Predicting the Future of Unstable Slopes / / by Jim Whiteley & Arnaud Watlet

on

Jim Whiteley and Arnaud Watlet are geophysicists working on the Multihazards and Resilience challenge area at Keyworth, focusing on using geophysical approaches to better understand landslide hazards. Along with other members of BGS, and partners at the Royal Observatory of Belgium, they were successful in winning a £55k grant from the Natural Environmental Research Council (NERC) in the form of a loan of seismometers, which will be used to monitor an unstable slope in the UK. Here, they first explain why geophysics can be useful in monitoring landslides, and secondly how they went about deploying their seismometers to a BGS site in North Yorkshire...
The Hollin Hill Landslide Observatory

In Part 1 of this post on landslide monitoring, we’re going to look at why the BGS uses geophysics for monitoring landslides, and in Part 2 we’ll see some recent developments that have been happening at the Hollin Hill Landslide Observatory in North Yorkshire. Here in the Multi-Hazards and Resilience Challenge Area, we’re pretty good at knowing where landslides are (see the fantastically comprehensive National Landslide Database), and we can work out why they fail, (in the UK it has a lot to do with the weather), but we still sometimes have difficulty in working out when exactly this might happen. This matters because landslides, even modest ones, affect when we travel, where we live, and how we manage risk. To predict when a landslide might happen, we need to understand the geology of an area, and we need to know how much rain falls on the slope and when; but we also need to know how much of that water is getting in to the ground, how quickly it is getting there, where it is going, and ultimately, how much this affects the slope stability. How can we work this out in complex environments like landslides? Geophysical monitoring can help, and is something that the Geophysical Tomography capability (‘GTom’ to our friends) has been working on for many years.
Figure 1: a 3D electrical resistivity tomography (ERT) model from the Hollin Hill Landslide Observatory. The clay-rich, saturated Whitby mudstone Formation (WMF, blue) is sliding over the porous Staithes Sandstone Formation (SSF, orange and red, creating landslide lobes in the mid-slope and a scarp at the top of the slope. These units overly the Redcar Mudstone Formation (RMF).

Geophysics: what, like on Time Team?

Well, yes, essentially. Using ‘active geophysics’ we generate signals, like radar pulses, seismic waves, or electrical currents, and record how these signals change once they’ve travelled through the ground. And work out what is in, or is happening in, the subsurface. These methods are great for making detailed images of the ground. For example, by injecting an electrical current in to the ground through electrodes (called ‘electrical resistivity tomography’, or ERT), we can look how resistive the ground is and how this changes across an area. This tells us about the relative porosity and lithology of the rocks beneath our feet, and using this, we can construct a 3D model (Figure 1).

Using ERT, GTom have developed a landslide monitoring system, able to image the ground in 4D. The ‘monitoring’ comes from the fact that the electrodes through which the current is injected are all buried in the landslide subsurface, and connected back to a central hub powered by solar panels. The hub sits on the landslide taking measurements every week, every day or every few hours, relaying the data back to BGS servers via mobile data connection.

Moving pictures

Once the data are processed back in the office, we can look at how the images change over time, and identify which changes are related to water movement, which normally occur over short time-scales. This can tell us a lot about how the landslide is behaving, even at times when we might not think it is doing anything at all. Using ERT, we can identify movements of water happening in the subsurface of the Hollin Hill Landslide in the middle of summer, a time when very little rain falls on the hill, but when important hydrogeological processes are still happening (Figure 2). The team and colleagues have been monitoring the Hollin Hill Landslide Observatory for over a decade and in that time have produced many research articles about the science happening there (see links below if you're keen!).

These detailed snapshots tell us a lot about the ground, but because one 3D image comprises several hundreds of measurements, there is always some time taken in acquiring all the data needed to make an image, and we can only collect so much data so often. When we look at our time-lapse images, it’s like running our thumb through a flick book, where the images change slightly from one page to another.

Figure 2: A difference image between two ERT datasets, showing the changes between two points in time. Using laboratory measurements, a relationship between resistivity measurements and gravimetric moisture content (GMC) has been established, showing where water is moving in the ground in the summer months.

Filling the gaps by adding a soundtrack

To fill in the gaps between our images, we need to start measuring signals that are happening all the time, such as the natural vibrations and tremors that occur in the ground, which we call ambient seismic noise. The Earth is constantly shaking and humming (no really, it’s a thing) and by using very sensitive seismometers, we can pick up all the tiny creaks and cracks that happen 24 hours a day, seven days a week, 365 days a year. Seismic signals can tell us a lot about landslides; they can occur more often or become larger as landslides start to move, and their frequency content can change depending on ground conditions, for example, with how much water is held in the rocks.

The BGS has operated a seismometer at the Hollin Hill Landslide Observatory for several years, and in 2016, GTom added two more seismometers to the site. Since then, it has become clear that recording seismic signals alongside our ERT measurements is an increasingly useful tool for monitoring. With this in mind, in late 2019, GTom applied to NERC’s Geophysical Equipment Facility (GEF) for the loan of seven more seismometers to be installed at Hollin Hill for a further two years. This expanded network of seismometers has the aim of monitoring seismic properties and processes at much higher spatial resolutions. The combination of spatial detail of our ERT images with the continuous observations from the seismic network will ultimately produce an integrated high-resolution landslide monitoring system.

In Part 2, we’ll look at the recent network installation that happened in March 2020, and what installing a seismometer entails (spoiler alert: there’s lots of digging), and what we hope to gain from burying all this equipment in the ground of a moving landslide!


References:
Uhlemann, S., Chambers, J., Wilkinson, P., Maurer, H., Merritt, A., Meldrum, P., Kuras, O., Gunn, D., Smith, A. & Dijkstra, T. 2017. Four-dimensional imaging of moisture dynamics during landslide reactivation. Journal of Geophysical Research: Earth Surface, 122, 398-418.

Further reading: 

Image credits:
Figure 1: Uhlemann et al., 2017 
Figure 2: Uhlemann et al., 2017

Comments

Post a Comment