Kelvin Waves and the QBO Rolando Garcia 2 Colorado Research Associates, Boulder, CO 1
Jan 03, 2016
Kelvin Waves and the QBO
Rolando Garcia
2 Colorado Research Associates, Boulder, CO
1
HIRDLS JAN 08 2
Synoptic mapping of long asynoptic data sequences• Example using TIMED/SABER temperature data
• Salby’s Fast-Fourier Synoptic Mapping
• Extension to arbitrarily long data sequences
Compare with models• WACCM3 with different convective parameterizations
• Compare wave forcing in model vs. observational estimates
HIRDLS data• Higher vertical resolution
• Ozone and water vapor (?)
HIRDLS JAN 08 3
SABER observations are asynoptic:
• Salby (1982) has shown that one can obtain a synoptic spectrum from the spectra of asynoptic observations referred to an “s-coordinate”:
(which is a hybrid of longitude and UT)
• Salby’s technique requires a regular observation sequence (which SABER–and HIRDLS–can provide)
• Fourier analysis of asynoptic data yields an asynoptic spectrum that corresponds closely to the synoptic one
• The technique has usually been applied to observations made in a single satellite “yaw cycle”
• But the technique can be extended to data sequences of arbitrary length
€
s =c0 λ − t
1+ c02
HIRDLS JAN 08 4
Interpolation to common s-coordinates
• At yaw maneuvers, the value of the s-coordinate shifts slightly, as shown on the right
• To analyze data sequences longer than one yaw cycle, the observations are interpolated to common ascending and descending s-coordinates
• The data are interpolated across gaps associated with yaw maneuvers
• The analysis then proceeds as usual; but it can be extended to arbitrarily long sequences of data. Here we will show results for 1-year sequences.
this shows the common s-cordinates for an
entire year’s sequence (>5000 orbits), referred to common sasc so the
entire sequence can be viewed clearly
absolute values of s
s relative to common sasc
yaw maneuver
sasc
sdesc
common sasc
common sdesc
common sasc
common sdesc
sasc
sdesc
HIRDLS JAN 08 5
Equatorial waves: Equatorial Spectra(z,) for m = 1
1. high-frequency Kevin waves2. low-frequency Kelvin waves3. diurnal tide4. non-sun-synchronous tide
Note sign convention: • positive frequencies <=> westward• negative frequencies <=> eastwardfrequencies are given in cycles/day (cpd)
2
stratosphere and mesosphere stratosphere
1 4 3
2
HIRDLS JAN 08 6
Examples of Kelvin wave structures1. m=1 Kelvin wave structure, ~12 d 2. m=1 Kelvin wave structure, ~3 d
0.5 K
2.5 K
These wave structures are obtained via Hayashi (1971) coherence analysis. Only locations with significance level > 1 are plotted. Note different latitude scales.
These waves are presumed to play a role in forcing
the QBO
HIRDLS JAN 08 7
Low-frequency m=1 Kelvin waves ( ~ 20 days) and the QBO
0402
03 05
06
2003
2002
~ 0.05 cpd => ~ 20 d i.e., c ~ 20 ms-1 for m=1
Spectra (z,), m=1, Equator
HIRDLS JAN 08 8
QBO West QBO East
2002
20052004
2003
Amplitude(t,z) of m=1 Kelvin waves at the Equator
note the clear relationship between wave amplitude and QBO phase at z ~ 2 – 5.5 s.h.
amplitude in frequency range –0.08 to –0.025 cpd, or ~ 12–40 days
eastward
HIRDLS JAN 08 9
Work with WACCM highlights the importance of the parameterization
of convection
HIRDLS JAN 08 10
Model calculations vs. SABER observations of temperature at
10 mb in the Tropics
• Spectra in WACCM and SABER observations are remarkably similar overall• But relative amplitude of east- vs. west-propagating (arrows) waves appears smaller in SABER compared to WACCM• Vertical resolution too coarse in SABER? Can HIRDLS data help here?
WACCM + Tietdke
HIRDLS JAN 08 11
Possible Applications of HIRDLS data
• Confirm SABER results for equatorial waves
• Does higher vertical resolution in SABER enhance the detectability of westward-propagating waves (inertia-gravity, Rossby gravity)?
• Use HIRDLS and SABER data to quantify wave driving; compare with model results
• Can ozone and water vapor data be used to look for the signature of the QBO in minor constituents?