Seismological
Research Centre

Geodetic Monitoring

The sensing system

The survey system for geodetic monitoring consists of two components: a satellite component and an in situ component.

The satellite component uses data acquired by radar systems mounted on satellites orbiting several hundred kilometres from Earth (Figure 1), capable of providing high-resolution microwave images. These systems are called “SAR”, which stands for synthetic aperture radar. Due to their longer wavelength than those of visible and infrared radiation, microwaves have the important property of penetrating through clouds, fog and possible coverings of ash or dust (for example, in the case of a volcanic eruption or collapse of a building); they are also independent of lighting conditions (day or night), thus allowing monitoring under any weather and environmental conditions.

The analysis of surface deformations in the Cornegliano Laudense area has been carried out using both archived SAR data, which allow observing the “past” history of ground movements in the area of interest, and recently acquired data, which monitor the current evolution of surface deformations.

As far as archived data are concerned, SAR images collected by the ERS-1/2 and ENVISAT sensors of the European Space Agency (ESA) over an area of approximately 95 km x 60 km, including the Cornegliano Laudense site, during the period 1993-2010, were used together. In particular, the deformation analysis was conducted using 141 images acquired by satellites along descending orbits (the satellite moves from north to south) and 76 along ascending orbits (the satellite moves from south to north).

With respect to monitoring, the analysis of surface deformations is carried out using data collected from the Sentinel-1 (S-1) constellation of the Copernicus European Programme, which is composed of twin sensors called S-1A and S-1B, which allow SAR images to be acquired every 6 days in the area of interest (Figure 1). In particular, 191 S-1 SAR data acquired from descending orbits and 200 from ascending orbits have been used over an area of approximately 190 km x 70 km that includes the site of Cornegliano Laudense from March 2015 through the present.

The analyses carried out based on satellite data are integrated with those conducted through the network of GNSS stations to bind the satellite measurements to some points located on the ground, for which the positions are known with precision.

Figure 1 - Sentinel-1 sensors acquire microwave SAR images every 6 days over the area of interest, allowing continuous monitoring under all weather and environmental conditions.

What is measured

The technique called “differential interferometry SAR” (DInSAR) allows the measurement of ground displacements (subsidence, slippage, etc.) with an accuracy on the order of centimetres and, in some cases, millimetres over very large areas (tens of thousands of square kilometres). The DInSAR methodologies basically “compare” pairs of SAR images acquired by one or more sensors from approximately the same position but at different times for a given area of the Earth’s surface. If something has changed in the time interval between the two acquisitions, i.e., if ground deformation has occurred between the two sensor passes, this can be measured and displayed by means of a false colour map with a correspondence between colours and ground movement.

In addition, using advanced DInSAR techniques, by appropriately combining a sequence of SAR images acquired for the same area over time, the evolution of the surface deformation can be measured in the time interval considered. One of these techniques, called the small baseline subset (SBAS), allows the generation of ground deformation maps at medium and high spatial resolutions (tens of metres/metres) much denser spatially than traditional geodetic methods (geometric levelling campaigns or GPS measurements) and with increasing temporal frequencies.

It should be noted that the maps generated by DInSAR analysis measure the ground deformation projected along the sensor’s line of sight (referred to in English as line of sight, LOS); in essence, these maps involve measuring how far a point on the ground moves away from or closer to the sensor along its LOS. However, taking advantage of acquisitions made from both descending and ascending orbits, the results of the respective analyses can be combined to generate both maps (Figure 1) and time series of deformation for the vertical and east-west components of the detected displacements.

Figure 1 - Average LOS deformation velocity map of the terrain, expressed in cm/year, generated using SAR data acquired by the Sentinel-1 sensor, in the period 2015-2019, over an area of approximately 190 km x 70 km that includes the Cornegliano Laudense site, superimposed on the optical image of the area. The green colour represents undeformed areas, yellow and red areas have subsided, and blue and blue areas have been uplifted. In the box at the top right, for example, are represented the historical series of deformation at the GPS station in Cremona measured using DInSAR satellite data (black triangles) and GPS data (red asterisks), respectively. Note that the DInSAR time series are calculated for each coloured point on the map.

What DInSAR analyses represent

The main results of DInSAR analysis are the time series (also called historical series) of the ground deformation, typically measured in centimetres, and the related average deformation velocity maps calculated in cm/year, related to the time interval during which the data were acquired by the radar sensor and with a time frequency given by the time interval between two successive observations of the selected area on the ground.

The average ground deformation velocity maps are geocoded maps; therefore, they can be superimposed on existing maps of the area and are represented in false colours that correspond to the ground displacements. In these maps, only the points where the deformation measurement can be considered reliable are shown, and for each of these points, the historical deformation series is calculated.

The final spatial resolution of the ground deformation maps obtained by DInSAR analysis of Sentinel-1 data is approximately 30 m x 30 m, while their spatial extent is approximately 190 km x 70 km; the time sampling of the related historical deformation series is currently 6 days.

DInSAR measurements are differential measurements, i.e., relative measurements with respect to a point considered stable that is chosen as a reference. In our case, the reference point (also called the docking point) is a pixel located in the urban area of Pavia near a corresponding GPS station, which, as evidenced by the related GPS measurements, is not affected by deformation (see Figure 1).

The DInSAR analysis natively generates the measurement of the ground deformation projected along the sensor’s line of sight (LOS) (see Figure 2 and Figure 3); however, by combining the DInSAR results produced by analysis carried out with acquisitions along ascending and descending orbits, it is possible to obtain measurements of the vertical and east-west components of the ground displacements (see Figure 4 and Figure 5).

Figure 1 - LOS strain rate map, geocoded and expressed in cm/year, superimposed on an optical image of the area of interest. The image reflects the processing of S-1 data acquired from descending orbits in the period March 2015 - May 2019. The green colour represents undeformed areas, yellow and red areas have subsided, and blue and blue areas have been uplifted. The graphs of the comparisons between the GPS deformation time series projected along the radar sensor’s LOS (red asterisks) and those obtained through DInSAR data (black triangles) are also shown at the 6 GPS stations marked on the map by white stars. The white rectangle indicates the area analysed in detail in the figures below.
Figure 2 - Enlarged view of the Cornegliano Laudense area of the map of average surface deformation velocity along the LOS obtained by DInSAR analysis of S-1 data acquired from ascending orbits in the period March 2015 - May 2019, geocoded and expressed in cm/year, superimposed on an optical image of the area of interest. The graphs show the temporal trend of the surface displacement along the LOS for three points located near Cornegliano Laudense (P1, P2 and P3) and one point located in the area of Turano Lodigiano (P4).
Figure 3 - Enlarged view of the Cornegliano Laudense area of the map of average surface deformation velocity along the LOS obtained through the DInSAR analysis of S-1 data acquired from descending orbits in the period March 2015 - May 2019, geocoded and expressed in cm/year, superimposed on an optical image of the area of interest. The graphs show the temporal trend of the surface displacement along the LOS for three points located near Cornegliano Laudense (P1, P2 and P3) and one point located in the area of Turano Lodigiano (P4).
Figure 4 - Enlarged view of the Cornegliano Laudense area of the average rate map of the vertical component of the surface deformation obtained by appropriately combining the LOS DInSAR results related to data acquired from ascending and descending orbits shown in the previous figures, geocoded and expressed in cm/year, superimposed on an optical image of the area of interest. The graphs show the temporal trend of the vertical component of the ground displacement for three points located near Cornegliano Laudense (P1, P2 and P3) and one point located in the area of Turano Lodigiano (P4).
Figure 5 – Enlarged view of the Cornegliano Laudense area of the average rate map of the east-west component of the surface deformation obtained by suitably combining the LOS DInSAR results relative to the data acquired from ascending and descending orbits shown in the previous figures, geocoded and expressed in cm/year, superimposed on an optical image of the area of interest. The graphs show the temporal trend of the east-west component of the ground displacement for three points located near Cornegliano Laudense (P1, P2 and P3) and one point located in the area of Turano Lodigiano (P4).

What is geodetic monitoring for

The activities of extraction/storage of hydrocarbons and re-injection of fluids into the sub-soil can induce surface deformation phenomena. These deformation effects provide important information about the characteristics of the sub-surface phenomena from which they originated and their temporal evolution. Geodetic monitoring aims to provide information on both the temporal trend of ground deformation and its spatial distribution in the analysed area, highlighting any variations with respect to the background deformation scenario, i.e., prior to the underground exploitation activities.

Further details

Precision GNSS measurement - station OL01

Station OL01, located near ‘Cluster A’ of the Cornegliano Laudense plant, houses both the seismic station and the GNSS station. GNSS instrumentation is intended to monitor the ground deformation at the surface and eventually to indicate potential anomalous phenomena.

Go to the dedicated section

Differential interferometry SAR

Satellite differential interferometry SAR for the measurement of ground displacements is a remote sensing technique that allows the measurement of surface deformations with an accuracy on the order of one centimetre and in some cases, even a few millimetres.

Go to the dedicated section

Geodetic data used

The dataset of the OL01 station, recorded in the RINEX format file, is available in daily sessions and with 30 second sampling (2880 observations daily).

Go to the dedicated section

Geodetic analysis and results

Below are the estimated time series for each GNSS LODI Network site for the time period 1/1/2017 - 31/05/2019. At the moment, the processing started on 12/12/2018 - does not show anomalous elevations of the ground.

Go to the dedicated section