In a groundbreaking development for environmental monitoring, a team of researchers led by Sheikh Fahim Faysal Sowrav from the National Oceanographic and Maritime Institute (NOAMI) in Bangladesh, and Institute of Water and Flood Management (IWFM), Bangladesh University of Engineering and Technology (BUET), and Bangabandhu Sheikh Mujibur Rahman Maritime University, has pioneered a semi-automated technique for assessing surface water quality in the Sundarbans. The research, published in ‘Heliyon’, which translates to ‘Healthy’, combines machine learning (ML) and remote sensing data to create a dynamic and adaptive model for continuous water quality monitoring.
The Sundarbans, a vast mangrove forest sprawling across Bangladesh and India, is a critical ecosystem known for its biodiversity and ecological significance. However, it’s also highly vulnerable to changes in water quality, which can be influenced by a myriad of factors, from climate change to human activities. Traditional methods of water quality assessment are often labor-intensive and impractical for large, remote areas like the Sundarbans. This is where the new technique comes in, offering a more efficient and scalable solution.
The research team, led by Sheikh Fahim Faysal Sowrav, used Google Earth Engine (GEE) and AutoML to create models that predict key water quality parameters such as Sea Surface Temperature (SST), Total Suspended Solids (TSS), Turbidity, Salinity, and pH. The models were then validated against field measurements, showing a strong alignment and effectively capturing spatial and temporal variations. As Sowrav puts it, “This integration provides a more efficient alternative to traditional methods, which are resource-intensive and less practical for large-scale, remote areas.”
So, what does this mean for the maritime sector? For starters, the ability to continuously monitor water quality can help maritime industries make informed decisions about their operations. For instance, shipping companies can use this data to plan routes that minimize environmental impact, while fishing industries can ensure they’re operating in areas with healthy water quality. Furthermore, the technique’s ability to identify deteriorating water quality trends can support climate resilience and policy-making, ultimately contributing to sustainable environmental management practices.
The research also opens up opportunities for commercial ventures in water quality monitoring technology. Companies specializing in remote sensing and machine learning could develop and market tools based on this technique, catering to the maritime sector and beyond. Moreover, the technique’s applicability to other ecologically sensitive areas presents a global market potential.
However, the study acknowledges some limitations, including data availability and uncertainties in ML predictions for dynamic water systems. But despite these challenges, the research represents a significant step forward in water quality monitoring, offering a valuable tool for protecting and preserving critical ecosystems like the Sundarbans. As the maritime sector continues to grapple with environmental challenges, innovations like this will be key in driving sustainability and resilience.
