渤、黄、东海同化TOPEX/POSEIDON高度计资料的半日分潮数值模拟

NUMERICAL SIMULATION OF SEMIDIURNAL CONSTITUENTS IN THE BOHAI SEA,THE YELLOW SEA AND THE EAST CHINA SEA WITH ASSIMILATING TOPEX/POSEIDON DATA

使用调和分析方法分析了渤、黄、东海中的TOPEX/POSEIDON沿轨高度计资料,并利用基于最优插值理论的混合法把交叉点处的两个主要的半日分潮(M2和S2)同化到动力模式中。同化结果显示,两个主要半日分潮的分布特征与前人动力模式结果比较一致,在同化高度计资料以后模式结果M2分潮与167个实测站的“距离”为17.2cm,S2分潮为8.9cm,比单纯的动力模式结果精度分别提高14.9%和23.3%。

After the launch of TOPEX/POSEIDON in 1992, there has been considerable improvement in the accuracy of the global tidal models. However, the application of global tidal model to the study on shallow sea tides was unsatisfactory in accuracy. M_2 and S_2 semidiurnal tidal elevations and currents in the study area including the Bohai Sea, the Yellow Sea and the East China Sea were derived by combining tidal harmonic analyses of TOPEX/POSEIDON altimeter data and a regional hydrodynamic model in an assimilation scheme based on optimal linear interpolation. The tidal harmonic analysis along satellite orbit was used to abstract tidal constituents from TOPEX/POSEIDON satellite altimeter data in the study area where 12 orbits situated. The cycle of the altimeter data is from 11 to 336.18 constituents, including 5 long-period (S_a, S_~sa , M_m, M_f and M_~Sf ) , 7 semidiurnal (N_2, M_2, S_2, P_2, μ_2, ν_2 and L_2), 4 diurnal(P_1, O_1, K_1 and Q_1), and 2 quarter-diurnal (M_4 and MS_4), were involved in the harmonic analysis. The results showed that the altimeter data could give a good distribution of tides in the Yellow Sea and the East China Sea. The internal coincidence accuracies of the amplitudes of M_2 and S_2 at 13 crossover points out of 14 are ~2.50cm and 1.67cm, and the accuracies of lags were 0.91° and 4.51° respectively. The POM model is a three-dimensional model that imbedded with sub-model of second moment turbulence closure, which provides vertical mixing coefficients. It uses a sigma coordinate in the vertical, a curvilinear orthogonal coordinate and an“Arakawa C” differencing scheme in the horizontal. The model has a free surface and a split time step. The external mode portion of the model is two-dimensional, and uses a short time step based on the CFL condition and the external wave speed. The internal mode is three-dimensional, and uses a long time step based on the CFL condition and the internal wave speed. It has been used to simulate the tide in this region correctly by other researchers. The computational grid size is10′×10′. The drag coefficient in the ocean bottom boundary layer was taken to be 0.0009 in the Bohai Sea and 0.0022 in the Yellow Sea and the East China Sea. The two major semidiurnal constituents were simulated simultaneously. Along the open boundaries, the heights of the water surface were given asζ=∑f_CH_Ccos[ω_Ct+(V_0+u)_C-g_C] (Eq.2), where H and g are harmonic constants for the amplitude and phase-lag respectively, the subscript C stands for either one of the constituents: M_2 or S_2;ω is the angular speed of the tidal constituents, f the nodal factor, u the nodal angle, V_0 the initial phase angle of the equilibrium tide. The tidal elevations of M_2 and S_2 derived from TOPEX/POSEIDON altimeter data at the crossover are assimilated into a dynamical model (POM) using blending method based on optimal interpolation methods. The weights matrix K come from the theory of optimal linear interpolation, which was used by Guoqi Han et al(1996,2000). Ideally, this matrix should be based on solid grounds, such as a statistical analysis of the differences between the model first-guess and the altimetric tidal data. Discussion on this topic is beyond the scope of this paper, and a relatively simple method,K=0.8 was assumed. The characteristics of the two semidiurnal tides agreed well with those from other researchers. The distance was 17.2 cm for the M_2 tide and 8.9cm for the S_2 tide when compared to the observations of 167 tide gauges distributed along the coastlines. The accuracy improvements were by 14.9% for M_2 tide and 23.3% for S_2 tide after the assimilation, respectively.