Overview of the SMOS Sea Surface Salinity Prototype Processor

Abstract

The L-band interferometric radiometer onboard the Soil Moisture and Ocean Salinity mission will measure polarized brightness temperatures (Tb). The measurements are affected by strong radiometric noise. However, during a satellite overpass, numerous measurements are acquired at various incidence angles at the same location on the Earth's surface. The sea surface salinity (SSS) retrieval algorithm implemented in the Level 2 Salinity Prototype Processor (L2SPP) is based on an iterative inversion method that minimizes the differences between Tb measured at different incidence angles and Tb simulated by a full forward model. The iterative method is initialized with a first-guess surface salinity that is iteratively modified until an optimal fit between the forward model and the measurements is obtained. The forward model takes into account atmospheric emission and absorption, ionospheric effects (Faraday rotation), scattering of celestial radiation by the rough ocean surface, and rough sea surface emission as approximated by one of three models. Potential degradation of the retrieval results is indicated through a flagging strategy. We present results of tests of the L2SPP involving horizontally uniform scenes with no disturbing factors (such as sun glint or land proximity) other than wind-induced surface roughness. Regardless of the roughness model used, the error on the retrieved SSS depends on the location within the swath and ranges from 0.5 psu at the center of the swath to 1.7 psu at the edge, at 35 psu and 15 • C. Dual-polarization (DP) mode provides a better correction for wind-speed (WS) biases than pseudofirst Stokes mode (ST1).

New methodology to estimate Arctic sea ice concentration from SMOS combining brightness temperature differences in a maximum-likelihood estimator

“…We correct T B for a number of standard contributions such as geomagnetic and ionospheric rotation and atmospheric attenuation (Zine et al, 2008). The galactic reflection is not significant at high latitudes, and no correction was applied.…”

Assessment of the SMOS inversion scheme for salinity and wind speed retrieval purposes

“…Other results are summarized in Tables 2 and 3, where the retrieval error analysis is carried out in terms of μ, RMSE and STD values by considering the optimal SSS and U 10 retrieval cost function configurations, respectively, with reference to a more realistic case study in which the simulation accounts for the different range of θ, N obs , and σ TB values according to several SMOS swath configurations. In particular, simulations are carried out at nadir and 300km cross-track location, as well as at the Alias-Free (AF) and the Extended Alias Free (EAF) Field of View (FOV) [Zine et al, 2008]. The results in Tables 2 and 3 refer to nadir simulations.…”

“…the SSS, the sea surface temperature (SST), and the sea surface roughness (SSR). As a matter of fact, the operational SMOS MIRAS TB measurements can be used in a multi-parametric inversion scheme to retrieve not only the SSS, but at the same time SST and sea surface roughness related parameters (P rough ), such as the sea surface wind speed (U 10 ) [Zine et al, 2008]. The sensitivity of SMOS TB measurements to U 10 substantially increases at high-wind conditions for which fair quality winds may be retrieved, as previously demonstrated for higher frequency radiometers such as the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) and the WindSAT one [Quilfen et al, 2007].…”

“…SST, U 10 , friction velocity or other roughness descriptors, etc.) is accomplished in the Data Product Generation System (DPGS) by means of the SMOS Level 2 Ocean Salinity processor (L2OS), which provides consistent retrievals of the above-mentioned parameters by efficiently processing geo-located TBs provided at the SMOS Level 1C (L1C) after the image reconstruction step, as described in [Zine et al, 2008]. The retrieval is conventionally accomplished by using a multi-parametric minimization algorithm based on the Levenberg-Marquardt (LM) procedure [Marquardt, 1963].…”

Resolving the Global Surface Salinity Field and Variations by Blending Satellite and In Situ Observations

“…Each viewing angle has different horizontal and vertical polarized surface TB responses to SSS, SST and wind. This information is exploited with a maximum likelihood estimate algorithm to derive SSS, SST and a wind parameter simultaneously [14], [15] and [16]). There is also a parallel effort to develop an alternative algorithm based on neural networks.…”