Determination of the New Tidal Parameters Obtained with a Superconducting Gravimeter at Station Wuhan/China

Abstract

The new tidal gravity parameters are determined precisely using 1997-99 registrations obtained with a superconducting gravimeter at station Wuhan/China. The new determinations are in agreement with those obtained using 1985-94 records with exhibit a small instrumental drift and a significant reduction of the site background noise. The atmospheric pressure and oceanic loading signals are removed from the tidal parameters and the free core nutation resonance parameters as of resonant eigenperiod and strength as well as the quality factor are determined. 1. History recall and introduction The Chinese superconducting gravimeter (SG) numbered T004 was installed since November 1985 at the tidal gravity laboratory (30.58•‹N, 114.36•‹E, 34m), Institute of Geodesy and Geophysics, Chinese Academy of Sciences in Wuhan/China. About 10-year continuous gravity data are accumulated except for some short time interruptions. For the purpose in accordance with the global geodynamics project (GGP) regulations (Crossley et al., 1999), the instrument was sent back to the GWR for upgrading in October 1996. The old SG was changed with a new unit sensing and an improved dewar mounting, it was given a new series number C032. It was installed successfully at new site (30.52•‹N, 114.49•‹E, 80 m) in November 1997 about 25 km away from center city. In the replacement of the old data acquisition system, the new one developed by the German group (Jentzsch, personal communication) was installed in November 1997 in connection to the GGP. The records with every 20 s interval are obtained from the original 1 s sampling and sent to a 6.0 digital voltmeter. The continuous data for more than two years are obtained for the purpose of GGP data exchange. This report will introduce the analysis results of the C032 observations in 1997-99 in term of the tidal gravity parameters (amplitude factors and phase lags), and the study for pressure and oceanic gravity signals as well as the determination of the Free Core Nutation (FCN) resonance parameters. * Supported jointly by Nature Science Foundation of China (49774223, 49925411), Chinese Academy of Sciences (KZ952-J1-411, KZCX2-106). 348 He-Ping Sun, Hou-Tze Hsu, Jian-Qiao Xu, Xiao-Dong Chen and Xing-Hua Hao 2. Calibration and Tidal Analysis In order to convert digital output in volt into tidal gravity unit in ƒÊgal, the calibration of the SG using a FG5 absolute gravimeter (AG) was carried out during a period of 3 days starting from at 04:00:00UT, January 29 to at 06:26:20UT, February 1, 1999. After applying for the correction of light speed, valid height and adjustment height, together with the SG output, the AG measurements are used to determine regression coefficient, it is given as of -84.05 ƒÊgal/N with a relative accuracy 0.36%. The tidal gravity and air pressure observations used in this study are for the period from at 19:00 December 20, 1997 to at 23:00 December 31, 1999. The data preprocessing based on the TSOFT (Vauterin, 1998) technique was carried out monthly before the tidal parameters are determined precisely. By using a slipped window function, the abnormal signals as jumps, tars and spikes are detected and eliminated. The hourly sequence of the tidal gravity is obtained using a remove-restore technique, it is carefully checked though a smoothing procedure that rejects short term perturbations. The missing data due to the power interruptions, earthquakes, refilling liquid helium are filled using a spline interpolation based on synthetic tide. Figure 1 shows the detected and corrected peaks, spikes and earthquakes in 1998 and 1999. Fig.2. Characteristics of high passed residuals in temporal (a) and frequency (b) domains 349 Determination of the New Tidal Parameters Obtained with a Superconducting Gravimeter Based on the Etema package (Wenzel 1996), the tidal parameters are precisely determined . The new determinations are in agreement with those obtained from 1985-94 records with exhibit a significant reduction of the station background noise and low instrumental drift . It is found that the precision of the main wave amplitude factors is as of 0.06% (O1) and 0 .04% (M2) respectively. The tidal gravity residuals are obtained after removing the synthetic tides . The high passed residuals in both temporal and frequency domains are shown in Figure 2. The monthly standard deviation is at about 0.2ƒÊgal level that is improved significantly comparing to the one as of 0.7ƒÊgal, obtained from old series (Sun et al ., 1998). It is found that the negative phase differences are mainly related to the Pacific oceanic tides . 3. Environmental Perturbation and Ocean Loading The influence of the air pressure on tidal gravity registrations gets more and more important, since the high precision SG can record simultaneously gravity signals induced by air pressure (Sun et al., 1998; Kroner and Jentzsch,1999). Therefore together with tidal residuals, the station air pressure is used to estimate regression coefficient as of -0.2724ƒÊgal/hPa. However, such coefficient is varied as time and frequency. It is due to the special meteorological condition as the (anti-) cyclones motion of the weather system. The amplitudes and power spectra density of gravity residuals are reduced at all frequency bands after removing pressure gravity signals. Since 1992, thanks to the most recent oceanic models developed by analysis of the precise measurements from Topex/Poseidon altimeters and as a result of parallel developments in numerical tidal modeling and data assimilation, the oceanic models are improved significantly (Meichior and Francis, 1996). Therefore it is important to study the various oceanic models, in order to select the one that fits best for tidal gravity data. Based on a direct discrete convolution method between ocean tides and loading Green functions (Sun, 1992), by using Schwiderski (Scw80) and the most recent models (PODAAC, 1999) as Csr3.0 (Eanes), Fes95.2 (Grenoble) and Tpxo2 (Egbert), the loading amplitude and phase (L,A,) for 8 main waves are calculated Table 1. Oceanic loading amplitude and phase (L,ƒÉ) for various global ocean models 350 He-Ping Sun, Hou-Tze Hsu, Jian-Qiao Xu, Xiao-Dong Chen and Xing-Hua Hao (Table 1). The results for the Oni and Ori96 (Matsumoto) models calculated by Francis and Mazzega (1990) are also listed in the Table. The observed residuals are obtained after subtracting the loading signals from observations and the tidal parameters are corrected. Table 2 shows the residual amplitude before (B, ƒÀ) and after (X, x) oceanic loading correction. It is found that the residual amplitudes are reduced significantly from 0.74 to 0.11 ƒÊgal (01), from 0.85 to 0.13 ƒÊgal (K1), from 0.84 to 0.16 ƒÊgal (M2) and from 0.27 to 0.02 ƒÊgal (S2). Table 2. Tidal gravity residuals before and after oceanic loading correction Table 3. Tidal gravity parameters before and after oceanic loading correction 351 Determination of the New Tidal Parameters Obtained with a Superconducting Gravimeter Table 3 shows the amplitude factors and phase differences before (ƒÂ , ƒ¢ƒÕ) and after (ƒÂ, ƒ¢ƒÕ') loading correction. Compared to the standard tidal model (Dehant ,1999), the discrepancy for 01 is reduced from 2.12% to 0.36% and that for M2 is reduced from 0 .86% to 0.20%, it shows the effectiveness of the loading correction. From the tables, it is found that although after pressure and loading corrections , there remains still the residual amplitude at some frequencies , which obviously induced by other kinds of perturbations, as the regional pressure , temperature, underground water and so on. The lateral heterogeneity of the Earth that is not yet included in tidal model may lead also the increase of the residual amplitude. 4. Diurnal Tidal Waves and FCN Resonant Parameters The dynamic influence of the Earth's core within a rotating , elastic and elliptical mantle will lead to a rotating eigenmode associated with the wobble of the fluid core with respect to the mantle. The FCN resonance processes the eigenperiod close to one sidereal day in the mantle reference frame and approximately 435 sidereal days in the space reference frame (Hinderer et al., 1993). In order to retrieve the resonance parameters , the influence of the ocean tides and pressure perturbations are removed for the first step as discussed in above mentioned. Then same procedures as Defraigne and Dehant (1994) are employed to determine the resonance parameters . This method is based on so called least squares fit of the tidal gravity data to a damped harmonic oscillator, the computation of the eigenperiod relates to the forcing frequency, the quality factor Q value relates to FCN frequency and attenuation factor. The corresponding results are given in Table 4. Table 4. FCN Resonance parameters (before and after applying for various oceanic loading correction) From Table 4, it is found that before ocean loading correction , the FCN resonance parameters are not true. When using Scw80 models, the eigenperiod is given as 435.8 sidereal days, it is about 1.8% reduced when including the local oceanic tides. The maximum discrepancy of the eigenperiod can reach to 3.0% when using various oceanic models. Comparing to those determined when using the 1985-94 series (406.8 sidereal days), it is found that a more realistic estimation of the FCN parameters is due to the low station background noise at new site. The analysis shows that the oceanic and barometric pressure influences are the main error sources for the determination of the FCN parameters . In addition to these, the errors of the phase differences can be responsible for a biased Q. 352 He-Ping Sun, Hou-Tze Hsu, Jian-Qiao Xu, Xiao-Dong Chen and Xing-Hua Hao