April 7, 2021

« All News

Jay Thompson to Present on the Use of Machine Learning in Environmental Forensics at the ACS Spring Meeting

Jay Thompson Ph.D., P.E. will present "Multi-site machine learning-enhanced source classification of PAH impacted sediments: Environmental forensics applications" to the Division of Environmental Chemistry at the American Chemical Society (ACS) Spring Meeting.  This virtual session will be broadcast live to conference attendees on April 8th and is available on-demand from April 19 – 30.

His co-authors Lauren Fitzgerald, Ph.D., P.E., and Amanda Hughes, Ph.D., Geosyntec.

Jay Thompson is an environmental engineer based in with technical expertise in environmental forensics, contaminant bioavailability, and contaminant fate and transport.  He has consulted on some of the largest sediment megasites across the U.S., advising his clients on nexus, allocation and managing potential environmental liabilities. 

Lauren Fitzgerald is an environmental engineer in the Baton Rouge office of Geosyntec Consultants with technical experience in contaminated site investigation and remediation.  Her experience includes environmental site characterization and forensics at numerous large sediment cleanup sites in the U.S. She has served her clients in litigation projects with contaminant and site historical forensics to evaluate liability and support allocation.

Amanda Hughes is an Indiana-based senior practitioner at Geosyntec Consultants. She provides remediation consulting services, and expert witness and litigation support to industrial clients and law firms across the country. Amanda is an expert in polychlorinated biphenyl (PCB) forensics, fate, and transport. Her experience includes liability allocation, investigations of inadvertently-generated PCB congeners, and critical reviews of PCB fate, transport and bioaccumulation models.

ACS is one of the world's largest scientific societies. With thousands of attendees across academia, industry, and government, its biannual national meeting is one of the premier scientific and technical venues for disseminating cutting-edge developments in all fields of chemistry.

Abstract

A primary objective of environmental forensics is connecting contaminant impacts in environmental media to potential sources. Experienced practitioners can recognize patterns in contaminant distributions (i.e., fingerprinting), to make such connections.  However, these analyses are time consuming and subject to bias. This work significantly reduces these disadvantages through the application of machine-learning algorithms.

A Random Forest model was trained on polycyclic aromatic hydrocarbon (PAH) sediment fingerprints compiled from 13 sites throughout the United States.  Human practitioners examined each fingerprint and identified one of three dominant PAH patterns – heavy pyrogenic, light pyrogenic, and petrogenic – and the model was trained on the labeled fingerprint data.  Overall, the trained Random Forest classifier concurred with the human practitioners in 92% of the test set sediment samples, with the majority of practitioner/model disagreement occurring in the ambiguous space between heavy and light pyrogenic categories.  The trained model correctly categorized 100% of pure phase reference samples.  No difference in model performance was observed between sites.  This modeling approach can be used at future sites and the method is generalizable to other compound classes.  This machine learning application will improve the efficiency and defensibility of environmental forensic evaluations.

More Information

About the event: https://www.acs.org/content/acs/en/meetings/acs-meetings.html
For consultation regarding environmental forensics or machine learning, contact Jay Thompson at This email address is being protected from spambots. You need JavaScript enabled to view it.; Lauren Fitzgerald at This email address is being protected from spambots. You need JavaScript enabled to view it.; or Amanda Hughes at This email address is being protected from spambots. You need JavaScript enabled to view it..
Learn more about Jay: https://www.linkedin.com/in/jaymthompson/
Learn more about Lauren: https://www.linkedin.com/in/lauren-fitzgerald-b0473331/
Learn more about Amanda: https://www.linkedin.com/in/amandashughes/