Satellite ground motion in a desk study: how to use the 2020 to 2024 European data

Your desk study can now cite a public satellite record of ground motion for 2020 to 2024, released by the European Ground Motion Service (EGMS) on 29 April 2026. The publisher sets the scale of use in its own FAQ, and that wording governs what you can write.

What the 2020 to 2024 release contains

The service is a Copernicus Land Monitoring Service product built from Sentinel-1 radar. Its latest release covers the reference period 2020 to 2024, announced in Product update: European Ground Motion Service 2024 on 29 April 2026.

The notice says the Explorer keeps the two most recent releases, 2019 to 2023 and 2020 to 2024, on direct access. The Baseline (2015 to 2021) and 2018 to 2022 datasets "have been pushed to the archive".

Each release is a five-year window. The European Environment Agency dataset record (revised 7 October 2026) says delivery is "on an annual basis, following a five-year moving window update strategy". A rate from this release describes those five years only.

The publisher's FAQ lists the United Kingdom and Ireland among the countries covered. Geological Survey Ireland notes on its natural hazards data page that the service "provides processed Sentinel-1 radar data to help detect and measure ground movements across Europe."

The British Geological Survey gives the UK context on its InSAR research page: "Coal mining and its legacy are one of the most prominent causes of ground motion in the UK." It adds that "active mining leads to subsidence and areas where mines have recently closed undergo uplift".

The three product levels

The service publishes three product levels, and a desk study should say which one each figure comes from. Basic and Calibrated are point data in the satellite's viewing direction. Ortho is a 100 m grid of vertical and east-west movement.

Table 1. The three EGMS product levels. Source: EGMS Product Description and Format Specification, version 3.0.2, sections 2.1 to 2.3; the "expert usage" wording is from the EGMS Product User Manual, version 1.6 (17 May 2022), section 3.4.

GNSS means global navigation satellite systems. The reference model behind Calibrated "contains average velocities in 3D (east, north, up) on a 50-km grid" (Product Description, section 2.4).

What does line of sight mean for a slope or embankment?

A radar satellite measures change in distance along its own viewing line, so a line of sight rate is not a settlement rate. The Product Description (section 2.1) defines it: "Satellite line-of-sight means that measurements are projected along the imaginary line which connects the satellite to the target". Such values "have to be reprojected to assume the meaning of vertical/horizontal displacement".

Basic and Calibrated each come as two datasets. One is from data "acquired orthogonal to the satellite ascending trajectory (south to north)". The other is from data "acquired orthogonal to the descending trajectory (north to south)".

So one slope carries two line of sight rates, one from each pass. One view alone does not give you the direction of movement.

The Ortho product resolves part of that. Section 2.3 says it "exploits the discrete look-angles provided by the Calibrated product to derive two further layers; one of purely vertical displacements, the other of purely east-west displacements".

North-south is the exception. The same section says the radar echoes "become less sensitive as the direction of any ground displacement approaches that of the satellite flightpath, i.e., north-south". Those movements are "reintroduced from the GNSS data".

The document judges that this "does not introduce a significant bias in the measurements, apart from area affected by strong spatial variations of north-south displacements". Note the aspect of any slope you report on.

Steep ground has a second effect. In high relief, "layover, foreshortening and shadow" prevent full coverage, and "In such areas, there are no EGMS Ortho measurements."

What scale and precision does the publisher state?

The publisher states the scale of use in its FAQ (checked 10 November 2026). The spatial resolution "is more appropriate to wider-scale (e.g. 1:50,000 scale) analysis, and is not appropriate to the analysis of individual buildings". It continues: "unless they are large enough to cover the ambiguity, or are located within a broader cluster of MPs". MP means measurement point.

The same answer describes "a metre-scale ambiguity as to the absolute xyz location" of any point. It adds: "These tolerances should always be taken into account when considering the exact location of any detected displacement."

Precision is specified in section 4 of the Product Description, version 3.0.2:

• Mean velocity standard deviation: "0.7 mm/year (1 σ) for MP with coherence greater than 0.7" for Basic and Calibrated, and "0.7 mm/year (1 σ)" for Ortho.
• Displacement standard deviation: "4 mm (1 σ)" for Basic and "8 mm (1 σ)" for Calibrated and Ortho.
• 3D geolocation accuracy: "Better than 10 m".

Coverage is uneven, which matters on a rural route. The FAQ says "typically vegetation causes decorrelation from one image to the next, and so no reliable data can be extracted". Maps are "discontinuous with most dense coverage in towns, cities and the built environment".

For linear assets the Product User Manual (section 5.2.2) is direct. "EGMS products are a baseline for the investigation of active movements along linear infrastructure", but "EGMS is not providing data for structural investigations". On landslides, section 4.1 says "Not all landslide types can be measured, and the motion rate must be smaller than 2 mm/day".

On isolated points the FAQ says: "The safest interpretation is when MPs are clustered and show some collective result."

How the data compares with levelling and GNSS

Published checks broadly support the specification, with stated exceptions. The EGMS Validation Report, version 3.0 (2 June 2025), concludes that "EGMS products meet the required technical specifications and are suitable for supporting various applications".

At its Jutland site in Denmark, "the velocity differences between the GNSS and the EGMS estimations are 0.8 mm/year on average" and "70% of all stations pass the velocity test".

A journal comparison in Germany found both agreement and disagreement. Even, Westerhaus and Kutterer (2024), PFG volume 92, compared EGMS with 1,095 levelling time series in Saarland and report "The overall standard deviation between EGMS and levelling is only 2.80 mm".

At the Epe cavern field the same paper reports "very considerable differences" from levelling and GNSS. The authors conclude that "The all-purpose models of the services are unable to describe the actual displacement satisfactorily."

Local processing can add points. A preprint on Swedish transport corridors (posted 20 August 2026, not yet peer reviewed) reports that a local analysis "consistently yields five times as many measurement points". It also reports that "Both datasets capture similar displacement patterns along all corridors".

With the levels, the geometry and the stated scale in hand, you can choose a product for the job.

Illustrative example

This route is invented and no figure below is a real reading. A desk study covers a 24 km pipeline route with a study strip 100 m either side of the centreline.

The strip is 24,000 m by 200 m, which is 4,800,000 square metres or 480 hectares. An Ortho cell is 100 m by 100 m, one hectare, so the strip holds 480 cells at most.

Suppose the Ortho download returns values in 120 of them. That is 120 divided by 480, or 25 per cent of the strip. The other 360 cells are mostly pasture and woodland.

Near kilometre 14 the route crosses an embankment with a crest 8 m wide. Three adjacent Ortho cells there show a mean vertical velocity of 3.0 mm/year downward, with east-west close to zero.

Set that against the published precision. The rate is 3.0 divided by 0.7, about 4.3 times the stated velocity standard deviation. Over five years it sums to 15 mm, against a displacement standard deviation of 8 mm for a single reading.

Now set it against the published position tolerance. Plus or minus 12 m east gives a band 24 m wide, three times the crest width. Position alone does not say whether the points sit on the crest, the toe or the field beside it.

The desk study therefore reports a cluster of three Ortho cells near kilometre 14, at corridor scale. It lists the location for the walkover and draws no conclusion about the embankment as a structure.

If your route crosses slopes, embankments or old workings, send us your route and we reply within two working days.

How to report it

A reader should be able to repeat your download from the paragraph alone. State these items each time.

• The release and its period, for example the 2020 to 2024 release, and the date you accessed it.
• The product level, and for Basic or Calibrated the geometry (ascending, descending or both).
• How many cells or points hold data in your study area, and where the gaps are.
• The threshold you used to call a cluster moving, and its basis in the published precision.
• The publisher's scale statement and position tolerance, quoted.
• The credit, and any change you made to the data. The EEA dataset record asks users to credit the source and to say clearly where a product has been modified.

Specimen paragraph. It describes no real site: replace every bracketed item with your own data and checks.

Satellite ground motion. European Ground Motion Service (EGMS) data for the 2020 to 2024 release were reviewed on [date] for a strip [width] m either side of the route. The Calibrated product (ascending and descending) and the Ortho product (vertical and east-west layers, 100 m grid) were used. Ortho values are present in [n] of [N] cells. Cells without values are gaps in the data and are not read as stable ground. [n] clusters show a mean vertical velocity beyond [threshold] mm/year and are listed in Table [x] with their line of sight rates. The publisher states that the data are "more appropriate to wider-scale (e.g. 1:50,000 scale) analysis" and gives position tolerances of North +/-8 m, East +/- 12 m and Vertical +/- 8 m. North-south movement is not measured directly. The data inform the scope of the walkover and ground investigation. They are not used here to assess any individual structure. Source: European Ground Motion Service, Copernicus Land Monitoring Service, accessed [date], [unmodified or modified as described].

What to do next

• Open the EGMS Explorer, linked from the Copernicus product page, and find your route. Note which release you are looking at.
• Download the Calibrated product in both geometries and the Ortho layers for your study strip. Count the cells that hold data.
• Mark clusters, not single points, and set them beside your records of mining, made ground, peat and drainage.
• Write the paragraph above into the desk study with the publisher's scale statement quoted in full.

The Vireo Corridor Evidence Pack is a desk read of a linear route before survey, across the UK and Ireland. Every figure in a pack carries its range, its source and its test label. Vireo can supply the surveys that follow: drone, aerial and ground, including ground-penetrating radar and magnetometry.

Quick answers

Sources

1. Copernicus Land Monitoring Service: Product update, European Ground Motion Service 2024, news item of 29 April 2026 (checked 10 November 2026)
2. Copernicus Land Monitoring Service: European Ground Motion Service FAQ (checked 10 November 2026)
3. Copernicus Land Monitoring Service: EGMS Product Description and Format Specification, version 3.0.2, cover date 7 August 2025 (checked 10 November 2026)
4. Copernicus Land Monitoring Service: EGMS Product User Manual, version 1.6, 17 May 2022 (checked 10 November 2026)
5. European Environment Agency: dataset record for EGMS Ortho, east-west component, revised 7 October 2026 (checked 10 November 2026)
6. Copernicus Land Monitoring Service: EGMS Validation Report, version 3.0, 2 June 2025 (checked 10 November 2026)
7. Even, Westerhaus and Kutterer: German and European Ground Motion Service, a Comparison. PFG, volume 92, pages 253 to 270, 2024 (checked 10 November 2026)
8. British Geological Survey: InSAR research (checked 10 November 2026)
9. Geological Survey Ireland: Natural (Geo) Hazards, data by environmental assessment and planning topic (checked 10 November 2026)
10. Aminjafari and others: Evaluating European Ground Motion Service against Local InSAR Analysis along Key Swedish Corridors, EGUsphere preprint posted 20 August 2026, not yet peer reviewed (checked 10 November 2026)
11. Copernicus Land Monitoring Service: European Ground Motion Service product page (checked 10 November 2026)

Ask the team

Written by the Vireo teamVireo Nature reads land before anyone surveys it. Every figure in our work carries its range, its source and its test.