Assessing intra and interannual variability of water quality in the Sundarban mangrove dominated estuarine ecosystem using remote sensing and hybrid machine learning models
Ismail Mondal,SK Ariful Hossain,Sujit Kumar Roy,Jay Karmakar,Felix Jose,Tarun Kumar De,Trinh Trong Nguyen,Ismail Elkhrachy,Nguyet-Minh Nguyen
DOI: https://doi.org/10.1016/j.jclepro.2024.140889
IF: 11.1
2024-01-28
Journal of Cleaner Production
Abstract:Coastal water bodies are susceptible to high levels of pollution as a result of population growth, rapid urbanization, the expansion of industrial activities, including ports, and expanding tourism-related activities. Untreated sewage disposal from municipalities, industrial wastewater, and numerous recreational and commercial activities along the coast not only deteriorates the water quality but also poses a significant threat to the marine ecosystem, food chain, and public health. Therefore, continuous monitoring of coastal water quality is essential for the sustainable upkeep of marine and aquatic ecosystems. The study aims to assess the intra- and interannual variability of coastal water quality in the Sundarbans estuarine system, straddling the borders of India and Bangladesh. The study employed Landsat 8 OLI satellite data and in-situ measurements of water quality parameters, including temperature, pH, dissolved oxygen, phosphate, total phosphate, nitrate, and nitrite. These measurements were obtained from 100 sampling locations across the entire Sundarban estuarine ecosystem. The data collection spanned the pre-monsoon, monsoon, and post-monsoon periods in the years 2021 and 2022. Several sophisticated machine learning models, including Random Forest (RF), K-Nearest Neighbor (KNN), Support Vector Machine (SVM), and Artificial Neural Network ( ANN ), were employed to precisely forecast various crucial water quality indicators. Cross-validation results showed that the random forest model is most promising, with R 2 values ranging between 0.50 and 0.96, followed by the artificial neural network (R 2 values ranging between 0.36 and 0.97), the support vector machine (R 2 values ranging between 0.29 and 0.97), and the KNN (R 2 values ranging between 0.26 and 0.94). Subsequently, a random forest algorithm was customized for the study area to compute the coastal water quality index on a seasonal scale for the years 2021 and 2022. Results reveal that the concentrations of dissolved oxygen, phosphate, total phosphate, nitrate, nitrite, pH, and water temperature were relatively higher during the pre-monsoon and post-monsoon seasons compared to the monsoon season. In situ and remote sensing data analysis confirmed that Sundarbans estuarine water quality was relatively good during the monsoon period as compared to the rest of the year. In addition, research also highlighted the significance of monsoonal rainfall in the watershed for maintaining coastal water quality. It is estimated that the coastal water was particularly vulnerable due to the combined effects of industrial waste water and municipal sewage discharge. Ultimately, the study demonstrates the effectiveness of machine learning models and remote sensing in predicting the condition of coastal waters. The findings of this research can offer significant perspectives to decision-makers in developing a sustainable strategy for the management of coastal water resources.
environmental sciences,green & sustainable science & technology,engineering, environmental