An Adapted Sediment Mask Based on FY-4B/AGRI Observations: Implications for Ocean Aerosol Retrieval

Sediments, including suspended sediments and salt precipitates, are commonly masked in aerosol retrieval algorithms over the ocean because they can introduce significant retrieval errors. The Moderate Resolution Imaging Spectroradiometer (MODIS) Dark Target (DT) algorithm applies a sediment mask based on a reflectance differential index between satellite-observed reflectance and power-law-fitted reflectance at 0.55 μm. However, many geostationary satellite sensors lack the 0.55 and 1.24 μm channels. This study proposes a spectrally adaptive sediment-mask framework for the Advanced Geosynchronous Radiation Imager (AGRI) onboard Fengyun-4B (FY-4B). By reconstructing the power-law–based sediment index under reduced spectral dimensionality, we demonstrate that the 0.65 μm channel preserves sediment-sensitive deviation behavior comparable to that of the traditional 0.55 μm channel. This reconstruction is achieved using a three-channel power-law fitting strategy based on the 0.47, 1.61, and 2.25 μm channels. An appropriate threshold for the reconstructed reflectance differential index was determined through threshold-sensitivity analysis. Pixel-level inter-sensor comparisons with MODIS- based reference masks were further conducted for regional and seasonal cases. The results indicate that the proposed method can identify major sediment-affected waters and capture the main spatial patterns of sediment plumes, showing reasonable overall agreement with MODIS-based reference masks, especially in highly turbid coastal and estuarine waters. However, uncertainties remain in weakly turbid waters, plume-boundary regions, and complex atmospheric conditions. The proposed method addresses the sediment-masking limitation of FY-4B AGRI caused by the absence of the 0.55 and 1.24 μm channels and has the potential to support geostationary ocean aerosol retrieval.