【作者】 张增海；

【导师】 王辉； 左军成；

【作者基本信息】 中国海洋大学 ， 物理海洋， 2006， 硕士

【摘要】 海-气间湍流热通量是研究海气相互作用的重要课题,深刻理解海气间湍流热交换有助于进一步耦合大气海洋模式,也有助于进一步研究海洋热量和淡水收支等科学问题。但目前较常用的几套潜热通量和感热通量资料在南海海域的质量状况如何并不十分清楚。为此,文章以基于船测的通量产品SOC为参照,比较了SOC、HOAPS2、NCEP1、NCEP2、ERA40和OAFLUX共六套数据集的潜热通量和感热通量在南海6年(1988-1993)平均场结构的差异。结果显示,南海多年平均的潜热通量呈北部高而南部低的结构。在南海的大部分区域OAFLUX和SOC符合得最好,HOAPS2、NCEP1、NCEP2和ERA40比SOC分别高估5~10、10~30、20~35和15~25Wm-2,高估的最大值中心在10oN~12oN的海域。对于感热通量,在南海的大部分海域,HOAPS2、NCEP1、ERA40以及OAFLUX的干热通量与SOC相比较分别高估出2~3Wm-2,2~5Wm-2,2~7Wm-2,1~4Wm-2,NCEP2在北部海域比SOC高估1~4Wm-2,南部低估1~4Wm-2。湍流热通量年平均场在整个南海海域随时间的变化显示WHOI OAFLUX的潜热通量和SOC最为接近,除个别年份外,HOAPS2和SOC的差别也不大,模式产品的潜热通量比SOC的都高,尤其是NCEP2和ERA40。对感热通量来说,各套资料差别的绝对值都不大。但因感热通量本身量值小,故相对误差很大。其中与SOC符合最好的是NCEP2和OAFLUX,较差的是ERA40。多年平均场纬向平均结果显示,无论是感热通量还是潜热通量都是北部高南部低,其中潜热通量在10oN处有个峰值区。为探讨通量数据集中误差的来源,文中以5次南海现场观测试验数据为参照,对NCEP2再分析资料中湍流热通量在南海西南季风期的精度进行了评估.结果表明NCEP2估算的潜热通量的平均值在试验Xisha2000、Xisha1998、Kexue1和Shiyan3期间分别高估了6(11%)、2(2%)、7(7%)和13Wm-2(16%),而在Xisha2002试验中低估了10Wm-2(11%).在5个试验中低估的感热通量分别为7(130%)、更多还原

【Abstract】 The sea surface turbulent heat fluxes, as the key subjects of air-sea interaction, are very important for understanding the coupled atmosphere-ocean models and oceanic heat and freshwater budget. The accuracies of latent heat flux (LHF) and sensible heat flux (SHF) in South China Sea (SCS) are needed to check firstly because of the poor quality of datasets. Using ship-based SOC (Southampton Oceanography Center) fluxes as reference, the mean fields of turbulent heat fluxes (1988-1993) from six datasets (SOC, HOAPS2, NCEP1, NCEP2, ERA40 and OAFLUX) are compared with each other to check the quality of datasets. All products reveal a similar pattern in the averaged fields of that the LHF are high in the north SCS while low in the south part. The 6-yr mean structure difference fields of latent heat fluxes between HOAPS2, NCEP1, NCEP2, ERA40 and SOC show that the overestimation are 5~10,10~30,20~35 and 15~25Wm-2,repectively. The region of highly overestimation is from 10oN to 12oN for the four datasets. WHOI OAFLUX shows the most similar pattern with SOC. For sensible heat fluxes, we found that the overestimations of HOAPS2, NCEP1, ERA40 and OAFLUX to SOC are 2~3Wm-2,2~5Wm-2,2~7Wm-2,1~4Wm-2,resptivly. NCEP2 overestimate the SHF 1~4 Wm-2 in the north SCS and underestimate 1~4 Wm-2 in the south. The yearly mean fields of turbulent heat fluxes over the whole SCS show that the LHF of OAFLUX is the most close to SOC and the differences between HOAPS2 and SOC are small except in 1990. The LHF products based on NWP outputs are all overestimated compared to SOC, especially NCEP2 and ERA40. As far as the time changes of SHF are concerned, the absolute differences are small for all products, but the relative differences are large. SHF of NCEP2 and OAFLUX hold the best agreement with SOC and ERA40 give the worst. The comparison of zonal averages of 6-yr(1988-1993) latent and sensible heat fluxes from all six datasets show the basic pattern of high value in the north and low in the south and there is a local maximum near 10oN.The widely used turbulent heat fluxes in the NCEP2 are evaluated using three更多还原