Skip to content
WorldNerve

Demonstration study

Urban subsidence monitoring with multi-temporal radar

How time-series radar observations reveal spatial differences in ground movement across a dense urban area — and how those differences translate into a prioritised review list.

Demonstration content — replace findings with real project figures

Context

Many cities sit on compressible soils, reclaimed land, historical mining zones, or aquifers under abstraction pressure. Ground movement in such settings is rarely uniform: one district may be stable while a neighbouring one settles steadily. Asset owners and insurers with urban exposure need to understand these spatial patterns without inspecting every building.

Challenge

Traditional approaches — levelling campaigns, GNSS stations, inspection visits — provide accurate point measurements but cannot economically cover thousands of buildings, and they rarely extend back in time. The challenge is to reconstruct multi-year movement histories across an entire urban area at the scale of individual blocks.

Data

Sentinel-1 SAR imagery acquired every 6–12 days over a multi-year period, in ascending and descending geometries. Dense urban fabric provides abundant persistent scatterers — building corners, roofs, and hard surfaces that reflect radar consistently over time.

Method

Persistent Scatterer InSAR (PSInSAR): stable reflective points are identified across the full image stack, atmospheric and orbital errors are estimated and mitigated, and a displacement time series is computed for each point relative to a reference area. Results are quality-checked and aggregated to building-block level.

Findings

In a typical demonstration processing of this kind, the majority of measurement points show velocities within a stable band, while localised clusters exhibit sustained movement of several millimetres per year, spatially coherent across adjacent structures. Time-series plots distinguish steady linear trends from seasonal signals and from step-like changes. (Demonstration description — replace with figures from a specific processed dataset.)

Business relevance

The output is a ranked spatial picture: which blocks moved, how fast, since when, and how that compares to their surroundings. For an insurer this supports portfolio review and underwriting context; for a municipality or asset owner it identifies where inspection budgets are best spent.

Limitations

PSInSAR measures movement of radar-reflective points, mostly on structures — not the ground beneath vegetation or water. Velocities are line-of-sight and relative to a reference. Cause attribution (groundwater, consolidation, mining, construction) requires supporting geotechnical and hydrogeological context.