TanDEM-X: use two radar viewpoints to map height

The Tandem-X spacecraft launched in 2010 to work with TerraSAR-X, which was already in orbit. The two similar spacecraft flew in close formation to observe terrain from slightly different positions. The TanDEM-X mission used this arrangement to build a consistent global elevation model, addressing gaps and mismatches among earlier regional mapping products.

The mission and the satellite bearing its name are not identical in scope: the mapping effort depends on both spacecraft, their relative geometry and a substantial ground-processing system. DLR managed scientific use, mission execution and elevation-model generation within a partnership with Airbus. A picture of one satellite alone would leave out much of the system needed to determine height.

A point on the surface lies at slightly different distances from the two radar viewpoints. Comparing the returned signals, with the baseline — open definition and observing geometry known, supplies information from which elevation can be estimated. Synchronization between the radar systems makes the comparison meaningful. The second spacecraft is therefore more than a backup camera taking an unrelated picture.

In a fictional demonstration, move one observer sideways while keeping the target fixed. The change in the two paths depends on the target’s position as well as the separation between observers. Without knowing that separation and the viewing geometry, a signal difference cannot be converted reliably into height. Formation control and precise data processing are integral parts of the measurement.

TanDEM-X: two distinct specifications

The original global model was produced from repeated acquisitions and completed in 2016. DLR describes products available on different horizontal grids, including 12, 30 and 90 metres. Grid spacing — open definition says how closely output samples are arranged across the map; vertical accuracy concerns the uncertainty in their estimated heights. These are different quantities and should not be substituted for each other.

Imagine comparing two fictional elevation maps of a glacier. Before calling their difference ice loss, check the acquisition periods, height reference, coverage and uncertainty. A finer grid alone does not prove a more accurate height change. TanDEM-X’s consistent acquisition and processing framework helps such studies, while the interpretation must still account for what each observation actually measured and how the final product was assembled.