OR/18/054 Known subsidence in Glasgow and InSAR data

From Earthwise
Jump to navigation Jump to search
Bateson, L, and Novellino, A. 2018. Glasgow Geothermal Energy Research Field Site - Ground motion survey report. British Geological Survey Internal Report, OR/18/054.

Due to a substantial industrial past, including underground mining, the Glasgow area is an obvious suspect for subsidence. In 2001 BGS and the British Space Agency commissioned an InSAR study for Glasgow and Newcastle using ERS data to reveal ground motion for the 1990’s. At the time InSAR was an immature Earth observation technique and this was one of the first studies of its type in the UK. Glasgow and Newcastle were chosen due to their coal mining history and the knowledge of BGS engineering geologists of subsidence problems in these areas. The results of the study indicated that Glasgow was largely stable from the InSAR point of view in the 1990’s (as does this current study). The Newcastle data obtained at the same time revealed interesting patters of motion associated with active coal mining amongst other processes (Gee et al, 2017[1]).

Subsequent InSAR studies on UK coalfield areas (Sowter et al, 2013[2]; Bateson et al, 2015[3]) have shown distinctive patterns of ground motion related to coal mining, and it has been possible to identify areas of active and abandoned mining activity via their ground motion signature. These motions cover a relatively large spatial area and usually relate to changes in the groundwater level, which are caused by the different phases of mining. During active mining the ground water level is depressed by pumping to allow for underground working, this causes a lowering of the ground surface. Once the mining has finished and the pumps are switched off, ground water levels recover and uplift is observed. Such patterns are observed in South Wales, Northumberland, South Yorkshire, Stoke on Trent and North Derbyshire. However, there is no evidence in any of the InSAR results of these characteristic ground motions in the Glasgow area. This is likely because mining under many areas of Glasgow is significantly older than the satellite data (for example in the vicinity of the GGERFS site, recorded mining ended in 1934) and therefore water levels have recovered prior to the 1990’s.

A web search for ‘Glasgow Subsidence’ returns several stories in the local press of subsidence events in Glasgow in the recent past. So, the question remains as to why we do not see these reported areas of subsidence within the InSAR data? The following reasons may account for this:

  1. Default display: The default way to view an InSAR data set is as coloured points on a map. The points are colour coded by the linear average velocity; this means motions which occur over a short time period can be masked out when the average for the longer processed time period is used to display the point.
  2. Small spatial scale of subsidence: The majority of subsidence events reported in Glasgow are associated with phenomena occurring over a short spatial scale. InSAR points are acquired via opportunistic measurements of persistent radar targets; it is not possible to choose which points will be measured. Therefore, an InSAR point is often too far from the event in question.
  3. Fast rates of subsidence: InSAR techniques are only able to resolve ground motions of less than half a wavelength between processed radar images. In the case of C-band radar (as used here) this is 2.3 cm (5.6 cm radar wavelength) and the revisit times are every 6 days for Sentinel 1 and 31 days for ERS and ENVISAT. Therefore if an area is moving faster than 2.3 cm between the image acquisitions then these areas are effectively moving too fast to be accurately represented.

With these limitations in mind we tested some of the subsidence events found in the media.

15th June 2018: Subsidence on High Street/Bell Street Junction

Figure 24    TRE ALTAMIRA SqueeSAR™ Vertical InSAR data for High Street/Bell Street junction. Contains © TRE ALTAMIRA 2018 data. Ordnance Survey data © Crown copyright and database right 2018, license number 100021290 EUL.

In Figure 24 the average velocity of the point nearest the junction shows that the area is stable (points appear green), time series plot also shows that the area is stable; there are no sudden changes within the time series. Note the report of subsidence is June 2018 and the time series only extend to November 2017. In this case the time series does not extend across the date of the subsidence, however the authors have access to demonstration InSAR data from a different processing company which does extend to June 2018 and we do not see significant motion in this area at the time in question.

Figure 25    Time series for an InSAR measurement point from a demonstration dataset that BGS have access to for the area in question in Figure 24, note the linear motion and lack of significant variations in June 2018.

594 London Road, Glasgow
13 January 2012

In 2012 extensive cracks appeared in the facade and internal structures of a London Road block (north of GGERFS) that was deemed unsafe and then demolished (www.eveningtimes.co.uk/news/13241269.Road_closed_after_sewer_collapses/).

Unfortunately this ground motion occurred during a period when there was no operational European Space Agency radar satellite, therefore we are not able to provide InSAR analysis using this free data source.

Figure 26    TRE ALTAMIRA SqueeSAR™ vertical InSAR data for London road, note time series step in March 2016. Contains © TRE ALTAMIRA 2018 data. Aerial photography © UKP/Getmapping Licence No. UKP2006/01.

Although this report was from 2012 (when we do not have suitable satellite imagery coverage) the time series for 2015–2017 shows an interesting 10 mm step in the motion history in March 2016. It is possible that this motion relates to remedial work carried out on the property in the years following the damage.

The two cases above represent motions that are challenging for medium resolution C-band InSAR to address; they both concern rapid motions over a small spatial scale and in each case the InSAR data does not cover the event itself. The first example highlights that it is not possible to guarantee a measurement point over the area of interest and the nearest point may not pick up such a localised motion. Data from higher resolution X-Band radar satellites, such as TerraSAR-X or COSMO-SkyMed, offer a better chance to detect such changes but at a financial cost. The second example highlights a data gap within the freely available ESA datasets; with the launch of Sentinel-1A in 2014 and Sentinel-1B in 2016 and the European Space Agency’s plan for a 12-year operational lifespan of systematic data acquisition the chance of such gaps is reduced. Later data over this site does show some motion which may relate to the original reported motions.


  1. GEE, D, BATESON, L, SOWTER, A, GREBBY, S, NOVELLINO, A, CIGNA, F, MARSH, S, BANTON, C, and WYATT, L. 2017. Ground Motion in Areas of Abandoned Mining: Application of the Intermittent SBAS (ISBAS) to the Northumberland and Durham Coalfield, UK. Geosciences, 7, 85. https://doi.org/10.3390/geosciences7030085
  2. SOWTER, A, BATESON, L, STRANGE, P, AMBROSE, K, and SYAFIUDI, N M F. 2013. DInSAR estimation of land motion using intermittent coherence with application to the South Derbyshire and Leicestershire coalfields, Remote Sens. Lett., 4(10), 979–987.
  3. BATESON, L, CIGNA, F, BOON, D, and SOWTER, A. 2015. The application of the Intermittent SBAS (ISBAS) InSAR method to the South Wales Coalfield, UK. International Journal of Applied Earth Observation and Geoinformation, 34, pp.249–257.