final_program_abstracts[1]

11 IMSC Session Program
Historical Southern Hemisphere Annular Mode variability from
statistical reconstructions and the IPCC AR4 simulations
Tuesday - Parallel Session 3
Julie M. Jones 1 , Ryan L. Fogt 2 and Martin Widmann 3
1
Department of Geography, University of Sheffield, UK
2
Department of Geography, Ohio University, USA
3
School of Geography, Earth and Environmental Sciences, University of Birmingham,
UK
Reconstruction of Southern Hemisphere climate presents a challenge due to the
smaller land-surface area than the Northern Hemisphere, and the lack of
measurements from the largest continent (Antarctica) before the mid twentieth
century.
Seasonal reconstructions of the Southern Hemisphere Annular Mode (SAM) index
back to 1865 using principal component regression with station sea level pressure
(SLP) data as predictors, have been derived. Two reconstructions using different
predictands were obtained, one (JW) based on the first principal component (PC) of
extratropical SLP from the ER40 reanalysis, and the other (Fogt) on the index of
Marshall (2003).
As the predictors are based on observational data, this provides an opportunity to
investigate reconstruction issues in a case where the predictors have a purely climatic
response to the variable being reconstructed, but as for many proxy-based
reconstructions, are based on a spatially inhomogeneous network.
The spatial structure of the SAM differs between seasons, and this influences the
structure of the statistical model and the reconstruction uncertainty. The spatial
structure of SLP anomalies in periods of strong positive and negative SAM index was
investigated, to determine how well the spatially inhomogeneous predictors capture
the SLP anomaly in each season. Many cases, such as the most recent peak in austral
summer, represent a full hemispheric SAM pattern. However a number of periods
project onto the SAM but are not canonical SAM-events. For example a number have
positive SLP anomalies in mid-latitude regions known to be preferential to blocking,
but not a zonally symmetric SAM signature. How well these events are captured then
depends on the location of predictor stations in relation to the anomaly regions. Hence
some events will be captured more realistically than others by a sparse predictor
network.
The reconstructed SAM indices were used to evaluate the SAM in simulations from
17 IPCC Fourth Assessment Report models from 1865-2005. The models capture the
recent (1957-2005) positive SAM trends in austral summer, which reconstructions
indicate is the strongest trend during the last 150 years; ozone depletion is the
dominant mechanism driving these trends. The models simulate too strong recent
trends in SON, indicating low ozone and greenhouse forcing on the SAM or that
natural variability overrides any trends.
Abstracts 128