Stochastic paleoclimatology: Modeling the EPICA ice core climate records

Area of Science:

• Paleoclimatology
• Climate Dynamics
• Statistical Modeling

Background:

• Pleistocene glacial cycles exhibit complex fluctuations across multiple timescales.
• Ice core data, such as from the European Project for Ice Coring in Antarctica (EPICA) Dome-C, provide crucial long-term records of climate variables.

Purpose of the Study:

• To analyze and model the stochastic behavior of paleoclimate time series.
• To assess the coupling and interrelationships between carbon dioxide, methane, nitrous oxide, and temperature during glacial cycles.
• To evaluate the stability of physical climate processes and model fidelity.

Main Methods:

• Multifractal time-weighted detrended fluctuation analysis applied to 800,000 years of EPICA Dome-C data.
• Modeling time series as one-dimensional periodic nonautonomous stochastic dynamical systems.
• Developing a four-variable model with intervariable coupling terms and response functions.

Main Results:

• Distinct near-red-noise and white-noise behaviors identified below and above the 100,000-year glacial cycle, respectively.
• Carbon dioxide and temperature were found to stabilize each other and methane/nitrous oxide.
• Methane and nitrous oxide were found to destabilize each other and carbon dioxide/temperature.

Conclusions:

• Glacial pacing is primarily driven by carbon dioxide and temperature.
• Terrestrial biosphere feedbacks, linked to methane and nitrous oxide emissions, modulate glacial cycles.
• The study confirms intervariable stability relationships consistent with paleoclimate data.