【摘要】 在地壳形变监测中,由于观测点相对较少,形变分析时常出现观测资料缺乏以及分布不均匀的情况,需要对某些未测点进行推估。考虑到同一区域变形一方面呈现一定的趋势性变化,另一方面又表现出局部的、随机的和不规则的特性,本文在基于可靠函数模型拟合的基础上,利用Kriging统计方法对区域未测点的变形量进行推估。顾及到观测数据中可能含有异常误差,不论是函数模型逼近,还是Kriging推估中半变异函数的拟合都将会受到影响,本文采用抗差估计的方法来抵制异常数据的影响。最后根据某实测地震网监测数据,通过实例验证该方法在未测点变形量估算中的有效性。更多还原

【Abstract】 As the observation information is often insufficient or uneven in crustal deformation analyzing,the deformations of some unobserved points often need to be predicted.The deformation in one area,on one hand,shows a general tendency and on the other hand,is local,anomalous and stochastic,a model synthesized by functional models and statistical methods is proposed.Considering not only functional approximating but also semivariogram function fitting in Kriging statistics are impacted by measurement outliers,robust estimation method is applied in this paper.Finally,a crustal deformation monitoring network is calculated and analyzed,and the result proves the effectivity of the proposed method in this paper.更多还原