Retrodiction of secular variations in deep-sea CaCO3 burial during the Cenozoic

Abstract

Deep-sea sediments record changes in oceanic carbonate chemistry and CaCO3 sedimentation rate through temporal variations in the total burial of CaCO3 and the position of the carbonate snowline, i.e., the ocean depth at which CaCO3-free sediments are first recorded. This paper links mathematically secular changes in snowline to those in the burial rate through a set of relatively simple equations. When the available Cenozoic deep-sea burial records are employed to predict secular variations in snowline, the process fails at some time in the past, indicating that these records are not consistent with each other. The burial records are more likely the source of this problem, as they involve far more uncertainties than the snowline records. As a consequence, we introduce a method for estimating carbonate burial through the use of a canonical CaCO3-depth profile, which can respond dynamically to secular changes in carbonate sedimentation and the positions of both the snowline and the carbonate saturation horizon. The resulting synthetic CaCO3 burial record is consistent with snowline records and indicates that the burial rates offered by Davies and Worsley (1981) are generally too high, with highly questionable maxima at 25 and 47 Ma BP. Our estimates of burial are more consistent with the range advanced by Mackenzie and Morse (1992); nevertheless, our results differ from the latter with respect to timing and magnitude of the variations. Our approach allows simultaneous calculation of the mean carbonate ion concentration of the deep sea. We find that carbonate-ion levels fell through the Cenozoic and are similar to those calculated by Tyrrell and Zeebe (2004), using a different model. Secular variations in CaCO3 burial are found to be primarily driven by changes in the Ca2+–CO32− ion product within the bottom-waters, with an increase in the sedimentation rate of CaCO3 of secondary importance over the Cenozoic.