A data-driven model of the global calcite lysocline

Gridded maps of sediment calcium carbonate (calcite) concentration and overlying water saturation state [Archer, 1996] are combined with maps of benthic oxygen fluxes and sediment accumulation rates from Jahnke [1996] and Cweink [1986] to drive a diagenetic model of calcium carbonate preservation in deep-sea sediments. The only model input for which we cannot draw a detailed map is the rain rate of calcite to the seafloor, so I use the model to calculate the calcite rain rate required to simulate the observed distribution of calcite concentration on the seafloor. The predictive power of the model can be checked by searching for input parameters to which the model is sensitive and comparing the model requirements with independent data. The model is sensitive to the ratio of organic carbon to calcite rain rates (the rain ratio) and, within the constraint of pelagic sediment trap rain ratio data, is unable to reproduce the calcite field without respiratory dissolution, the promotion of calcium carbonate dissolution by the oxic degradation of organic carbon. However, relative to variability in sediment trap data, the required model rain ratio is insensitive to the extent of anoxic respiration, the stoichiometric ratio of O:C during respiration, the bioturbation rate, the dissolution rate constant, the effect of berate chemistry, and small offsets in the saturation state for CaCO3. The model is sensitive to the accumulation rate of non-CaCO3 material and predicts an Atlantic/Pacific difference in non-CaCO3 rain rate which is consistent with observations. The model predicts that the dissolution flux of CaCO3 from sediments is 25-40 x 10(12) mol yr(-1), roughly half of the total deep sea CaCO3 dissolution rate estimated from the water column alkalinity distribution [Broecker and Peng, 1987; Archer and Maier-Reimer, 1994; Wollast, 1994]. The model also predicts that only 20-30% of the flux of CaCO3 to the seafloor globally escapes dissolution.