Geo-information about the spectral variability of the water-leaving radiance is the key to the validation of satellite and in situ measurements and the development of regional algorithms. In this study, using cluster analysis, five trends were identified that are characteristic of various phenomena in the northwestern part of the Black Sea (summer and winter blooms, turbid waters, river runoff). Typical values of the remote sensing reflectance coefficient are calculated for each case. Additionally, the standard values of the color indices for each cluster are calculated. It is shown that the ratio of the color index CI(412/443 nm) is slightly variable and equals 0.8±0.07.This can be a reference point for recovering incorrect (negative) satellite Rrs(λ) values in the shortwave region. In a similar way, the color index was calculated according to the MODIS and VIRS data, it was shown that on days with a turbid atmosphere (high AOT values), the standard deviation of the color index is 30%. For days with a clear atmosphere, the color index is close to the in situ results.