(Table 2) Vertical fluxes of organic carbon and sediment matter from the photic zone in the Arctic Basin

On the basis of analysis of satellite and field data collected in Russian Arctic Seas maps of distribution of primary production for different months of the vegetation period were compiled. These maps were used to estimate annual primary production of organic carbon: 55 million tons in the Barents Sea; about 20 million tons in the Kara Sea; 10-15 million tons in the Laptev Sea and in the East Siberian Sea, 42 million tons in the Chukchi Sea. In the central and eastern parts of the Barents Sea during the vegetation period values of primary production decreased by factor >5 (from >500 to <100 mg C/m**2/day). By reviewing results of studies with sediment traps vertical fluxes of organic carbon in different regions of the Arctic Basin were estimated. Significant temporal variability of Corg fluxes with maxima during phytoplankton blooms (by 830 mg C/m**2/day) was noted. Typical summer fluxes of Corg are 10-40 mg C/m**2/day in the southern Barents Sea, 1-10 mg C/m**2/day in the northern Barents Sea and in the Kara Sea, and up to 370 mg C/m**2/day in the zone of marginal filters of the Ob and Yenisey rivers.

(Table 2) Weight percent organic carbon from ODP Site 138-847

As the length of marine cores increases and sampling intervals decrease, the need for rapid and inexpensive means of determining sediment composition has become apparent. In this study we examine one potentially useful technique for assessing compositional changes in marine cores, diffuse reflectance spectrophotometry. We examined near-ultraviolet, visible, and near-infrared reflectance spectra from five data sets. Each data set consists of calibration samples and test samples. The calibration samples' spectra were related to a sediment component using multiple linear regression. The resulting regression or calibration equations were then evaluated using the test samples. Calibration equations were written relating spectra to several sediment components incduding carbonate (Atlantic and east Pacific Rise ODP Site 847), organic carbon content (Atlantic and east Pacific Rise), and opal content (east Pacific Rise). The correlation coefficients for the regression equations ranged from a high of 0.99 for carbonate and opal at ODP Site 847 to a low of 0.97 for Atlantic carbonate indicating that spectral variations are highly correlated to sediment composition. All of the equations include a substantial number of variables from shorter visible and longer near ultraviolet wavelengths suggesting that these wavelengths are especially important for devices designed specifically to scan marine cores. Although equations for estimating organic and carbonate content appear independent of other sediment components, the opal equation is strongly dependent on carbonate content indicating that opal concentration is correlated to carbonate content. Tests of the calibration equations indicated that all our equations reasonably estimate the pattern of changes, either down core or in surface sediments. Where our spectral estimates have difficulty is with absolute values, frequently over or underestimating observed values by a substantial amount. Within these limitations diffuse reflectance spectrophotometry can be a useful tool for characterizing marine cores and as our understanding of the relationship between spectra and mineralogy improves so will estimates of absolute values.