Conference Papers
Conference presentations and proceedings, including work shared at the AGU Fall Meeting.
Ground Water Quality Assessment using Quantum Machine Learning [Journal Link]
Published in 2026 17th International Conference on Information, Intelligence, Systems & Applications (IISA), 2026
Groundwater contamination by arsenic, nitrate, and uranium poses significant public health risks across the United States, and accurately predicting exceedance of maximum contaminant levels (MCLs) is critical for risk assessment and monitoring, particularly under sparse and heterogeneous sampling. This work develops a hybrid quantum-classical machine learning framework that classifies contaminant presence above or below a threshold (80% of the MCL) using a large-scale, curated dataset of groundwater observations and environmental covariates. Principal Component Analysis selects 12-14 informative features, which are encoded into quantum states using parameterized feature maps (ZZ and Y-ZZ embeddings) and combined with classical features to train a neural network classifier. Evaluated against classical machine learning baselines on nationwide data, the hybrid quantum models achieve comparable predictive performance. Gains are modest, but quantum feature embeddings offer competitive accuracy and a pathway toward richer representations of nonlinear environmental interactions. The study is presented as the first large-scale evaluation of hybrid quantum ML for groundwater quality assessment.
Unlocking the Potential of Wastewater Treatment Plants for Phosphorus Recovery: Identifying Optimal Locations for Technology Implementation and Fertilizer Production
Presented at American Geophysical Union (AGU) Fall Meeting, 2022 and 2023
Microbial Risk Assessment Informed by De Facto Reuse at Public Drinking Water Systems Across Contiguous United States [Link]
Presented at American Geophysical Union (AGU) Fall Meeting, 2021
De facto potable reuse (DFR) occurs when treated wastewater discharged to surface waters is present in downstream drinking water sources, and pathogens such as Cryptosporidium, Giardia, and enteric viruses that resist conventional disinfection can persist in treated effluents. Direct Potable Reuse guidelines set log inactivation targets for these pathogens to achieve an annual risk benchmark of 1 infection per 10,000 people per year, and earlier quantitative microbial risk assessments predicted that public water systems (PWSs) under DFR scenarios (roughly 10% or more treated effluent in the source) exceed this benchmark. Using the DRINCS geospatial model to assess DFR at PWSs of all sizes nationwide, the results predicted that more than 3% of U.S. PWSs could pose high public health risk relative to the benchmark, while only 8% of these affected systems employ advanced treatment technologies. About 12.3 million people were estimated to be exposed to microbial risk in source water (at DFR above 10%), with risks rising seasonally or during droughts as in-stream dilution varies. DRINCS includes a proximity analysis of upstream discharges and downstream intakes and accounts for in-stream pathogen die-off. The authors propose using DFR predictions to rank relative risks for communities, target epidemiological and risk assessment studies, and prioritize infrastructure investments in disinfection.