Ground Model Development

The ground model of Hollin Hill, which has been generated through the merging of results of many investigative methods, is presented as Figure 3.16. This section aims to provide a summary of the contribution that each method provided to the process of ground model development.

Surface characterisation was performed through the interpretation of airborne LiDAR remote sensing data. Remote sensing data permitted the

interpretation of surface morphology, breaks in slope, and as a result identification (and location) of landslide type, such as the back scarp and back-rotated blocks indicative of rotational slumping. Visualisation of airborne LiDAR and production of geomorphology map allowed the spatial distribution of landslides to be determined as a series of rotational slumps towards the top of the slope giving way to a number of flow deposits in the mid-slope region. By combining airborne LiDAR with aerial photography the resulting DEM can be used to identify which areas of the landslide are most recently active by looking for surface features such as partially vegetated slopes/areas and abrupt or smoothed breaks in slope.

Low resolution ERT determines the overall structure of the hill slope, at the formational scale, from beyond the back scarp to the relict flow deposits nearing the base of the slope. The low resolution ERT survey picked out three lithological formations present at the field site (WMF, SSF and RMF), identified the nature of the flow deposits positioned over SSF and determined the regional dip of the formations.

High resolution ERT was performed to gain more information about active flow deposits and in particular their internal structure and lateral persistence. High resolution surveying identified the resistivity contrasts which exist between flow deposits as a result of lithological variation. The blue, low resistivity uppermost flow deposits in contrast with deeper, higher resistivity flows. Core logging and index testing of boreholes performed at the field site shed further light on the reasons behind resistivity variation both between flow deposits and between lithological formations represented. ERT permits the interpolation of interpreted borehole information which can aid determining the landslide structure laterally.

Core logs were interpreted on the basis of mass movement type, so whether the soil cores were flow or slump deposits or in-situ material. The task of differentiating between flow deposits and slump deposits required identification of features associated with each deposit type, for example rotated bedding planes and presence of rip-up clasts is indicate of slumping. Slip surfaces associated with flow deposits were identified as thin (~5mm)

bands of light brown clay between layers of highly disturbed dark brownish grey silty, sandy clay or slity sand.

Analysis of core soil samples allowed core samples and ERT surveys to be correlated. For example, core resistivity measurements can be used as a confirmatory tool when interpreting ERT surveys. In order to differentiate between clay and sand flow deposits in the high resolution ERT survey core resistivity measurements were utilised. Core resistivity measurements were then related to particle size analyses, the resistivity variation could then be explained in terms of lithology. CEC versus core resistivity plot can inform about potential similarities in resistivities between lithological formations. In our example differentiating between sand flow, slump and in-situ SSF could be problematic and as a result any interpretation took this into account. For this reason the differentiation between in-situ SSF and slumped SSF was impossible and attributed to there being little to no lithological – and therefore electrical property – variation between the two. XRD results also indicate the mineralogical similarities between slumped and in-situ SSF, and dissimilarities with clay and sand dominated flows of WMF.

Implementation of peg displacement results along with ERT surveys and geomorphology studies made it possible to define active landslide regions, types of movement (flow or slump) and rate of displacement. With the addition of inclinometer results the active shear surfaces were identified and incorporated into the ground model.

Order of investigative technique application

If one were to implement the techniques utilised and described in detail during this investigation in the context of landslide site investigation and ground model development, performing the methods in the following order would give the most beneficial outcome. To characterise the surface expression, which may provide an indication of subsurface structure, the whole site both quickly and at high resolution, visualisation of either airborne or static LiDAR should be the first technique applied.

Figure 3.16. Ground model of the Hollin Hill study site based on geophysical, geomorphological and geotechnical investigations.

Upon gaining an insight into both surface expression and potential internal, subsurface feature, either several site-scale 2D ERT profile or 3D ERT volumetric image should be performed with the aim of ascertaining large-

scale subsurface features such as lithological boundaries. 3D ERT is preferred to several disparate 2D ERT surveys as the majority of geological features are three-dimensional in nature, therefore the former method is most appropriate and reduces the need for guess-work later on in the investigation. The two main draw backs of performing three dimensional ERT are that the method can be time-consuming relative to 2D ERT when surveying large areas of terrain, and pre-inversion data processing can be demanding if quality, intelligible survey notes are not available for reference. Once large scale ERT has been performed at the site, it would be wise to perform smaller scale ERT surveys of the areas of the landslide system which look either most active or may reveal the most structural information. The outcome of small scale ERT surveys should then dictate the direction of the intrusive investigation, i.e. where to drill and also which features could potentially be encountered. An ERT survey performed either in undisturbed landslide material or off the landslide system can act as a control and aid better understanding landslide physical properties such as soil desiccation. Cores should firstly be analysed and interpreted in terms of lithology and later, in conjunction with core petrophysical and geotechnical information, such as cation exchange capacity, core resistivity and moisture content, as well as structural core observations be interpreted in terms of landslide deposit type (as is outlined in (3.4.3 and 3.4.4 of this chapter, see Figure 3.8). By linking core petrophysical, geotechnical and index testing information the process of interpreting 3D ERT surveys is more straightforward as any trends observed in field ERT surveys will be directly visible in core log results. Links can be identified between cation exchange capacity and respective core resistivity measurement, confirming that, differentiation between lithologies and deposit types can take place solely based on core electrical properties. This technique is most useful where formations have distinct differences in lithologies and therefore electrical properties, such as between sand- and clay-dominated formations.

Finally, performing these techniques in the order described above produces a detailed landslide ground investigation and subsequent ground model as each technique informs and directs the next technique to be utilised. It is generally good practise to start with surface characterisation through LiDAR

visualisation and geomorphological map production followed by subsurface characterisation through application of 3D geophysics and intrusive investigation.