Master Thesis Defense Katharina Theresa Scheidt

Title: Modeling Paleotemperatures from Stable Nitrogen Isotopes with the Community Firn Model

Abstract:
Stable oxygen isotopes from water molecules in polar ice (δ18Oice) and stable nitrogen isotopes enclosed in gas bubbles of ancient ice (δ15N) are quantifiable indicators for paleotemperatures.  Paleotemperatures can be reconstructed by forcing firn densification models coupled to heat diffusion and firn air diffusion models with past surface temperature and snow accumulation rate histories and fitting the output δ15N to measured δ15N data. The firn community uses a variety of firn models suitable for various ice core sites. However, different firn models predict different surface temperatures. In order to estimate the uncertainty in reconstructed temperatures resulting from using different firn models, various firn models provided by the Community Firn Model (CFM) - an open-source framework providing several modules for the simulation of firn physics processes - are inverted. An automated inversion procedure was implemented with commonly used SciPy minimizers in order to avoid manual parameter search for minimization. Paleotemperatures were reconstructed from δ18Oice and δ15N at NorthGRIP, Greenland for different combinations of firn densification and gas diffusion models. Furthermore, run times of the CFM were reduced and a second spin run was implemented within the CFM to make the inversion procedure more feasible. In addition, this thesis provides an overview of the architecture of the CFM and the main modules used for this project.

Supervisor: Vasileios Gkinis