Date of Completion


Embargo Period



Julie Granger, Craig Tobias

Field of Study



Master of Science

Open Access

Open Access


The cause of the initial rise in atmospheric CO2 during the last deglaciation remains unknown. Coincident with the rising atmospheric CO2, the δ13C of atmospheric CO2 decreased by ~0.3‰ during Heinrich Stadial 1 (HS1: 14.5-17.5 kyr BP), which requires the input of carbon from an isotopically light reservoir. The light carbon signal in the atmosphere occurred concurrently with the carbon isotope minimum, or a decrease in surface ocean δ13C of ~0.5‰, suggesting the two phenomena are related. The leading hypotheses explaining the δ13C minimum are (1) enhanced ventilation of 13C-depleted abyssal water in the Southern Ocean which in turn caused low δ13C values in the surface ocean and atmosphere, and (2) a reduction in the Atlantic Meridional Overturning Circulation (AMOC) weakened the efficiency of the ocean’s biological pump, thereby increasing the concentration of light carbon in the surface ocean. In order to evaluate these two hypotheses, we compiled 70 published, globally-distributed planktonic foraminiferal δ13C records and enhanced the sampling resolution of three low resolution records from the western tropical Pacific (WTP). The HS1 δ13C anomaly, or the relative difference in δ13C between the LGM and HS1, was calculated for each record, and we compared the spatial patterns between ocean basins and within the tropical Pacific and Southern Oceans. We find that the average δ13C anomaly is similar in all ocean basins. We also find similar δ13C signals in the eastern equatorial Pacific (EEP) upwelling regime and the WTP convergence zone. In the Southern Ocean we find a latitudinal trend of δ13C anomalies decreasing in magnitude progressing towards higher latitudes and the region of abyssal upwelling. Because the Southern Ocean hypothesis implies that the δ13C signal should be largest in the Southern Ocean and in upwelling regions, our results are inconsistent with a Southern Ocean driver. Our findings are instead consistent with a recent modeling study that simulated the effects of a weakened biological pump, which produced an excess of isotopically light carbon in the surface ocean and atmosphere, similar to observations. We conclude that our results are broadly consistent with a biological pump mechanism, suggesting that the initial rise in atmospheric CO2 was driven by biogeochemical processes in the upper ocean as opposed to upwelling of light carbon from the abyss.

Major Advisor

David Lund