Tracking the jets as Lagrangian objects

Various algorithms developed to track the synoptic evolution of the subtropical jets have proved useful in diagnosing variability and trends. However, consensus about trends remains low, and issues in jet detection persist. Notably, algorithms:

• Frequently employ a variety of climatological parameters, making them unsuitable for long-trend analyses;
• Rely on instantaneous meteorological fields (or Eulerian-averaged fields), thereby overlooking the temporal coherence of jet features. This results in the inability to systematically separate the true axis of the jets from underlying waves, affecting the characterization of variability known to affect the mean position of the jets —and long-term trends.

To address these limitations, we define the jets as Lagrangian Coherent Structures; persistent features that resist synoptic variability and thereby shape the atmospheric circulation. Using this Lagrangian definition, a new algorithm named JetLag is developed and applied to the ERA5 reanalysis. JetLag employs 2 parameters –a time scale and a spatial scale, both set by the Rossby wave dispersion relation– and is virtually insensitive to changes in those parameters, within physical bounds. Compared to wind-based methods, we show that JetLag:

• Locates jet features with better temporal coherence;
• Has enhanced capabilities for detecting weak and highly variable jets;
• Produces a different seasonal cycle of mean jet position;
• Produces different decadal to multi-decadal variability, with implications for trend detection.

We also present a new jet axis dataset for use by the community.