LOESS四维客观分析在中国近海的应用

Application of LOESS Four-dimensional Objective Analysis to the China Sea

利用全国海洋综合调查资料和世界大洋数据库(WOD)资料,运用局部加权回归法(LOESS)四维客观分析方案对中国近海的温度场进行了客观分析,将散布在空间中的温度资料点插值到均一的网格中。在分析中充分考虑了中国海曲折复杂的海岸特点,将底深调整系统(TAR)和障碍调整系统(BAR)整合到四维插值系统中,使得资料点空间权重的计算更为合理。计算结果显示,LOESS客观分析方案充分考虑了空间因素(经度、纬度、深度)和时间因素,能客观反映出中国近海一年四季的温度时间-空间结构。通过与Kriging插值方法的对比,发现LOESS方法能更精确地把中国近海温度结构反映出来。

In this paper, the National Marine Comprehensive Investigation Data and World Ocean Database （WOD） are used to analyze the temperature field in the offshore China Sea using the locally weighted regression （LOESS） four-dimensional objective analysis scheme, and the scatteringly distributed temperature data are then interpolated onto uniformly spaced grids. To take the complicated coastline of the China Sea into full consideration, the topographic adjusted relief （TAR） and barrier adjusted relief （BAR） are integrated into the four-dimensional interpolation system to make the spatially weighting calculation of data points more reasonable. It is shown from the analysis results that the LOESS objective analysis scheme has taken the spatial and temporal factors into full consideration, and can objectively reflect the seasonal spacetime temperature structure in the offshore China Sea. Compared with the Kriging interpolation scheme, the LOESS sheme can reflect the temperature structure in the offshore China Sea more accurately.