This dataset represents the remote sensing and Secchi Disk depth data that were used to calibrate and validate a global Mixture Density Network algorithm. Remote sensing reflectance, water quality, and satellite data associated with this dataset were compiled from several providers and institutions in global waters between 1990 and 2022. Data were compiled to support water quality remote sensing applications and the development of a Secchi Disk Depth algorithm. Information about the methods used in each dataset and how they were collected is available in the referred publications.

Files in the Simulated Dataset.zip folder (rrs_tm_v3.xlsx, rrs_etm_v3.xlsx, rrs_oli_v3.xlsx, and rrs_hyper_v3.xlsx) represent the in situ datasets. The first three sheets are the simulated Rrs for the specific sensor (e.g., TM, ETM+, or OLI) and the last one (rrs_hyper_v3.xlsx) is hyperspectral Rrs. Band simulation was performed using the bandSimulation R package available here: https://github.com/dmaciel123/BandSimulation. Detailed metadata for these files are available in the simulated dataset metadata.XLSX file.For the satellite matchups (ACOLITE_Matchups.ZIP), the files tm_acolite_maciel_v3.xlsx, etm_acolite_maciel_v3.xlsx, and oli_acolite_maciel_v3.xlsx are the matchups between in situ measured Zsd and Landsat data for TM, ETM+, and OLI, respectively. Detailed metadata description for this dataset is available in acolite metadata.xlsx.