An inverse modeling approach to investigate deep ocean ventilation from radiocarbon records

Submitted in partial fulfillment of the requirements for the degree of Master of Science at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution September 2022. The goal of this project is to investigate the relationship between paleoceanographic radiocarbon records...

Full description

Bibliographic Details
Main Author: Duffy, Faith
Other Authors: Marchal, Olivier
Format: Thesis
Language:English
Published: Massachusetts Institute of Technology and Woods Hole Oceanographic Institution 2022
Subjects:
Online Access:https://hdl.handle.net/1912/29354
Description
Summary:Submitted in partial fulfillment of the requirements for the degree of Master of Science at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution September 2022. The goal of this project is to investigate the relationship between paleoceanographic radiocarbon records and the ventilation history of deep oceanic basins during the last 40 kyrs. Deep ocean ventilation changes, especially changes in Atlantic meridional overturning circulation (AMOC), are often invoked to explain the deglacial rise in atmospheric carbon dioxide (CO2) concentration as inferred from ice core records. Much of our current understanding regarding ventilation of the deep ocean during the deglaciation comes from records of the radiocarbon concentration of benthic foraminifera and deep-sea corals (paleo-Δ14C data). Here, we combine a global compilation of paleo-Δ14C data for the past 40 kyrs with a 16-box model of the world ocean (except the Arctic Ocean) to address two key questions: (1) To what extent can the paleo-Δ14C data be explained by atmospheric Δ14C variations when deep ventilation rates are fixed to modern ocean estimates? and (2) To what extent can the paleo- Δ 14C data be explained by atmospheric Δ 14C variations when the ventilation rates are allowed to vary? To address these questions, the box model is fitted to the paleo-Δ14C data using the following sequential methods of optimal estimation theory: the linear Kalman filter, the Extended Kalman Filter, the Rauch-Tung-Striebel (RTS) smoother, and a linearized RTS smoother. We find that 62–76% (depending on the assumptions made about air-sea 14CO2 exchange) of the paleo-Δ 14C data for the past 40 kyrs can be explained by the modern flow rates as represented in the box model, if the model is forced with the IntCal20 reconstruction of atmospheric Δ14C. When the flow rates in the model are allowed to vary with time, 74-89% of the data can be explained by the model. Here, the range in data that can be explained reflects the different assumptions ...