论文部分内容阅读
利用径向基函数的多尺度分析方法可以对地球重力场进行分解,提取出细节的地球物理信息.目前通常采用离散积分法进行解算,但信号分解并重构后,所得信号并不能与原始重力场信号完全符合,分解过程中出现了信号泄露.针对这一现象,本文在最小二乘算法和方差分量估计的基础上,提出了新的在各个尺度上直接解算基函数系数的算法(直接法),有效的减少了重力场分解过程中的信号泄露.以南海地区DTU13重力异常数据为例,分别运用离散积分法和直接法对重力异常进行多尺度分解,结果显示:直接法的5个尺度上的信号泄露误差相对离散积分法减小约39%~79%;直接法总的泄露误差为±1.12 m Gal,明显小于离散积分法的±4.04 m Gal,直接法具有更优的效果.
The multi-scale analysis method based on radial basis function can decompose the earth’s gravitational field to extract the details of geophysical information.Currently, the discrete integral method is usually used to solve the problem.But after the signal is decomposed and reconstructed, the signal can not be compared with the original In this paper, based on the estimation of least square method and variance component, we propose a new algorithm to directly solve the base function coefficients at all scales ( Direct method) to effectively reduce the signal leakage in the process of gravity field decomposition.Using gravity anomaly data of DTU13 in the South China Sea as an example, using the discrete integral method and the direct method respectively to decompose the gravity anomalies at multiple scales, the results show that: The error of signal leakage on a scale is reduced by 39% -79% compared with the discrete integral method. The total leakage error of the direct method is ± 1.12 m Gal, which is obviously less than ± 4.04 m Gal of the discrete integral method. The direct method has better effect .