Japan Sea thermohaline structure and circulation. Part III: autocorrelation functions

The autocorrelation functions of temperature and salinity in the three basins (Ulleung, Japan, and Yamato Basins) of the Japan/East Sea are computed using the U.S. Navy’s Master Oceanographic Observational Dataset for 1930–97. After quality control the dataset consists of 93 810 temperature and 50 349 salinity profiles. The decorrelation scales of both temperature and salinity were obtained through fitting the autocorrelation function into the Gaussian function. The signal-to-noise ratios of temperature and salinity for the three basins are usually larger than 2. The signal-to-noise ratio of temperature is greater in summer than in winter. There is more noise in salinity than in temperature. This might be caused by fewer salinity than temperature observations. The autocorrelation functions of temperature for the three basins have evident seasonal variability at the surface: less spatial variability in the summer than in the winter. The temporal (spatial) decorrelation scale is shorter (longer) in the summer than in the winter. Such a strong seasonal variability at the surface may be caused by the seasonal variability of the net surface heat flux. The autocorrelation functions of salinity have weaker seasonal variability than those of the temperature field. The temporal and horizontal decorrelation scales obtained in this study are useful for designing an optimal observational network. 1