15 September 2015
Live landslide monitoring data from Nepal
One of the great uncertainties in the aftermath of the April 2015 earthquake in Nepal was the likely behaviour of damaged slopes. It is well documented from other earthquakes, such as the 2005 Kashmir (Pakistan) earthquake and the 2008 Wenchuan (Sichuan) earthquake that many landslides occur after the main earthquake. Unfortunately our ability to forecast numbers of events and their locations is poor, mainly because of both a lack of understanding of the mechanics of these earthquake damaged slopes and of course the huge uncertainties about the likelihood of the meteorological events that trigger them. This is an image from after the 2005 Pakistan earthquake, showing just some of the cracked slopes:
Amazingly, most of the slopes above have not failed since the earthquake and indeed most of the cracks have “healed” up and are now invisible.
My former colleagues at Durham, led by Dr Nick Rosser, have set up a project in Nepal to monitor slopes damaged by the 2015 earthquakes, funded by the UK research council NERC. Working with a range of partners in Nepal, they have installed landslide monitoring instruments in ten slopes in the earthquake affected areas. Interestingly, they are streaming the data live to the web via a graphing package. You can view the data on each of the instruments on their website. As the instruments are newly installed the data is mostly noise at present, though there are some interesting daily patterns that might or might not be important. It will be worth keeping an eye on this data in the coming months, especially if Nepal were to suffer a late monsoon rainfall event.
On their blog, the team have provided a description of the aims of the landslide monitoring project:
Our overall aim is to determine how hillslopes damaged during the 2015 Nepal EQS progressively deform in response to stress perturbations resulting from precipitation and continuing seismicity. Our proposed study site is the Upper Bhote Koshi (UBK), in Sindupalchowk District. The UBK was damaged during eqG, was < 10 km from eqK, and satellite and aerial imagery shows considerable widespread damage to slopes as a function of the close by (< 5 km) eqD. The critical Arniko Highway to Tibet, several major HEP plants, plus extensive roadside settlements, are all within the UBK. Our initial assessments from imagery indicate that EQ-triggered landsliding from eqK is characterised by small-scale but widespread rockfalls, and from eqD large-scale rockslope failures from ridgelines and reactivation of existing deeper-seated landslides, each with a clear directional tendency relative to the three epicentres). Colleagues in the UBK report extensively-cracked valley slopes, as seen in reconnaissance aerial photography from colleagues at USGS/USAID. Our data will most importantly quantify the relationship between stress perturbations and landsliding, but will also provide ground-truth for remotely-sensed deformation.