【作者】 黄艳松；

【导师】 宋金宝；

【作者基本信息】 中国科学院研究生院（海洋研究所） ， 物理海洋， 2011， 博士

【摘要】 海气界面通量对海气相互作用、海气耦合模式、海洋与大气动力环境预报等具有非常重要的作用。本文采用三重循环拟合的方法,提出了一种获取超声风速仪和浮标坐标系之间角度偏移矩阵及观测风速平台晃动校正的方法,并以黄海北部连续14天的浮标观测资料为例,研究了晃动校正对湍流能谱密度、协谱函数、Ogive曲线及海气通量的影响,结果表明:晃动校正不仅对动量通量有很大的影响,对感热通量、潜热通量、海气交换(如:水汽通量和二氧化碳通量等)也有较大的影响。晃动校正后,涡相关法和块体法计算的通量(动量和感热通量)结果之间的相关系数显著提高,且不同高度上涡相关法计算的水汽和二氧化碳通量十分接近。基于黄海北部第一次连续14天的浮标观测资料,采用多尺度分解法确定了海气通量涡相关法计算中的截断时间尺度,并分析了该截断时间尺度的特征及其对感热通量计算的影响。研究表明:由多尺度分解法获得的湍通量截断时间尺度可将总通量中湍通量和中尺度通量分离开来,截断时间尺度随着湍流强度或水平风速的增加而增加,且感热通量的截断时间尺度大于动量通量的截断时间尺度;在弱湍流状态下,动量通量和感热通量的截断时间尺度主要分布在100s左右,湍通量和总通量之间差别较大;在湍流较强时,动量通量和感热通量的截断时间尺度主要分布在800s左右,湍通量和总通量之间差别较小;通过计算和分析不同截断时间尺度下感热通量的平均值和不确定性,可以看出截断时间尺度对感热通量的平均值的影响较小,但对感热通量的不确定性有很大的影响,这两种影响之间的差别会随着湍流强度的增加而减小。对黄海北部第二次定点浮标观测资料和美国国家环境预报中心公布的NCEP1、NCEP2再分析资料进行了比较和分析。结果表明:NCEP再分析资料中的海表气象参数是可信的,NCEP2与NCEP1相比要更接近于浮标观测值,而净辐射通量则是NCEP1更接近于浮标观测值;使用更合理的块体公式计算出的湍流热通量来代替NCEP数据库中已计算好的湍流热通量来驱动海洋模式是很有必要的;与浮标计算结果相比较,NCEP1的再计算净热通量低估了42%,而NCEP2的再计算净热通量高估了5%。更多还原

【Abstract】 Air–sea fluxes play important roles on air-sea interactions, coupled ocean–atmosphere modeling, dynamic environment prediction of marine and atmosphere.In this paper, a flux system deployed on the buoy has been described which is capable of directly estimating the air-sea fluxes after removing the contamination in the signal due to the buoy motion. A triple loop fitting method has been demonstrated for determining the three angular offsets between measurement axes of the sonic anemometer and motion pack. The data collected during the experiment in the North Yellow Sea is used to discuss the effect of buoy motion on the turbulence spectra, frequency-weighted cospectral density function, Ogive curve and fluxes. Results show that the motion correction not only greatly improved the estimation of the momentum flux but also has large impact on calculated sensible heat flux, latent heat flux and gas exchange (such as water flux and CO2 flux). The correlation coefficients of fluxes calculated by the eddy correlation method and bulk aerodynamic method are greatly improved after motion correction. The water vapor and CO2 fluxes at two levels also become similar after motion correction.Based on a total of 14 days’data obtained from the 1st moored buoy observation located at the Yellow sea, the cutoff time scale (CTS) used in the calculations of the air-sea fluxes by eddy covariance method is determined by the multiresolution decomposition method, the behaviors of the CTS and its effect on the sensible heat flux are analyzed then. Results show that the cutoff time scale estimated by the multiresolution decomposition method can be used to separate the real contribution from turbulent and mesoscale fluxes to the total air-sea fluxes efficiently. The CTS increases with the turbulence intensity or the wind velocity and the CTS of the sensible heat flux is always longer than that of the momentum flux. When the turbulence intensity is less than 0.2 m/s, most of the CTS of momentum and sensible heat fluxes are around 100 s and the difference between the turbulent and total fluxes is large; when the turbulence intensity is larger than 0.3 m/s, most of the CTS are around 800 s and the turbulent flux is similar to the total flux. It is found that the CTS has small effect on the flux average value while large effect on the flux uncertainty when we calculate and analyze the average and uncertainty of sensible heat flux. The difference between the two effects will become smaller as the turbulence intensity increases. The NCEP (NCEP1 and NCEP2) reanalyzed marine meteorological parameters and sea surface heat fluxes are compared with the 2nd moored buoy observation in the north-western Yellow Sea. The analysis shows that: The NCEP reanalyzed marine meteorological parameters are credible over the Yellow Sea. The marine meteorological parameters and turbulent heat fluxes from NCEP2 are more close to that from the buoy while the net radiation flux computed from NCEP1 is more close to that from the buoy. A more appropriate bulk algorithm to recalculate surface heat fluxes is recommended. The NCEP1 recalculated net heat flux is underestimated by 42% while the NCEP2 recalculated net heat flux is overestimated by 5%.更多还原