Aurelien Meray (Computer Science)

Aurelien Meray (Computer Science)

Aurelien Meray, a DOE Fellow and Ph.D. student in computer science at Florida International University’s Applied Research Center (FIU-ARC) has successfully published a peer-reviewed journal paper in the Environmental Science and Technology (ES&T) Journal. ES&T is a groundbreaking environmental science and technology research journal that publishes rigorous and well-researched articles for a broad and diverse audience of scientists, policymakers, and the general public.

The paper titled “PyLEnM: A Machine Learning Framework for Long-Term Groundwater Contamination Monitoring Strategies”, was authored by Aurelien Meray along with Savannah Sturla, Masudur R. Siddiquee, Rebecca Serata, Sebastian Uhlemann, Hansell Gonzalez-Raymat, Miles Denham, Himanshu Upadhyay, Leonel E. Lagos, Carol Eddy-Dilek, and Haruko M. Wainwright, a team of scientists from Florida International University, University of California Berkeley, Lawrence Berkeley National Laboratory, Savannah River National Laboratory, and Massachusetts Institute of Technology.

The paper highlights mainly the development of a Python package called PyLEnM (Python for Long-term Environmental Monitoring) that has a complete set of machine learning (ML) functions for long-term groundwater contamination monitoring. Among the most significant ML innovations are:

  • Time series clustering to identify groups of wells with similar groundwater dynamics and to guide spatial interpolation and well optimization.
  • Evaluating several regression models for spatial interpolation, with automated model selection and parameter tweaking.
  • A proxy-based spatial interpolation technique using spatial data layers or in situ measurable variables to predict contaminant concentrations and groundwater levels.
  • A well optimization approach to identify the most effective subset of wells for long-term monitoring.