Gleixner, Stephanie (2012) Southern Annular Mode response to volcanic eruptions in the MPI-ESM. (Master thesis), Christian-Albrechts-Universität, Kiel, Germany, 94 pp.

Preview

Text Msc_Thesis_Stephanie_Gleixner_KORR.pdf - Published Version Available under License German copyright act UrhG. Download (7MB)

Abstract

Large explosive volcanic eruptions represent a natural external forcing on climate with the strongest direct radiative impact in the stratosphere. This study examines the indirect circulation response to such eruptions in the Southern Hemisphere (SH), by analyzing the Southern Annular Mode (SAM) in ERA-40 reanalysis data and the simulation ensemble of the CMIP5 historical experiment of the Max-Planck Institute Earth System Model (MPIESM). In contrast to earlier modeling studies on the volcanic impact on SAM, this high-top coupled atmosphere-ocean model has a well resolved stratosphere and mesosphere. The CMIP5 historical experiment are hindcasts, which include observed anthropogenic and natural forcing. A surface SAM based on sea level pressure and a stratospheric SAM based on geopotential height at 50hPa are defined using EOF analysis. The SAM response to 10 volcanic eruptions with global aerosol optical depth (AOD) of more than 0.02 within the time period of 1850 to 2000 is investigated in three ensemble simulations, while only the latest 5 eruptions are covered by the ERA-40 data. For hemispheric comparison, Northern Annular Mode (NAM) after volcanic eruptions is also examined. A comparison of temperature and wind climatologies and their seasonal evolution showed good agreement between MPI-ESM and ERA-40. The first spatial EOF pattern at the surface and in the stratosphere in the SH, which the SAM indices are based on, also agree well between the model and the reanalysis data. The MPI-ESM CMIP5 historical runs successfully captured the observed positive SAM trend reported in previous publications. As a case study temperature changes after the eruption of Mount Pinatubo are analyzed. Model and reanalysis both show a warming of the lower stratosphere in the subtropical latitudes due to the increased aerosol loading. In contrast to the MPI-ESM, in the reanalysis the warming of the equatorial stratosphere is delayed by approximately one year due to the easterly phase of the Quasi-biennial Oscillation (QBO). The surface cooling is stronger in the MPI-ESM than in ERA-40. In general, it is difficult to identify volcanic signals in the annular modes due to their strong unforced variability. In the reanalysis data a shift to a negative stratospheric SAM can be found after all 5 eruptions except for Agung, but only after the eruption of Fuego and Mount Pinatubo these changes are significant. In the first winter after all eruptions the surface SAM is negative, leading to a significantly negative SAM in the composite mean. These "observed" SAM changes are in sharp contrast to the simulated SAM changes. A significant positive stratospheric SAM is found in the SH spring seasons after the eruptions of El Chichon in 1982 and Mount Pinatubo in 1991. These two eruptions showed the largest AOD forcing in the 20th century and the SAM responses to these eruptions in the MPI-ESM integrations support findings of previous modeling studies. Composites of post-volcanic SAM changes with regard to the strength of the eruptions' AOD signals highlight the fact, that only the composite mean of the strongest eruptions considered in this study evoke a significant SAM response. There is no significant surface SAM response after the eruptions in the MPI-ESM. In the Northern Hemisphere (NH), the positive NAM responses to large volcanic eruptions, that have been found in reanalysis data in previous publications, are strongly underestimated in the ERA-40 results in this study. The NAM is positive in the first winter after 4 of the 5 observed eruptions, but the signals are not significant. This result highlights the sensitivity of volcanic response studies to the anomaly calculation. In contrast to the SH, in the NH no significant volcanic impact on the annular mode can be detected in the MPI-ESM CMIP5 historical runs. Overall, the findings about the simulated SAM response to large volcanic eruptions of this study are very similar to previous modeling studies, suggesting that the inability of models to simulate observed annular mode changes after volcanic eruptions is not only due to an insufficiently resolved model stratosphere. The eruptions considered in this study are likely to be too weak to cause extra-tropical circulation changes that exceed unforced variability. The differences between ERA-40 and MPI-ESM results are assumed to be due to non-volcanic forcing terms, since the MPI-ESM CMIP5 historical experiment runs do not include the observed El Ni~no/Southern Oscillation (ENSO) phases, QBO or inter-annual ozone variability. These effects need to be investigated in future studies.

Actions (login required)

View Item