22 May 2009

Sinorock conference report – the Wenchuan Earthquake session

Posted by Dave Petley

Yesterday there was a special session of the Sinorock Conference in Hong Kong, focused on the Wenchuan Earthquake. This involved six presentations and then a lively round-table discussion. This is a brief review.

The first talk was given by colleagues from Chengdu University of Technology on the topic of the landslides triggered by the earthquake. This was visually spectacular, emphasising the scale, density and impact of the landslides that were triggered. The CDUT team have completed an impressive amount of analysis of the data that they have collected now. An interesting addition was some small scale shaking table tests to investigate whether the morphologies that they observed can be replicated in the laboratory – the conclusion was that they can – and the videos made good viewing, although great care is needed in the interpretations due to the well-known problems of scaling the stresses and displacements.

Next up was Li Yong, also from CDUT, who talked about the mechanics of the earthquake itself. Interestingly, he highlighted the work that he and his co-workers had done with my colleague at Durham Dr Alex Densmore, in particular noting that their 2007 paper on the Longmenshan fault system had highlighted before the earthquake the threat posed by this active fault system. Dating of the sediments excavated from trenches along the fault suggest that the last major rupture was about 950 years before the present. However, his final note was quite chilling – this is that there are three large parallel faults in the Longmenshan system, but only two ruptured…

The third presentation was by the ever-impressive Chen Zuyu, who gave a wonderful talk about the mitigation of the Tangjiashan barrier lake, of which most readers will be very familiar. Prof. Chen focussed on the pressures that were on them during the successful drainage operation, the unanswerable questions that they faced, and the scientific data that they collected during the breaching. They measured the discharge as the breach proceeded and have a paper in with the ASCE at the moment. He pledged to release the data once the paper is published – so please can the ASCE make a quick decision!

An interesting issue that he highlighted was the state of Tangjiashan now. How noted that a debris flow blocked the river again on 24th September, forcing a 7 m rise in the lake level. The channel has now been widened and deepened ahead of the monsoon (this is what they were doing when I was there). More worrying, he noted that above the back scarp of the landslide scar there are tension cracks. Monitoring suggested very substantial amounts of ongoing movement, but this has now been terminated due to a lack of funds. This must be a very serious concern given the impending monsoon.

Finally before the break, He Chuan reviewed damage to lifeline engineering. Most of this went over my head, but I was interested in the fact that they examined the behaviour of rockslope reinforcement and retaining walls. In the case of rockslope reinforcement they noted that rock anchors and grids worked much better than mesh and shotcrete (not a surprise actually). The case of retaining walls they noted that walls on convex bends in the road performed poorly, whilst those on straight sections did OK. I am not sure whether this is because of the underlying materials (maybe walls on bends are holding up large colluvial bodies) or something to do with the dynamics of the earthquake shaking, or something else (presence of water?), but it is an intriguing result.

After coffee there were two brief presentations. The first was by Alexander Strom, who compared the impact of the earthquake in China with a potential event in the Tien Shan, noting that the presence of similar earthquake landforms (fault scarps, offset terraces, large landslides) suggest that an area that is considered to be of low seismic hazard may actually be quite risky. The second was by Xu Wenjie, who considered the ways in which the Xiajiapiao landslide blocked a river, and the ways in which it was breached. The fact that 14 tonnes of explosives were used to drain it sticks in the mind. I was also good to see some outcomes of modelling of the slides, but this did highlight the concerns about using models with rigid blocks to simulate materials that can fracture and fragment. The use of rigid blocks generates very high local stresses that cause ballistic behaviour of debris within the lodel. This does not mean that the the landslide did this (to be fair the author did not suggest that it did, but the audience might misinterpret the plots).

The discussion focused upon the availability of seismic data (apparently it has now been released in a book – I will try to get a copy), models of landslide initiation, topographic amplification, and the ways in which we understand the behaviour of coarse materials in landslide dams. It was all pretty interesting and there was some disagreement, which was good.

Overall I learnt a great deal but I am increasingly concerned that the opportunities presented by the event to gain scientific knowledge about landslides might be getting away from us. This is not the fault of our Chinese academic colleagues – who are impressively proactive and skilled – more of the global scientific system. This is very frustrating!

Incidentally, the Great Firewall of China has now blocked access to blogger across the whole of China. This is a great shame, and a retrograde step in terms of sharing knowledge.