The along-track InSAR technique for direct surface current measurements at high spatial resolution from aircraft was first demonstrated by scientists of NASA-JPL in the late 1980s. An along-track InSAR has two SAR antennas separated by some distance in along-track (flight) direction, which acquire images of the same scene with a corresponding short time lag on the order of milliseconds. Phase differences between the two complex SAR images are equivalent to Doppler shifts of the backscattered microwave signal and can thus be converted into line-of-sight velocities.

We have participated in the development of this technique since 1993. A theoretical model for the contributions of surface currents and wave motions to the detected along-track InSAR phase has been developed, and numerical simulations of wave-current interaction effects and SAR / InSAR imaging artifacts have been carried out in order to study the nonlinearities of the along-track InSAR imaging mechanism of spatially varying currents and to determine ideal parameters of airborne and spaceborne InSAR systems as well as measuring accuracies and spatial resolutions that can be achieved under various conditions.

Within the framework of the project EURoPAK (1998-2001), two experiments with the airborne X band InSAR system of the German SME Aero-Sensing Radarsysteme GmbH (now Intermap Technologies GmbH) were carried out at test sites near the islands Heligoland and Sylt in the German Bight of the North Sea. Main objectives were the development, validation, and demonstration of robust techniques for current measurements by along-track InSAR and for the retrieval of bathymetric maps from measured spatially varying current fields. Fig. 1 shows the aircraft of Aero-Sensing GmbH in the configuration used during the experiment at Sylt.

The test areas at Heligoland and Sylt are characterized by a strong modulation of the tidal flow by an underwater reef and by a tidal channel between two islands (Sylt and Rømø), respectively. In both cases, four flights at different tidal phases were performed within two days (Heligoland: 27-28 April 1999, Sylt: 21-22 May 2001). During each flight, four images of the test area were acquired from four different directions: Two images from orthogonal look directions are required to compose a fully two-dimensional vector current field; four images permit a better separation of stationary and unstationary contributions, thus a better noise reduction or higher spatial resolution of the final data product.

Fig. 3 shows the four InSAR phase images obtained from the first flight at Heligoland. The pixel size of the original data is 2.0 m × 2.0 m. After some filtering, conversion of phases into horizontal velocities, the composition of a vector current field, and iterative corrections for contributions of wave motions based on numerical simulations, the InSAR-derived current field of Fig. 4 is obtained. The pixel size of this final data product is 20 m × 20 m; a comparison with reference data from other sources has shown that the effective resolution of small-scale current variations is on the order of 100 m with an rms error of measured currents of about 0.1 m/s.

"Track 2"

"Track 3"

"Track 4"

"Track 5"

+π -π

0.8 m/s 0.0 m/s

Our current retrieval algorithm as well as the airborne InSAR system of Aero-Sensing GmbH (now Intermap Technologies GmbH) are readily available for further scientific experiments or routine operations. Promising applications are, for example, bathymetric monitoring in coastal waters and the detailed exploration of potential sites for electric power generators driven by tidal currents. The numerical SAR / InSAR imaging model suite M4S, which has been developed and used for our InSAR simulations, is publicly available. It has also been used for theoretical investigations on spaceborne InSAR and on  the ocean. For more information see our other InSAR-related webpages or the corresponding articles listed in the publication list of Roland Romeiser and in our general publication list.