Sporadic low salinity signals (SLSS) in the mixed layer observed by the KEO buoy: identification and characterization

To improve one-dimensional turbulence closure models (hereafter 1D-model) used in OGCMs, we have analyzed the consistency of the vertical profiles of temperature and salinity and air-sea fluxes as observed by the Kuroshio Extension Observatory (KEO) buoy. The advantage of the KEO buoy data is the successful measurement of extreme weather events, such as typhoons and storms with hourly time-resolution in the past 14 years. We have selected 7 typhoons for comparing the KEO-buoy data and the results of 1D-model experiments. This 1D-model incorporates the effects of both mesoscale eddies and internal waves by referring to the observed undulation of the seasonal thermocline. Comparisons of the buoy data and the model results indicate that the 1D-model cannot reproduce sporadic low salinity signals (SLSS) in the mixed-layer as observed by the KEO buoy. SLSS often appear at the period of typhoon passages. We have confirmed that the precipitation data of the KEO buoy are statistically consistent with the satellite precipitation data. This indicates that SLSS are caused by processes other than on-site precipitation. SLSS may be regarded as a metric for three-dimensional processes in the mixed-layer such as interaction between submesoscale eddies and barrier layer. For the concept of SLSS to be useful, we propose a definition for them based on a wavelet analysis for the time-series of 10m-depth salinity with a band of 48-96 hours. The wavelet transform enables to identify sporadic signals that are present only in salinity and not in temperature. Then we have revisited the 14-year KEO buoy data to identify the statistical characteristics of SLSS, that provides a basis for the development of working hypotheses to simulate SLSS in numerical models.