Testing the Forensic Predictive Geolocation Significance of Human-made Particulates in Urban Surface Soils

  • Mathijs van Gijtenbeek

Student thesis: Master's Thesis


Forensic geolocation is of increasing interest to the soil forensics and geoscience community. Not much is known about the forensic geolocation significance of anthropogenic materials in soils. As a result, the goal of this study was to investigate the potential forensic predictive geolocation significance of human-made particulates in urban surface soil samples using quantitative data.

In total, 78 samples have been collected from 2 urban areas (Enschede, the Netherlands & Cardiff, Wales) and 1 rural area (Cornwall, England). Multiple analyses were done to differentiate between the soil samples. The soil colour, particle size and particle sorting were determined for every soil sample. In addition, subsamples were taken and analysed of 34 samples (32 urban samples and 2 rural samples) to determine the presence and abundance of the inorganic anthropogenic materials using automated scanning electron microscopy with linked energy dispersive spectrometers (SEM-EDS) (using a QEMSCAN system) and fluorescence microscopy analyses to determine the number of fibres with a human-made origin in the samples. Furthermore, potential plastic particles were handpicked from 10 samples for fourier-transform infrared spectroscopy(FTIR)analysis.

Then hand-picked particles were analysed with the FTIR and the spectra were compared with two FTIR databases for identification. Only 1 particle was identified as a plastic (polyethylene), the other analysed particles could not be reliably identified using the 2 different databases. However, based on the spectra, it could be concluded that there were different types of potential plastic particles analysed. More research is needed to determine if plastics could be used to distinguish between soil samples in forensic casework.

This study was unable to reliably distinguish between human-made and natural fibres and therefore also unable to quantify the number of fibres with a human-made origin in urban surface soil samples. Further research is required to determine the forensic significance of fibres for forensic geolocation.

The results of the automated SEM-EDS analysis showed that there is a significant variation between soils based on the percentage of anthropogenic compositional groups in each sample. Distinctions could be made based on the overall percentage of human-made phases between samples for (semi-)rural, urban and industrial areas. Differentiations could also be made between samples based on the individual human-made compositional groups. However, the differences based on the individual human-made compositional groups were less significant when comparing samples within the same area and therefore also harder to use to distinguish between soil samples within the same area. In addition, 10 samples taken within the same street spaced about 5-10 meters apart from each other also showed significant variation, both based on the overall percentage of human-made phases and based on various individual human-made compositional groups. As a result, human-made compositional groups can be used to predict the general geolocation of a forensic soil sample, for example, to predict if the sample originated from a (semi-)rural area, urban area or an industrial area. However, further research is required to predict the exact geolocation of a forensic soil sample.
Date of Award2022
Original languageEnglish
Awarding Institution
  • University of South Wales
SupervisorDuncan Pirrie (Supervisor), Sorcha Diskin (Supervisor) & Zella Parry (Supervisor)

Cite this