OR/18/020 Ground motion

R S Ward1, G Allen2, B J Baptie1, L Bateson1, R A Bell1, A S Butcher1, Z Daraktchieva3, R Dunmore4, R E Fisher5, A Horleston6, C H Howarth3, D G Jones1, C J Jordan1, M Kendall6, A Lewis4, D Lowry5, C A Miller3, C J Milne1, A Novellino1, J Pitt2, R M Purvis4, P L Smedley1 and J M Wasikiewicz3. 2018. Preliminary assessment of the environmental baseline in the Fylde, Lancashire. British Geological Survey Internal Report, OR/18/020.

Introduction

There is speculation about whether the potential exists for shale-gas operations at depth to cause surface ground deformation. Conventional oil and gas operations have on rare occasions been shown to result in subsidence above compacting oil and gas reserves (Geertsma, 1973) and a recent study suggests that surface uplift in eastern Texas was due to fluid injection, which was distinguished using satellite remote sensing (Shirzaei et al., 2016). These studies do not imply that shale-gas operations at depth will cause ground motion. Nevertheless, undertaking objective and authoritative monitoring of the ground surface at operation sites and surrounding regions is advisable (a) to determine if there are any impacts on the ground surface and (b) to reassure the public that appropriate independent monitoring of all potential environmental impacts is being undertaken. Knowledge regarding the baseline ground motion conditions, compared with the current situation, would enable the provision of impartial and objective information on whether shale-gas operations have affected the status of the landscape.

Firstly, it is necessary to define ‘ground motion’ in the context of this baseline monitoring study. The first clarification is that the term does not refer to seismicity, which is the frequency, intensity and distribution of earthquakes (induced or otherwise) in an area. We use ‘ground motion’ to mean the motion of the surface of the landscape upwards (uplift), downwards (subsidence) or sideways (lateral motion). The motion is measured as the average velocity per year (in millimetres) along with profiles of motion at each point/pixel for each time period a measurement was captured. The measurements are derived by processing satellite radar data using the Interferometric Synthetic Radar (InSAR) technique, with the motion generally described in terms of Line of Sight (LOS) from the satellite or as absolute motion (vertical/horizontal displacement). Satellite-based InSAR interrogates the differences in phase from a series of radar images to generate results of surface motion. The motion we measure does not take account of ground acceleration, i.e. peak ground acceleration (PGA).

The key monitoring question is whether shale-gas operations are altering the earth-surface processes and stress conditions that are operating at the site. We cannot assume that an area is stable prior to shale-gas operations. When considering a monitoring system, it is important to account for the dynamic nature of the earth’s surface i.e. there may be some pre-existing displacement due to either natural or induced factors. Therefore, a baseline survey is vital to determine the pre-existing conditions of the site including displacement such as upwards motion (uplift), downwards motion (subsidence) or horizontal/lateral motion, and ongoing monitoring during any operations is required to characterise the current situation.

The investigation in this work package is designed to monitor surface ground motion (subsidence, uplift or stability) of the target area using LOS InSAR prior to any permitted unconventional gas production in Lancashire. InSAR is considered an appropriate technique for ground motion monitoring because:

1. archive radar data (acquired by satellites since 1992) are available and can be utilised to ascertain a baseline of motion (or lack of motion) prior to any permitted gas operations;
2. data from currently-orbiting satellites such as Sentinel-1 can be analysed to acquire information about the ongoing surface ground motion conditions in a region;
3. the analysis produces measurements over areas measuring up to thousands of square kilometres rather than at a point location, which other techniques such as GNSS provide.

InSAR can provide millimetric measurements of surface ground motion from satellite platforms such as ENVISAT, ERS1&2, RADARSAT, Sentinel-1A/B, TerraSAR-X and CosmoSKY-Med. The technology has been validated by the British Geological Survey (BGS) in projects such as TerraFirma and was used by BGS in projects including PanGeo, SubCoast and EVOSS to develop and demonstrate viable services (e.g. Cigna et al., 2015; Jordan et al., 2017. InSAR has also been successfully used in CO₂ sequestration monitoring projects in locations such as In Salah where Mathieson et al. (2011) stated that ‘perhaps the most valuable, and initially surprising, monitoring method so far has been the use of satellite based Interferometric synthetic aperture radar (InSAR) to detect subtle ground deformation’.

Table 12 provides a guide to the advantages and limitations of remote and in-situ systems for ground motion monitoring.

Table 12    Comparison of remote and in situ ground surface motion monitoring systems.

Monitoring technique	Advantages	Limitations
InSAR	Measurements are made remotely (non-invasive). Measurements can be made using historic data to gain a baseline prior to operations. Imagery can cover a large area simultaneously. Entire deformation field can be imaged, rather than isolated points.	Conventional techniques have difficulty in vegetated areas. High magnitudes of motion (greater than the satellite detected phase difference) cannot be measured. Temporal and spatial resolution is limited by satellite set up and orbital parameters. Affected by steep topography (shown not be an issue in most of the UK).
GNSS	High precision. Does not require line of sight between benchmarks. Continuous site can operate without frequent human interaction.	Equipment can be stolen/vandalised/damaged. Sampling of deformation field is limited to individual points; several points are required. Requires at least 4 satellites in view simultaneously.
Tiltmeters	High precision. Does not require line of sight between benchmarks. Continuous site can operate without frequent human interaction.	Equipment can be stolen/vandalised/damaged. Sampling of deformation field is limited to individual points. Complex installation (e.g. in boreholes) — several tiltmeters are required.
Total Stations	High precision. Continuous sites can operate without frequent human interaction.	Requires line of sight between benchmarks. Generally they are operated manually, requiring repeat site visits to operate the system.

To date, the InSAR process has not been applied to monitoring energy operations in the UK because of the challenge of gaining coherence over non-urban areas. To resolve this challenge, we processed the data using the conventional SBAS (small baseline subset) process to gain precise results over urban areas and subsequently utilised the ISBAS (intermittent small baseline subset) process to acquire results over the non-urban areas (Cigna and Sowter, 2017).

BGS has experience of applying InSAR to several ground surface monitoring applications in the UK e.g. utilising 55 ERS-1/2 images between 1992 and 1999 to investigate ground motion linked to ceased mining operations in south Wales (Bateson et al., 2015). The deliverable in this ground motion work package is to provide ‘an analysis of satellite (InSAR) data’. In order to achieve this, the following steps were followed:

• Obtain stacks of satellite SAR data;
• Process the data using the SBAS InSAR techniques, thereby deriving results primarily for urban areas;
• Process the data to ISBAS level, thereby extending the results to non-urban areas;
• Provide an analysis of the InSAR results.

The work package utilised the ISBAS technique of InSAR analysis as it has been found to provide results in non-urban areas where other InSAR techniques fail. The conventional SBAS technique requires that the target shows coherence in every image of the stack, while the ISBAS technique utilises coherence that is intermittent throughout the stack. Both SBAS and ISBAS processing and analysis was undertaken on each stack of radar images to provide results in urban and non-urban areas.

Data selection

This BGS-funded research project utilised archive radar images from the ERS-1/2 satellite for the period 1992–2000 (Table 13) to assess the efficacy of InSAR for ground motion monitoring as part of an integrated baseline monitoring programme. The stack of radar data consisted of 63 images that were analysed using SBAS and ISBAS InSAR techniques, i.e. two sets of analyses were undertaken and completed within this ground motion work package. Data from the ENVISAT and Sentinel satellites (covering the time period from 2007 to the present day) have not yet been utilised in this study.

Table 13    ERS-1/2 image metadata.

Track	No. of scenes	Dates of image acquisition
T409	63	04/07/1992–11/07/2000

Ancillary data

A selection of ancillary datasets (listed below) were utilised in order to interpret the InSAR process i.e. to determine/understand the potential causes for the motion:

1. Bedrock geology (incl faults)
2. Surficial geology (incl. compressible ground)
3. Historic mining information/plans
4. Seismic records
5. Groundwater abstraction records
6. Borehole records
7. Geohazard information (e.g. landslides and shrink/swell)
8. Landcover information
9. Historic topographic maps
10. Aerial photography
11. Digital elevation models
12. Digital terrain models

Data processing

All of the InSAR processing was undertaken by the BGS Earth Observation Team. Our approach to acquiring, processing and interpreting the InSAR data is briefly outlined in this section. The methodology to effectively undertake the monitoring programme of ground-motion conditions using InSAR techniques is illustrated in Figure 55 with the associated actions listed below.

1. Search of catalogue satellite radar data to confirm that suitable stacks of images were available for the study area;
2. Download the stack of image datasets covering the geographic area and the time period(s) of interest;
3. Process the imagery for the region using SBAS and ISBAS InSAR technique(s);
4. Ensure that the outputs from the InSAR processing match the quality required e.g.:

Suitable density of spatial coverage in the area of interest

Suitable temporal coverage in the area of interest

Assess output statistics to gauge if the results are fit-for-purpose

5. Interpretation of the InSAR outputs. This is a key stage because the outputs from the InSAR image processing are dependent on the quality of the interpretation. There are two fundamental components;

Ensure that interpretation is undertaken by sufficiently-experienced personnel. For shale gas applications the interpretation should be done by experienced geoscientists who compile and integrate a suite of geoscientific information

The interpretation was reliant upon access to a comprehensive range of ancillary data, listed in the section above

Figure 55    Flowchart of the approach and data utilised for the ground motion InSAR monitoring work package.

Results of Fylde InSAR analysis

The processing of the InSAR data has provided the first baseline assessment of land-surface deformation in the region. The analysis has not covered the full time period from 1992 to the present day as this was a preliminary study. In the Fylde, the ERS-1/2 radar data have been analysed to produce InSAR results for urban and non-urban areas. The results indicate a maximum velocity of +15.8 mm/year and a minimum velocity of -13.6 mm/year. The SBAS InSAR analysis comprises 140k points while the ISBAS analysis comprises 890k points.

The results of the ERS-1/2 InSAR analysis for SBAS and ISBAS are shown for the regional area in Figure 56 and Figure 57 respectively. Green areas are considered stable, red are subsiding on average over the time period, and blue are undergoing uplift.

A larger area than the Fylde was processed; the results highlight the potential for InSAR to detect the range of motion in the region including discrete areas of subsidence and uplift, as well as confirming the stability of large areas (Figure 58, Figure 59).

Outside the Fylde, the discrete area of uplift (blue points) north-west of Salford is likely due to the rise in groundwater levels following cessation of water pumping in abandoned coal mines. Minewater pumping data have not been evaluated to assess this hypothesis. There is an area of subsidence to the south-west of the uplift, in the Bickershaw-Goldborne-Leigh region. This is likely due to mining activity in the three collieries including water abstraction (Arrick, 1995 and formation of the Pennington Flash, illustrated in Figure 60.

Figure 56    InSAR processing using SBAS technique of ERS-1/2 data from 1992 to 2000. The red box outlines the extents of the Fylde study area.

Figure 57    InSAR processing using ISBAS technique of ERS-1/2 data from 1992 to 2000. The red box outlines the extents of the Fylde study area. © Crown Copyright and/or database right 2018. Licence number 100021290 EUL.

A detailed view of the InSAR results for the Fylde are seen in Figure 60. The results in this time period (1992–2000) contain discrete areas of subsidence indicating that the Fylde area is undergoing some ground motion. Sufficient resources were not available in this preliminary study to validate these with ground surveys.

Figure 58    Areas of subsidence identified (in red) on the InSAR data in Leigh (outside of the Fylde study area) between 1992 and 2000. Black circle outlines the detailed time series results in Figure 59. © Crown Copyright and/or database right 2018. Licence number 100021290 EUL.

Figure 59    Time-series profiles of motion for 1996 to 2000.

Figure 60    Areas of subsidence identified (in red) on the InSAR data in Blackpool between 1992 and 2000. © Crown Copyright and/or database right 2018. Licence number 100021290 EUL.

These areas of subsidence correspond to an area of ‘peat and blown sand’ on the published geological maps. Boreholes from the area indicate the presence of ‘sand and peat’ at the top of the stratigraphy (Figure 4), suggesting that the subsidence may be caused by the existence of compressible ground.

Additional processing of more recent satellite imagery (as conducted for the Vale of Pickering area) would provide a more complete representation of the baseline of ground motion prior to or during hydraulic fracturing operations.

Discussion of results

The preliminary Fylde ground motion InSAR analysis entailed processing one stack of ERS-1/2 (covering the period from 1992 to 2000) using SBAS and ISBAS techniques (i.e. two levels of analysis in total). The assessment indicates that zones within the wider region covered by the satellite image stack underwent both uplift and subsidence, while the majority of the region was stable. It is suggested that the uplift and subsidence in the Manchester area may be related to coal mining, while the subsidence in the west of the Fylde is thought to be related to compressible ground. These examples provide an indication of the ground motion which this monitoring technique can detect. Work has not yet been undertaken to confirm the cause of the motion.

The recently launched Sentinel-1A and 1B constellation is now orbiting the Earth and acquiring data every six days. These satellites offer the opportunity to extend ground motion monitoring for the Fylde from 2015 onwards, once a sufficient stack of data become available. Preliminary InSAR results have been obtained from these satellites for the Vale of Pickering baseline study. They indicate that higher concentrations of measurement points can be achieved using both the SBAS and ISBAS techniques when compared to ERS and ENVISAT InSAR results.

Summary

It was apparent at the public engagement event in the Fylde (and the Vale of Pickering) that there is some confusion between seismic activity and ground motion. Many of the attendees link the two and presume that if there is seismic activity there must be ground motion and vice versa. It is therefore important to communicate the situation regarding baseline ground deformation and also provide evidence regarding the opportunities for monitoring in order to address public concerns. Part of this is the establishment of ground motion baselines along with monitoring of the situation throughout any shale-gas operations. This baseline allows an understanding of how the natural (and anthropogenic) processes lead to small-scale ground motions. The baseline provides evidence that small-scale motions are occurring continually and may not normally impact on day-to-day life. It also offers comfort to the public that there is a record of the existing conditions so that if operations start, there is a baseline with which to compare the up-to-date information.

The unique characteristics of satellite-based InSAR have proven it to be a valuable technique in the establishment of a baseline of ground motion for the Fylde prior to any exploitation of shale gas. There are three main benefits of using InSAR to derive ground motions:

1. In common with most remote sensing techniques, InSAR offers a regional view of the phenomena being measured. Ground deformation points are generated for the entire radar scene; this offers the opportunity to not only focus on ground motions for the immediate area surrounding the shale gas site, but also the wider area. This wider view allows an understanding of the processes, which drive the movement of the ground.
2. C-band satellites have been orbiting the Earth, and imaging the UK, since 1992–1993. These data have been archived. It is therefore possible to process the archive data and ‘look back in time’ and retrospectively establish the patterns of ground motion for an area. This is simply not possible with other techniques such as GNSS where the survey equipment must be located onsite with knowledge of the phenomena to be measured.
3. InSAR processing results in a dense network of opportunistic measurement points. For techniques such as SBAS the greatest densities are found over urban areas where the built environment acts as a good radar scatterer. However, recent advances in processing such as ISBAS increase the density of measurements, especially in rural areas, such as the Fylde. Each measurement point has an average velocity but also a time series. This offers the opportunity to understand how the ground at that point has moved through time, thereby enabling the interpretation.