Temperature reconstruction based on stable nitrogen and argon isotope measurements on Greenland ice cores

Döring, Michael (2020). Temperature reconstruction based on stable nitrogen and argon isotope measurements on Greenland ice cores (Unpublished). (Dissertation, Universität Bern, Philosophisch–naturwissenschaftliche Fakultät, Physikalisches Institut, Abteilung für Klima– und Umweltphysik)

Text doering20phd.pdf - Other Restricted to registered users only Available under License BORIS Standard License. Download (48MB)

Stable isotopes of the inert gases nitrogen (δ15N) and argon (δ40Ar) measured on ancient air extracted from polar ice cores provides the opportunity to reconstruct past temperature variations. Thereby, the temperature reconstruction is conducted by inverting firn densification and heat diffusion models to fit the δ15N and δ40Ar data. The main focus of this work is to develop automated procedures for solving that inverse problem. Automated fitting procedures have the advantage that manual tuning of parameters is avoided which leads to a better reproducibility and comparability between different studies. Furthermore, the use of automated algorithm saves working hours, which should not be underestimated. Another aim of this thesis is the reconstruction of the Holocene temperature history of Greenland summit using one of the automated methods developed in this study.
First, an introduction into the topic is given in Chapter 1 containing a motivation of Holocene climate reconstructions, and an overview about a variety of reconstruction methods and studies. In section 1.2 the physical fundamentals needed for the understanding of this work are presented and in section 1.3 an overview about different methods of firn model inversion techniques are provided.
Chapter 2 presents the automated firn model inversion methods developed in this work. In section 2.1 the suitability of a 2-dimensional polynomial transfer function for describing a firn model is investigated for using the model in static or dynamic behaviour. This was done in order to visualize the non-linearity of the problem. Section 2.2 presents an algorithm for the automation of the δ18Oice calibration technique. In sections 2.3 and 2.4 the fundamentals and implementations of a Monte-Carlo based automation technique independent from δ18Oice is described and evaluated on synthetic data for Holocene reconstructions and on Glacial data.
In chapter 3 the results of using the Monte-Carlo based algorithm (sections 2.3 and 2.4) for fitting already published GISP2 (Greenland Ice Sheet Project Two) Holocene δ15N, δ40Ar and δ15Nexcess data with two different firn models are shown and analysed. Additionally, the gained temperature solutions are evaluated against each other and are set in context to two other studies, where the same data but different temperature reconstruction strategies were used. Also, the performance of two different firn models for fitting Holocene gas isotope data was investigated and an uncertainty estimation for the reconstructed temperature estimates was performed.
Chapter 4 presents the comparison of the most robust temperature estimate (T(δ15N)) – obtained in chapter 3 – with a variety of marine and terrestrial temperature proxies from the North Atlantic region (section 4.1), with GISP2 accumulation rate (section 4.2), and GISP2 stable water isotope (δ18Oice) data (section 4.3). The comparison of T(δ15N) with the marine and terrestrial temperature proxies reveals comparable signatures for multi-centennial to multi-millennial signals. Furthermore, evidence for a coherence of AMOC variability and Greenland summit temperatures during the Holocene was found and it is concluded that Greenland summit temperature variability is predominantly driven by changes in North Atlantic circulation patterns and AMOC variability for multi-centennial to multi-millennial oscillations. The comparison of T(δ15N) with GISP2 accumulation rate data shows moderate correlations for the general trends and multi-centennial oscillations especially in the early Holocene. The results show that a regime change took place at about 6.5 kyr b2k, which altered the dependency between T(δ15N) and GISP2 accumulation rate data from correlation to anti-correlation behaviour. The comparison of T(δ15N) and GISP2 δ18Oice data shows that the general trends in both variables are highly correlated over the Holocene. However, only weak correlations were found for multi-centennial to multi-millennial oscillations.
Chapter 5 contains two publications where the author of this study contributed as co-author. The first study (section 5.1) investigates the role of volcanic forcing as a potential driver of centennial to millennial-scale temperature changes in Greenland using ice core data and climate model simulations. The role of changes in overloading pressure on permeation processes inducing δAr/N2 fractionation in the firn is presented in the second study (section 5.2).

Interest & Impact

Downloads

1 since deposited on 19 Jun 2024
1 in the past 12 months

Citations

5 Citations in Web of Science ®
6 Citations in Scopus

Search

in Google Scholar™

Services

Actions (login required)

Edit item

Actions (login required)

Edit item