Our statistical method identifies all locations where the meridional gradient of isentropic potential vorticity (PV) is negative (reversed) as possible mixing zones, without distinguishing these events by dynamical cause. This method has the advantage of being applicable to all levels in the atmosphere, unlike other studies which focus on one particular potential temperature or PV layer, and identifies many different types of reversals. These include reversals near the poles, where β (and therefore the meridional PV gradient) is weak, midlatitude reversals caused by wave activity, and reversals in the deep tropics of the upper troposphere, where the meridional PV gradient is weak. Our methodology combines a statistical approach with empirical examination of individual events and Fourier analysis of the PV field to determine the dynamical cause of the PV reversals. We characterize the strength of a reversal by the meridional PV gradient at that location. The statistics available are the strength of reversals, PV at reversals, and frequency of reversals. Statistics are compiled for each month from January, 1979 to December, 2004 from daily NCEP reanalysis data that has been regridded to a 2.5 degree latitude and 5 degree longitude grid and interpolated to potential temperature levels between 320 and 850 K. Statistics are then compiled by season and phase of the QBO and compared to the PV field.
We find that the PV reversal frequency tends to be larger at locations where the climatological meridional PV gradient is weaker (more negative) and that reversals tend to be stronger at locations where the daily variability is larger. Thus, in the deep tropics (with the exception of the upper troposphere), there is a consistent minimum is PV reversals at the equator, and maxima at the poles. However, this only describes the PV reversal frequency to first order, and there are important exceptions that will be shown. Since Rossby waves cannot penetrate easterly wind regimes, the summer stratospheric easterlies coincide with PV reversal minima and the stratospheric westerlies coincide with PV reversal maxima; both of these structures are vertically coherent. These structures, as well as the dynamical origins of these reversals, will be shown.
The QBO modulates the tropical PV field and the PV reversal frequency. However, there is a strong seasonal dependence to this modulation, which does not always follow the gradient-frequency correlation: since there are more Rossby waves propagating upward from the northern hemisphere than the southern hemisphere during the winter, the QBO affects the PV reversal frequency differently. During southern winter, the modulation of the PV reversal frequency follows the modulation of the meridional PV gradient, with less frequent PV reversals occurring within the westerly QBO jet (where the gradient is stronger) and more frequent reversals to the north and south. During the northern winter, there is stronger modulation of both the meridional PV gradient and the PV reversal frequency. However, while the PV field modulation is symmetric and similar to that seen in the southern winter, the modulation of the PV reversal frequency is asymmetric, with less frequent reversals occurring at the equator but more frequent reversals occurring to the south (summer hemisphere) only. This finding agrees with other studies (e.g., Ortland, 1997; O'Sullivan, 1997; Chen and Huang, 1999) which found that westerly winds at the equator tend to "guide" Rossby waves away from the winter hemisphere and across the equator, where they break as they approach the easterly wind regime.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner