Review of the Image Reconstruction Techniques Used in Smos Data Processing

The Soil Moisture and Ocean Salinity (SMOS) mission by ESA has been delivering Level 1 data to the Level 2 users since 2010. The Level 1 processing is divided in three levels: Level 1a, comprising the conversion of the raw data from MIRAS into calibrated visibilities; Level 1b, where the calibrated visibilities are corrected from the effects of unwanted sources (for example, the Sun) and are transformed into Brightness Temperatures; finally, in the Level 1c the Brightness Temperatures are geolocated over the Earth in a discrete grid and several auxiliary parameters are computed and annotated in the L1c final product. The algorithms responsible for the Calibration (L1a), Image Reconstruction (L1b) and Geolocation (L1c) have been validated during the In-Orbit Commissioning Phase that took place in the first half of 2010. However, as the knowledge on the performance of MIRAS increases, it has become necessary to set in motion a review on the current algorithms.This paper focuses on the review of the Image Reconstruction algorithms being conducted by DEIMOS Engenharia and IRAP and it is divided in the following sections: in Section 1 the theory for the proposed alternatives with respect to the baseline algorithm is demonstrated, along with several examples to ilustrate them. All the results have been obtained using the baseline algorithm or the alternatives that have been implemented in the Level 1 Prototype Processor (L1PP); in Section 2 the impact on the Level 1 Operational Processor (L1OP) is assessed, using as reference the L1PP results; and finally, in Section 3 the main conclusions found throughout this paper are summarized.