Behl and Kennett (1996) have correlated oxygenation events in the Santa Barbara Basin to Dansgaard/Oeschger warming events found in Greenland ice cores. Lund and Mix (1998) in contrast reported that carbon isotope heavy excursions in the deep-water Core W8709-13 matched stadial episodes in North Atlantic records. The Northern California stacks shown in Figure 11 exhibit events that are reminiscent of North Atlantic instability and can be directly compared in Figure 13.
Within the interval between 75 and 10 ka we observe a similarity between high CaCO3 events and Dansgaard/Oeschger interstadials as recorded in the GISP-2 ice core (k.y.-scale warming events in Greenland; Grootes et al., 1993; Stuiver et al., 1995; time scales of Meese et al., 1994, and Sowers et al., 1993), although there are significant differences between the two records. We have not adjusted the records to improve the match, except perhaps indirectly through the use of Core W8709-13 as the primary age model in most of this period (Lund and Mix, 1998).
The most obvious differences in the records are the amplitudes of events around the carbonate peak CC 3-3 (30-45 ka) and on both the MIS 5/4 glaciation and on the MIS 2/1 deglaciation. The largest CaCO3 peak in the MIS 5/4 glaciation interval, CC 5-2, has no counterpart in the Greenland oxygen isotope record, but a series of events from CC 5-1 (75.1 ka) through CC 3-5 (49 ka) mimic the Dansgaard/Oeschger interstadials. Events in the ice-core record may appear in the northern California CaCO3 stack in the late MIS 3 and MIS 2 interval, but the amplitudes are clearly much lower. The early deglaciation around CaCO3 peak CC 2-1 (18-12 ka) is made up of at least two oxygen isotope events in GISP-2 rather than the one large event we observe, but the younger Dryas (~13 ka) can be detected in both records. Before 85 ka there is little similarity between the two records, perhaps because only near-glacial conditions trigger a coherent response in both time series.
Clearly, the two records have some forcing in common but other factors limit the coherence between the time series. Determining these factors must in part depend upon determining the processes that produce the carbonate record. As always, the question arises whether the CaCO3 record is produced by production or dissolution. Production implies a surface ocean link and a possible atmospheric connection to the North Atlantic. Dissolution, in contrast, implies a deep ocean connection perhaps linked to fluctuations in production of North Atlantic Deep Water.
The comparison between the CaCO3 and Corg MAR time series (Fig. 14) shows no strong relationship between them, as would be expected if CaCO3 burial were driven by a simple link to productivity. Corg percentages are generally higher in interglacials but the MAR does not show a clear trend, indicating that the percentage data is affected by dilution with terrigenous detritus. In contrast, both CaCO3 MAR and percent are typically higher in glacials.
Strong CaCO3 MAR events occur sometimes when there is moderately high Corg MAR (e.g., CC 3-3 at 36.5 ka). At other times CaCO3 and Corg MAR events alternate in strength (e.g., MIS 5/4 boundary, 70-85 ka; MIS 2/1 interval, 25-8 ka). Despite the lack of a simple link between CaCO3 burial and productivity, we have other evidence that has convinced us that the k.y.-scale events are driven by production, not dissolution.