重力数据网格化方法比较

COMPARISON AMONG METHODS FOR GRAVITY DATA GRIDDING

使用Shepard曲面拟合法、加权移动趋势面拟合法及普通Kriging算法对重力异常数据进行空间内插实现数据网格化,从计算速度、网格化后数据图形的连续性、结果精度及适合应用范围等方面对3种方法进行比较,结果证明:1)同一地区的离散数据越均匀、起算数据越多,得到的网格化结果越好;2)普通Kriging方法插值精度最好,但是计算速度慢,适合应用于数据变化大的区域;其他两种方法计算速度快,只适合应用于数据变化小的区域。

Kriging method By using Shepard surface fitting method, weight moving trend surface fitting method and ordinary , the anomalous gravity data were interpolated for gridding and then the gridded data results were compared with each other in calculating speed, continuity of gridded surface, the accuracy and the suitable fields of application . It is verified that： 1 ） the accuracy of gridding result; would become nice if the discrete data points are even in the same area and the initial data are enough; 2） the Ordinary Kriging method has the best interpolation accuracy, but calculating speed is slow, and it is suitable for great change areas; the other two methods have the fast speed, but they are suitable for little change areas.