13B.8 The Southern Hemisphere Main Centers of Action Associated with Teleconnection Patterns in Results of CPTEC/COLA GCM Simulation

Wednesday, 7 April 1999: 10:45 AM
Cassiano D'Almeida, CPTEC/INPE, Cachoeira Paulista, SP, Brazil; and I. F. A. Cavalcanti

Climatological data from a integration of 11 years using CPTEC/COLA GCM (T42L18) showed similar features to observed and reanalyses data. The model also showed satisfactory results in seasonal precipitation simulations over Brazil, specially during winter and autumm. In order to broaden the understanding of the global aspects of the model, we attempt to evaluate in the current work, the GCM ability of reproducing the teleconnection patterns observed in the Southern Hemisphere.

Correlation analyses and orthogonal empirical functions were performed with the purpose of obtaining teleconnectivity patterns. The analyses were applied to the model results of a long integration (11 years), initially with the Southern Hemisphere seasonal streamfunction data at 200hPa. Pentadal means were performed to eliminate the high frequency oscillations from the data series and anomalies were obtained, subtracting the climatological pentads from each individual pentad.

The preliminary correlation patterns show significant and seasonally dependent wavetrains representing tropical-extratropical interactions and teleconnections between the high level circulation over Pacific and Atlantic Oceans. Anomalies close to southeast of South America show correlations with anomalies over South Pacific and South Atlantic. Correlation patterns are similar to those found in observational studies, i.e. wavetrains from tropical South Pacific to south of South America and then, northeastward reaching the Tropical South Atlantic. EOF analyses were applied to the model streamfunction data in order to determine the dominant patterns. Filtered data are also used to focus the analyses in the intraseasonal band. The interannual variability is evaluated analysing the time series of EOF amplitudes. Periods of maximum positive and negative principal components are analysed related to outgoing long-wave radiation data to identify the linking between convection and wavetrains on the model results.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner