Figure 7. L2 Triple point analysis spatial error variance for SMAP V5.0(a,b), SMOS 700(c,d), and EN4 in situ profiles (e,f) for the minimal flags(a,c,e) and all flags(b,d,f) scenarios.

The triple point method has been applied to evaluate errors for SMAP V5.0 satellite salinity as part of product validation efforts. Spatial error variances for equal area aggregated and monthly binned data are shown above. Blank locations are regions where the sensitivity of the datasets is more than the dataset variance. Errors are smallest at low latitudes where waters are warm. SMAP error variances are ~2 times larger north and south of 55° as they are at the equator. SMAP has smaller errors compared to SMOS at all latitudes as well as in the global average. The effect of applying all flags does not reduce errors drastically in each individual region but has a large effect on the global average by removing regions with large errors next to the coast and at high latitudes.