Speakers
Description
SMOS has been in orbit for more than 8 years, which has allowed to learn significantly on the reconstruction process of the synthetic aperture radiometer and the influence of the correspondent calibration parameters. One of the remaining artefacts to be corrected is the spatial ripple that can be observed in the brightness temperature images even if in perfect calibration conditions and perfect knowledge of the antenna patterns. Its origin is due to the fact that in the reconstruction process there are more unknowns than observables and is basically affected by the spacing between the antennas and the similarity between them. It has been demonstrated that closer antennas, would, first, enlarge the alias-free region, and, second, would reduce the level of the undesired spatial modulation that affects the images. In the same way, more similar antenna patters will also significantly reduce the level of reconstruction errors and hence of this modulations.
In the first case, there is a clear limitation on the distance that the L band antennas can be located due to the physical sizes. An improvement from the current SMOS design could be achieved if located from the current 0.875 lambdas separation in SMOS to the closer 0.767 lambdas proposed for the follow-on SMOS-Ops instrument. Regarding antenna similarity, it has been seen that there is a clear dependency of the patterns on the electromagnetic environment that affects the antennas once mounted on the arms of the instrument, which have a strong shape dependency on the geometry of the arm, and hence on boundary conditions imposed by the geometry itself and the coupling between antennas located in different positions with respect to the polarization axis for each of the arms.
The objective of this project is to experimentally demonstrate the improvement on image quality reconstruction by reducing the distance between the antennas and minimizing the pattern dissimilarity. This last goal is achieved surrounding the antennas by other ones identical (dummies), properly arranged on the hexagonal instrument structure, which is the one proposed for a future SMOSops follow on. Three different array configurations have been considered to characterize the embedded antennas: 1) a reference array, where 13 antennas are located in two contiguous segments of a hexagonal array following the SMOSops geometry surrounded by a ground plane. 2) An array, where the previous 13 antennas are surrounded by dummy matched antenna elements separated 0.767 lambdas. A total number of 87 radiating elements is considered, since the experience gained during the antenna ground characterization of MIRAS showed that the radius of the scattering area of one particular element, and hence the mutual coupling, spans up to 3 times the element spacing, which sets the amount of dummy antennas that is needed to surround every active element to have the same boundary conditions. 3) An extended target array, a modification of configuration 2 by adding and extra element, to confirm the size of the scattering region.
| ESA Technical Officer | Manuel Martín-Neira. |
|---|
