Applications of Remote Sensing in Water Quality Monitoring: A Comparative Study of Empirical Algorithms and Machine Learning Models
Volume Title: ICASGE2025
Paper ID : 1163-ICASGE (R1)
Authors
Teaching assistant ,Transportation department, faculty of engineering, Alexandria University
Abstract
Monitoring water quality is essential for safeguarding human health, aquatic ecosystems, and economic stability, particularly in coastal regions that are at risk from pollution and climate change. In addition to the effects of climate change, human activities like agriculture and urban runoff pose a threat to coastal zones, which are areas where freshwater and saltwater mingle. Large-scale monitoring is challenging due to the labor-intensive, costly, and time-consuming nature of traditional in situ measures, which provide precise data. An alternative is satellite remote sensing, which offers broad temporal and spatial coverage for identifying alterations in water quality, like sediment plumes or algal blooms. Although water quality metrics can be estimated from satellite data using empirical techniques.
Because it can evaluate vast datasets and identify intricate correlations between water quality metrics and reflectance, machine learning (ML) has emerged as a possible option. Better predictions and ongoing performance improvement are made possible by ML models' increased accuracy and adaptability. The purpose of this study is to evaluate the efficacy of machine learning models and empirical techniques for monitoring water quality in coastal areas, where climate change is making environmental circumstances more unpredictable.
This study makes a substantial contribution by comparing the effectiveness of empirical techniques with machine learning models for monitoring water quality in coastal areas. Additionally, the growing popularity of machine learning and deep learning applications in geographic artificial intelligence for water quality monitoring is being made possible by the availability of open-source software, which is helping to monitor water contamination.
Because it can evaluate vast datasets and identify intricate correlations between water quality metrics and reflectance, machine learning (ML) has emerged as a possible option. Better predictions and ongoing performance improvement are made possible by ML models' increased accuracy and adaptability. The purpose of this study is to evaluate the efficacy of machine learning models and empirical techniques for monitoring water quality in coastal areas, where climate change is making environmental circumstances more unpredictable.
This study makes a substantial contribution by comparing the effectiveness of empirical techniques with machine learning models for monitoring water quality in coastal areas. Additionally, the growing popularity of machine learning and deep learning applications in geographic artificial intelligence for water quality monitoring is being made possible by the availability of open-source software, which is helping to monitor water contamination.
Keywords
Subjects