A new study published in the journal “Nature” as part of the “Scientific Reports” series, Moroccan researchers, in collaboration with teams from the UAE, Saudi Arabia, and the USA, have utilized deep learning artificial intelligence tools to “map nitrate groundwater pollution in northwest Morocco,” marking a pioneering Moroccan study to enhance water resource management and improve understanding of “nitrate groundwater pollution” in the coastal aquifer of Larache.

This study aims to understand nitrate (-NO₃) pollution in the coastal sandy aquifer located in northwest Morocco, a region heavily impacted by agricultural activities and chemical fertilizer use.

To analyze this phenomenon, researchers used AI and deep learning models to predict nitrate concentrations in groundwater based on various environmental indicators such as Normalized Vegetation Index, Electrical Conductivity, Chlorides, Organic Matter, and Fecal Coliforms.

The results showed that the “TabNet model” was the most accurate and efficient, achieving about 82% success, outperforming the MLP-ANN (Multilayer Perceptron Neural Network) model.

The research analysis, reviewed in full by Hespress, indicated that fecal coliforms and electrical conductivity are the most influential factors in elevated nitrate concentrations, suggesting that irrigation with polluted water and intensive fertilizer use are the main causes of groundwater quality degradation.

Maps generated from the models showed that forested areas in the north contain cleaner water, while agricultural zones in the central and southeastern parts record the highest pollution levels due to intensive farming and poor irrigation water management.

One of the key highlights of this study is that the use of artificial intelligence represents a “powerful tool for monitoring and analyzing groundwater quality and predicting pollution development, helping decision-makers identify the most vulnerable areas and take effective protective measures,” as emphasized by one of the research supervisors, Professor Mourad Aknoui, from the Faculty of Science and Technology in Errachidia, in an interview with Hespress.

According to available data, this work, published on October 7, is the result of scientific collaboration between Moroccan and international laboratories and institutions, including Abdelmalek Essaadi University in Tetouan, the National Center for Energy, Sciences and Nuclear Techniques in Rabat, the Scientific Institute at Mohammed V University, the American University of Sharjah (UAE), Princess Nourah bint Abdulrahman University (Saudi Arabia), the University of Mississippi (USA), and the Genome and Technology Research Team at Moulay Ismail University.

This work is part of the PhD thesis of researcher Mourad Shahid, supervised by Professor Jamal Eddine Stitou Al-Massari from the Faculty of Science in Tetouan at Abdelmalek Essaadi University, with joint supervision from Dr. Ismail Hilal from the National Center for Energy, Sciences and Nuclear Techniques in Rabat, and Professor Mourad Aknoui from the Faculty of Science and Technology in Errachidia at Moulay Ismail University.

The study confirms that employing modern smart technologies can effectively contribute to protecting water resources from pollution and supporting sustainable environmental decisions, ensuring balanced development that considers both humans and nature.

The researchers concluded by proposing the integration of additional data in the future, such as climatic data, agricultural data, and isotopic analyses, to enhance model accuracy and improve understanding of nitrogen sources and its transfer mechanisms within the water system, thereby contributing to sustainable groundwater management in sensitive coastal areas.