eSQG方法在南海垂向流速诊断中的应用研究

本文利用OFES模式0.1°×0.1°高分辨率温盐、流速和海面高度（SSH）数据，分析eSQG方法在南海垂向流速诊断中的适用性和南海垂向流速的时空变化特征。结果表明，基于SSH和eSQG的诊断垂向流速ω_eSQG与OFES模拟“真值”垂向流速ω_OFES整体量级基本相当，为O（10?5 m/s），空间上呈现北强南弱。深海盆（水深大于1 000 m）ω_eSQG与ω_OFES的水平空间分布的相关系数rs在次表层150 m左右达到最大，高于南海全海域的空间相关系数，表明eSQG在远离边界的深水区更有效。季节上，ω_eSQG总体夏强冬弱，与ω_OFES的相关系数rs冬季大、夏季小，表明eSQG更适用于冬季南海垂向流速的诊断。eSQG诊断垂向流速在台湾西南部和越南以东适用性更高，与ω_OFES的时间相关系数rt超过0.6；海盆南部和西北部陆架区诊断效果较差，时间相关系数rt多小于0.2。同一区域ω_eSQG和ω_OFES空间相关系数存在18～55 d的周期变化。分析显示，海面高度与海面密度同相位分布时，ω_eSQG与“真值”ω_OFES更接近。时间分辨率对eSQG诊断效果几乎无影响，空间分辨率降低到0.25°时rs整体增大，说明eSQG在中尺度诊断效果更好。

Application of eSQG method in vertical velocity diagnosis in the South China Sea

Using 0.1°×0.1° high-resolution temperature, salinity, velocity and sea surface height (SSH) data from the ocean general circulation model for the earth simulator (OFES) model, this study analyzes the capability and applicability of the eSQG (effective Surface Quasi-Geostrophy) method in vertical velocity diagnosis in the South China Sea (SCS), as well as the spatiotemporal variation of vertical velocities. The diagnosed vertical velocities ω_eSQG from SSH with the eSQG method are of the same order of 10?5 m/s as the “true” vertical velocities ω_OFES from the OFES model. ω_eSQG shows spatial variations with higher values in northern basin. The correlation coefficients of the horizontal distribution of ω_eSQG and ω_OFES (rs) are greater in deep basin than that in the whole SCS, suggesting that the eSQG method is more efficient in vertical velocity diagnosis in deep water far from boundaries. Vertically, the correlation coefficient has maximum values occurring in the subsurface layer at about 150 m. ω_eSQG is stronger in summer and rs show seasonal variation with higher values in winter, indicating more efficient in eSQG diagnosis in winter. ω_eSQG is reliable in the regions southwest of Taiwan and east of Vietnam, where the temporal correlation coefficients of ω_eSQG and ω_OFES (rt) exceed 0.6, while ω_eSQG is poorly correlated to ω_OFES in the shelf regions in the southern and northwestern SCS with rt mostly under 0.2. rs in the same region is varying at periods of about 18?55 d. ω_eSQG performs better as the distributions of the SSH and the sea surface density are in same phase. ω_eSQG varies little as the temporal resolution of SSH varies, while rs increases as spatial resolution reduced to 0.25° in mesoscales.