Machine Learning, Groundwater Modeling & Software Solutions for Hydrogeologic Decision-Making

We help environmental consulting firms and project teams interpret groundwater conditions, evaluate alternatives, and plan further investigation.

Fontaine dbMax Solutions brings together experience in hydrogeology, numerical groundwater modeling, applied machine learning, and software development to support your project decisions.

View Selected Projects Discuss Your Project

How We Can Support Your Team

Groundwater Modeling & Hydrogeologic Analysis

Connect site observations and conceptual site models with numerical groundwater flow and contaminant transport modeling.

Site characterization and conceptual site model development
Groundwater flow and contaminant transport modeling
Scenario evaluation and investigation planning

Applied Machine Learning

Explore patterns in groundwater data and develop predictions that can be evaluated alongside hydrogeologic interpretation.

Groundwater risk assessment and spatial prediction
Hybrid plume characterization and monitoring design methods
Prediction evaluation and model interpretation

Software, Databases & Modeling Workflows

Organize complex project information and build practical tools that make analyses easier to review, update, and communicate.

Database design, data integration, and quality checks
Repeatable processing and modeling workflows
Custom analysis tools and GIS outputs

Featured development study

Hybrid Machine Learning for Groundwater Plume Characterization and Monitoring Design

We are developing and evaluating a workflow that combines monitoring observations, hydrogeologic information, and physics-informed predictors to interpret plume concentration patterns and identify candidate locations for additional monitoring.

The study uses a controlled synthetic test bed to compare predictions with a known reference plume across three monitoring periods, including migration after source cutoff. Evaluation includes withholding monitoring locations and comparing spatial patterns and concentration thresholds.

Current status: Development and evaluation are ongoing, with potential field applications under consideration.

Ask About the Study →

Comparison at 1,278 days: reference plume, kriging of log-transformed concentrations, and hybrid machine learning plume, with validation statistics for 27 monitoring locations — Reference plume, kriging, and hybrid machine learning at 1,278 days (3.5 years) in the controlled synthetic test bed. Select the figure to view the maps, statistics, and validation note at full size.

Hydrogeologic Experience Behind the Tools

Founded by Russ Fontaine, Fontaine dbMax Solutions draws on his experience as a principal hydrogeologist in site characterization, conceptual site model development, and numerical groundwater modeling. His software and database background supports the integration of monitoring data, modeling, and GIS results.

Russ also serves on the Interstate Technology and Regulatory Council's AI/ML committee, contributing to its use-case subgroup.

About Russ and Our Approach →