A Review on Artificial Intelligence for Water Quality Prediction in Amazonian Countries

Water quality prediction plays an important role in environmental monitoring and ecosystem sustainability in the Amazon. Therefore, this review focuses on determining the advances in the scientific production of artificial intelligence in water quality prediction in the Amazon, as well as the limitations and perspectives compared to water quality indexes (WQI). In this sense, Boolean operators were applied, using the following terms: “artificial intelligence”, “machine learning”, “water quality,” and “Amazonia” The databases were Scopus, web of Science, Springer, and IEEE. In this study, 14 scientific articles published during the period 2000-2024 focused on Amazonian countries were evaluated. Although in the Amazon low scientific production was evidenced and is led by Brazil, the highest scientific growth was for 2021, and 93% belongs to the Scopus database, with a compound annual rate of 12.16%. The IA is characterized by using data from governmental institutions and is only limited to parameters such as Total Suspended Solids (TSS), Total Organic Carbon, Turbidity, and Chlorophyll, using satellite imaging techniques, and the most commonly used algorithm was the Clustering Algorithms. In this context, AI applications are still very low in Amazonian countries compared to other European countries. Its limitations are in the accuracy and the limited amount of physicochemical and microbiological data used for predictions. However, AI is a tool that will replace the water quality indexes used manually.