In the heart of India, a silent crisis has been brewing beneath our feet—groundwater contamination. Arsenic and nitrate, two insidious pollutants, have been seeping into our water sources, posing significant risks to both the environment and public health. But a beacon of hope shines through the work of Gift Mbuzi, a researcher from the Department of Computer Science and Engineering at SRM University-AP in Amaravati, Andhra Pradesh. His recent study, published in the Mehran University Research Journal of Engineering and Technology (translated from Urdu as “Mehran University Research Journal of Engineering and Technology”), is harnessing the power of artificial intelligence to predict and mitigate this growing threat.
Mbuzi’s research is a game-changer, applying machine learning (ML) and deep learning (DL) models to predict groundwater contamination trends. By mapping a five-year historical dataset of key physicochemical parameters, his study compares various ML and DL models to provide data-driven insights into groundwater quality. “We found that Biochemical Oxygen Demand (BOD), total dissolved solids (TDS), and conductivity are important predictors of arsenic contamination,” Mbuzi explains. “Agricultural and industrial activities, on the other hand, dictate nitrate contamination.”
The study’s temporal analysis revealed intriguing trends. Arsenic levels showed a decrease post-2019, likely due to dilution effects and regulatory measures. Nitrate contamination, however, fluctuated region-wise, highlighting the need for localized solutions. After hyperparameter tuning, the XGBoost model emerged as the most predictive, outperforming traditional regression analysis. “XGBoost achieved an R² value of 0.70, indicating a strong predictive capability,” Mbuzi notes.
The implications of this research are vast, particularly for the energy sector. Groundwater contamination can impact energy production, from cooling water for power plants to water used in hydraulic fracturing. By enabling real-time monitoring and better mitigation of contamination, Mbuzi’s AI-based systems can contribute to sustainable resource management planning. “Our findings indicate the potential of predictive models based on AI in groundwater monitoring,” Mbuzi says. “This can enable better mitigation of contamination and support sustainable resource management planning.”
The study also caught detailed non-linear relationships among water quality parameters using Partial Dependence Plots (PDP), providing a more nuanced understanding of the complex interactions at play. This research is a significant step forward in the field of hydrogeochemistry and water resource management, offering reliable AI-based systems for real-life cases.
As we look to the future, Mbuzi’s work paves the way for more sophisticated and localized solutions to groundwater contamination. By leveraging the power of AI, we can better understand and mitigate this silent crisis, ensuring a healthier environment and more sustainable resource management for all. The energy sector, in particular, stands to benefit from these advancements, as reliable water sources are crucial for various energy production processes. Mbuzi’s research is a testament to the power of AI in driving positive change and shaping a more sustainable future.