Friday, January 25, 2008

Rock avalanche in New Zealand and Sodbury landslip in the UK

In the last week or so there have been two contrasting examples of landslides that demonstrate the poor relationship between size and impact.

First, there was this rather interesting and quite large rock avalanche in Aoraki-Mount Cook National Park, New Zealand:

The 13th January 2008 Aoraki-Mt Cook landslide
Picture from: New Zealand Herald

There is a really nice set of photos in Flash form on the New Zealand Herald website here and a good video overview of the landslide here. This is quite a large slide, with a scar (source) size of about 120 metres high, 80 m wide and 30 m deep, giving a volume of about 300,000 metres cubed by my calculation (or about 1 million metres cubed by some reports). It ran out over a glaciaer for a distance of about 2 km, probably because the slopes are steep and the debris has flowed across ice. However, no-one was killed or injured as this happened in an uninhabited area. Interestringly, there was no obvious trigger to the event.

On the other hand, rail services in SW England have been seriously disrupted by this landslide:


The 24th January 2008 Chipping Sodbury landslide
Picture from: Sky News

In comparison with the Aoraki-Mount Cook this landslide is tiny - probably jest a few hundred cubic metres. However, the landslide, which occurred as a result of the prolonged heavy rain that the UK is suffering at the moment, occurred on the SW mainline of the UK rail network. Although it didn't reach the tracks, it has damaged signal equipment. As a result, train services were cancelled for a day and will be heavily delayed and disrupted for a while. I expect that the whole of the embankment will now need to be inspected.

Tuesday, January 22, 2008

2007 fatal landslide map

Here is the map of fatal landslide locations in 2007. Each black dot represents a single fatal landslide.

2007 fatal landslide map.
Click
on the map for a large version. Copyright Prof. David Petley

A number of clusters are clearly evident:
1. The Himalayan Arc: this cluster of landslides extends along the whole of the southern edge of the Himalayan Arc and down into Bangladesh. Landslides here occur mostly in the period of the summer monsoon. This will be the subject of a future post.
2. Central China and N. Vietnam. There s a clear cluster extending through central China ans into the northern part of Vietnam. I am not quite sure why this cluster is preseny. I will post on this area again.
3. The Philippines: 2008 saw a large number of landslides occurring within the Philippines. Tropical cyclones may be a key factor here.

There are also smaller cluster is:
a. Java
b. SW India
c. Haiti and the Dominican Republic
d. Central America through Mexico to Colombia.

Interestingly of course the Asian monsoon was strong in 2007 but tropical cyclone activity was weak. Thus the occurrence of landslides in areas affected by tropical cyclones is probably below average.

Wednesday, January 9, 2008

UPDATED: January 2008 landslide map

Here is the fatal landslide map so far for January 2008, as per 16th January. So far I have recorded 13 fatal landslide events with 91 fatalities. I will update this through the month as more events occur. The average number of fatalities in January for the last five years is 159. Four of the recorded landslides have been failures into quarries, which seems to be a particular problem into Vietnam at the moment. I wonder if this is a sign that the rampant global demand for resources is allowing the opening of increased numbers of small, dangerous sites?


January 2008 landslide fatality map: Click on the map for a full sized view. Note that there are two points almost superrimposed in Vietnam.

Thursday, January 3, 2008

Landslides in Northwest USA

The New York Times today carries an interesting article here about destructive floods in Washington State resulting from a large storm in December. For me the interesting thing is an argument that appears to be raging as to whether the floods have resulted from measures to protect salmon fisheries, which have purportedly led to blockages forming in river channels, or from landslides and erosion that have resulted from logging. The stakes have been raised by this Seattle Times image (the image is well worth a look - there is actually a better version on the New York Times web page here).

So what to make of this. First, there is plenty of evidence that logging increases landsliding and erosion dramatically. It is clear that this slope has been recently deforested, and thus there is a logical link. The adjecant forested slopes appear to be unaffected by this level of landslides. The landslides themselves are shallow, which is exactly the type that trees stop so well. The landslides have certainly reached the channel and thus might well be responsible for releasing sediment associated with the floods. Note though that some nearby slopes without trees have not suffered landslides like this.

Whatever the cause, which is hard to determine from a single image, the need to change forestry practise in this area seems to me to be clear.