| ||Format||Pages||Price|| |
|PDF (280K)||16||$25||  ADD TO CART|
|Complete Source PDF (6.3M)||278||$65||  ADD TO CART|
This paper describes a geostatistical case study to assess TCE contamination from multiple point sources that is migrating through the geologically complex conditions with several aquifers. The paper highlights the importance of the stationarity assumption by demonstrating how biased assessments of TCE contamination result when ordinary kriging of the data that violates stationarity assumptions. Division of the data set into more homogeneous geologic and hydrologic zones improves the accuracy of the estimates. Indicator kriging offers an alternate method for providing a stochastic model that is more appropriate for the data. Further improvement in the estimates results when indicator kriging is applied to individual subregional data sets that are based on geological considerations. This further enhances the data homogeneity and makes use of stationary model more appropriate. By combining geological and geostatistical evaluations, more realistic maps may be produced that reflect the hydrogeological environment and provide a sound basis for future investigations and remediation.
Geostatistics, environmental contamination, kriging, second order stationarity
Assistant Professor, Colorado School of Mines, Golden, CO
Professor, Colorado School of Mines, Golden, CO