Analysis of atmospheric circulation from climate model big data -Current approaches and future challenges

• Main Authors: Handorf, Dörthe, Dethloff, Klaus, Jaiser, Ralf, Rinke, Annette
• Format: Conference Object
• Language: unknown
• Published: 2017
• Subjects: Arctic; North Atlantic; North Atlantic oscillation; Sea ice
• Online Access: https://epic.awi.de/id/eprint/52733/; https://epic.awi.de/id/eprint/52733/1/handorf_etal_2017_ws_data_science_awi_20171108.pdf; https://hdl.handle.net/10013/epic.4d64a3d9-dd7f-4f61-a448-27f59253eab8; https://hdl.handle.net/

A large part of low-frequency variability in the climate system on sub-seasonal to decadal timescales can be described in terms of preferred atmospheric circulation patterns, often called circulation regimes. Such recurring and persistent, large-scale patterns of pressure and circulation anomalies span vast geographical area and are closely related to atmospheric teleconnection patterns like the famous North-Atlantic Oscillation (NAO). Within the conceptual framework of circulation regimes, low-frequency variability can be observed as a result of transitions between the distinct atmospheric circulation regimes. Moreover, the frequency of occurrence of preferred atmospheric circulation regimes is influenced by the external forcing factors such as other components of the climate system and anthropogenic forcing. This determines, at least partly, the time-mean response of the atmospheric flow to the external forcing. In this framework, one of our research foci is to advance the understanding of past, recent and future changes in the spatial/temporal structure of atmospheric circulation regimes and to assess the impact of internal climate dynamics versus external forcing. To tackle these questions, we exploit large global gridded data sets either from different reanalysis data sets or from model simulations with state of the art climate models mostly performed in the framework of CMIP (Coupled model intercomparison project) initiatives. We introduce and apply a hypothesis-driven approach, in particular to study the impact of sea-ice changes on atmospheric circulation patterns. The hypothesis-driven approach consists in three (iterative) steps: (i) Application of statistical methods for pattern recognition on reanalysis and climate model data, (ii) development of a hypothesis about underlying dynamical mechanisms of the impact of sea-ice changes on atmospheric circulation patterns, (iii) testing of the new hypothesis by performing new well designed climate model experiments and new model data analysis. By applying this approach, we identified tropospheric and stratospheric dynamical pathways which explain, how Arctic climate changes, in particular sea-ice changes, influence the weather and climate in mid-latitudes.