In this paper we present a first demonstration of spaceborne L-band SAR Tomography over a forested scenario. To this aim, we process a SAOCOM stack of nine acquisitions with across-track baselines theoretically achieving a vertical resolution around 22 m. To obtain a correct TomoSAR imaging, we phase calibrate the stack by properly accounting for forest volumetric structure, which drives the choice of calibrating with smaller baselines. Finally, we ground steer calibrated TomoSAR to achieve a consistent reconstruction of forest structure above ground and derive a forest height map. Comparison with reference canopy height gives a very good overall agreement, proving the effectiveness of our approach and making the experiment an interesting test-bed in view of future long-wavelength SAR missions.

In this work we present a first spaceborne L-band forest SAR Tomography (TomoSAR). We process a stack of 9 real SAOCOM acquisitions over Amazonas, a scenario mostly consisting of tropical rainforest with average tree height of about 25 m to 35 m. We phase calibrate the stack with Phase Linking algorithm, in order to obtain a consistent reconstruction of forest over terrain and compare TomoSAR results with different spectral estimators. We achieve good results with MUSIC spectral estimator, coping with spatial and temporal baselines and compare the upper TomoSAR envelope with LiDAR forest height. The overall agreement is quite good, confirming the effectiveness of our approach.

Along-track constellations of SAR satellites are a promising technology capable of imaging large areas with fine resolution. The enhanced performance is achieved by reducing the sampling frequency of each spacecraft, and the caused aliasing is later solved in processing by combining data from several sensors. This work presents a flexible and robust method to combine the channels based on back-projection. A secondary step further suppresses residual azimuth ambiguities to obtain the proper high-resolution image. The method allows for the incorporation of additional calibration steps before the final signal reconstruction. Testing and validating were accomplished with simulated data using point targets and a realistic scenario.

Formations of small satellites are gaining momentum as an alternative to single Synthetic Aperture Radar (SAR) systems, allowing them to improve performance and reduce costs. Each sensor operates with a low Pulse Repetition Frequency (PRF), so swath size can be kept large, and the strong azimuth ambiguities are later resolved by combining acquisitions from multiple satellites. For proper imaging to be successful, careful calibration of the system and geometric parameters is performed based on the single-channel images. However, ambiguities jeopardize the retrieval of parameters from data; therefore, it is crucial to identify those areas that are inherently ambiguity-free. This paper handles acquisitions that are highly affected by azimuth ambiguities, as in the case of the formation. First, we model the replica’s structure in the focused image and then propose an effective yet simple method to detect such disturbances. Lastly, we demonstrate an improved approach for estimating the along-track baseline by incorporating the aforementioned ambiguity detection scheme. The proposed analysis is validated with simulated data, considering a realistic reflectivity map derived from a COSMO-SkyMed stack.

Along-track (AT) formations of synthetic aperture radar (SAR) satellites are widely discussed in the literature for their ability to solve the trade-off between resolution and coverage. It is usually assumed that all the satellites in the constellation follow the same orbit with a negligible orbital tube since cross-track (XT) baselines introduce major complexity to high-resolution wide-swath (HRWS) imaging. However, addressing the XT issue is crucial for the feasibility of future formation missions, as realistic orbit control conditions and collision risk considerations will surely impose such baselines. This work discusses HRWS imaging with nonzero XT baselines and derives an algorithm for combining the different channels. The approach differs from traditional methods because the combination is done after focusing, where each target is compressed to several pixels. In this manner, one can address the varying elevation locally. Further performance improvement is obtained by a data-driven definition of the forward model, which utilizes the Matching-Pursuit algorithm to identify the significant ambiguities in each area. The proposed method is highly flexible, as it locally adapts the solution to the actual backscatter and elevation of the scene. Validation and testing were achieved by simulations of X-band acquisition, showing promising results.

Coherent compact Single-Input-Multiple-Output (SIMO) SAR formations have been studied as an attractive alternative to the standard monostatic SAR systems in terms of resolution and swath size. Significant gain in performance can be achieved given that the along-track (AT) spacing is well calibrated. We propose an innovative solution to the problem of AT position knowledge, using ambiguities of isolated point targets. A theoretical model of the interferometric ambiguity phase is provided, showing the direct relation to the AT baseline. Focusing the ambiguity allows us to increase its intensity, making the algorithm feasible even in the presence of clutter. The achievable accuracy is discussed theoretically and based on simulation.