5.2. Algorithms

In comparison with the previous issue [6], the modified algorithms for calculation of chlorophyll and suspended matter concentrations were used. They were derived by using new in situ data obtained in the summer expeditions of 2008-2009 in the northern part of the Caspian Sea [37].

Chlorophyll concentration

The regression equation for calculation of chlorophyll concentration from SeaWiFS data is similar to the previous one [6]:

Chl = 0.732·[Lwn(555)/Lwn(510)]3.40, (15)

where Lwn(555) and Lwn(510)] are the normalized water-leaving radiances at 555 and 510 nm (n = 55, r2 = 0.708, the error is equal to 0.33 mg/m3).

The algorithm to derive chlorophyll concentration from MODIS-Aqua data was simplified in comparison with [6]:

Chl = 0.573 (Rrs(488)/Rrs(551)-2.39. (16)

where Rrs(488) and Rrs(547) are the above-surface remote-sensing reflectance at 488 and 547nm, respectively (n = 55, r2 = 0.638, the error is equal to 0.41 mg/m3).

Particle backscattering and TSM

First, the particle backscattering coefficient bbp was derived from the remote sensing reflectance corrected to the bottom reflection [38]. TSM concentration was calculated by using the regression equation between TSM and the particle backscattering coefficient bbp :

TSM = 67.1 bbp + 0.554, (17)

where TSM is in mg/l , bbp in m-1 (n=42, R2 = 0.835, the regression error is 0.81 mg/l).

Yellow substance absorption

The values of ag were calculated with the semi-analytic algorithm as described in 4.2. The algorithm worked stably only for the May-September period so the mean monthly distributions of the ag values were calculated only for this season.