HR: 11:50h
AN: A12C-07 [Abstracts]
TI: Why the West Wind Wobbles: Stochastic Potential Vorticity Forcing of Annular Modes
AU: * Goodman, J C
EM: jgoodman@whoi.edu
AF: Woods Hole Oceanographic Institution, Mail Stop 21, Woods Hole, MA 02346
United States
AU: Czaja, A
EM: czaja@ocean.mit.edu
AF: Imperial College, Space and Atmospheric Physics Group,
The Blackett Laboratory,
Prince Consort Road, London, MA SW7 2BW
United Kingdom
AB:
We propose a useful tool for understanding the variability of the Northern and Southern Annular Modes (NAM/SAM). These modes
can be thought of as polar pools of potential vorticity (PV), surrounded by an annular midlatitude pool of opposite sign.
Meridional PV fluxes caused by synoptic weather stochastically transfer PV to and fro between the polar and midlatitude
pools, gradually increasing or decreasing the strength of the annular mode.
In a 3-layer quasigeostrophic model constrained to reproduce the observed winter climatology, we define the ``model North
Annular Mode'' (mNAM) index as the amplitude of the first EOF of daily streamfunction data. We compute the model's daily PV
flux through 58 north, the latitude line which separates the polar and midlatitude centers of action of the mNAM. We then
use a least-squares technique to find the auto-regressive (AR1) model with meridional PV flux as input and mNAM as output
which optimally links the two timeseries, using a least-squares approach. That is, we test for the existence of a linear
relationship between PV flux and the rate of change of the mNAM index. The best-fit AR1 model output has a correlation of
0.86 with the mNAM index. This correlation is much less when the PV flux through a latitude other than the mNAM's
polar/midlatitude boundary is used as the AR1 input.
We perform a similar analysis with NCEP-NCAR reanalysis data. We use the observed North and South Annular Mode indices
(NAM/SAM), and the isentropic mass flux through a latitude line as a proxy for PV flux (ignoring the relative vorticity part
of the PV for simplicity). As with the QG model, an AR1 model with PV flux as input explains much of the NAM/SAM index
variability (correlation 0.62 and 0.55, respectively), with peak correlations for PV fluxes at the boundary between the
annular modes' polar and midlatitude centers of action.
Thus, we demonstrate that much of the variability of the annular modes in both a 3-layer quasigeostrophic model and in
NCEP-NCAR reanalyses can be explained by a simple stochastic process, in which the annular mode index represents the strength
of a polar PV pool, forced by random exchanges of PV through the pool's boundary by synoptic eddies.
UR: http://www.mit.edu/~goodmanj/pvflux_annular_mode.pdf
DE: 1620 Climate dynamics (0429, 3309)
DE: 3319 General circulation (1223)
DE: 3367 Theoretical modeling
SC: Atmospheric Sciences [A]
MN: Fall Meeting 2005