Sign in Subscribe
Genetic research

Hybrid Approach for Predictive Mapping of Arsenic Pollution in Groundwater Resources

Strategic Insights

— 1 min read

Researchers at the University of Guilan in Iran have developed a new methodology for modeling and mapping arsenic contamination in groundwater. The approach uses a coupled supervised self-organizing map (SSOM) and genetic algorithm (GA) to identify areas with high arsenic concentrations. By analyzing groundwater samples and variables that introduce arsenic to the aquifer systems, the researchers established a relationship between arsenic concentration and affecting factors. The final optimal model showed a high capability in both training and testing phases, and the predicted arsenic concentration map was verified by comparing measured and predicted values.

Key Takeaways:

  • The researchers used a coupled SSOM and GA to model and map arsenic contamination in groundwater, demonstrating a high level of accuracy in both training and testing phases.
  • The study identified population density, distance from industries, and nitrate concentration as the top factors correlating with arsenic concentration in groundwater.
  • The high correlation between arsenic and nitrate concentrations is attributed to excessive use of pesticides and fertilizers in agricultural lands as well as industrial activities.
  • The predicted arsenic concentration map can be used for managing water quality and safeguarding public health.
  • The study's findings provide a strategic guide for land-use planning, particularly in deciding the optimal locations for establishing industries.
  • Vahid Gholami, a researcher at the University of Guilan, led the study and is available for further information.

Statistics:

  • R-squared value for the optimal model in the training phase: 0.99
  • Mean squared error (MSE) for the optimal model in the training phase: 0.004
  • R-squared value for the optimal model in the testing phase: 0.9
  • MSE for the optimal model in the testing phase: 1.9
  • Correlation between arsenic and nitrate concentrations: N/A
  • Study area: The entire region covered by the alluvial unconfined aquifer in Iran
  • Sample size: Not specified
  • Timeframe: The study was conducted in 2024 and published in 2025

Sources:

  • Exposure and Health, 2024;16(3):775-790
  • NewsRx, Findings on Life Science Reported by Investigators at University of Guilan (A Hybrid Approach of Supervised Self-organizing Maps and Genetic Algorithms for Predictive Mapping of Arsenic Pollution In Groundwater Resources). Genomics & Genetics Weekly. August 29, 2025; p 1033.

Read more