Reconstructing sea level pressure variability via a feature tracking approach

Abstract

The aim of this paper is to quantify the contribution of synoptic transients to the full spectrum of space-time variability of sea level pressure (SLP) in middle latitudes. In previous work by the authors it was shown that tracking cyclones and anticyclones in an idealized atmospheric model allows one to reconstruct a surprisingly large fraction of the model's variability, including not only synoptic components, but also its large-scale lowfrequency component. Motivated by this result, the authors performed tracking of cyclones and anticyclones and estimated cyclone and anticyclone size and geometry characteristics in the observed SLP field using the 1948-2008 NCEP-NCAR reanalysis dataset. The reconstructed synoptic field was then produced via superimposing radially symmetrized eddies moving along their actual observed trajectories. It was found that, similar to earlier results for an idealized model, the synoptic reconstruction so obtained accounts for a major fraction of the full observed SLP variability across a wide range of time scales, from synoptic to those associated with the low-frequency variability (LFV). The synoptic reconstruction technique developed in this study helps elucidate connections between the synoptic eddies and LFV defined via more traditional spatiotemporal filtering. In particular, we found that the dominant variations in the position of the zonal-mean midlatitude jet are synonymous with random ultralow-frequency redistributions of cyclone and anticyclone trajectories and, hence, is inseparable of that in the storm-track statistics.