In this paper, the admittance function between seafloor undulations and vertical gravity gradient anomalies was derived. Based on this admittance function, the bathymetry model of 1 minute resolution was predicted from vertical gravity gradient anomalies and ship soundings in the experimental area from the northwest Pacific. The accuracy of the model is evaluated using ship soundings and existing models, including ETOPO1, GEBCO, DTU10 and V15.1 from SIO. The model＇s STD is 69. 481m, comparable with V15.1 which is generally believed to have the highest accuracy.
Using bathymetry and altimetric gravity anomalies, a 1°×9 1° lithospheric effective elastic thickness（Te） model over the Louisville Ridge and its adjacent regions is calculated using the moving window admittance technique. For comparison, three bathymetry models are used： general bathymetric charts of the oceans, SIO V15.1,and BAT_VGG. The results show that BAT_VGG is more suitable for calculating T e than the other two models. T e along the Louisville Ridge was re-evaluated. The southeast of the ridge has a medium Te of 10–20 km, while Te increases dramatically seaward of the Tonga-Kermadec trench as a result of the collision of the Pacific and IndoAustralian plates.
The mean sea surface （MSS） model is an important reference for the study of charting datum and sea level change. A global MSS model named WHU2013, with 2′ × 2′ spatial resolution between 80° S and 84°N, is established in this paper by combining nearly 20 years of multi-satellite altimetric data that include Topex/Poseidon （T/P）, Jason-1, Jason-2, ERS-2, ENVISAT and GFO Exact Repeat Mission （ERM） data, ERS-1/168, Jason-1/C geodetic mission data and Cryosat-2 low resolution mode （LRM） data. All the ERM data are adjusted by the collinear method to achieve the mean along-track sea surface height （SSH）, and the combined dataset of T/P, Jason-1 and Jason-2 from 1993 to 2012 after collinear adjustment is used as the reference data. The sea level variations in the non-ERM data （geodetic mission data and LRM data） are mainly investigated, and a combined method is proposed to correct the sea level variations between 66°S and 66°N by along-track sea level variation time series and beyond 66°S or 66°N by seasonal sea level variations. In the crossover adjustment between multi-altimetric data, a stepwise method is used to solve the problem of inconsistency in the reference data between the high and low latitude regions. The proposed model is compared with the CNES-CLS2011 and DTU13 MSS models, and the standard derivation （STD） of the differences between the models is about S cm between 80°S and 84°N, less than 3 cm between 66°S and 66°N, and less than 4 cm in the China Sea and its adjacent sea. Furthermore, the three models exhibit a good agreement in the SSH differences and the along-track gradient of SSH following comparisons with satellite altimetry data.
The Kuroshio Extension （KE） plays an important role in climate and environmental change in the North Pacific. In this paper, more than 20 years of merged absolute dynamic topography and merged sea level anomaly products from satellite altimetry are used to analyze the stability of the KE system. By analyzing the annually averaged sea surface topography, the variations of inter-annual path and annually averaged eddy kinetic energy at the KE region, the KE＇s two dynamic states are given as： the relatively stable state during 1993 1995, 2002-2005, and 2010-2012, and the unstable dynamic state among 1996-2001 and 2006-2009. During the stable state, the KE spindle had a shorter path length and smaller time-varying amplitude, as well as a trend to move northward. While during the unstable state, the KE spindle had a longer path length and an integral southward transport trend, and was observed to oscillate significantly over time. The analysis on the KE＇s upstream and downstream region gives the same variations, indi- cating that they are significantly affected by the El Nino events. The power spectrum of the mean latitudinal position variation of the KE＇s upstream and downstream shows significant quasi-decadal oscillation characteristics and strong annual signals. Furthermore, the correlation of the strength vari- ation between the southern RG and the KE＇s upstream is calculated to be 0.50 after low-pass filtering, and that of the mean latitudinal position variation between the southern RG and the KE＇s upstream/ downstream are 0.75/0.69 after low-pass filtering, respectively. The strong correlations demonstrated that the southern RG and the KE are closely linked.
