中文摘要：

利用动态PPP对香港12个GPS测站2007—2012年的数据反演了海潮负荷位移,通过与7个全球海潮模型、1个区域模型和静态PPP反演的结果比较发现,相对于另外几个模型,动态PPP反演结果与TPXO.7.2、EOT11a、HAMTIDE和NAO99Jb模型的结果符合得更好。与静态PPP的结果比较发现其RMS与各模型的RMS大体上一致,只是在S2、K2和K1的E方向和M2、S2的N方向稍有增加。此外,除K2和K1潮波外,动态PPP与模型的RMS值在水平方向上均小于1 mm,在垂直方向上均小于2.5 mm,能达到和静态PPP相当的精度。本文反演的结果与NAO99Jb模型值存在明显的系统偏差,当去除系统偏差后,所有潮波的RMS值都有明显的减小,尤其在K1的垂直方向RMS从16.4 mm减少到1.3 mm。此外,通过将香港2012年验潮站数据反演的潮波参数与模型的结果进行比较发现,其结果同样与TPXO.7.2、EOT11a、HAMTIDE和NAO99Jb这4个模型更为符合,这进一步验证了动态PPP反演海潮的有效性,同时说明这4个模型比较适合香港区域。

英文摘要：

In this study, 12 GPS stations＇ continuous observations from 2007 to 2012 were utilized to determine ocean tide loading （OTL） displacements by kinematic PPP. Through comparison of estimates from kinematic PPP and results predicted by 7 global ocean tide models, one regional model and static PPP, TPXO.7.2, EOT11a, HAMTIDE and NAO99Jb were found having smaller RMS misfits than other models. Comparing the RMS misfits of static PPP results with that of models, we found that they have similar patterns except for the east direction of S2, K2 and K1 and north direction of M2 and N2, in which the static PPP＇s RMS misfits are a little bit larger than models＇. Furthermore, apart from K1, andK2, the RMS misfits between Kinematic PPP and model estimates are less than 1 mm and 2.5 mm in horizontal and vertical directions, which reach the same precision of static PPP. Obvious system bias was found between the results of kinematic PPP and NAO99Jb model. After subtracting the system bias, all constituents＇ RMS misfits decreased evidently, especially for K1 constituent＇s vertical component, reducing from 16.4 mm to 1.3 mm. Moreover, by comparing the harmonic parameters estimated from tide gauge data and ocean tide models, we found results from gauge data also agree best with TPXO. 7.2, EOT11a, HAMTIDE and NAO99Jb, which demonstrates that kinematic PPP is capable of determining OTL displacements and also indicates that these models are more applicable for Hang Kong.