Recovery And Application Of Gravity Field Model Based On GOCE

The research of Earth’s gravitational field model is a core in geodesy all the time. With the launching of CHAMP (CHAllenging Minisatellite Payload), GRACE (Gravity Recovery And Climate Experiment), and GOCE (Gravity field and steady-state Ocean Circulation Explorer), the study of gravitational field model become a hotspot in geodesy. Fortunately, the data of three LEO (Low Earth Orbit) satellites has been freely released for users. GOCE is the first satellite that both GPS (Global Position System) receiver and EGG (Electrostatic Gravity Gradiometer) are carried. So, the GOCE has two survey modes, one is SST-HL (Satellite-to-Satellite Tracking in High-Low mode), another is SGG (Satellite Gravity Gradiometry). The long-medium wave information of Earth’s gravitational field can be recovered by SST data, and the short-medium wave information of Earth’s gravitational field can be recovered by SGG data. A high precision and resolution model can be derived by combining SST data and SGG data.The main work of this dissertation is that recovery of static gravity field model based on GOCE data. A time wise method is developed by combining GOCE orbit and gravity gradient, which is based on predecessors’ research results. Besides, author has developed a program independently, which is specialized for GOCE satellite. Author has put forward 3 schemes that the effect of gravity field model will be taken into account in road engineering survey. Finally, the static gravity field model is applied to simulation test of road engineering survey.The main work and contributions of this paper is following:1. The definitions of coordinate system and detailed transform formulas relate to GOCE satellite have been provided. Recursion formulas of Legendre series function and Legendre integral function are studied, and backward recursion has been applied to resolve the problem that integral calculation is non-stationary in polar area. In order to avoid the singularity of gravity gradient computation in polar area, the recursion formulas of specialized Legendre function PS1 and PS2 is deduced. The detailed simulated formula of gravity gradient in earth-fixed reference system has been deduced. Some perturbation forces are calculated based on GOCE orbit, planet ephemeris file DE405 and FES2004 ocean tidal model, including lunisolar gravitational perturbation, solid tide perturbation, ocean tide perturbation and pole tide perturbation. Intel MKL functions, which are highly optimized and paralleled, are applied to improve computational efficiency.2. GOCE precise kinematic (PKI) orbit data has been preprocessed and 3 gravity field models have been recovered by energy balance approach, short arc integral approach and average acceleration approach respectively. In addition, advantages and disadvantages of 3 recovered approaches are evaluated by calculated results. The research results show that recovery capability of GOCE orbit is up to 120 degree and order. GOCE satellite’s ambition is recover static gravity field model, but author come up with that detection of time vary information of gravity field model using GOCE orbit. An improved fan filter has been proposed based on error characteristics of GOCE model. GOCE monthly model has correlated errors, which is different from the correlated errors of GRACE monthly model. The quality change of Antarctic ice sheet is detected by about 1 year GOCE orbit, and results show that the variation at point (-75°,250°) is about -11.30cm/a. Furthermore, the detected results by GOCE orbit are verified using contemporaneous GRACE RL05 monthly models.3. GOCE gravity gradient data has been preprocessed and three filtered approaches are designed based on the principle of gravity gradient measurement and colored noise characteristics. Moreover, three approaches are evaluated by their filtered results, and the approach that filtered result is best has been used for filtering gravity gradient all the time. Author applied gravity gradient to recover gravity field model by time-wise approach and space-wise approach respectively. In that GOCE orbit has not covered global completely, the recovery computation of GOCE is ill-conditioned. In order to recover only-GOCE information model, regularization has been applied to restrain ill-condition. Primitive functions of 4 kinds of trigonometric function have been deduced, which are kernel of spherical cap regularization. The results show that spherical cap regularization is superior to standard regularization. A time-wise approach has been deduced by combining GOCE orbit and gravity gradient. A gravity field model named SWJTU-GO2013 is derived using 1 year GOCE orbit and gravity gradient. The results show that the space resolution of SWJTU-GO2013 is 100km half-wavelength, and geoid error is about ±20.32cm. SWJTU-GO2013 has same precision with the time-wise model, which is released by ICGEM.4. Software named SWJTU-GOCE has been designed and programmed by author independently, which is specialized for GOCE data. SWJTU-GOCE’s software interface is compiled by C# and kernel functions are compiled by Fortran. In order that time consuming computation can be done on personal computer, Fortran code has called some highly optimized and paralleled Intel MKL functions. In the term of visualization, the program can generate Matlab or GMT script file. Software modules of SWJTU-GOCE contain whole flow that from preprocessing to accuracy evaluated.5. Traditional road engineering survey has a problem that GPS system takes normal as reference, but total station system takes vertical as reference, two survey modes have not the same reference. However, it is common that GPS and total station are mixed use in road engineering and system errors are ignored. Author proposes three schemes that the influence of gravity field is taken into account in road engineering survey. The three schemes are designed based on survey department and road design department respectively. Author deduces a set of formula and accuracy evaluated formula that GPS horizontal coordinates are applied to recover road engineering gradual change coordinate. Finally, gravity field model SWJTU-GO2013 is applied to simulation compute and simulation results is verified by EGM2008.