We begin with a discussion of the lineage of the SMAP SSS algorithms and model functions from Aquarius, showing excellent agreement for the ocean surface roughness correction. Next we discuss positives and negatives of a conically-scanning radiometer such as SMAP as compared to a push-broom platform like Aquarius. Then, we give an overview of some newly developed algorithms that are only relevant to the SMAP system, such as a new galaxy correction and a land correction enabling SMAP SSS retrievals to be done up to 40 km from coast, which have been implemented in the JPL version 3 SSS algorithm. The JPL SMAP SSS data, validated with in situ ARGO, tropical moorings, and ship-based observations, are available from the JPL Physical Oceanography Distributed Active Archive Center (PO.DAAC).
Version 4 of the JPL SMAP SSS data product contains a number of recent algorithm improvements. We have implemented and validated an expected SSS uncertainty estimate for each L2B retrieval and L3 SSS retrieval. This will aid efforts to optimally blend SMAP SSS with other SSS such as from SMOS. We have modified the ancillary numerical weather products used in the SSS algorithm and have significantly improved performance in high-wind conditions. Other changes include an extended range of SSS retrievals up to 45 PSU, retrievals in land-locked inland lakes, improved land correction, and reduction of biases at high latitudes.
Finally we discuss the various L2B and L3 datasets available from SMAP and where they may be obtained.