Multidimensional SAR tomography for complex non-stationary scenes: COSMO-SkyMed urban and P-band forest results

Abstract

Multibaseline (MB) SAR tomographic (3D) elevation beamforming, i.e. spatial spectral estimation, is a promising technique in the growing field of advanced interferometric SAR methods for sensing complex scenarios with multiple (layover or volumetric) scatterers mapped in the SAR cell. Recently, the Tomo concept has been integrated with the differential interferometry concept, producing the new "differential tomography" (Diff-Tomo, "4D") processing mode. Advances in the experiments of these new frameworks are presented for complex multiple scattering scenarios, also with temporal signal variations, both from scatterer deformation motions and temporal decorrelation. Results are reported of Tomo/Diff-Tomo single-look superresolution and light-burden processing in urban areas, with the new generation high-resolution COSMOSkyMed data. Moreover, new results are shown concerning the innovative capability of Diff-Tomo of analyzing volumetric forest scenarios, based on the original concept of the "space-time" signatures of temporal decorrelation. E-SAR P-band analyses are reported about separation of the canopy and ground temporal decorrelation mechanisms.

Forest temporal decorrelation: 3D analyses and processing in the Diff-Tomo framework

“…Also, the first large-scale P-band separation of height-varying temporal coherences [15,16] is expanded to several hectares of the boreal area airborne data. This extended "stratified" analysis is obtained with the new semi-parametric (AR temporal model) [15] unique Diff-Tomo functionality of separation of temporal decorrelation levels of canopy and ground employed in [14][15][16]. Moreover, Diff-Tomo separation of temporal coherences is extended to multiple polarizations, and a phenomenological interpretation is discussed consistent with typical scattering mechanisms.…”

“…This new coherent data combination framework allows, through proper spatialtemporal spectral estimation, the production of "space-time" signatures of multiple scatterers inside the SAR cell. In particular, it enables joint resolution of multiple heights and * UniPI patent (subsidence) deformation velocities of layover scatterers; this has got many successful applications to monitoring garbled urban areas [2,13,14].…”

Three-dimensional and higher-order imaging with tomographic SAR: Techniques, applications, issues

“…4-D (3-D + time, multidimensional) SAR imaging. To handle frequent heavy ambiguities from the very sparse baseline-time sampling, 2D superresolution methos have been developed, Capon [21], MUSIC, also at singlelook [22], and CS [23]. Typical Diff-Tomo application is the reconstruction and monitoring of urban/infrastructure deformating scenarios in presence of layover [24], [25], [23], [22].…”

“…While for the case of multiple sparse point-like scatterers in layover affected by (possibly different) uniform deformation motions the space-time spectrum to be identified by 4-D Diff-Tomo processing is discrete, the differential tomographic sensing concept is not intrinsically limited to such scenarios. More generally, Diff-Tomo can identify scattering component distributions in the domain of spatial (height) and temporal frequency of harmonics in which a signal from a scatterer with temporal decorrelation [21], [22] or non-uniform motion [24], [22] can be decomposed, also avoiding their misinterpretation in tomographic processing [22], [26], and potentially allowing new mapping and monitoring possibilities. An example of 4-D differential tomography output and a validation of 4-D extracted information are reported in Fig.…”

“…To handle frequent heavy ambiguities from the very sparse baseline-time sampling, 2D superresolution methos have been developed, Capon [21], MUSIC, also at singlelook [22], and CS [23]. Typical Diff-Tomo application is the reconstruction and monitoring of urban/infrastructure deformating scenarios in presence of layover [24], [25], [23], [22]. While for the case of multiple sparse point-like scatterers in layover affected by (possibly different) uniform deformation motions the space-time spectrum to be identified by 4-D Diff-Tomo processing is discrete, the differential tomographic sensing concept is not intrinsically limited to such scenarios.…”