InSAR can resolve millimeter-scale changes in the distance between a satellite and a radar target over time, across thousands of points per square kilometer. Getting full value from that sensitivity means understanding what the numbers represent. Not every point delivers an exact vertical ground movement measurement, and results depend on geometry, reference assumptions, noise, coherence, processing choices and the time period assessed. Buyers should distinguish between measurement sensitivity, precision, accuracy and confidence in a decision.
What does the satellite actually measure?
Interferometric measurements are made along the radar line of sight, the direction between a ground target and the satellite. A point can move vertically and horizontally and produce the same line-of-sight displacement as a different combination of motions. As a result, a single viewing geometry generally cannot identify full three-dimensional movement. You can read more about how the method works in SkyGeo's introduction to InSAR.
For a subsiding urban area, assumptions about the main direction of movement can sometimes support a vertical estimate. On a steep mine slope, horizontal movement can be important and a vertical-only interpretation may be wrong. Clear reporting states whether a value is line-of-sight displacement, an assumed vertical component or a solution that combines several views with independent information.
Precision is not the same as accuracy
Precision describes how repeatable or tightly grouped the estimates are. Accuracy describes how close they are to the true quantity, relative to an appropriate reference. A smooth time series can look precise while still containing systematic effects from atmospheric delay, reference point motion or processing choices. Equally, a noisier point can still support a meaningful long-term trend if its uncertainty is estimated correctly.
Claims of millimeter-level measurement should state the metric, time frame and conditions. The uncertainty of a long-term average velocity is not the same as the uncertainty of a single observation or a sudden displacement event.
Why the reference matters
InSAR measurements are usually relative to a reference area or point. If the reference itself moves, some apparent movement elsewhere may simply reflect that choice. Differences between datasets can also come from different references, satellite tracks, geocoding or processing windows. Before treating a numerical difference as a physical change, check the reference framework. You can find answers to related questions on SkyGeo's InSAR FAQ page.
GNSS, leveling and stable reference areas can help test an interpretation. However, validation must compare like with like: the same location where possible, compatible dates and the same component of displacement.
Ask for evidence, not a single headline number
Ask your supplier for uncertainty or quality indicators, coverage gaps, the satellite look direction, the method used to select the reference and representative comparisons with independent measurements. Where a result matters for an engineering decision, interpret the uncertainty alongside the asset's allowable movement and the consequences of getting it wrong.
A very dense dataset can provide useful spatial context, but thousands of points do not remove bias or establish the condition of a structure. The most useful result is one whose uncertainty and meaning are clear enough to support the decision at hand.
Can two InSAR providers report different movement rates?
Yes. Different image archives, time windows, points, reference areas and modeling choices can produce different estimates. The right response is to reconcile the methods and compare equivalent time series, not to assume one dataset must be wrong. At SkyGeo, these processing choices are configured for each project and explained as part of the interpretation.
Frequently asked questions
-
No, and the difference matters. Precision is about repeatability. Accuracy is about how close a measurement is to the actual displacement under the stated reference and conditions. A provider should explain both.
-
Radar phase measures changes in distance along the viewing direction. Estimating vertical or horizontal components requires additional geometry, assumptions or independent observations.
-
Not from a single viewing direction. A fuller estimate of motion may require multiple viewing geometries and external geodetic data.
Clarify your measurement requirements
If geometry or uncertainty matters for a planned monitoring decision, let's discuss the site and evidence requirements before you select a product. Contact SkyGeo.
