Synthesizing high-resolution satellite salinity data based on multi-fractal fusion

The spaceborne platform has unprecedently provided the global eddy-permitting (typically about 0.25°) products of sea surface salinity (SSS), however the existing SSS products can hardly resolve mesoscale motions due to the heavy noises therein and the over-smoothing in denoising processes. By means of the multi-fractal fusion (MFF), the high-resolution SSS product is synthesized with the template of sea surface temperature (SST). Two low-resolution SSS products and four SST products are considered as the source data and the templates respectively to determine the best combination. The fused products are validated by the in situ observations and intercompared via SSS maps, Singularity Exponent maps and wavenumber spectra. The results demonstrate that the MFF can perform a good work in mitigating the noises and improving the resolution. The combination of the climate change initiative SSS and the remote sensing system SST can produce the 0.1° denoised product whose global mean standard derivation of salinity against Argo is 0.21 and the feature resolution can reach 30–40 km.