Forensics receptor models, such as polytopic vector analysis, are used to differentiate contaminant profiles and sources. But how accurate and sensitive are they when dealing with complex, real-world environmental datasets that may have extra information and varying numbers of samples?

We are pleased to announce that Nicholas Rose, Tim Negley, and Dr. Michael Bock of TIG Environmental (a Verdantas company) recently published an article in Environmental Forensics on the Predictive Accuracy of Polytopic Vector Analysis in Environmental Forensics: Sensitivity to Seeding Methods, Random Noise, and Sample Size.

The team artificially generated 435,600 unique datasets with between two and seven sources of PCBs or PFAS that approximate real-world environmental datasets. After evaluating the datasets, they found that certain seeding methods are more accurate when sources are similar to each other or are present in small proportion. The results also show that increasing sample sizes up to 25 samples can improve predictive accuracy.

The article is published online in the journal Environmental Forensics here: Predictive Accuracy of Polytopic Vector Analysis in Environmental Forensics: Sensitivity to Seeding Methods, Random Noise, and Sample Size: Environmental Forensics