20 January 2011

A review of the Brazil landslides – Nova Friburgo

Posted by dr-dave

The lack of posts on this blog will not have passed unnoticed I am sure, and I can only apologise.  Unfortunately, my day job is insanely busy at the moment, and I have no time for research, let alone blog posts.  This will ease within a fortnight, if not before, so please be assured that the blog is not on the way to oblivion.

During my unfortunate lull in posts an extraordinary landslide disaster has of course unfolded in Brazil, killing at least 700 people, and possibly more than a thousand.  The trigger has been very heavy rainfall associated with the current La Nina conditions, although the underlying cause is likely to be a combination of people living in the wrong places and damage to the environment as a result of human activities.  The epicentre has been the three towns of Nova Friburgo, Teresopolis and Petropolis, located to the north of Rio de Janeiro.  So let’s take a look at Nova Friburgo.

This is the Google Earth perspective view (dated March 2009) of Nova Friburgo – it is interesting because the landscape is not particularly high, though it is steep, and there are surprisingly few fresh landslide scars:

Inevitably the news reports and photos have focussed on the damage caused by the disaster rather than on the causes, but these news images are helpful:

Source: AP

Source: AP

Source: AP

Source: Reuters

Source: Reuters

The news images tell us a fair amount about the landslides in Nova Friburgo at least.  First, there are many (perhaps not surprising).  Second, there appears to be two types of accidents occurring here.  Some of the losses have been caused by direct hots by the landslides themselves.  Most of these slides are long and narrow, extending most of the way up the slope.  In rainfall induced landslides this is characteristic of failures that involve static liquefaction, which leads to rapid failure and high rates of movement. This is supported by the second to last image, in which the runout distance of the debris at the foot of the slope is quite long.  It is unsurprising that the loss of life has been so high in areas affected by such slides.  The second type is indicated by the very first image, which suggests multiple flow-type landslides in close proximity.  Note here that the slides are both large and in some cases quite deep.  However, there is almost no debris accumulated at the toe of the slopes, suggesting that it has moved rapidly down the watercourse.  It is likely to have been this mechanism that created the devastating debris and mud flows that caused so much damage downstream.

There is also a rather impressive set of images of the disaster on this presentation (unfortunately I cannot find a way to embed authorstream files as yet – I am working on it!).