Abstract
Oceanographic datasets, which are arranged for standard depths, have many applications for various users. However, oceanic observations are not always conducted exactly at standard depths, especially in the case of historical bottle observations. Therefore, interpolation for density data is usually calculated from interpolated temperature and salinity data. However, in areas such as the Kuroshio–Oyashio mixed water region, east of Japan, the oceanic conditions are extremely complex, and pseudo-density inversions are often generated during the interpolation procedure. The authors have designed a new scheme based on parabolic or cubic interpolation, to avoid the generation of pseudo-density inversions, that uses temperature and density as independent variables. If the generation of a density inversion in the parabolic or cubic interpolation is found, it is replaced by a linear interpolation. The salinity calculated from interpolated temperature and density sometimes lies outside the range between the salinities at the observed levels above and below the standard depth; therefore, the scheme is designed so as to avoid the generation of such unacceptable salinity values.
Corresponding author address: Dr. Sachiko Oguma, Marine Information Research Center, Japan Hydrographic Association, 5-3-3 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan. Email: oguma@mirc.jha.jp
