关于地球化学数据处理方法的讨论-以大兴安岭喜桂图地区1/20万区域地球化学数据为例

针对地球化学数据中存在的离群数据点在地球化学图面上出现的等量线畸变或“牛眼点”现象,本文使用多种数据网格化插值方法和参数对比以消除“公牛眼”现象,结果表明,克里格、径向基本函数、改进的谢别德、最小曲率、移动平均等方法均不能较好地消除该现象。各种网格化算法编制的地球化学等量线图和交叉验证结果证明,以改进的指数距离幂倒数加权网格化方法效果最好,并且网格参数选择中以取样间距为网格间距,以2.5倍网格间距为搜索半径和幂指数因子的网格化参数既能消除等量线畸变现象,又能对原始数据进行较接近的表达。上述各种地球化学数据网格化插值方法的对比及结果可为地球化学数据处理中遇到同类问题提供信息。

Discussion on the Geochemical Survey Datum Processing Method:An Example from 1/20 0000 Geochemical Survey Datum of the Xiguitu Area,Da Hinggan Mountains

To eliminate the contour distortion or"bull's-eye point"existing in the geochemical map of the outlier data points, this paper used different data grid interpolation methods and parameter models. The results show that the kriging, radial basis function, modified shepard’s method, minimum curvature and moving average method can’t eliminate the phenomenon. The geochemical contour map and cross validation results with demonstrate var-ious grid algorithm programming demonstrate that the distance exponential reverse weighted method is the most ideal method,and choosing to sampling spacing as the grid spacing, 2.5 times the grid spacing as the search radi-us and distance power exponent can eliminate the contour distortion or"bull's-eye point"phenomenon. This grid method and parameters can also express closer to the original data. The comparison result above can provide in-formation for the similar problems during the datum processing.