Intraseasonal variability in a two-layer model and observations

• Main Authors: Christian L. Keppenne, Steven L. Marcus, Masahide Kimoto, Michael Ghil
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: 1999
• Subjects: Pacific; North Atlantic; North Atlantic oscillation
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.117.6931; http://trs-new.jpl.nasa.gov/dspace/bitstream/2014/17856/1/99-1306.pdf

A two-layer shallow-water model with R15 truncation and topographic forcing is used to study intraseasonal variability in the Northern Hemisphere's (NH) extratropical atmosphere. The model's variability is dominated by oscillations with average periods near 65-70 and 40-50 days. These periods are also found in 13.5 years of daily upper-air data from January 1980 to July 1993. The spatial variability associated with these oscillations is examined by compositing the streamfunction-anomaly fields of the model and the observations. The model's 70-day oscillation is strongest in the Euro-Atlantic sector, where it bears a close resemblance to observed streamfunction composites of the North Atlantic Oscillation. The observed 70-day mode exhibits similar features in the Euro-Atlantic sector, accompanied by a north-south "seesaw " over the Pacific and Eurasia. Previous authors, in their analysis of geopotential height observations, also found these features to be present in an empirical orthogonal function that contains aspects of both the North Pacific and North Atlantic Oscillations. The 40-day oscillation is characterized, in both the model simulations and observed data, by a zonal wavenumber-two pattern anchored over the NH topography. This pattern undergoes tilted-trough vacillation in both the model and observations. This midlatitude vacillation is strongest in the Pacific-North American sector, where it resembles a 40-day oscillation in the UCLA general circulation model that is largely driven by mountain torques over the Rockies. Comparisons with observational data show a clear separation between a tropical 50-day oscillation, not present in our model results, and a 40-day NH extratropical oscillation which resembles the topographically induced oscillation that arises in our two-layer model.