Methods
In our studies, we use different models of the Earth system as well as statistical methods to analyse instrumental and proxy data. Analysing paleoclimate records and model results in tandem allows for an improved understanding of climate transitions and the identification of forcing and feedback mechanisms in the climate system.
Reconstructions vs. model: Meridional distribution of C13 concentration in per mille based on the standard value PDB in the glacial Atlantic 20,000 yrs ago, T. Hesse
Paleoclimate simulations with complex Earth System Models
State-of-the-art Earth System Models explicitly resolve the three-dimensional fields of atmospheric and oceanic dynamics and include all major climate components such as land vegetation, the cryosphere, and the marine biosphere. Such models provide a laboratory for the numerical simulation of various past climates as well as for future climate changes.
The climate simulations enable a separation of the externally forced climate signal from internal climate variability. In addition, simulations of past climate transitions, e.g. during the last glacial-interglacial cycles, allow examining thresholds and feedbacks in the climate system that might be relevant for the future.
In our work, special attention has been paid to the validation of the model results by comparison to terrestrial, marine and ice core data. Furthermore, we focus on an improved data-model comparison by explicitly simulating key proxies used for paleoclimate reconstructions.
Scheme of the complex Earth System Model COSMOS
Statistical analysis of observations, simulations, and proxy data
The statistical analysis of climate observations, simulations, and proxy data combines paleoclimate reconstructions and climate simulation results. It provides estimates of natural climate variability, which is relevant for the detection of the anthropogenic influence on climate change.
The simplest approach comparing paleoclimate data and model results is to subsample the model output fields, picking out data only from these locations and seasons for which the paleoclimate reconstructions exist. Modelled signal-to-noise-ratios may be used to evaluate locations where climate phenomena can be detected. Comparison can also be made with different simulations, in order to assess whether external forcing factors raise the levels of multi-decadal to centennial variability to a similar extent in both the simulated and reconstructed climates.
Typical circulation patterns, e.g. ENSO or NAO, can be evaluated with the help of simulation results, as well as instrumental and historical data. Historical and reanalysis data sets are used to find the associated atmospheric circulation and sea surface temperature patterns. One objective of such studies is to understand the teleconnections that control low frequency variations in different climate proxy records.
One of our tools is the use of correlation and composite maps on sea level pressure, sea surface temperature, and sea ice fields on different time scales.
Dynamical system theory and conceptual models
In addition to our work using complex climate models and statistical analysis, it can be of great utility to reduce the climate system to low-order, box, and conceptual models. This complementary approach has been successfully applied to a number of questions regarding feedback mechanisms and the basic dynamical behaviour of the climate system. It can help differentiating between fast-response and slow-response climate states and variables.
Alternatively, one can also try to derive a phenomenologically based climate theory. This concept is based on the distinction between observed fundamental climate modes relying on their own physical mechanisms and derived modes that emerge from the interaction of other modes. Such a deterministic concept can further improve our understanding of different observed climate modes at interannual to millennial time scales.