Figure 1: Time series of the root-mean-square error of the modeled SSS field compared with the Argo data for the four experiments from 2011 to 2013. (a) Global average, (b) average within 20°S–20°N, (c) Northern Ocean average within 20°N–60°N, and (d) Southern Ocean average within 60°S–20°S. The unit is pss. From Lu et al., 2016

 

A study recently published in JGR ocean by Lu et al. (2016) attended to answer how can SMOS SSS data benefit ocean model performance? Previous studies found contradictory results. In this work, the authors assimilated SMOS-SSS data into the LASG/IAP Climate system Ocean Model (LICOM) using the Ensemble Optimum Interpolation (EnOI) assimilation scheme. To assess and quantify the contribution of SMOS-SSS data to model performance, several tests were conducted.

 

The key topic of their study is the examination of whether SMOS-SSS data can improve model simulations, i.e., whether the error in model simulations can be reduced after including SMOS-SSS data. To answer this question, four experiments were implemented within the period of January 2011 to December 2013.

1.       Non_assim: this was a model control run from January 2011 to December 2013 without assimilating any data. The experiment indicated the performance of the model for ocean state simulations.

 

2.      Assim_sss: in this experiment, we assimilated only the SMOS-SSS data (CEC) into the model. Compared with Non_assim, the experiment indicated whether the SMOS-SSS data could improve model simulations (especially for salinity).

 

3.      Assim_other: we assimilated the three different data sets: SST, SLA, and T/S profiles into the model. These data are introduced in section 2.2, including OISST, SLA, and ocean subsurface temperature and salinity profile data (in EN4). Most of the available data assimilation systems in previous studies included those three ocean data sets. Therefore, another important task was to investigate whether SMOS-SSS data can play a complementary role in the current data assimilation system when traditional observational data are assimilated.

 

4.      Assim_all: we assimilated all of the available data, including the SMOS-SSS, SST, SLA, and T/S profiles. Compared with Assim_other, this experiment indicated whether the inclusion of SMOS-SSS data can further improve the model simulation of a traditional data assimilation system.

 

 The results indicate that the CECOS/CATDS 2010.V02 SMOS-SSS product can significantly improve model simulations of sea surface/subsurface salinity fields. This study provides the basis for the future assimilation of SMOS-SSS data for short-range climate forecasting. 

 

 Lu, Z., L. Cheng, J. Zhu, and R. Lin (2016), The complementary role of SMOS sea surface salinity observations for estimating global ocean salinity state, J. Geophys. Res. Oceans, 121, doi:10.1002/2015JC011480.