Representation of the Antarctic Oscillation and related precipitation patterns in the MPI Earth System Model

• Main Authors: Babian, Stella, Rust, Henning W., Grieger, Jens, Prömmel, Kerstin, Cubasch, Ulrich
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Freie Universität Berlin 2016
• Subjects: Antarctic Oscillation; MPI-ESM-MR; precipitation; southern hemisphere; 500 Naturwissenschaften und Mathematik550 Geowissenschaften, Geologie551 Geologie, Hydrologie, Meteorologie; Antarctic; The Antarctic; Antarc*; Antarctica
• Online Access: https://dx.doi.org/10.17169/refubium-23669; https://refubium.fu-berlin.de/handle/fub188/20366

The Antarctic Oscillation (AAO) is the dominant mode of atmospheric variability in the southern hemisphere. It is obtained via a principal component analysis (PCA) for geopotential height anomalies. Being the southern hemisphere's dominant mode, an adequate representation in earth system models is desirable. This paper evaluates to what extent the AAO and related precipitation is represented in the Max Planck Institute's Earth System Model (MPI-ESM). To this end we compare AAO spatial patterns (empirical orthogonal functions, EOFs), spectral properties of the associated principal components (PCs) and AAO-related precipitation patterns of MPI-ESM to three reanalyses: the ECMWF's ERA-40 and ERA-Interim, and the NCEP/NCAR 40-year reanalysis project. Differences between MPI-ESM and ERA-Interim leading EOFs reveal that the three typical centres of action are less pronounced and slightly shifted in the model. Spectral density estimates of the associated PCs show reduced variability in the MPI-ESM for periods between 4 to 5 months. The relation between AAO and southern hemispheric precipitation is assessed via composites and correlation analysis. In both, model and reanalyses, a negative AAO index leads to a general increase of precipitation between 30° S and 50° S and a decrease south of 50° S. Differences between maps of correlation for AAO and precipitation are most prominent near Indonesia and Antarctica probably due to a lack of pressure around Antarctica in the model. Altogether the MPI-ESM underestimates the relation of AAO and southern hemispheric precipitation but gives the correct sign and spatial distribution of correlation values.