Non-linear Modeling of Chlorophyll-A Dynamics in the Indian Ocean Around Western Sri Lanka

A 23-year analysis of phytoplankton biomass variability off Sri Lanka’s western coast (78.75°E–80.5°E, 5.5°N–10°N) was presented using MODIS-Aqua ocean color data (January 2002–May 2025). This region encompasses the biologically productive Sri Lanka Dome upwelling feature. Generalized Additive Models (GAMs) to capture non-linear relationships between chlorophyll-a concentrations and environmental drivers—an approach essential for tropical marine systems where linear models fail to represent threshold behaviors and curved responses were employed. The GAM framework explained 71.2% of pigment variance (R2 = 0.698), with sea surface temperature contributing 45% through non-monotonic thermal responses, precipitation 23% via interactive nutrient-dilution effects, and biogenic carbonates 12%. Cross-validation confirmed robust predictive performance (RMSE = 0.34 mg m−3). These findings advance mechanistic understanding of phytoplankton drivers in the tropical Indian Ocean, supporting ecosystem assessment and fisheries management applications.