Weather in the earthquake affected areas

It is now clear that very substantial numbers of landslides have been triggered by the earthquake in China. Continued landslide activity will be a very major problem during periods of heavy rain. Unfortunately the forecast is not good for the next few days. Xinhua reports:

"Three rains are forecast to sweep earthquake-hit southwest China's Sichuan, southern part of Gansu and Shannxi provinces and Chongqing Municipality in the next seven days, said the China Meteorological Administration on Tuesday. Light and moderate rain would hit much of the regions on Tuesday and Wednesday. The southern part of Sichuan jolted severely in the earthquake measuring 7.8 on the Richter scale Monday was likely to see heavy rainfalls and even thunderstorms, possibly increasing difficulties of the disaster relief work, said the national meteorological observatory. The latest weather report added that another two moderate showers were forecast in the regions on Saturday and next Monday. Yang Guiming, chief forecaster of the Central Meteorological Station (CMS) warned that lingering rains and thunderstorms may trigger geological disasters like landslides as rocks and the earth were pretty loose after the earthquake."

This is supported by the forecast from the National Meteorological Agency for Tuesday:

Chengdu

Thundershower

SoftBreeze 3mph

75°F

Moderate rain

SoftBreeze 3mph

61°F

Unfortunately, thunderstorms are bad for landslides. This area of wet weather is quite clear on this satellite image, from Yahoo.

Landslides from the 12th May 2008 Sichuan Earthquake

12th May @ 07:27 UT: News reports this morning are suggesting that there has been a large, shallow earthquake in Sichuan Province, China. Initial estimates from the USGS are that this was a M = 7.5 (NB this has now been corrected to MW=7.9) event located at just 10 km, although clearly this will need updating. If so, it is reasonable to assume that this earthquake will have triggered large numbers of landslides as this is a very landslide-prone area. The Google Earth image below shows the earthquake affected region with the initial epicentre marked. The landscape is highly mountainous.

Fig. 1: Google Earth image of the earthquake-affected area showing the location of the epicentre. If the earthquake is shallow and large, as initial indications suggest, then the number of landslides triggered in this region is likely to be very large.

The image in Fig. 2 shows the population density of China, based upon the GPW dataset available here. The black circle indicates the approximate location of the earthquake. It appears that this is a quite densely inhabited area. If the initial reports on this earthquake are correct then its impact could be fearsome.

Fig. 2: GPW map of the population density of China. The black circle indicates the area affected by the earthquake. The population density is clearly high in this region

Update 1 08:06: Inevitably there is not much news as yet, but the reports are now confirming that the magnitude is about M = 7.5. For info, the map below (Fig 3) shows the fatal landslides that we recorded in 2007 in China and its environs (Note not the landslides from this earthquake). I post this map to show how Chinese landslides are focused on the area affected by this earthquake, which suggests that landslides will have been a significant impact this time as well.

Fig. 3: Map of fatal landslides recorded in 2007 in China. Note the concentration of landslides in the epicentral zone of this earthquake, suggesting that many landslides will have been triggered by this event.

Update 2 11:40: It is now becoming apparent that this is a very substantial event, with large numbers of fatalities. Given the size (MW=7.8) and depth (10 km), the number of landslides is likely to be large. For info, the area affected by landslides can be estimated using the classic graphs compiled by Dave Keefer back in the 1980's. Figure 4 (taken from my lecture notes) shows the empirical relationship between earthquake magnitude and area affected by landslides. Given the mountainous terrain and the sensitivity of the environment to landslides, plus the shallowness of the source, it is likely that this will be at the upper boundary of the curve - i.e. that the area affected will exceed 50,000 square kilometres.

Fig. 4: Keefer (1984) graph of the relationship between earthquake magnitude and area affected by landslides. The graph indicates the likely area affected by landslides in this earthquake.

The graph in Fig.5, also from Keefer (1984), indicates how far from the epicentral zone landslides are likely to occur. For an earthquake of this magnitude this can be several hundreds of kilometres.

Fig. 5: Keefer (1984) graph of the relationship between earthquake magnitude and maximum distance to landslides that it triggered. The graph indicates that some landslides may have occurred over 200 km from the epicentre

Finally, the photograph in Fig. 6 shows landslides triggered by the 2005 Kashmir earthquake in Pakistan, which is an area of similar topography and level of development. The impact of these landslides was catastrophic. Note that the response should consider two elements:
1. Substantial numbers of people may have been buried in the landslides;
2. The landslides will impose a substantial constraint on the delivery of aid as they will have seriously damaged the transport infrastructure.
The fact that it is raining in the earthquake affected area makes this problem worse, especially as the wet season is only weeks away. This means that there will be a need to mobilise helicopters into the affected area as soon as possible.
There are more images of landslides associated with the Kashmir earthquake available here and here.

Fig. 6: Landslides triggered by the 2005 Kashmir earthquake. A similar level of landslides is likely in this earthquake.

Update 3 20:10: It is now becoming apparent that landslides are indeed a major issue in the earthquake affected areas. For example, Xinhua is reporting that:

"Monday's strong earthquake in southwest China's Sichuan Province have caused multiple landslides and collapses along railway lines near the provincial capital Chengdu, leaving 180 trains stranded on the rails. Thirty-one passenger and 149 cargo trains were stranded on the Baoji-Chengdu line, the Chengdu-Kunming line, the Chengdu-Chongqing line and their branch lines linking Chengdu with the rest of the country. At least 15 cases of landslides and collapses had so far been reported along rail tracks, with 34 railway stations on the Baoji-Chengdu Railway losing power supplies due to the earthquake, Wang Yongping, spokesman of the Ministry of Railways said Monday night."

I suspect that we have not seen any indication yet of the level of destruction in the upland areas around the epicentre, where the landslides are likely to have been triggered. Expect the death toll to increase significantly as information from these regions emerges.

Landslide sizes (numbers killed) around the world

One of the interesting aspects of our landslide fatality database is that it is possible to produce maps of the size of the impact of landslides (which in this case means the numbers of people killed) in each event around the world. I tried to do that for my EGU presentation, but my colleague Nick Rosser has produced a rather more attractive version (Fig. 1) showing the 2007 dataset.

Fig 1: Map of the size of the impact of landslides, as measured by the number of fatalities in each event, for 2007. Click on the map to open a larger version.

The variation in size is quite interesting. Note that in Europe, N. America and Japan the recorded landslides tend to be comparatively small, almost always resulting in less than five fatalities. South Asia and Central America on the other hand have a scatter of landslide sizes from small events through to the very large. China also has a high frequency of large landslides, but appears to have comparatively few small ones. Africa has surprisingly few recorded fatal landslides, but those that do occur tend to be quite large.

So what to make of this? Well, first and foremost it shows the different levels of mitigation of landslides around the world. In the more developed world (i.e. Europe and N. America), large landslides tend to be very well mitigated (i.e. engineering measures have been used to stabilise them). The few fatal landslides that do occur tend to be small and in fairly remote locations, resulting in low fatality counts. On the other hand, in South and South-East Asia, mitigation is much less likely and there is a high density of people in affected areas, meaning that fatality counts are much higher. There might also be an effect of the both the tectonic setting (i.e. the high mountains of S. Asia in particular might mean that the fatal landslides are larger) and the climate (i.e. torrential monsoon rainfall might tend to trigger very energetic landslides, which are more likely to kill, perhaps). Finally, the data also perhaps show that the current database still misses smaller landslide events in China and in particular in Africa. We are trying to improve this situation.

As ever your comments are really welcome.

Black Ven landslide in Dorset

NOTE: see follow-up blog here, outlining why this landslide is not as large or significant as was initially reported.

-------------------
Update: ITV have a low res video of the landslide site here. This confirms the general location and shows that the failure is indeed large.
-------------------
Update 2: The BBC now have some good aerial footage of the landslide here. It is slightly intriguing as it appears that the slide is wide but not very long. From a preliminary view there should be real concern about what might happen here in the next heavy rainfall event.
-------------------
Update 3: Based upon the BBC footage, the failure area appears to be the lower part of the slope highlighted by the red box on Fig. 3 below. The landslide appears to be a 400 m wide section of this lower cliff, extending about 100 m inland. This is a really intriguing development. It will be interesting to see how the sections upslope now behave as they have lost their toe support. This will be a fascinating landslide to watch over the next few months. I suspect that we have not seen the last movement at this site as yet.
-------------------

Original text:
Timed to perfection to coincide with the launch of the Durham University Institute of Hazard and Risk Research tomorrow, news reports indicate that there as another major landslide in the UK overnight. This time the failure occurred on the well-known Black Ven landslide system in Dorset, the site of a great deal of landslide research over the last few years.

Black Ven is shown by the arrow on the Google Earth image to the left, located on the south coast of England, just east of the town of Lyme Regis in Dorset. It is a big, active landslide complex consisting of an upper section in a glauconitic sandstone known as Foxmould. The lower part consists mainly of a series of shales and marls with thin limestone bands that form benches in the landscape. This is quite well shown in the cross section in Fig. 1 (Koh 1992). NB Click on any image for a better view.

Figure 1: Indicative cross section of Black Ven landslide in Dorset, showing the permeable sandstones of the Foxmould overlying the impermeable shales and marls (Koh 1992). NB Click on any image for a better view.

The landslide is clearly shown in the picture from the 1980's shown in Fig. 2 (Brunsden et al. 1988). The yellow rocks at the top is the Foxmould, and the grey rocks lower down are the shales. The landslide usually activates because of a deep seated rotational failure in the Foxmould. The rotated blocks then break-down over time to form flows that cross the benches below, eventually flowing out into the sea.

Figure 2: Photograph of Black Ven landslide in Dorset, showing the permeable yellow sandstones of the Foxmould overlying the impermeable shales and marls (Brunsden et al. 1988). Rotational failures in the Foxmold transition to flows that run across the lower benches. NB Click on any image for a better view.

The reports suggest that the landslide occurred in the area known as the Spittles, although I have yet to see any pictures to confirm this. Koh et al. (1992) indicated the location of the Spittles on the sketch map of the landslide shown to the left. This is just to the left of the landslides shown in the photograph in Fig. 2. This is an area that has been showing very considerable slope distress for many years, with lots of evidence of deformation as the toe of the landslide slips away. This area is quite clear on the Google Earth Imagery in Fig. 3 - I have put a faint red box around the general area usually known as the Spittles. Zooming into the upper part of the area delineated in red, the Google Earth image very clearly shows the active deformation that has characterised this zone for a while, with abundant evidence of tension cracks and small scarps (Fig. 4). These have been developing over a long period, so failure should not be a surprise. The landslide is large - the BBC report that it is about 400 m, which appears to be the width. If so this is a notable failure.

Figure 3: Google Earth image of the Black Ven landslide. The red area highlights the zone known as the Spittles, which appears to be the area that has failed in early May 2008. Update: The BBC film footage suggests that the failure has occurred in the lower left side (i.e. SE corner) of this box. NB Click on any image for a better view.

Figure 4: Google Earth image of the Black Ven landslide zoomed in on the upper part of the Spittles. Note the multiple tension cracks and scarps indicating that failure was likely. NB Click on any image for a better view.

The timing of the failure is quite interesting - it appears that the UK experiencing an exceptional level of landslide activity at the moment (e.g. Cayton Bay). Interestingly, amateur weather stations do not suggest that the weather of late has been exceptionally wet in this area - for example, an amateur station just along the coast at Dawlish suggests that the rainfall total in April was 64.6 mm, which is lower than the totals for both 2006 and 2007. The first few days of May have been essentially dry. The only notable rainfall event of late occurred on 29th April, when 18.5 mm of rain fell. Although high, this is not really exceptional. Rainfall in 2007 was recorded as 916 mm against a long term average of 850 mm (i.e. about 8% above the mean value), but this was below average for the last four months of the year. This pattern continued in January and February 2008. What is interesting is that March was unusually wet (103.5 mm against an average value of 68.1 mm - i.e. 93% above the long term average). I assume that the movement we are seeing is a delayed response to this rainfall period.

I would really welcome any comments, especially from anyone near to the site. Please email me or leave a comment below.

References:
Koh, A. 1992. Black Ven. In Allison, R.J. The Coastal Landforms of West Dorset. Geologists Association, Essex, 67-79.
Brunsden, D., Gardner, R., Goudie, A. and Jones, D. 1988. Land Shapes. Channel 4.

The Taihape Landslide monitoring project

Last summer I was lucky enough to spend a couple of days with my friends from GNS Science who are undertaking a monitoring project on the Taihape landslide in New Zealand (see Google Earth image right). Taihape, which is located as -39.68°, 175.80°, is described on Wikipedia as "a small, picturesque town near the middle of the North Island . It services a large rural community and lies on the main north-south route through the centre of the North Island." It has a population of about 2000 people and is in general a very nice little place. Unfortunately, the area around the town is somewhat landslide prone (actually, mapping landslides here would probably involve trying to find those few areas that have not been affected by mass movements!).

Perhaps not surprisingly, part of the town is built on a landslide, which moves (slowly) during periods of wet weather. The landslide is certainly not new - dating suggests that it first moved over 1,800 years ago. Nonetheless, the level of recent movement, although slow and gentle, can cause some damage to buildings and roads in the affected area of the town. The landslide is a translational block slide with a shear surface at about 25 - 35 m depth. It's pretty big - over 200 houses and a school are sited on it. Movement appears to be associated primarily with rainfall. Details of the landslide are available here.

Hence for the last year or so the hugely impressive chaps from Geonet have been working with GNS Science to monitor the movement of the landslide. In my view this is about the most impressive monitoring project I have seen to date. The data are collected by an automated laser monitoring system that sights onto prisms located on the landslide. This is supplemented by two rain gauges and four borehole piezometers. Excitingly, the data are available in graph form online in real time, so that anyone can see how the landslide is behaving. The data are available here, which uses a Java graphing tool. It takes a few minutes to load, but stick with it as it is fantastic. The results are great fun. You can plot side by side the movement recorded by different prisms and relate this to the groundwater level and the amount of rainfall. Although there hasn't been much rainfall since the monitoring started, it is already possible to see the link between rainfall, groundwater and movement. It is really worth playing with the graphing function.

The strength of this project is in its thoroughness and its transparency. The monitoring system being used is about as good as is possible using current technology, and the team are taking great care to quantify and deal with the potential errors. The fact that the results are online is hugely impressive. I can only wish that more landslide movement data was available in this way.

April 2008 landslide map

The map below shows the distribution of fatal landslides for April 2008.

The statistics are:
Number of fatal landslides: 12
Number of fatalities: 33

As I predicted in March, April was once again very substantially below the average for 2003-2007, which is 96 fatalities per year. The distribution is very random, with absolutely no clear pattern. Two fatal landslides occurred in the Himalayan Arc, which is a sign of things to come over the next few months. La Nina conditions do still prevail, but the event continues to weaken. Typically the number of landslides starts to increase around early May. This is shown by the cumulative totals - for example, at the end of April 2004 I had recorded just 339 fatalities for the year to date (this years total is even lower - 301), but by the end of May 2004 this had jumped to 1136. It could be an interesting month!

April 2008 fatal landslide locations (click for a larger version)

Some patterns are starting to emerge in the annual map. Note in particular the emerging clusters in the Himalayas, western Central America, the Philippines and Indonesia. These are the normal locations for fatal landslide clusters. The cluster in the Himalayas is focused on the west side at the moment as this area gets heavy rain in the early part of the year. It will spread eastwards over the next three months.

2008 fatal landslide locations up to the end of April (click for a larger version)

Serious landslide in Indian Kashmir - at least 20 people killed

Updated: It appears that May has started with a terrible rockfall in Indian Kashmir (Figure 1). Reports are still rather sketchy, but it seems that the landslide occurred in the Chatroo area Kadiyah,which is about 220 kilometers from Jammu. The rockfall which happened on a road linking Kishtwar and Islamabad, seems to have been caused by blasting. Greater Kashmir online reports:

Eyewitnesses said that the incident took place when some of the labourers were busy in blasting a hanging part of the mountain for widening of the highway this morning. Two tippers, carrying labourers working with the Border Roads Organisation, crossed the same spot from opposite sides when a large portion of the mountain slid down, burying the two tippers and men under several feet of earth.

Figure 1: Greater Kashmir online image of the 1st May landslide, showing one of the buried vehicles. The chances of survival appear to be very limited.

Reports vary on the number of victims, but at least 20 and maybe as many as 30 seems to be the ballpark figure. The outlook doesn't look promising for the victims judging by Figure 1. Understandably, recovery is proving difficult - Greater Kashmir online reported:

“We need to blast the large stones that cannot be moved otherwise,” Kishtwar’s additional deputy commissioner, Riyaz Ahmad Zargar, told Greater Kashmir. “It is difficult to tell how many more are trapped. Fears are that the number could be in excess of 20. Exact details of the causalities can be ascertained only when the rescue operation ends." Zargar said that screams could be heard from under the rocks and efforts were being made to rescue men who could still be alive under the debris.

In an earlier post I highlighted the danger of mountain roads to buses. In a sense these trucks were being used for the same purposes, once again with tragic consequences. Sadly of course with good supervision during construction such rockfalls can be mostly avoided.