Databases for Validation

Critical to insuring confidence in the satellite derived data is the need for high quality in situ matchup data. Listings of various water quality databases can be found elsewhere on this website. Historically, the widely-used SeaBASS database has been primarily used to store oceanic and coastal data, including data as far back as 1930. The most common parameters found in SeaBASS are total absorption coefficient, R_rs, and conductivity, temperature and depth (CTD) measurements. The least frequent parameters on SeaBASS are SPM, primary productivity, phycocyanin, and total scattering coefficient (Werdell et al., 2003).

For inland waters, the main available database is the LIMNADES, which has approximately 39,794 data measurements taken from 3,547 stations that are available either through request or download at the current time. Measurements stretch back nearly 30 years, with most being recorded between 2000 and 2023. The most common parameter within LIMNADES is Chl-a concentration and total suspended solids (same measurement previously referred to as SPM in the SeaBASS dataset)._.

Because of the scarcity of R_rs data in inland and coastal waters (8% of data records in LIMNADES), a recent effort on collecting data from both SeaBASS and LIMANDES led to the GLObal Reflectance community dataset for Imaging and optical sensing of Aquatic environments (GLORIA), which includes over 7,000 hyperspectral remote sensing reflectance measurements from both inland and coastal waters (Lehman et al. 2023). While the data needs from oceanic, coastal and inland waters are different, there is also a geographic need for data. The figure below shows the distribution of the data within the GLORIA database and the parameters of calibration and validation data on SeaBASS, AERONET-OC, AERONET-MAN, and MOBY databases. From these maps we observe that the eastern coast of the US is extremely well studied, as well as most of the western coast of the United States and the coast of Europe. There is a lack of data in the Indian Ocean, the Greenland Sea/Arctic Ocean, as well as the Western Pacific, Antarctic Ocean and Southeastern Atlantic (Figure 4B). For inland and coastal waters, most of the data comes from the Eastern United States, Western Europe, and a few areas in China, Japan, Australia and New Zealand. There is a clear lack of data in the Global South, especially in Latin America and the Caribbean, and in Africa. These maps highlight the need for support to conduct studies in these regions.

References

Werdell, P.J., Bailey, S., Fargion, G., Pietras, C., Knobelspiesse, K., Feldman, G., McClain, C., 2003. Unique data repository facilitates ocean color satellite validation. EoS Transactions 84, 377–387. https://doi.org/10.1029/2003EO380001

Lehmann, M.K., Gurlin, D., Pahlevan, N., Alikas, K., Conroy, T., Anstee, J., Balasubramanian, S.V., Barbosa, C.C.F., Binding, C., Bracher, A., Bresciani, M., Burtner, A., Cao, Z., Dekker, A.G., Di Vittorio, C., Drayson, N., Errera, R.M., Fernandez, V., Ficek, D., Fichot, C.G., Gege, P., Giardino, C., Gitelson, A.A., Greb, S.R., Henderson, H., Higa, H., Rahaghi, A.I., Jamet, C., Jiang, D., Jordan, T., Kangro, K., Kravitz, J.A., Kristoffersen, A.S., Kudela, R., Li, L., Ligi, M., Loisel, H., Lohrenz, S., Ma, R., Maciel, D.A., Malthus, T.J., Matsushita, B., Matthews, M., Minaudo, C., Mishra, D.R., Mishra, S., Moore, T., Moses, W.J., Nguyễn, H., Novo, E.M.L.M., Novoa, S., Odermatt, D., O’Donnell, D.M., Olmanson, L.G., Ondrusek, M., Oppelt, N., Ouillon, S., Pereira Filho, W., Plattner, S., Verdú, A.R., Salem, S.I., Schalles, J.F., Simis, S.G.H., Siswanto, E., Smith, B., Somlai-Schweiger, I., Soppa, M.A., Spyrakos, E., Tessin, E., Van Der Woerd, H.J., Vander Woude, A., Vandermeulen, R.A., Vantrepotte, V., Wernand, M.R., Werther, M., Young, K., Yue, L., 2023. GLORIA – A globally representative hyperspectral in situ dataset for optical sensing of water quality. Sci Data 10, 100. https://doi.org/10.1038/s41597-023-01973-y