Thermophysical Properties of Surface Seawater for Sustainability in Estuarine Systems

This study investigates the thermodynamic feasibility of establishing desalination plants along the Arabian Sea coast of Goa (India). Based on the experimental values of ultrasonic velocity (u) and density (u) at varying temperature and salinity conditions, isothermal compressibility (βT), adiabatic compressibility (βs), acoustic impedance (Z), and molecular free length (Lf) were derived. Utilizing the real seawater data, we developed robust machine learning models, including random forest (RF), gradient boosting (GB), AdaBoost (AB), and stack ensemble machine learning (SEML), to predict these thermodynamic properties solely on the basis of temperature and salinity. Our models exhibited high accuracy, enabling reliable predictions that inform energy and resource management strategies for desalination in India’s coastal regions, highlighting the importance of understanding the estuarine system.