南海风生正压环流动力机制的数值研究

NUMERICAL STUDY ON THE DYNAMICS OF WIND DRIVEN BAROTROPIC CIRCULATION IN THE SOUTH CHINA SEA

利用ECOM si模式 ,1 0′× 1 0′水平分辨率 ,垂向 2 0个σ层 ,由H/R( 1 983)气候学月平均风应力场和开边界流量驱动 ,模拟了南海风生环流的季节变化 ,并针对南海冬夏季风生正压环流的动力机制进行了数值实验。实验中考虑以下动力因子对南海冬夏季环流的影响 :1 )开边界入流和出流 ;2 )风应力旋度 ;3)地形 ;4)惯性效应 ;5 ) β效应。数值实验表明 ,通过开边界进入南海的流量与风应力在南海内部引起的流量量值相当 ,特别是冬季两者对北部陆坡边界流和南海西边界流均有重要贡献 ;冬季南海海盆尺度气旋式流圈主要是由风应力旋度引起的 ,但平均风应力可以加强卡里马塔海峡的出流 ,而北部反气旋风应力旋度可引起南海暖流 ;陆坡地形使得海盆尺度冬季气旋式流圈中心限制在深海区 ,南海北部陆架的存在大大削弱了南海暖流的强度 ;惯性效应对南海环流的整体结构无明显影响 ,但使得黑潮入侵和台湾西南的流套变弱 ;深海海盆环流中 β项是与风应力旋度平衡的基本项 ,且

Wind-driven circulation and its dynamic mechanisms in the South China Sea(SCS) were simulated using ECOM-si model with 10′×10′ horizontal resolution on 20 vertical sigma levels. The present work mainly focuses on barotropic circulation in winter. The wind stress of the Hellerman-Rosenstein (1983) climatology was adopted to drive the circulation. Previous studies have shown that the circulation in the intermediate to deep levels in the SCS is relatively weak and is not closely related to the wind force, the water depths were thus truncated at 400m depth. The model was first forced by annually cyclic wind stress to produce monthly mean stream function for January. Then the model was run again forced by steady January wind stress to reach a stationary stream function. The open boundary conditions used in these two cases were the same on the basis of the existing estimates. Very little difference was revealed by the comparison between these two model products. This indicates that the time-dependent terms in governing equations are not significant in the SCS monthly mean barotropic circulation. Thus in the following experiments only steady wind stress fields were employed. The above run with steady wind stress was regarded as a standard experiment. The importance of outer-ocean forcing was examined by two experiments. The experiment with open boundaries closed but retaining the original H/R wind stress yields rather similar circulation pattern in the interior of the SCS. The experiment in which the wind stress was removed but the open boundary condition was the same as the standard experiment, resulted a boundary current along continental shelf of the northern SCS and west shore of the SCS. The maximum values of the stream functions from these two experiments are close, indicating that the outer-ocean forcing and the local wind forcing are of similar importance in regard to the strength of the upper-ocean circulation. To examine the role of the wind stress curl in the generating of SCS circulation, the H/R wind stress field was averaged horizontally over the SCS to yield a uniform wind stress field. When uniform wind stress was imposed, the model-produced stream function was quite similar to that without wind forcing. When the wind stress curl was doubled, the resulting stream function was significantly strengthened. Two experiments with flat seabed were carried out to examine the topographic effects. The model-produced stream functions display a large cyclonic gyre in the south and a small anticyclonic gyre in the north. The difference between the circulation patterns from the flat bottom experiment and the standard experiment implies that the continental shelf in the northern SCS restricts the extension of the anticyclonic gyre, and thus weakens the SCS Warm Current. The existence of the Sunda Shelf forces the cyclonic gyre to retrogress within deep basin. The Nansha Islands may generate meso-scale eddies. The experiment with the advection terms removed from the governing equations yields almost the same circulation pattern to that of standard experiment, except that the loop current southwest of Taiwan is slightly enhanced. In case that the β effect was removed, that is, the meridional variation of the Coriolis parameter was ignored, the model produced quite different circulation pattern from that of the standard experiment. Firstly, without β effect the Kuroshio almost no longer intrudes into the SCS. Secondly, the intensification of the stream function toward the northern slope and western boundary no longer presents. Thirdly, the cyclonic gyre has a significantly increased value. The first and second differences indicate that the β effect plays an essential role in the westward intensification of the SCS circulation. The third difference implies that the β term is a fundamental term in the balance with the wind stress curl. In conclusion, the numerical study showed that the transport induced by inflow/outflow across the open boundaries is comparable to that induced by local wind stress curl over the SCS. T