|
Kenneth Pye Associates Ltd, Crowthorne Enterprise Centre, Crowthorne Business Estate, Old Wokingham Road, Crowthorne RG45 6AW, UK k.pye{at}kpal.co.uk
Analysis of geochemical data has become an important tool used in forensic comparisons of soils and sediments. The combined use of inductively coupled plasma optical emission spectrometry and mass spectrometry (ICP-OES and ICP-MS) instrumentation allows the abundance of up to 50 elements to be routinely determined in small samples (typically >0.1g). However, key issues concern the extent to which analysed subsamples are representative of the parent material from which they are taken, the best means of comparing datasets for different samples, and interpretation of the significance of the results obtained. ICP measurement precision for most elements is good, but it is important to understand the degree of variation that can arise due to subsampling procedures and selective transfer mechanisms relating to forensic exhibits, as well as the extent of spatial (and sometimes temporal) variability which exists in nature. Although analysis of several different size fractions is often helpful where sufficiently large samples are available, analysis of a standardized <150µm fraction separated from a bulk sample in many cases provides adequate discrimination between samples and provides the most practical method for mass-screening of samples. Where possible, duplicate or triplicate sample preparations should be made, and several analytical determinations made on each prepared subsample. However, the additional time, cost and sample size requirements involved need to be weighed against the benefits of undertaking additional types of analysis of the samples. In order to obtain maximum information from multielement geochemical data, the dataset should be evaluated using a variety of numerical, statistical and graphical procedures. This paper discusses a number of options for such data evaluation using a simple dataset example. Casework experience and experiments have shown that, even in complex situations, multi-element geochemical data can provide very sensitive environmental indicators. However, such data should normally be used in combination with the results of other analyses when making forensic comparisons of soils and sediments.
