Diese Seite ist derzeit leider nur auf Englisch verfügbar.
Coordinator: K. Fraedrich
Overarching Questions
Variability: The largest contribution to climate variability is given by a smooth background variability whose spectra are white or scale with enhancing intensity for low frequencies. In some key regions, the sea surface temperature reveals a non-stationary 1/f-power spectrum up to centennial time scales, which inhibits the assessment of anthropogenic climate change scenarios. The background variability is caused by non-linear processes leading to instabilities, which limit weather predictability to a few weeks. Atmosphere and ocean interact with the land and biosphere as reservoirs of energy, water and other substances, with the hydrological cycle as a most prominent exchange process. The reservoirs modify climate variability and enhance the memory of the climate system and predictability. The variability of the hydrological cycle, exhibiting droughts or wetness, has severe impact on soil, vegetation, agriculture, and water supply which is the major reason for the interest in monitoring extreme events.
Achievements and problems: A prominent example of climate prediction on a seasonal time scale is ENSO forecasting, which has become possible since oceanographic data and assimilation are accessible. Although dynamic atmosphere-ocean models have been successfully applied to simulations of much longer time scales and palaeo-climates, problems related to radiation, oceanic mixing, the biosphere, the hydrological cycle, and clouds remain unresolved. Possible solutions for complex parameterisation problems can be obtained with novel optimisation approaches (for example maximum entropy production). In particular, a deeper understanding of atmospheric and oceanic dynamics including their feedback processes is necessary to increase the confidence in anthropogenic climate change scenario simulations.
Challenges: Less exploited predictability lies in troposphere-stratosphere interactions, with slowly downward propagating wind anomalies. In addition, long term memory, which has been observed and correctly simulated by dynamic models, offers promise for predictability. This, however, needs further analysis since long term memory may also lead to difficulties in empirical and dynamical predictions. The most challenging application in climate prediction is the prediction of extreme events. At present, statistical theories can predict recurrence times and expected threshold crossings, which are both affected by low frequency variability clustering extreme events. The ability of dynamic models to predict individual extremes, for example, a season of drought or wetness in particular regions, has to be assessed.
Goals of RA-B
The aim of RA-B is the investigation of predictable elements of the coupled climate system (the terrestrial climate and its components atmosphere, ocean, land, and biosphere and in coupled settings), the performance of best possible coupled IPCC– like (Intergovernmental Panel on Climate Change) climate scenario runs and the performance of pilot decadal predictions of climate components. The most relevant application is climate prediction for anthropogenic impacts given by greenhouse gas emission scenarios and land-use changes. The studies require an assessment of natural variability, including studies on long term memory and extreme events, which restrict predictability.
Accordingly, goals include:
- Analyses of climate variability and the identification of responsible mechanisms. Main aspects are long term memory, extremes, instabilities, feedbacks, thresholds and risks involved.
- Development of conceptual frameworks of predictability including verification of predictions by available observations and proxy data. Development and implementation of prediction tools using appropriate dynamic and empirical models in a hierarchy spanned by different spatial and temporal resolution, complexity of processes, parameterisations, and number of compartments.
- Understanding of fundamental problems of variability from decades to centuries and millennia, which includes parameterisation of small scale processes (for example oceanic mixing) and biosphere representation and interaction using novel approaches like maximum entropy conditions.
- Predictability of extreme events, which is a highly relevant and interesting research issue due the outstanding impact of singular climate excursions. Predictability in the presence of long term memory since this delivers a unique and hitherto unexploited reservoir for the skill of climate predictions.
- Quantification of uncertainties in climate projections and risk assessment.
Methods
- Analysis of coupling mechanisms based on the data base provided by RA-A.
- Analysis of predictable elements of the climate system through use of a hierarchy of coupled models with varying complexity, resolution and geographical extent.
- Further development of coupled models.
- Development of coupled model initialisation for prediction purposes.
- Investigation of decadal predictability, by using the optimal estimates obtained through data assimilation in RA-A, as the initial conditions for experimental long-term climate prediction experiments. The combination of data assimilation and coupled modelling expertise gives Hamburg a near-unique role worldwide in this area.
- Pilot prediction of climate indices and uncertainties
Relations with other RA's
These goals require intense cooperation among meteorology and oceanography (ZMAW) working in the parts A, B, and D. Part A aims at data assimilation as a prerequisite for the analysis of natural climate variability. Parts A and B produce boundary conditions for the regional part D. The predictability studies in B depend on the work on feedbacks in part C. Like RA-A, RA-B will contribute to the public common of this CoE by providing, on a regular basis predictions of climate indices and through the provision of climate forecast outputs.
Participating Scientists
- UniHH: R. Blender, K. Fraedrich, E. Kirk, F. Lunkeit, M. Claussen (Meteorological Institute), M. Hort (Institute for Geophysics), D. Quadfasel, D. Stammer (Institute for Oceanography), M. Funke (Institute for Macroeconomics and Economic Policy), M. Kalinowski (ZNF), M. Köhl (Centre for Forestry and Forest Products)
- MPI-M: M. Giorgetta, S. Hagemann, D. Jacob, J. Jungclaus, J. Marotzke, E. Meier-Reimer, E. Roeckner, J.-S. von Storch
- GKSS Research Centre: H. von Storch
