Bootstrapped spatial statistics: a more robust approach to the analysis of finite strain data

Standard spatial statistics involves exploratory data analysis (EDA) and the computation of a semi-variogram prior to kriging from spatial data such as the finite strain data from a thrust sheet. However, the main shortcoming of this method is that uncertainties in the computation of EDA and the semi-variogram parameters cannot be estimated; standard EDA allows the computation of the uncertainties associated with only the sample mean. We use resampling statistics to overcome this and use previously published data from the Sheeprock thrust sheet in the Sevier fold-and-thrust belt in western USA to illustrate Bootstrapped spatial statistics. The use of Bootstrapped spatial statistics in place of standard spatial statistics also simplifies the computation of the best semi-variogram resulting in the computation of a more robust semi-variogram. Kriging estimates obtained using the Bootstrapped semi-variogram indicate that the results are practically insensitive to the uncertainty associated with the estimation of parameters of the semi-variogram used in the Sheeprock example.