Machine learning (ML) models are increasingly being used to analyze complex environmental variables like soil salinity, water quality, and crop suitability, especially in vulnerable coastal zones. For soil salinity, ML algorithms such as Random Forest, Support Vector Machines, and Extreme Gradient Boosting have been effectively used in combination with remote sensing data (e.g., Landsat, Sentinel-2) and field observations to create accurate salinity maps and monitor salt-affected regions over time (Omonov et al., 2023, Zhou & Li, 2024, Nguyen et al., 2025). For water quality, hybrid models that integrate remote sensing data with optimization techniques like Particle Swarm Optimization have been used to predict salinity intrusion in river systems, aiding coastal water management strategies (Bui et al., 2025). Regarding crop suitability, ML models assess multiple soil and climate parameters, such as nutrient levels, pH, and water availability, to recommend optimal crops for specific regions, helping farmers improve yields and adapt to environmental constraints (Deshmukh et al., 2023, Sathya et al., 2024). Collectively, ML enables more efficient, scalable, and data-driven approaches to managing agricultural risks in fragile coastal ecosystems.

Machine learning models analyze soil salinity, water quality, and crop suitability in coastal zones by processing remote sensing and environmental data to predict conditions and classify risk areas. This enables efficient, real-time monitoring and informed decision-making for sustainable land and water management.