A high-quality spectral class dataset for global inland and coastal water bodies spanning 1984–2021

We built a 30-m resolution spectral class dataset for global inland and coastal waters spanning 1984 to 2021, using high-quality Landsat 5 and 8 Collection 1 surface reflectance products in combination with our previously developed OCN (omni-spectral classification and scoring with near-infrared) method. The resulting spectral class database provides a powerful tool for advancing large-scale assessment of inland and coastal aquatic ecosystems over the past four decades, with substantial implications for water quality management and global conservation initiatives.