Year
1994
Abstract
Use of statistical methods in environmental restoration programs is critical to help strike the balance between environmental and public health protection versus allocation of scarce clean-up resources. Statistical methods provide contamination estimates with predefined accuracy and precision that control the management risks of sampling results that might lead to wrong decisions (e.g., the decision that no residual contamination exists when in fact it does). This paper presents examples of geostatistical methodologies which have been successfully applied to environmental restoration and related programs over the past several years. These methods include characterization of spatial autocorrelation hi contaminant levels via semivariogram analysis, estimation and mapping of contaminant levels via ordinary kriging, and estimation of geologic heterogeneities via indicator simulation. These and related methods have been successfully applied in different environmental media, including soil, groundwater, and air, and at different Department of Energy and Department of Defense installations.