CO as an important high-altitude tracer of dynamics in the polar stratosphere and mesosphere

We present new ground-based measurements of polar stratospheric and mesospheric
CO, made with a millimeter-wave spectrometer at Thule, Greenland (76.5 N, 68.7 W).
Almost daily measurements were made between 17 January and 4 March 2002 and again
between 5 January and 22 February 2003. We stress here the retrieval and analysis of CO
mixing ratios in the 50–80 km altitude range, though it can be monitored at lower altitudes
as well. Since CO exhibits a strong positive latitude gradient from the summer to the
winter pole, it is an excellent tracer for poleward transport from lower latitudes. Moreover,
the mixing ratio of CO increases rapidly from 40 km to at least 100 km at midlatitudes,
providing a good tracer for high-altitude vertical transport as well. Our profiles indicate
that in winter near the poles the CO mixing ratio decreases above 70 km because of
subsidence of air and minimal high-altitude photoproduction at high latitudes. Our data
also show large variations in mixing ratio and vertical distribution, yielding a good picture
of stratospheric and mesospheric dynamics-induced changes on a scale of hours to
days. These observations verify that CO serves as an excellent tracer of vortex-related
dynamics in the 30–80 km altitude range, where other information, particularly above
40 km, may be sparse, unreliable, or nonexistent. Our results are in general agreement
with analyses of 1991–1992 Improved Stratospheric and Mesospheric Sounder (ISAMS)
satellite data by Lopez-Valverde et al. [1993, 1996] and by Allen et al. [1999, 2000].
We show the contrast between CO over the summer pole and CO over the winter pole with
the aid of trial observations made at the South Pole during the austral summer of
1999–2000. Our Thule data indicate that large concentrations of CO generally exist in
winter just outside the vortex boundary. The large rapid variations in vertical profile that
are found in our data in 2002 appear to be well correlated with vortex position in the lower
stratosphere. In 2003 this correlation appeared to be much weaker, but early 2003 was
also a period of vortex splitting in the Arctic on three occasions during our observation
period, accompanied by generally more complex vortex dynamics.