Der Ozeanboden

Decoding feedbacks and thresholds: Objective causality analysis in Earth-system simulations

Important: Please read in detail the instructions provided in the umbrella advertisement (linked here).

Do not send applications directly to the contacts on this page.

Project Description

Understanding cause-and-effect relationships is a key motivation for using numerical models in Earth-System science. Comprehensive climate or ecosystem models have reached a level of complexity that render traditional approaches to analyze causality (e.g. by means of sensitivity experiments) practically impossible. Feedback mechanisms and the threshold behavior in such models are therefore often only interpreted with regard to assumed underlying mechanisms instead of analyzing the system thoroughly. At the same time and from the dynamical systems perspective, model experiments provide a complete description of the state of the system and its changes. The project will employ state-of-the-art methods (e.g. convergent cross mapping, transfer entropy) from non-linear dynamics and network analysis to infer causality in model simulations. A key goal is to infer causality based on objectively analyzing model results instead of seeking for patterns along assumed cause-and-effect chains. Model experiments for the project will be available. Initially, the project will focus on self-sustained oscillations of the large-scale ocean overturning circulation. These quasi-periodic oscillations occur in different settings of a climate model and have periods ranging from centuries to millennia. In a second step, experiments with transitions between stable states (e.g. sea-ice coverage in the Arctic) will be analyzed in terms of the feedbacks underlying the threshold behavior. For this part of the project, output from climate models as well as from marine ecosystem models will be used. It is expected that the intended causality analysis is a major enabler for Earth system and ecosystem modeling, as it empowers us to uncover the true mechanisms behind environmental changes.

Further Reading

• Prange, M. et al. (2023) A multicentennial mode of North Atlantic climate variability throughout the Last Glacial Maximum. Science Advances, 9(44), eadh1106. doi: 10.1126/sciadv.adh1106 (for initial model experiments)

Modalities

The doctoral project will be supervised by Bernd Blasius (University of Oldenburg) and Michael Schulz (University of Bremen). Depending on the personal expertise of the doctoral candidate, it will be supervised within the working group “Mathematical Modeling” (Oldenburg) or the working group “Geosystem Modeling” (Bremen). The remuneration corresponds to pay group TV-L 13 with 75% of the weekly working hours. The positions is for a fixed-term period of 4 years, starting January 1st 2026 the earliest and until at most December 31, 2029 (according to § 2 WissZeitVG) with the aim of obtaining a doctorate.

The project mentoring team also includes: Alexey Ryabov (Oldenburg),
Matthias Prange (Bremen), and Ute Merkel (Bremen)

Main Tasks

• Scientific research in the field of:
  applied mathematics and non-linear system dynamics
• Analysis of:
  numerical model experiments
• Writing of scientific publications
• Participation in international conferences

Formal Requirements:

Completed scientific university degree (Master/University diploma or comparable) in physics, mathematics, meteorology or related discipline.

Further questions?

Questions about the project can be addressed to Bernd Blasius ([Bitte aktivieren Sie Javascript]) or Michael Schulz ([Bitte aktivieren Sie Javascript]).