Research On Monitoring Of Terrestrial Water Storage Change And Its Load Deformation Combining With GNSS And GRACE Data Processing

Terrestrial water is an important part of water resources,its accurate measurment of the spatio-temporal variation of regional terrestrial water and its load deformation,which is of importantly practical and scientific significance for revealing terrestrial water cycle,understanding the geodynamic process of the crustal nonlinear motion,and establishing and maintaining a high-precision earth reference frame.With the development of space geodetic technology,measurement data is characterized by diversification and high precision.The Global Navigation Satellite Systems(GNSS)has become an important means of monitoring crustal movements due to its high precision,real-time and all-weather advantages.The nonlinear changes in crustal movement mainly reflect the combined effects of non-tidal ocean loads,atmospheric loads,hydrological loads,and geophysical effects such as glacial equilibrium adjustments.After subtracting the effects of atmospheric and non-tidal ocean loads from the linear change of time series,the residual sequence can be used to study the terrestrial water load.The evolution of the Earth's gravitational field over time caused by terrestrial water load migration,based on the physical mechanism of time-varying gravitational field and surface mass change,the terrestrial water storage(TWS)and its load deformation can be quantitatively inverted.The time-varying gravity field model from Gravity Recovery and Climate Experiment(GRACE)and GRACE-Follow On(GRACE-FO)can provide an effective technical means for continuously monitoring regional surface quality changes.In this paper,we study the theory and method of integrating GNSS and GRACE data to monitor the deformation of terrestrial water load in the region,and combine the geophysical model and measured data to analyze the regional TWS changes and the spatio-temporal characteristics of load deformation.The main work and research results of this paper are as follows:1.The development and application of GRACE satellite gravity measurement technology are reviewed.The methods and research progress of TWS change and its load deformation based on gravity field model and GNSS inversion are reviewed.Some key problems of GRACE and GNSS data in inversion are summarized and analyzed.It is pointed out that the filtering method and leakage error are important factors affecting the accuracy of GRACE inversion.Improving the accuracy of GNSS inversion model is the key to improve the reliability of GNSS inversion results.The joint inversion method is described in the deformation of water load.The role and significance of joint inversion method in water load deformation research is expounded.2.The basic theory of GRACE inversion of TWS and its load deformation calculation is introduced.The spherical harmonic coefficient method for inversion of surface mass variation based on GRACE time-variable gravity field model is studied.The effect of single filtering method and combined filtering method on south-north band noise is compared and analyzed.The calculation formula of regional time-variable gravity field extraction algorithm is given.The physical mechanism of glacial equilibrium adjustment is described.The ICE-5G model is used to correct the GIA effect.The conversion relationship of the first-order load LOVE number of different earth reference frames is completely derived.3.Verify the GRACE leakage error correction method and its application.The theoretical formula of third-filter kernel function method based on leakage error correction of large-scale river basin is deduced.The reliability of the new algorithm is verified by numerical simulation and measured data.The results show that the new method is relative to the time series from bias correction method,multiplicative correction method and scale factor method,increasing the accuracy of leakage error correction by 15%,37%,and 35%,respectively,which is in good agreement with the WGHM hydrological model and the measured data of the Three Gorges Reservoir.The scale factor method was used to quantitatively invert TWS change in small-scale areas.Taking Guanzhong area as an example,the GRACE leakage error was reduced by 8%after the scale factor correction.The inversion results were in good agreement with the measured groundwater level and precipitation data.The forward modeling method is used to quantitatively estimate the changes of TWS in medium and long-space scales.The numerical simulation results show that the accuracy of corrected signal is improved by nearly 30%.The change of TWS after leakage signal recovery in the Tibet Plateau is closer to that from WGHM results.Nearly,the signal energy in the low frequency domain is significantly enhanced.4.The processing optimization scheme of GNSS observation data based on singular spectrum analysis(SSA)is systematically studied,including gross error detection of GNSS elevation direction sequence,noise reduction,missing data interpolation,and signal extraction.The results show that compared with the traditional interpolation method,the SSA iterative interpolation method has the highest interpolation precision and the accuracy in the elevation direction is better than 5mm.The method still has good stability with the increase of missing data in the elevation direction sequence.Compared with the wavelet transform compared with FFT and SSA filtering method,the signal-to-noise separation effect of the observation data is the best,which is the closest to the real value.Compared with the least squares fitting method,the time series of the GNSS periodic term obtained by the SSA method is more accurate,which can achieve a correlation coefficient of 0.7 or more comparable with GRACE deformation after leakage error correction and GAC addition.After deducting the GRACE mass load term,the WRMS value of the same station is also reduced,and the contribution rate of terrestrial water load to GNSS vertical sequence is reduced,which is within the range of 1.6%?17%.5.The GNSS data downscaling features are extracted and the observation of GNSS inversion model is refined.A GNSS data downscaling method based on Multi-channel Singular Spectrum Analysis(MSSA)is proposed to improve the observation accuracy of GNSS inversion model and local high-frequency signal effect.The analysis on the elevation sequence from 38 CORS reference stations shows that the time series accuracy of MSSA method reconstruction is generally higher than that of FFT multi-cycle reconstruction method.The accuracy of 92%of the stations were improved by MSSA method to some extent,which was improved by 2.01%to 16.89%.The removal-recovery method is used to optimize the calculation process of atmospheric load and sea level change load,which reduces the truncation error caused by the original grid data during spherical harmonic expansion.The results show that the influence of atmospheric load on the vertical displacement of the reference station is on the seasonal scale,the maximum displacement is 12.4 mm,and the sea level change load causes the maximum vertical displacement of the station to be about±3 mm.6.Verify the method reliability for regional TWS change monitored by GNSS elevation direction displacement.For the construction of inversion model,the problem of rank defect of the equation is solved by the method with constraints,which avoids the difficulty of selecting the ridge parameter in the regularization method and improves the calculation efficiency.The selection of integral radius and boundary scale expansion in the inversion model are analyzed in detail.Based on the crustal load elastic deformation theory,the numerical simulation results of the GLDAS model are obtained to evaluate the stability of the GNSS multiple iterative inversion algorithm.The results show that for the area with dense sites,the results are in good agreement with the analog signals.However,the signal deviation in the local area where the point is sparse is relatively large.In the surrounding area far from the coverage of the base station,the result deviates significantly from the analog signal,which shows that the stability of the GNSS vertical displacement inversion method is related to the intensity of the reference site.Compared with the direct solution algorithm,the multiple iterative inversion results are consistent with the local features of the analog signal,which effectively validates the stability of the multiple iterative inversion method;further reveals the regional TWS and its vertical load deformation spatio-temporal characteristics based on GNSS inversion,the spatio-temporal commonality with GRACE monitoring results is mainly discussed from a qualitative perspective.7.The method of integrating GNSS and GRACE data monitoring regional terrestrial water load deformation is studied.The joint inversion results show that the vertical deformation caused by regional terrestrial water load has obvious seasonal characteristics,and the displacement deformation value is about-12mm?12mm.In 2015,2016,2017 year,from January to March each year,the terrestrial water load migration drives the earth surface to produce upward displacement;in May,July,and August each year,the earth surface produces a downward displacement under the action of terrestrial water load;compared with the monitoring results of a single means,after the joint inversion,the spatial signal of the study area has undergone new changes.A relatively ideal physical signal appears at the boundary away from the coverage area of the CORS station,which makes up for the deficiency of GNSS inversion in the far field of the integral;the large vertical deformation is concentrated in the coverage of the CORS station,which mainly reflects the signal characteristics of the GNSS results.The spatial signal distribution has higher spatial resolution than the GRACE results.In order to verify the superiority of joint inversion method compared with GNSS inversion,the calculation formula of Mascon equivalent water height to Mascon deformation displacement is derived.The results show that the vertical deformation of the regional terrestrial water load obtained by the joint inversion method is more consistent with the Mascon deformation solution.8.The GRACE-FO time-variable gravitational field using CSR,JPL and GFZ reveals the spatio-temporal characteristics of water storage in China over the past two years.It is proposed to use the improved hard model and PREM model to analyze the impact of water load vertical deformation derived from GRACE-FO inversion in China's mainland caused by regional earth structure difference.The results show that the relative difference of the 60-order vertical displacement load is close to 4.27%,and the maximum difference of the first 60-order vertical load deformation is located in the Lancang River Basin in Yunnan,with an amplitude of about 0.7mm/a.An SSA iterative method for predicting changes of TWS during the connection between GRACE and GRACE-FO was constructed.The prediction accuracy of TWS in six experimental areas in China on the short-term,mid-short term,mid-term and long-term scales is analyzed.The results show that the prediction accuracy is almost higher than that of the ARMA model.The prediction trend is consistent with the Mascon solution and GLDAS model.