grained material, with some blocks. I hope that the core is rather more coarse-grained as preventing rapid erosion of the dam will not be easy if the material is as shown above.

Friday, 29 January 2010

New images of the level of destruction in the Cusco area of Peru

The Spanish language blog El Caminerito has been covering the magnitude of the rainfall, flood and landslide disaster in Cusco, Peru. They have put together a very helpful map showing the locations of serious damage:


Ver Cusco en Emergencia en un mapa más grande

The scale of the disaster, which is still being ignored by the western media in favour of coverage of the tourists at Machu Picchu, is well-illustrated by these images of the Huacarpay region:


Meanwhile, via the Typeboard site, the Spanish language site Peru.com reports that the village of Zurite was severely damaged by a landslide, which sounds to be a debris flow (Google translation):

" the landslide of mud and stones covered the Plaza de Armas, the town’s Church, the main streets of the city and damaged 500 houses. The incident occurred about 3 pm on Thursday after they noticed a crack on a hill and began to take appropriate action, emergency services were able to evacuate the entire population."

This is the Google Earth imagery of Zurite:


There is no shortage of landslide scars on the hillside above the town, plus the quarries, one of which appears to have excavated out the toe of one of the scars.

Here is an image of the central square via the Panoramio site:

Thursday, 28 January 2010

Earthquake-driven coastal erosion (or a coastal lateral spread) in Haiti

The Discovery Channel has news of a very interesting example of rapid coastal erosion driven by the earthquake in Haiti. The site is at Petit Paradis to the west of Port-au-Prince. Eye-witness reports suggest that the town was struck by a highly localised tsunami in the earthquake, apparently killing 20. This is interesting in part because it is quite likely such a localised event would have been caused by a (submarine?) landslide - not at all unusual in earthquakes. However, the before and after satellite images show the magnitude of the changes on the coastline in this area.

Before:


And after:


At first glance the change does not look so dramatic, but take a look at the location of the coastline in the before image in relation to the white building that I have highlighted below:


Actually, it is a little more interesting than you might initial suppose. First, note that the section of coast that has "disappeared" lies only between the two yellow dots that I have marked on the image above. To the east and the west the beach is intact. Second, take a look at this CNN video:



The key aspect is the picture of that lone tree standing upright in the ocean. A USGS report has suggested this is the site of a lateral spread (the same type of landslide that is evident in the port area of Port-au-Prince), which has caused the coastline to slip into the sea. As lateral spreads are essentially translational, a tree can remain upright. Unfortunately, as the coastline is now no longer protected by a beach, further erosion is likely. An interesting hypothesis is of course that this not-insubstantial slip caused the localised tsunami - or could it be that there was no wave at all, just the appearance of one to those people standing on the land as it slipped below the waves?

Wednesday, 27 January 2010

On narrow-minded press coverage

Southern Peru has for the last few days suffered extremely heavy rainfall. The Living in Peru blog reports upon the impact for local people in the province of Urubamba, Cusco. This includes
10 people killed; 2,000 collapsed houses leaving 10,000 people homeless; and crops, cattle and roads swept away. The district of Yucay is isolated due to the floods, the Vilcanota river has broken its banks on both sides, causing many local residents houses to collapse, and the police station has also been swept away.

So how does the international press report on the disaster? Like this from The Times:

Food and water dwindles as backpackers scramble to escape flood-hit Machu Picchu
British backpackers were among 1,500 tourists scrambling today to escape from Peru’s ancient Inca citadel of Machu Picchu, which has been cut off by floods and landslides since the weekend. As food supplies dwindled and hostels ran out of space, many were sleeping in the train station and the town’s main square, fighting for the few seats on rescue helicopters. “The situation is about to erupt,” Rudy Chalco, a tour guide with a group of elderly Europeans, told the Peruvian daily paper El Comercio. “We don’t have any more food, disorder is starting to reign, the soldiers and police that are here don’t know what to do or how to organise the help that has arrived, people are getting desperate and no one is taking charge.” Some tourists were prepared to pay up to $500 (£300) for a seat on one of the rescue helicopters, he said.


And so it goes on for a few more paragraphs. There is not a single mention of the plight of the local people, even though (as the article states) the authorities have declared a state of emergency.

This is the same in news reports in many other newspapers from around the world. Shameful!

Tuesday, 26 January 2010

The San Antonio retaining wall collapse - did the developer have a permit?

Reports suggest that ground movement has ceased in the Rivermist subdivision, which suffered a fairly spectacular retaining wall failure over the last few days. Attention is now focussing upon that retaining wall - and in particular on why it failed. This is a pretty key issue given the potential liabilities and its magnitude - the wall is 1000 feet (>300 m long).

Press reports note two interesting aspects of this accident:
  1. The City of San Antonio has released a statement that says "The hillside collapse yesterday within the River Mist subdivision that damaged several homes was a result of an improperly constructed retaining wall, as well as improper compaction of fill on which homes were constructed by the developer, Pulte Homes. The developer did not obtain the required City construction permits for the retaining wall that collapsed."
  2. Some reports suggest that this is the second time a retaining wall has failed at this site: "One neighbor who was among the first homebuyers in the subdivision set among rolling hills on the outskirts of San Antonio said he was initially told no homes would be built on the crumbling ridge because it was too steep. Romeo Peart, 32, said one retaining wall failed several years ago before the current one was built and homes were constructed above it" (Washington Post).
I should also add that I am not sure that this is a particularly simple retaining wall failure, based upon the image below (from here):

In particular, it appears that the slope / small retaining wall below where the man in the green jacket and hat is located has also failed. I wonder why?

Finally, I thought I'd embed this video, at least in part for the picture shown before the video starts, which is a somewhat surprising image in my opinion:



The video itself provides an interesting overview and a verification that a section of the wall had needed rebuilding on a previous occasion.

Fatal rockfall near to Munich in Germany


Last night a fatal rockfall occurred at Stein an der Traun in Southern Bavaria. According to this report, which is also the source of the remarkable image above, the block was the "size of a house" (probably something of an over-estimate), falling from the 15 m slope behind the building. There were four people in the house at the time, of which two were killed and two were injured. There is no obvious trigger for the failure.

It is of course too early to speculate about the causes of this sad accident, but in my experience the combination of an unfortunately-orientated discontinuity in the rock (often a joint) and cutting of the toe of the slope to create space for the building are often to blame. It will be interesting to find out what caused this accident.

Monday, 25 January 2010

Ongoing massive retaining wall failure in San Antonio, Texas, USA

Updated here

In the United States there is an interesting developing story about a large-scale retaining wall failure that is causing the collapse of a housing estate in San Antonio, Texas. The housing development, called the Rivermist subdivision, is the one shown in the Google Earth image below. It is still under construction. The marker shows the approximate location of the slide, as far as I can tell:


According to the media reports the slide started moving on about Friday is now sliding at about 4 inches (10 cm) per hour. There is a pretty clear video of the slide here (but note that this is definitely not a sinkhole as the report suggests).

The slide is far from trivial. These images, from this site, appear to have been taken a day or so ago:



More recent reports and images suggest that the slide has moved a great deal since these pictures were taken. More than 80 houses have been evacuated.

The area of the failure is, as far as I can tell, shown on this Google Earth image. This is two years old and was taken before the houses in this area were built. I have highlighted a rather peculiar feature that runs across the road on this image:

Interesting! The worst-affected houses in this area already look beyond recovery. I wonder how many more will be affected. Personally, I would not allow people to stand or to wander around anywhere near the toe of that wall, but maybe I am overly-cautious. This is clearly a tragedy for the families involved.

Updated here

Continuing concerns about the landslide dam at Attabad in Hunza, Pakistan

The Pamir Times continues to provide excellent coverage of the state of the landslide dam at Attabad in Hunza, N. Pakistan (see earlier posts here, here, here and here). First, they have posted another remarkable image of the landslide in action:

Second they have provided various reports on the state of the efforts to mitigate the rising water level and to reopen the highway. It is clear that the road remains blocked, which must be creating incredible hardship in the valley upstream of the slide. It appears that the Chinese engineers who are planning to reopen the road have only just arrived on site. Promisingly, the report states that another group of Chinese engineers have arrived to attempt to drain the lake, presumably working with the Pakistan Army. One can only hope that these are the same group that performed extraordinary Tangjiashan barrier lake drainage operation.

The magnitude of the task is well-illustrated by this report, also from the Pamir Times. The lake is now 8.9 metres long and is filling at a rate of about 3.6 feet (1.1 metres) per day. The Pakistan Army has promised to release the water within 45 days. The images show the nature of the problem even more clearly. This image, taken from a helicopter, shows the landslide dam deposit:


It is clear that the the lake is now rising quite rapidly towards the natural spillway. This image shows the size of the lake itself:

It is clear that the water is now starting to encroach onto farmland in a serious way. Finally, this image shows the state of the lake on the ground:

Sunday, 24 January 2010

New, remarkable animation video of the Po Shan Road landslide in Hong Kong

Thanks to David Kwok of the Geotechnical Engineering Office in Hong Kong for highlighting this video to me, and for the image below.

On 18th June 1972 Hong Kong suffered an extraordinary landslide at Po Shan Road, in the Mid-Levels area:


The slide, which had a volume of about 40,000 cubic metres, induced the collapse of two large buildings, killing 67 people and injuring a further 20. The landslide understandably caused considerably soul-searching within Hong Kong, not least because major stability problems at the site had been identified nine months before the failure. A recommendation of the Royal Commission (Hong Kong was a British dependency at that time of course) was that a government agency should be established to manage slopes. This led to the formation of what is now called the Geotechnical Engineering Office (GEO), which has worked tirelessly and with considerable success to reduce landslide losses in Hong Kong.

In 2007 as part of its 30th anniversary celebrations, GEO produced this brochure. Much more information about their work, and about slope issues in Hong Kong, can be found on their slope safety website. There is little doubt that the approach has been successful (and should be repeated elsewhere - for example in Taiwan). This graph, taken from the brochure above, shows the trends in landslide losses before and after the establishment of GEO:


However, through all of their work they have not forgotten the importance of the Po Shan Road incident. Recently, they have produced a short video describing the event. Most interesting is a set of animations of the occurrence of the landslide, including its impact on the buildings. This is an impressive and useful illustration of the destructive power of even comparatively small landslides when they occur in urbanised areas, and of the need to maintain vigilance.

The video is available from the GEO download site at the following address:

http://hkss.cedd.gov.hk/hkss/eng/download/poshan_incident_english.wmv

I have embedded the video here - you should be able to play it by clicking on the video screen below:

video

Do take a look - it is well worth it - and download the original from the GEO download site.

Thursday, 21 January 2010

More on the earthquake damage to Haiti port

Google have now released an even higher resolution set of satellite images of Port-au-Prince, this time collected using the IKONOS instrument. These images have an extraordinary level of resolution - about 15 cm - meaning that the images are spectacular. This allows us to get a better understanding of the liquefaction damage at the port, the subject of my earlier post. So, this is a close up of the damage to the main container wharf, with the crane in the water:


It is pretty clear from this that the dock support has failed and the deck has slipped into the water, taking the crane with it. That will not be at all easy to clear. Further along the wharf the scale of the liquefaction damage is pretty clear:


Perhaps the most surprising thing is the lack of effort going into re-opening the port. This is a good illustration of just how difficult it has been to get the aid operation underway in Haiti.

Reports of landslides in the rural earthquake-affected areas of Haiti

As the Haiti disaster moves from the rescue phase into stabilisation and the initial component of recovery, the media are running out of miracle survival stories and tales of looting. This means that they are now looking for other stories to tell, and in particular are starting to focus on both the plight of communities outside of Port-au-Prince and on detailed eye-witness accounts. Interestingly, this is starting to suggest that although the number of landslides is smaller than we might have expected, in the upland areas they have caused some problems. So for example, Sphere reports about the road to Jacmel, based upon the account of David Belle from the Cine Institute in Jacmel:

"Belle was finally able to drive from Port-au-Prince to Jacmel on Tuesday morning. "The drive over was very alarming. The town of Leogane is very badly hit; the small mountain communities are badly hit from landslides, homes taken out by dirt," he said. The film school plans on expanding its reporting to neighboring towns soon, to make sure they also get the help they need."

Meanwhile, Reuters reports about the upland areas close to the epicentre:

"Two hours drive west of Port-au-Prince, in the banana-growing hills where the epicenter of Haiti's earthquake tore chunks out of hillsides, hurled boulders and cracked roads, survivors with festering wounds sleep by their wrecked homes, unseen by aid workers.
..

By Tuesday, foreign medics were finally at work at a field hospital at the dirt-poor farming town of Leogane, by the quake epicenter. But nobody in the rock fall-plagued hilltop hamlets seems to know they are there, and the medics do not have the personnel to send teams out to look for patients.

Here at the core of the violent 7.0 magnitude quake, lush green hills have been ruptured and split. Locals have sawn through trees sticking out of fallen earth on the roads and they point to where truck drivers at a sand quarry were crushed when a giant chunk of it collapsed, redrawing the landscape. "I was inside bathing when it started shaking. I ran out and I saw that where there had been a hill there was empty space," said Seraphin Sonel, 14, who lives by the destroyed quarry.

Finally, AP has this image of the aftermath of rockfalls on the road to Jacnel:

Meanwhile, images are also starting to appear of the damage to shanty towns, such as this one, also from AP:

It is clear from this image that, unsurprisingly, some of the damage in these poor areas has been caused by shallow slips under the houses. This does not bode well for the rainy season.

Tuesday, 19 January 2010

Landslides from the Haiti earthquake - not many!

NASA has today released an image, partly obscured by cloud, of the upland area to the south of Port-au-Prince. This is the area that received the highest levels of shaking and, given the terrain, was most likely to have suffered slides. The image is available here:

http://photojournal.jpl.nasa.gov/catalog/PIA12495

and a comparison with an image taken in 2008 is here:

http://earthobservatory.nasa.gov/IOTD/view.php?id=42333


Note that this area suffered heavy rainfall in the 2008 hurricane season.

This is a close-up of a part of the area, close to, and just to the south of, the fault:

Clearly there are landslides visible, but the number is comparatively low and for the most part the slides are small. Note that NASA has, understandably, marked them as being "potential landslides". The terminology used here might be interpreted as indicating the possible existance of a landslide, rather than an area in which a future landslide may occur. The big unknown is whether there is a higher likelihood of sliding in the next heavy rainfall event. This can only be addressed with fieldwork, but I am not sure who will do this (maybe the USGS?). It is also essential that refugee camps are located away from potential mudflow and debris flow tracks (the worse structure to be in during a debris flow is a tent - they offer no protection, but the ropes and pegs make rapid escape difficult). I hope that this will be taken into account as the situation stabilises.

Finally, servir.net has produced an erosion potential map for the earthquake-affected area, which is available here:

Monday, 18 January 2010

Haiti Earthquake - video of the liquefaction damage to the main dockyard in Port-au-Prince

Following up on my post yesterday on the nature of the liquefaction damage to the main container port in Port-au-Prince, CNN has a video from the site. The video is here:

http://edition.cnn.com/2010/WORLD/americas/01/15/haiti.harbor/index.html


The recording appears to have been shot in this area (this is the post earthquake Google Geoeye imagery):


The video clearly shows the liquefaction features, the collapsed wharves and the damaged cranes and notes how much work is needed to reinstate all of this. There is no sign on the video that this work has started.

Very small earthquake kills seven in landslides in China

AFP is reporting that an earthquake struck Guanling, Zhenfeng and Zhenning districts in Guizhou province at 09:35 UT yesterday (Sunday), triggering two landslides that have killed seven people and left one person missing. Other news agencies are reporting that one of the landslides struck a boat on a river.

The peculiar thing about this is that the earthquake appears to have been very small indeed. The USGS does not have a report of the event on its global earthquake web site, meaning that it presumably did not reach their minimum size threshold. The Chinese Earthquake Administration report is here. This confirms the occurrence of the two landslides and the seven fatalities, but also notes that the earthquake had a magnitude of just 3.4 (local magnitude ML=4.0) and a depth of 7 km. This is an usually small earthquake to trigger landslides.

Potential landslides in California and the ongoing crisis in Hunza, Pakistan

Amongst the (appropriate) focus upon Haiti at the moment, we must of course remember that landslides do not stop elsewhere. So, here are two important ongoing stories:

1. Potential landslides in California this week:
The US National Weather Service has issued a series of warnings for very heavy rainfall in California this week, with the pontential for landslides, flash flood and debris flows. For example: ...Flash flooding and debris flows possible for recent burn areas In southwestern california late monday morning through monday night... .the first in a series of powerful pacific storms is expected to Move into southern california tonight into monday. Initially...rainfall Is expected to be fairly light...with rainfall coverage and intensities Expected to increase significantly by late monday morning through Monday evening. Rainfall rates between one half inch and three Quarters of an inch per hour will be possible monday afternoon Into monday evening. Local rates over one inch per hour will be possible Near thunderstorms and across favored south facing slopes...including The station and morris burn areas. Rainfall totals with this first Storm system through monday night are generally expected to range Between one and three inches across coastal and valley Areas...with three to five inches in the foothills and Mountains...except local amounts up to 6 inches possible across Favored south facing slopes. For the station burn area...rainfall Totals of 3 to 6 inches can be expected through monday night. Additional periods of heavy rain will be possible later in the Week as future storm systems move into the area...which may Require additional flash flood watches to be issued.

And:
...Wet weather to continue through much of the week... A series of storms...no fewer than six...are lined up across the Pacific all the way to asia. The first of these systems was moving Through the central california interior this afternoon...and will Spread light to moderate rain to the central and southern san Joaquin valley and the adjacent sierra foothills through tonight. Rain also will spread into the high deserts of kern county by late Afternoon and tonight. A second system is developing west of the California coast...and will move into california monday. It is This system that will spread occasional rain...heavy at Times...into the region. For the rest of the week...it is still too early to say when and How much each individual system will impact central california. But there will be periods of rain...heavy at times. Preliminary Indications are for another strong surge of moisture during the Day tuesday...again wednesday...and yet another thursday. Some Decrease in activity is now seen by the weekend...or at least Becoming more showery at times.

There is particular concern about the potential for debris flows in the Station Fire burn zone in the San Gabriel Mountains, as shown in this NASA satellite image:


However, as the La Conchita landslide showed five years ago, many areas in California have the potential for a landslides during very heavy periods of rainfall. It will be an interesting week.


2. The Attabad landslide in Hunza
The blockage of the Karakoram Highway by the Attabad landslide in northern Pakistan continues to cause major problems to that area. Meanwhile, the lake is continuing to fill. The Pamir Times continues to provide excellent coverage of this serious event. Over the weekend they posted a series of images of the slide. First, the lake:


Second the contact between the water and the dam sediment:


And finally the flow path once over-topping is achieved:

In some places the materials look worryingly fine grained and even muddy, but it is impossible to get a proper idea from these images. The report says that the water level is currently 10-15 m from the dam crest, which is of increasing concern. Various reports suggest that there is little government action to date in terms of creating a channel, although clearly this is unverified. It would be unwise to allow the water to flow over the top of the dam without being controlled. However, a report from the ICRC in Pakistan, dated 14th Jan, states that "According to the local government, a dam burst is very highly unlikely to occur."

Saturday, 16 January 2010

Earthquake-triggered liquefaction damage to the docks at Port-au-Prince in Haiti

UPDATE: there is a video of this site here.

One of the major impacts of large earthquakes in coastal areas can be the damage to dock facilities caused by liquefaction. The process of seismic liquefaction is well-explained elsewhere, so I won't reprise that here. However, if liquefaction does occur then high levels of damage can occur to the very facilities needed to bring aid into the affected area. Given the Google has provided wonderful free GeoEye imagery of the earthquake affected area, allowing a before and after analysis, I thought that it would be interesting to look at the state of the facilities:

It should be stated here that reports from before the earthquake suggested that the port was not in a great shape. The main port area appears to be this one (you can click on this image, or any other featured here, for a better view in a new window). This image was collected on 4th March 2008.


OK, so lets take a look at the main wharf with the warehouses in the centre of the image above.

Before:


After:


At first glance it may appear that there is little change, but look closely and you will see some extraordinary damage. I have annotated the image below to help:

The main impact is that the main wharf frontage has collapsed into the water and has now disappeared. The container unloading crane that sat on the dockside appears to be upright in the water (there appears to be another smaller crane in the water too). The white material covering much of the surface of the land around the docks is, I think, material released by liquefaction. This is best shown by this zoomed image, which shows the crane plus the collapsed wharves. If you look carefully you can see containers in the water in a number of places:


Clearly this part of the docks is not going to be usable until some considerable work has been completed. It is impossible to estimate from thid imagery how long this might take.

To the south the "before" image shows a jetty with ships alongside:


The "after" image shows that a large section of this jetty has collapsed:


The short section that is intact is probably not usable as the ground at the start of the jetty appears to have suffered liquefaction. There also appears to be a ramp structure here that has collapsed and is now floating:


It may be that the damaged ground here can be repaired temporarily to get the docks at least partially functional again.

Further up the coast to the north there is another area of wharf. In the before image this looks to be old and in poor condition, with many old vessels tied alongside. This could be a ship breaker, but nonetheless if intact it could offer some possibility to bring aid in:


Unfortunately there is no such luck. This wharf has also collapsed due to liquefaction. The warehouse here has also collapsed:



A little further to the north there is another jetty area, which appears to be quite new in the before imagery:

Unfortunately this one has also collapsed (as has part of the building in fact):

Round towards Carrefour there appears to be an oil unloading area. Unfortunately, as this before and after shows, this has also collapsed due to liquefaction, leading to a large oil leak:

Before:
After:

In summary, as the above images show, the port infrastructure is in a dire condition. The upshot is that at the moment most of the aid needs to be brought in by air, which is an extraordinary logistical challenge when there are 3 million people in need of help. A very high priority must be to get some of these docks reinstated. Some reports have suggested that this is already underway.

UPDATE: there is a video of this site here.

Comments and thoughts welcome.

Friday, 15 January 2010

Why the Haiti earthquake takes us into new territory for disaster response

The Haiti earthquake rightly continues to dominate the news around the world, with the situation on the ground looking increasingly desperate. The news media are already reporting on the growing frustration amongst the population about the lack of aid, not uncommon in large disasters actually, but probably magnified in this case. Although every rapid onset disaster is different, in a number of ways this event takes us into new territory. This will make the short, medium and long term mitigation of this event very difficult. These are the key issues as I see them:

1. Haiti has no army - and therefore effectively no national capacity to deal with the immediate aftermath
In recent years we have seen large earthquakes in other less developed countries - Kashmir in 2005 and Sichuan in 2008 spring to mind. In both cases, as is generally the case, the immediate response is framed by the national army, who undertake the initial rescue and logistics operations, followed quickly by the international organisations and agencies. Once the immediate rescue phase is over the army usually plays a key role in co-ordinating the response, in particular with the logistics of getting medical aid, food, water and sanitation to the people who need it. Haiti has no army - the military was demobilised in 1995. This is a key factor in the lack of coordination that you can see in the television pictures.

2. The earthquake struck the capital city
Point 1 is hugely exacerbated by what was in effect a direct hit on the capital city. It is clear that many government buildings collapsed and many civil servants were killed. In addition many of the key staff from international agencies were also killed. In a moment a huge component of the in-country organisational capacity was destroyed. This was not the case in other comparable disasters, where the national capability remained intact.

3. Population concentration is a key issue
The earthquake has struck a very densely populated urban area. Port-au-Prince is a natural amphitheatre in which almost all usable level ground has been quickly developed. Finding space to locate large refugee camps and medical facilities will be a great challenge.

4. Haiti is on a key hurricane track
I have used the NOAA Historic Hurricane Track viewer to produce the map below, which shows the hurricanes that have passed within 200 km of Port-au-Prince in the period since 1980:

You will probably have noticed that there are a worryingly high number. Hurricanes bring strong winds, storm surges and, more importantly in this context, very intense rainfall. A direct hit from a strong hurricane would be exceptionally hazardous for those in temporary camps, would test damaged drainage systems, and could trigger extensive landslides (especially mudflows) on slopes weakened by the shaking.

The hurricane season starts on 1st June (less than five months away). This is going to place an extraordinary level of urgency on operations over the next few months. Unfortunately, the December forecasts (which do have a comparatively low level of skill) for the 2010 hurricane season suggest that we should expect an above average season:

"We foresee an above-average Atlantic basin tropical cyclone season in 2010 and anticipate an above-average probability of U.S. and Caribbean major hurricane landfall." For Haiti, the Caribbean and C. America Landfalling Hurricane Probability Project estimates that there is a 49% probability of a named storm tracking within 50 miles.

Synthesis
Given the above, I suspect that there will be a need for the international community to direct considerable resources into Haiti over a prolonged period. I wonder whether this will be possible as the story fades from the news screens, and given the pressure on western government budgets at present.

Landslide potential in the aftermath of the earthquake
Finally, a word on the future landslide potential. In a nutshell we just don't know what will happen and, whiteout proper field investigations, including mapping and modelling, there is no way to estimate the likely future effects. It may be that there will be a large number of slides in the first very heavy (probably hurricane-induced) rainfall event, or it could be that there will be no more than usual. Note here in recent years these have killed hundreds and even thousands of people, but mostly in the area around Gonaives, not Port-au-Prince. The image below, from here, shows a before and after IKONOS satellite image of Gonaives in the aftermath of Tropical Storm Jeanne in 2004. The mud was released by landslides on the hills above the town. Over 3000 people died.

I must stress here that at present we have no way of knowing whether this could happen in the earthquake affected area or not. This requires attention with some urgency.

Thursday, 14 January 2010

Google Earth imagery of the Haiti Earthquake

With admirable speed Google have released two high quality images of the earthquake affected areas:

http://www.gearthblog.com/blog/archives/2010/01/imagery_layer_for_haiti_earthquake.html

At the moment the imagery is just for the Carrefour and Port-au-Prince areas (i.e. not the rural upland areas), but it contains the first good news to emerge from the area. This is that on this imagery at least the number of landslides appears to be small and, perhaps most interestingly, there are few signs of slope failures under the shanty towns on the edge of the urban areas, which we feared could be the cause of substantial loss of life.

There are some interesting other aspects of the damage, not least a large oil spill leak into the sea:

The main roads appear to be mostly open.

If landslides have occurred (and this would be the norm for an earthquake of this size) then they are most likely to be in the hills outside the imagery area. However, it could be that the slopes are so denuded and stripped, and prone to hurricane rainfall so frequently, that the material that would normally be available for sliding has already been stripped off.