Calibration Strategy for Compact Polarimetric GNSS-R Instruments

This work presents the bases of the Polarimetric Global Navigation Satellite System – Reflectometry (GNSS-R) calibration. Polarimetric GNSS-R is being proposed by different teams as a solution to directly estimate soil moisture by analyzing polarimetric ratios without considering any further compensation. Up to date, two approaches have been proposed to estimate soil moisture, using the ratio between the right-hand and left-hand circularly polarized received signal, and using the ratio between the linear horizontal and vertical component at a set of incidence angles, known as Hybrid Compact Polarimetric (HCP) GNSS-R. In this manuscript, the necessary calibrations of a received HCP GNSS-R signal are presented for computing the Stokes parameters. A methodology to calibrate the receiver effects, scene-antenna polarimetric miss-alignment, Faraday rotation, and transmitter non-idealities is proposed. Finally, the calibration performance is evaluated using several statistical parameters and polarimetric ratio models based on two different soil moisture products. Results show a correlation coefficient increase of 4.7% with respect to the model derived from the Soil Moisture Active Passive (SMAP) soil moisture product, and a 6.5% improvement with respect to the model derived from the European Center for Medium-Range Weather Forecasts (ECMWF) Reanalysis v5 (ERA-5) soil moisture product. Furthermore, the proposed calibrations show an unbiased root-mean-square error reduction of 6.3% and 6.8% for the SMAP soil moisture model and the ERA-5 model, respectively. Due to the SMAP antenna and pointing design, SMAP-R is implicitly insensitive to most corrections. More significant corrections should be expected for general polarimetric instruments.