Towards operational retrieval of chlorophyll-a in inland waters using optical water types

Monitoring inland water quality has become increasingly critical due to the escalating effects of climate change and anthropogenic activities on aquatic ecosystems. Remote sensing of water quality parameters in optically complex inland water bodies remains challenging due to spatially diverse inherent optical characteristics, which limits operational products. The currently adopted approach of two-step optical water type classification and class-specific coefficient modeling tends to provide improved retrieval accuracies in complex inland water conditions. This paper aims to evaluate different models applicable to various orbiting satellites to identify the best-suited model for an operational algorithm. GLORIA in-situ hyperspectral dataset was used to generate simulated satellite reflectance spectra by convolving it with various sensor-specific spectral response functions. Several indices and band ratio-based algorithms were evaluated, and accuracies were assessed using various metrics. Results show a significant increase in retrieval accuracies in all algorithms when using the optical water class-specific coefficient model. This work aims to provide a robust framework for delivering operational remote sensing-based water quality products for understanding long-term changes in inland water bodies.