Australian geoboffins have zoomed in and enhanced CSI-style soil analysis: Bung it in a machine, find the crime scene

New approach could lead to a system for geolocating dirt samples

The University of Canberra and Australian Federal Police have detailed a new method for digging up dirt on crims – and it could mean an end to literally digging up dirt on crims.

Anyone with half an eye on the screen when CSI or the like is blaring away will be familiar with the concept of soil analysis: samples of dirt found on or around a suspect or victim can be used to narrow the scene of a crime down to a particular area.

In the real world, though, carrying out soil analysis takes a little more than slipping a slide under a microscope and pressing a button on a flashy Visual Basic interface pretending to be a database. Typically, it involves having to go to potential sites and take fresh soil samples for comparison – and it's this stage scientists are looking to simplify.

Using geochemical analysis through Fourier-transform infrared spectroscopy (FTIR), X-ray fluorescence (XRF), magnetic susceptibility, and mass spectrometry, the team tried to pinpoint the origins of three blind samples within a 260km² region broken down into a 1km²-resolution grid.

The result was an immediate exclusion of 60 per cent of the search area. "Much of forensics is about elimination, so being able to rule out 60 per cent of an area is a substantial contribution toward successfully locating a sample," explained Patrice de Caritat, principal research scientist at Geoscience Australia.

"You can reduce the time, risk and investment of the ongoing investigation. The more parameters we look at, the more accurate the system is. We have reached 90 per cent detection in some cases, although we think that would involve too many factors for real-world crime detection."

While the experiment used simulated soil samples, they were prepared to an agreed standard by the Australian Federal Police to accurately reflect the samples that might be gathered during an investigation. Performance was variable between samples, however, with the researchers suggesting differences in samples taken for geochemical mapping and for evidence of a crime may hamper real-world use of the approach.

There's hope, however, and a major breakthrough may lie ahead in the next part of the team's research. "Most developed countries have existing soil databases, used for such things as mineral exploration or land use decision support," said de Caritat. "We're plugging our methods into these databases to see if we can locate samples from the database information, rather than needing to collect samples specifically for each investigation.

"Conventional soil analysis has already been used in Australia to identify and prosecute criminals. For example, soil analysis was used to identify the movements of a man who carried out a sexual assault on a young girl in Adelaide. There are several such examples. We now want to take this further."

If successful, de Caritat and colleagues would have developed the closest thing yet to the CSI ideal: an automated system for pinpointing, as precisely as possible, where a given soil sample came from – and without detectives ever having to break out the bucket and spade.

The new approach was highlighted in a keynote presentation by de Caritat at the Goldschmidt2021 conference on geochemistry and related subjects this week; the paper has been published in the Journal of Forensic Sciences. ®