Aspects of the zonally asymmetric Southern Hemispheric Circulation and Representation in the MPI Earth System Model

Abstract

Due to the characteristic land-sea distribution of the southern hemisphere (SH) its atmospheric circulation is generally more zonally symmetric than its northern hemispheric equivalent. Distinct deviances from its symmetric nature, however, play an important role with regard to weather and climate on the regional scale, e.g. on the Antarctic sea ice distribution over the Amundsen Sea (Lefebvre et al., 2004). Thus the SH represents an ideal research area for the investigation of different aspects of the zonally asymmetric components of the circulation. As the Antarctic Oscillation (AAO) is the leading mode of tropospheric variability an adequate representation in earth system models is essential. Therefore the first goal of this thesis is evaluating to what extent the AAO and its related precipitation patterns are represented in the MPI Earth System Model (MPI-ESM). This is done by comparing the spatial patterns of the AAO (given as EOFs) with their associated principal components (PCs) and precipitation patterns of the MPI-ESM to three different reanalyses. A comparison between the leading EOFs of the MPIESM and ERA-Interim indicates shifted and less pronounced centers of action in the MPI-ESM. Similarly the spectral density estimates of the associated PCs in the ESM show reduced variability for periods between 4 to 5 months. The relation between AAO and SH precipitation is assessed via composite and correlation analysis. Altogether the MPI-ESM underestimates the relation of AAO and SH precipitation but exhibits the same sign and spatial distribution of correlation values. These findings suggest a lack of El Nino-Southern Oscillation (ENSO) variability in the MPI-ESM and related teleconnections towards Antarctica, which accounts for the reduced variability over the Amundsen Sea. One of the most prominent asymmetric features of the SH circulation is the split jet over Australia and New Zealand in Austral winter. Previous studies have developed indices to describe to what degree the upper-level mid-latitude westerlies are split. Furthermore, these studies have investigated the relationship of the split jet to the AAO and ENSO. The results of these studies, however, are fairly inconsistent in their message so that the relationship between the SH wintertime split jet and climate variability indices remains unclear. The scope of this thesis is a more thorough investigation of this link. So far, all established split jet indices are based on a definition that is focused on the specific region in which the jet split is recognizable. In this thesis, the split jet is considered to be of a hemispheric nature rather than of a regional. Therefore a new, hemispherically defined index based on the principal components (PCs) of the zonal wind field for the Austral winter is proposed. A linear combination of PC2 and PC3 (PSI = PC2 - PC3) of the anomalous monthly (JAS) zonal wind is used to identify the split jet condition. The newly defined index indicates a strong coherence with the AAO. A regression analysis with the Multivariate ENSO Index indicates a non-linear relationship between PSI and ENSO, i.e. split jets occur during strong positive and negative phases of ENSO but hardly under "normal" conditions. The 2nd and 3rd PCA-mode of the geopotential height variability in 500 hPa are defined as the Pacific South American patterns. They exhibit only a weak correlation with the PSI, but they show a significant correlation with the individual components (PCs) of the PSI, thereby uncovering an indirect influence on the SH split jet variability. This leads to the conclusion that a positive AAO phase, as well as both flavors of ENSO and the PSA-1 pattern, produce favorable conditions for a SH split event. The MPI-ESM's ability to reproduce the zonally asymmetric components of the SH circulation was further evaluated by an investigation of the split variability in the model. Modes larger than the first order in the zonal wind anomalies over the SH in Austral winter are found to be not distinguishable according to North's "rule of thumb". Nevertheless, PC1 and PC3 of the 200 hPa zonal wind field are found to contain the "model-intern" split variability. A composite analysis reveals, that the split variability is simulated insuffciently in the MPI-ESM. It is deduced that the split variability of the higher order PCs is mixed by the model. A likely cause for this difference compared to the reanalysis could be a low-pressure system over the Amundsen Sea, i.e. the Amundsen Sea Low (ASL). The model lacks ASL variability due to insuffciently simulated ENSO and its teleconnections which are known to reach and impact the ASL region. This can be confined by considering the climatologies and standard deviations of two associated fields (U200 and Z500). Furthermore, it was analysed to what extent the model is able to reproduce the zonally asymmetric component by comparing the deviations from the zonal mean in both fields (U200 and Z500). The zonal asymmetric part of the total variability is underestimated in the model especially over the ASL region and south of Australia, i.e. where the split jet is located. A composite analysis of the zonal asymmetric AAO component (i.e. the transient eddy portion) reveals that the model lacks substantial parts of the asymmetric AAO component. The composite's spatial distribution of significant values resembles approximately the PSA patterns in the reanalysis which are traditionally seen as the primary mechanism for the poleward transport of tropical signals. As this spatial structure is largely absent in the model, it is deduced that the deficiencies of the model in simulating the zonal asymmetric part of the SH circulation originate from the inadequate representation of ENSO variability including the high-latitude teleconnections to the ASL region. Consequently, improving the ability of the model to simulate the discussed variability modes would improve the representation of the SH split variability (and likely other asymmetric aspects of the SH atmospheric circulation) in the MPI-ESM.