Primary Analysis Of Influence Of Large Mass Anomaly On Recovered Gravity Field Based On Energy Principle

It is a very effective way to recovery gravity field using satellite data, but the satellite data will be contaminated by noise inevitably. Many methods of recovering gravity field using contaminated satellite data are based on Least-Square principle, thus the error is uniformly distributed globally, and the sun square of the error is least. It is a very interesting problem whether larger gravity signal will introduce larger error. This paper does the following work based on above-mentioned motivation.In section 1, the application and recovering methods of gravity is introduced, and points out that it is an efficient method to recovering the Earth gravity field using energy balance approach. In section 2, the time systems and reference systems are introduced, and the transformation relationships among them are given. At the same time, forces from all kinds of sources exerting on satellites in the complex space environment are presented, and expression of some forces is given. In section 3, we firstly derived the kepler orbits of satellites and Gauss equation and Lagrange equation that describes the moving patters of satellites when conservative and non-conservative forces exert on them, then numerical integration that can calculates a series of positions and speed of discrete points is studied, including the single-step method that could self-starts and the simple multi-step methods. Energy balance approach was introduced that could recoveries gravity fields using satellite data in the rotating and inertial frames. Then we derived the conjugate gradient method that could be used to solve over-determined equation efficiently and presented methods that could improve convergence rate. We studied error of the gravity model deriving from satellite simulation data where local large mass exists. The model we used is a uniformdensity sphere whose radius is the same as the mean radius of the earth, and we put a mass on the surface of the sphere. The simulation results shows that the error of the recovered gravity model in most area of the globe is around 1 magl when decompose the mass to sphere harmonic coefficients to degree 60, and only the edge of the mass shows error that is around 4 mgal. The error of the recovered gravity model decreases too when the mass decreases, but the relative error remains approximately the same.