Soil and Water Pollution

In landfills, non-invasive resistivity and Induced Polarization (IP) geoelectrical surveys effectively define contaminant leachate plume geometry, depth, and continuity beyond the limitations imposed by monitoring wells.
Our geoelectrical surveys at two closed landfills in Ontario delineated contaminated groundwater which, when integrated with hydrogeochemical data, informs environmental monitoring, risk assessment, and remediation.
This approach is transferable across diverse vulnerable environmental settings and contaminated sites.

At a northern Ontario First Nation affected by a nearby industrial landfill, our geophysical surveys delineated leachate-impacted zones, supporting transient multi-decadal modelling of contaminant transport and plume migration to improve forecasting and remediation planning.
Our transport models incorporate site hydrostratigraphy, precipitation records, and long-term hydrogeochemical datasets compiled from environmental assessment and water-quality reports.
Together, these analyses provide a robust framework for evidence-based remediation planning in vulnerable communities.

In landfill sites leachate accumulation, moisture, ionic content, and biogeochemical activity define a subsurface Waste Stabilization Zone (WSZ) that controls methane emissions.
At a municipal landfill in southern Ontario, we used resistivity and IP proxies of the WSZ as antecedent variables in a supervised machine-learning model to infer landfill-scale trends in surface methane emissions, with surface biogas measurements as labelled variables.
These results provide the basis for extending this research to identify cross-media contaminant exchange pathways in other site-specific contexts using complementary numerical approaches, such as probabilistic machine-learning mapping.

Unlike single-metric pollution indices that quantify only contamination intensity, our machine-learning analysis of leachate plumes preserves their multivariate hydrogeochemical structure, allowing the identification of pollution regimes, mixing zones, and migration fronts within the aquifer.
Integrating ML-derived hydrogeochemical facies with 3D resistivity/IP sections provides complementary, mutually constraining interpretive evidence, resolving plume transport, lithological constraints, and matrix-controlled attenuation processes.

Our recently initiated environmental magnetism project with the University of Edinburgh aims to characterize toxic anthropogenic fine and ultrafine airborne particulate matter (PM) at the nanoscale in environments where conventional air-quality monitoring is limited.
The project uses First-Order Reversal Curve (FORC) magnetometry on PM-loaded topsoil samples and applies the AI-based inversion framework FORCINN, developed by the Edinburgh team, to recover magnetic nanoparticle size and shape distributions using micromagnetically modelled libraries of synthetic FORC diagrams with known granulometry.
When integrated with complementary chemical data, these magnetic fingerprints provide a non-invasive, cost-effective approach for air-quality monitoring, source attribution, and linking PM size distributions to health outcomes, transferable to rural and underserved regions.

We developed a preliminary numerical framework, tested with synthetic matched case-control data, combining average treatment effect and uplift modelling to assess exposure-AML relationships, identifying a causal effect only for indoor benzene exposure.
We further applied Random Forest, Bayesian Additive Regression Trees, and Rule Ensemble to synthetic data generated by the previous regression model to estimate heterogeneous treatment effects, capturing nonlinear exposure response relationships, subgroup-specific responses, and policy-relevant exposure thresholds.
In a subsequent phase, these computational frameworks would be applied in an Indigenous First Nation using community-specific environmental and health monitoring data.

We advance a novel, community-driven model that integrates environmental engineering with Indigenous epistemologies through sustained dialogue, shared interpretation, and the ethical co-production of knowledge.
Rather than positioning communities as passive data recipients, this approach engages Indigenous communities in shaping research questions, guiding field activities, and interpreting results through an Indigenous Integrated Knowledge Translation (IIKT) strategy that fosters trust, transparency, meaningful participation, and local leadership in environmental stewardship.
Engineering activities are embedded within culturally grounded processes, such as Talking Circles, where Indigenous Knowledge is recognized as an equally valued system guiding the research agenda and framing environmental risk in relation to land, health, culture, and well-being.
Capacity building occurs through hands-on training in field monitoring and data interpretation, generating accessible, visual, community-owned environmental data that strengthen locally led environmental assessment and support informed decisions on contamination, health risks, and remediation priorities.