Satellite observations make it possible to construct long-term series of data covering vast ocean areas. Data from the ocean color sensors enable us to derive a set of bio-optical parameters such as chlorophyll concentration, the particle backscattering and yellow substance absorption coefficients which manifest spatial and temporal variability of three important seawater constituents – photosynthetic phytoplankton, particulate and colored organic matter. Such work was made for the seas surrounding Russia with SeaWiFS data in 1998-2001 [1-2], 1998-2002 ,1998-2004 , 1998-2006 , 1998-2008 .
The new issue continues the long-term series of data on bio-optical characteristics of the seas, surrounding Russia, with the addition of data 2009-2010, as compared with the previously issued .
Two problems had to be solved when calculating bio-optical characteristics for the new atlas. The first was Reprocessing 2009 and 2010, which changed the data from satellites scanners SeaWiFS and MODIS, used in our computations. The changes from Reprocessing 2009 resulted from sensor-specific calibration and characterization updates, algorithm revisions, updated ancillary data sources. Majorchanges in the algorithms include new aerosol models and model selection methodology, new Rayleigh tables, and many other changes (http://oceancolor.gsfc.nasa.gov/REPROCESSING/R2009/).
In 2011 the MODIS Aqua ocean color products were reprocessed for the latter part of the mission, starting from 1 January 2009. The calibration before January 1st 2009 was unchanged (http://oceancolor.gsfc.nasa.gov/WIKI/OCReproc20100MA.html).
To account for changes related to the results of the above mentioned reprocessing, we had to re-compute all the previously calculated bio-optical characteristics.
The second problem was joining SeaWiFS and MODIS-Aqua data together . Because of the final termination of SeaWiFS activity in the end of 2010, it was decided, in preparation for the new atlas, to convert the SeaWiFS monthly averages to MODIS-like monthly averages (by using the regression equations MODIS-Aqua vs. SeaWiFS). It was done for the period from January 1998 to June 2002; from July 2002 MODIS data were used. This procedure applied to all considered seas, except the White Sea, where all computations were performed only on MODIS-Aqua data (since July 2002).
The regional algorithms, used for calculation of bio-optical characteristics for the seas of Russia, were derived on basis of the ship measured data. For this reason, the reprocessing of satellite data has not changed them. The exceptions are the algorithms for the White Sea which are based on the regression equations between thein situ measured values of the suspended matter concentration TSM (or chlorophyll concentration) and the remote sensing reflection derived from satellite data (see 3.2).
The algorithms for the Caspian Sea were also changed, because there were new data of in situ measurements were obtained and used to refine the previous bio-optical algorithms (see 5.2).
The daily Level-3 values from MODIS-Aqua data were generated by averaging the Level-2 values within a given bin of 3 x 3 km (for the Barents Sea the size of a bin was 6 x 6 km). The monthly or other Level-3 files were created by averaging the daily Level-3 values over the corresponding period.
The monthly and annually averaged data on SST from MODIS-Aqua were added in this issue to provide an opportunity for analysis of connection between changes of bio-optical characteristics and SST[8, 9]. The SST data were downloaded from the NASA web site http://oceancolor.gsfc.nasa.gov. The daytime level 2 MODIS Aqua files were used. The quality level qual_sst varies between 0 and 4, where 0 indicates best quality and 4 indicate complete failure (usually land). When averaging, the pixels with values of qual_sst exceeding 2 were filtered out.