Skip to main content

Advanced Geostatistics Based Cost Effective Sampling Strategy for the Investigation and Remediation of Land Contamination

Supervisor: Dr Kang Li

There is increasing pressure on developers to reuse brownfield sites rather than develop greenfield sites.  The changes in environmental legislations are putting demand on land contamination investigation, sampling and remediation.  Cost effective and real-time contaminated land management and remediation has become more important.  The big challenge in contaminated land management is, given limited budget, how to improve the quality of sampling data, which is affected by the quality of sample analysis as well as the sampling material.

In the literature of sampling design, the basis of most sampling methods is the concept of random or probabilistic selection of the sample to be collected and the sub-sample that is to be analysed.  Properly designed sampling plans based upon the laws of probability are claimed to be able to provide a means of making decisions that have a sound basis and are not likely to be biased.  However, various errors are likely introduced in sampling.  To reduce various sampling errors, other methods have been introduced such as sub-sampling, double or multi-phase sampling, composite sampling, and systematic sampling.  In particular, the systematic sampling plan is an attempt to provide better coverage of the soil study area than could be provided with the simple random sample or the stratified random sampling plan.  It is in reality a stratification based upon spatial distribution over the site.

The application of Geostatistics and geostatistical concepts in sampling is relatively new.  Among various such techniques, Kriging is perhaps the most popular one.  Other geostatic methods include Bayesian statistics and Radial Basis Functions.  For example there are reports of using Bayesian statistics and compositing of samples as a means of detecting the size of contamination size and Bayes statistics allows he investigator to make use of prior information to guide in the design of the next phase  of sampling.  The Radial Basis function defines optimal relevance of nearby data points when the grid node is interpolated and there is report on using Radial Basis Functions to interpolate solute concentration profiles.

Conventional geostatistic methods like Kriging however can only perform mapping using interpolation, not extrapolation. This severely restricts its ability to predict the distribution of the contamination in the sampling area.  Moreover, due to their mathematical complexity and the lack of effective support tools to integrate the new techniques systematically, Geostatistical methods have not yet been widely used in field investigations.

In sampling design, apart from sampling methods which have been briefly discussed above, other issues are equally important, these include the number of samples, the filed sample size, the location of samples, and the cost of sampling.  All these factors are closely related to the sampling method, and are subject to the sampling purpose as well as the available budget for the investigation. Therefore, to produce a cost effective sampling strategy for the investigation and remediation of land contamination becomes of paramount importance for problem holders and industry at large.

The aim of this project is to develop an advanced real-time modelling method to support cost-effective and reliable contaminated land investigation, using advanced geostatistics.  By properly incorporating the advanced geostatistics, i.e. Bayes statistics and Radial Basis Function (RBF), into the sampling design using existing sampling protocols, the proposed method can effectively predict 'where' to sample, given available historical information, thus improve the overall sampling confidence for reliable decision making.  The addition of real-time to advanced geostatistics provides an extra dimension to the site investigation, hence allowing the development of fast and cost-effective sampling strategy.