【作者】 颜梅；

【导师】 左军成；

【作者基本信息】 中国海洋大学 ， 应用海洋学， 2008， 博士

【摘要】 本文利用TOPEX/Poseidon(T/P)卫星高度计资料,分析了全球海平面变化规律,利用Ishii(2005)的温度、盐度资料计算了热比容、盐比容海平面高度,分析了比容海平面的变化规律及它们在不同时间尺度上对海平面变化的贡献。利用POP模式做了六个数值试验,模拟研究了在不同热力强迫条件下的海平面季节变化特征,探讨全球海平面季节变化的热力学机制。主要结论如下:北半球,T/P海平面的季节变化主要由比容变化引起,二者季节变化的振幅和位相几乎相同,比容对T/P高度的季节变化贡献为92%;北大西洋盐比容海平面从2000年后有显著的下降趋势,北印度洋盐比容海平面有明显的季节变化。从长期趋势来看,T/P海平面的上升速率为2.22 mm/yr,比容海平面为0.97 mm/yr,比容对T/P海平面上升趋势的贡献为44%,其中热比容海平面上升速率为1.48 mm/yr,盐比容海平面为-0.52 mm/yr。南半球,季节尺度上比容海平面和T/P海平面基本同位相,但比容海平面年振幅大于T/P海平面,非比容因素与比容因素对总的海平面变化的作用相反,且贡献较显著;长期尺度上T/P海平面的上升速率(3.85 mm/yr)大于比容海平面(1.30 mm/yr),比容海平面上升速率的贡献(34%)小于北半球,其中热比容海平面的上升速率为1.1 mm/yr,盐比容海平面为0.21 mm/yr。热带海区,T/P海平面和比容海平面都表现出显著的年际变化,97/98年间两者都有明显的下降,具体体现为在太平洋下降而在大西洋和印度洋海平面是上升的。热带海区年周期和半年周期变化也显著,T/P海平面的年周期振动在位相上滞后比容海平面3-4个月,振幅大于比容海平面,但半年周期的振动则相反,位相上T/P海平面超前比容海平面2-3个月,比容海平面的振幅上也大于T/P海平面。1993-2003年全球T/P海平面的线性上升速率为2.89mm/yr,热比容海平面的上升速率1.22mm/yr,盐比容海平面接近0,比容对T/P海平面上升的贡献为42%。太平洋,比容变化是海平面变化的主要影响因子,其中热带太平洋比容的贡献最大(72%);北大西洋,热比容海平面的上升速率(2.49 mm/yr)同盐比容(-2.48 mm/yr)量级相当,但作用相反;非比容因素对海平面趋势变化的贡献在大西洋及印度洋的南半球海域要大于比容作用。从比容变化角度来看,除北大西洋外,热比容的作用比盐比容作用大得多。1993-2003年间比容海平面的年变化振幅的空间分布和T/P海平面较一致,北半球的T/P海平面年振幅大于南半球,最大振幅出现在黑潮及其延伸体附近,其核心值为15cm。1993-2003年T/P海平面变化经验正交函数分解(EOF)的第一模态(24.4%)是季节模态,第二模态(18.1%)为ENSO模态,比容海平面的EOF第一模态(29.3%)也是季节模态,第二模态(13.9%)是ENSO模态。1950-2003年的热比容海平面具有20年左右的显著周期变化,1970年前后开始明显的上升。全球平均的热比容的上升速率为0.29 mm/yr,盐比容为0.03 mm/yr;其中北大西洋的热比容上升最快,为0.78 mm/yr。过去50年全球热比容海平面的第一模态(EOF1)(29.5%)为ENSO模态,第二模态是PDO模态,与副热带环流的变化有关。利用POP模式研究了不同海表热力强迫对海平面季节变化的影响。通过与T/P海平面以及Ishii温盐资料计算得到的比容海平面季节变化进行分析对比,发现在海表风应力、热通量和淡水通量强迫下,海平面及比容海平面的季节变化特征较显著,与T/P海平面的标准差在中低纬度有较大差异,在北半球的中高纬度一致性较好。没有海表热通量的强迫,海表温度会变低,海表盐度严重失实,海平面的季节变化特征不显著。如果海表热通量没有季节变化特征,海平面的季节变化特征也不显著,说明海表热通量季节变化特征决定着海平面季节变化。海表淡水通量对海平面季节变化也有一定影响,但不及海表热通量显著,其季节特征对部分海域海平面的季节变化有一定影响,对太平洋影响不大。海表温度升高后,对北半球高纬海域的海平面季节变化影响非常显著。更多还原

【Abstract】 Global sea level change is estimated using TOPEX/Poseidon (T/P) satellite altimeter measurements. Thermosteric and halosteric sea level are calculated from seawater temperature and salinity of Ishii (2005), steric sea level variation and its contribution to sea level at different time scales is evaluated. Six experiments are made by using POP model to stimulate the characters of seasonal sea level variation under different conditions of thermal forcing, and to discuss the thermal mechanism of the seasonal variation of global sea level. Conclusions could be drawn as below,In the Northern Hemisphere, the seasonal variation of sea level is mainly caused by variation of steric sea level, the amplitude and phase of seasonal variation of both of them are almost the same, steric contribution to sea level at seasonal time scale is up to 92%. Halosteric sea level is obvious decreasing from 2000 in the North Atlantic, and display significant seasonal characteristic in the North India Ocean. In long term, T/P sea level has a rising rate of 2.22 mm/yr, while steric sea level has a rising rate of 0.97 mm/yr, rising trend of steric sea level accounts for 44% of rising trend of T/P sea level, in which thermosteric sea level has a rising rate of 1.48 mm/yr, while halosteric sea level has a rising rate of -0.52 mm/yr.In the Southern Hemisphere, at seasonal time scale, phase of steric sea level is approximately identical with the one of T/P sea level, but amplitude of steric sea level is larger than the one of T/P sea level, which means non-steric and steric factors have opposite effects to sea level, and non-steric effect is significant. In long term, the rising rate of T/P sea level(3.85 mm/yr) is larger than that of steric sea level (1.30 mm/yr), which accounts for a smaller percentage of T/P sea level in the South Hemisiphere(34%) than the one in northern hemisphere, and in which thermosteric sea level has a rising rate of 1.1 mm/yr, while halosteric sea level has a rising rate of 0.21 mm/yr. ,In tropical area, both T/P and steric sea level display significant inter-annual variation, between 1997 and 1998 both of them decline obviously, which is embodied by the phenomena of the decline of the Pacific and the rising of the Atlantic and the Indian ocean,; there are also annual and semi-annual oscillation in tropical area. At the annual scale, T/P sea level is 3-4 months later than steric sea level in terms of phase, and has lager amplitude than steric sea level; at the semi-annual scale, however, T/P sea level is 2-3 months earlier than steric sea level, and also has lager amplitude than steric sea level.The linear rising rate of global mean T/P sea level is 2.89 mm/yr from 1993 to 2003, rising rate of thermosteric sea level is 1.22 mm/yr, while halosteric sea level is nearly zero; rising of steric sea level accounting for 42% of rising of T/P sea level; In the Pacific steric change is the major factor of sea level variation, among which the contribution in tropical Pacific steric is maximal (72%); In the North Atlantic, the magnitude of the rising rate of thermosteric(2.49 mm/yr)is the same as that of halosteric(-2.48 mm/yr), but their effect are opposite. The contributions of non-steric factor to sea level change in the Atlantic and tropical Indian Ocean southward region are larger than that of steric ones. In terms of variation of steric sea level, effect of thermo-steric is greater than the one of halosteric expect for in northern Atlantic.From 1993 to 2003, annual amplitude of steric sea level has approximately identical spatial pattern with T/P sea level, which is lager in the Northern Hemisphere than that in the Southern Hemisphere. The maximal amplitude appears near Kuroshio and its extension, which is 15 cm.T/P sea level from 1993 to 2003 is analyzed by EOF and the results show that the first mode is seasonal mode, which accounts for 24.2% of the variance; the second mode (18.1%) is the ENSO mode. The first mode of steric sea level (29.3%) is seasonal mode, and the secend mode is ENSO mode, which spatial-pattern is much different from the second mode of T/P sea level.Steric sea level from 1950 to 2003 have about 20-years prominent period, and there is an obvious rising trend from 1970. The global rising rate of thermosteric is 0.29 mm/yr, halosteric is 0.03 mm/yr, the rise of thermosteric in the Noth Atlantic is maximal, up to 0.78 mm/yr. Thermosteric sea level in the past 50 years is analyzed by EOF and it is found that the first mode (29.5%) is ENSO mode, and the second mode is PDO mode, which is related with subtropical circulation.Six experiments are made by using POP model to study effects of different sea surface thermal forcing on seasonal variation of sea level, and their results are compared with the characters of seasonal variations of T/P altimeter data and steric sea level from Ishii temperature and salinity data, it is found that seasonal variations of sea level and steric sea level are significant under forces of sea surface wind stress, sea surface heat flux and freshwater flux; their standard deviation has greater difference with the one of T/P sea level in mid-low latitudes, while has better coherence in mid-high latitudes. Sea water temperature will get lower, salinity will be seriously inaccurate and seasonal variation of sea level will be insignificant without forcing of sea surface heat flux. If sea surface heat flux do not have seasonal character, seasonal variation of sea level will also be insignificant, which indicates that seasonal variation of sea surface heat flux decides the one of sea level. Sea surface freshwater flux has some effects on seasonal variation of sea level, but not as much as sea surface heat flux; its seasonal variation has some effects on variation of sea level in certain individual area, but not much in the Pacific. If sea surface temperature is rising by 0.6℃, there is a significant effect on the seasonal variation of sea level in the high latitudes of the Northern Hemisphere.更多还原