Multidimensional SAR Imaging: Studies in the Framework of LIMES Project
Mario Costantini(1), Gianfranco Fornaro(2), Fabrizio Lombardini(3), Matteo Pardini(3), Francesco Serafino(2) and Francesco Soldovieri(2)
(1) Telespazio S.p.A., Via Tiburtina, 965, 00156 Rome, Italy
(2) CNR IREA, Via Diocleziano, 328, 80124 Naples, Italy
(3) University of Pisa, Via G. Caruso, 16, 56122 Pisa, Italy
In Earth Observation, unique capabilities are associated with the use of SAR and, particularly, with the extensions of SAR to interferometric modes and more generally to the joint use of coherent multiple acquisitions.
By using phase interference of two (SAR Interferometry) or more (multibaseline SAR interferometry) views, accurate DEMs can be generated. Furthermore, acquiring images at different time intervals (multitemporal or multipass SAR) precise tracking of the velocity of ground deformations at an accuracy of the order of mm/yr can be achieved. Known techniques essentially use only the phase information. Moreover, to work properly, they assume the scattering mechanism to be of a basic nature: i.e., dominated by a permanent scatterer or distributed on the ground surface. However, when the radiation penetrates under the surface, a situation that may even occur with existing sensors over specific surfaces, or ground topography is steep enough to generate critical projection of the scatterers in the slant imaging geometry (layover), or the imaged area is characterized by a high spatial density of strong scatterers, the signal received in a pixel may contain the superposition of responses from multiple scatterers. The latter two conditions are frequent when data are acquired over complex scenarios such as urban areas or large infrastructures. Precise target height estimation, and thus precise target geolocation, as well as the maximization of the number of tracked scatterers may be issues of primary importance to improve monitoring over complex structures.
SAR 3D Tomography is a way of overcoming limitations of standard algorithms for target height determination by achieving focused fully 3D images, starting from multibaseline data. In this framework, amplitude information is useful in addition to phase, to exploit beating phenomena in the separation of the multiple signal components, and to enhance statistical accuracy even for a single scatterer. With this regards, use of proper monodimensional inversion techniques of the acquired data in the baseline domain allows mitigating difficulties arising from non-uniform baseline sampling.
Recently proposed Differential SAR Tomography, i.e. extension to 4D (elevation/velocity) SAR imaging, may also allow estimating possible relative deformations between multiple layover scatterers. In this framework, proper two-dimensional inversion techniques of the sparse data in the baseline-acquisition time domain are of paramount importance for the new technique to get unambiguous and accurate results.
This work, performed in the framework of the FP6 LIMES project, aims first of all to present the current status of the 3D and 4D (MultiD) SAR imaging from the space and to describe the results obtained by processing real ERS-1/2 data, with a particular emphasis to the demonstration of the possibility of localizing precisely targets and of separating multiple scatterers starting from the analysis of the reconstructed scattering profiles in the direction of elevation, and in the elevation/velocity domain. Latest algorithmic developments that we have obtained will be also presented in the full paper.
It is expected that potentials and applications of these techniques will grow with the ongoing development of new high resolution SAR sensors and satellite clusters, allowing rich and frequent data acquisition. We thus also discuss perspectives in this framework by presenting the results of an analysis, based on simulations and analytical performance bounds, that aims at describing the potentiality related to the use of such techniques with respect to future satellite clusters operating in multimonostatic or multistatic mode and that thus will allow acquiring simultaneous multibaseline data and repeated in time. In particular in this analysis we focus on the study of characteristics of the imaging, and particularly to the control of sidelobes, with respect to the number and location of the antennas that simultaneously observe the scene.
 P. Pasquali, C. Prati, F. Rocca, M. Seymour, J. Fortuny, E. Ohlmer, A.J. Sieber, “A 3D SAR Experiment with EMSL Data,” Proc. 1995 IEEE Int. Geosci. and Remote Sensing Symp, pp.784-786, Firenze, Italy, 1995.
 P.A. Rosen, S. Hensley, I.R. Joughin, F.K. Li, S.N. Madsen, E. Rodriguez, R.M. Goldstein, “Synthetic Aperture Radar Interferometry,” Proc. of the IEEE, vol. 88 (3), pp. 333-382, March 2000.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,