Related Experiment Video
Updated: May 16, 2026

In Silico Modeling Method for Computational Aquatic Toxicology of Endocrine Disruptors: A Software-Based Approach Using QSAR Toolbox
Published on: August 28, 2019
Integrated machine learning-bayesian receptor modelling framework for uncertainty-aware source apportionment in a
Sanjit Kumar Sahu1, Annadasankar Roy2, Aswini Kumar Panigrahi1
1Department of Environmental Science, Fakir Mohan University, Balasore, Odisha, India.
Groundwater fluoride contamination in Indian coastal aquifers is mainly from natural rock leaching, especially before the monsoon. After the monsoon, recharge and land use significantly alter fluoride sources, requiring advanced methods for management.
Area of Science:
- Hydrogeology
- Environmental Chemistry
- Water Resource Management
Background:
- Coastal aquifers face complex fluoride contamination due to seawater intrusion, geology, recharge, and human activities.
- Seasonal variations and overlapping hydrogeochemical processes make evaluating fluoride sources challenging.
- Existing methods struggle to accurately apportion fluoride contributions in dynamic aquifer systems.
Purpose of the Study:
- To develop and apply an integrated, uncertainty-aware framework for identifying dominant groundwater fluoride processes.
- To quantify the relative contributions of different sources to fluoride contamination.
- To link fluoride sources to lithological and land use factors in a coastal aquifer.
Main Methods:
- A Bayesian-Absolute Principal Component Scores-Multiple Linear Regression (APCS-MLR) receptor model was integrated with Self-Organizing Maps (SOM).
- Data-informed priors, prior-posterior checks, and No-U-Turn Sampler (NUTS) inference were used for uncertainty-aware source apportionment.
- Seasonal monitoring across 171 stations in an Indian coastal aquifer system provided the dataset.
Main Results:
- The Bayesian APCS-MLR model identified geogenic leaching as the primary fluoride source (89.14%) in the pre-monsoon season.
- Post-monsoon season showed a significant decrease in geogenic control (48.86%) with increased contributions from recharge-driven pH effects (16.13%) and ion exchange (3.83%).
- Lithology and land use were identified as key regulators, with fine-textured formations and croplands showing persistent geogenic influence, while aquaculture settings exhibited stronger post-monsoon anthropogenic impacts.
Conclusions:
- The study presents a robust, uncertainty-aware methodology for source apportionment in complex hydrogeochemical environments.
- Seasonal dynamics significantly alter fluoride sources, with recharge and land use playing crucial roles post-monsoon.
- The findings support informed groundwater management strategies for addressing fluoride contamination in coastal aquifers, aligning with Sustainable Development Goal 6.
Related Concept Videos
Model Approaches for Pharmacokinetic Data: Distributed Parameter Models
The distributed parameter models are specifically designed to account for variations and differences in some drug classes. This model is particularly useful for assessing regional concentrations of anticancer or...
Model Approaches for Pharmacokinetic Data: Compartment Models
Two primary types of compartment models are recognized: mammillary and catenary. The more...
