9 December 2011

AGU days 2 and 3: some notable talks

Posted by Dave Petley

The sharpest amongst you will have noticed that I have fallen badly behind with blogging.  Unfortunately I have found myself so busy from dawn into the evening that reserving time to blog has been a struggle.

So to try to address that here are some notable talks that I have attended in the last two days:

Thorne Lay gave a really excellent keynote talk on great earthquake ruptures in the age of seismo-geodesy – an event that was perhaps slightly marred by extreme over-crowding in the room, and some slightly heavy-handed treatment by the convention centre staff of those without a seat.  Thorne’s thesis was that the combination of high-resolution (both in terms of time and space) geodetic measurements and excellent seismological data is starting to allow us to really get an understanding of the nature of great (i.e. M>8 earthquakes).  Nonetheless, he emphasised that every great earthquake of the last decade (and he noted that this has been a period of a high rate of large events) has generated a scientific surprise.  I guess we will start to really understand earthquakes when this is no longer the case.  The component that I found most interesting was a set of observations about coupled earthquakes – i.e. events where one seismic shock triggers another, sometimes with quite a different fault mechanism.  Thereseem to be more of these than one might expect.

One of the first questions after the presentation was whether Thorne felt that there was a possibility of a Magnitude 9 earthquake in the Himalayas.  The answer was short – yes.  This is interesting in the context of this news report about work being undertaken by Roger Bilham in this region – the paper will be presented tomorrow.  The message is clear about N. India and Pakistan:

…The zone likely to rupture when a quake eventually happens could be 200 kilometres wide, rather than about 80 kilometres, as was previously thought. The zone would encompass the Kashmir Valley – including the city of Srinigar, home to some 1.5 million people. If slippage occurs over a length of 300 kilometres, as is possible, a megaquake of magnitude 9 is the likely result. Given building codes and population in the region, that could mean a death toll of 300,000 people. What Bilham can’t predict from his GPS results is when such a disaster might happen.

It is interesting that Bilham is emphasising rightly the threat posed by landslides in such an earthquake:

Worse still, Bilham fears that such a major quake is likely to trigger landslides that could dam the Jhelum River, which drains from the Indian-controlled Kashmir Valley into Pakistan. That could put the Kashmir Valley under water within three months – and would also threaten disastrous flooding in Pakistan if the waters were released too quickly.

All the warning signs are there for a major catastrophe in the Himalayas at some unknown point in the future.  I wonder whether we are heeding the warnings?

Later in the day it was on to a lecture by Tad Pfeffer, less well-attended (but far from empty), on delivering predictions of sea level change.  He rather nicely pulled together the a range of estimates of future sea level rise associated with melting land ice (not from other sources).  There are several things to note here.  First that Antarctica, Greenland and alpine type glaciers are all losing ice mass quite rapidly, and that this is contributing to sea level rise.  Second, he noted that whilst the last IPCC report rather hedged on this component of sea level rise, simply because the data was not good enough at that time, the picture is now much clearer due to improved science.  Interestingly though, the resulting estimates ae very similar to the very first IPCC estimates from 20 years ago, which shows that actually people knew what they were doing back then.  Third, he noted that planners want shorter term estimates than we normally provide as scientists, but that it is incumbent upon us to provide that data.

Tad also highlighted the really shocking imagery and videos of ice loss that have been collected by James Balog and his colleagues, which are online here.  Unfortunately I cannot embed the video, but if you do nothing else today take a look at this clip – it is only a minute and is narrated by Tad.  I challenge you not to be shocked by just how rapid the rate of ice loss actually is on this clip.  Sadly this is being repeated in many other glaciers.  In my view ice loss is the canary warning us about the magnitude of the climate change problem. I should say that at this meeting, I have not seen one presentation or one poster that suggests that anthropogenic climate change is a reality.  There are about 20,000 scientists at this meeting.

Yesterday, there was another landslide session, which was mostly good.  One of the invited lectures, about the SAFELAND project, was somewhat marred by the fact that neither of the two authors turned up.  A colleague, Carl Fredrik Forsberg gave the presentation in their stead, and did an admirable job in the hardest of circumatnces.  It was clear that he was deeply uncomfortable though, and he could not answer questions about the analysis (for obvious reasons), which makes the whole process of giving a paper a mockery.  Frankly if authors are going to choose not to attend to present a prestigious invited paper then they should withdraw and allow someone else the opportunity.  There was so much good stuff that was given only as a poster that this is at best deeply regrettable.

On a more positive note, I would like to highlight a couple of really interesting presentations.  First Naoki Sakai presented an amazing paper on their large landslide model tests, using an artificial slope and a rainfall simulator.  This was essentially an investigation of the validity of the so-called Saito effect for landslide prediction – and he could show that it worked.  Amazing data were accompanied by videos of the landslides occurring.  Dr Sakai is keen to collaborate on this work, and their facility is genuinely unique.    Second, Brian Collins of the USGS presented an initial monitoring dataset for shallow landslides in the San Francisco bay area.  This work is in an early stage but looks likely to generate an amazing dataset.  Personally I’d monitor ground displacement as well hydrology and weather – they just might catch a collapse event, but that is a minor quibble.

At the end of the day yesterday I also attended a lecture by Michael Manga on the ways in which earthquakes promote changes in hydrology, even at quite large distances from the rupture.  These changes include increased river discharge, increased water levels in boreholes and increased spring discharges.  He argued strongly that these changes are mainly due to increased vertical permeability caused by the earthquake shaking.  He then went on to look at the LUSI / Sidoarjo mud volcano in Indonesia (if you do not know about this then take a look at the Wikipedia page – it is incredible).  In the Wikipedia image below, the ruined houses were all destroyed by the flow; the steam in the centre is the area of the vent, and the levees have all been built to try to hold in the erupted mud.

Controversy has raged about whether this was triggered by a gas well that was being drilled near to the site when the eruption started or by a moderately large earthquake about 240 km away.  Manga argued strongly that the cause was the drilling, and in my view this is the most probable explanation (though that is contested by others).    Interesting, he also estimated the likely duration of the eruption, and concluded that there is only a 50% chance that it will stop in 50 years!

I should add that this summary covers about 20% of the presentations that I have seen – there is just too much going on!