Optimized multi-phase sampling for soil remediation surveys

Abstract

We develop an algorithm to optimize the design of multi-phase soil remediation surveys.The locations of observations in later phases are selected to minimize the expected loss incurred from misclassification of the local contamination status of the soil. In contrast to existing multi-phase design methods, the location of multiple observations can be optimized simultaneously and the reduction in the expected loss can be forecast. Hence rational decisions can be made regarding the resources which should be allocated to further sampling. The geostatistical analysis uses a copula-based spatial model which can represent general types of variation including distributions which include extreme values.The algorithm is used to design a hypothetical second phase of a survey of soil lead in Glebe, Sydney. Observations in this phase are generally dispersed on the boundaries between areas which according to the first phase either require, or do not require, remediation. The algorithm is initially used to make remediation decisions at the point scale but we demonstrate how it can be used to inform over blocks.

Potential of integrated field spectroscopy and spatial analysis for enhanced assessment of soil contamination: A prospective review

“…With the ability to make estimates in real-time in the field, it should be possible to adapt the sampling to the next location which maximises the reduction in uncertainty integrated over the study area (Marchant et al, 2013), and so on, until some overall quality criterion or financial constraint has been reached.…”

Optimising sampling designs for the maximum coverage problem of plume detection

“…If this is not the case and if autocorrelation ranges are known, McBratney and Minasny (2013) proposed space deformation to obtain stationarity. Then regular sampling in the deformed space transforms back to sampling with density relative to spatial variability.…”

“…A common application is minimisation of the kriging variance as by Heuvelink et al (2013) in a space-time setting. SSA has also proven its fitness for adaptive sampling: Delmelle (2013) searched for locations where kriging variance was high and more information was needed as neighbouring samples gave different information; Marchant et al (2013) used it to add samples to delineate areas that require soil remediation, based on predicted contamination distributions. Lark (2011) aimed at a nested sampling where distances between samples cover a wide variety, to estimate variability of the modelled phenomenon on various scales and at the same time to estimate the parameters of this variability accurately.…”

Cost-Effective Strategy for the Investigation and Remediation of Polluted Soil Using Geostatistics and a Genetic Algorithm Approach

“…An adaptive cluster sampling (ACS) approach, for example, was proposed for additional sampling to reduce Kriging errors and misclassification rates [19] [24]. The remediation cost uncertainty or expected total loss from misclassifications of the contamination status of the soil has also been used as performance criteria for additional or later-phase sampling [32] [33]. Many studies have focused on conditional simulation (CS) or sequential Gaussian simulation to generate probability maps for the assessment of soil pollution [18] [22] [23] [32]- [37].…”