UARS MLS Limb-Tracking Observations
 During limb-tracking observations MLS usually tracks the ~18 km tangent height where most of the 63 GHz radiances are saturated, producing a good data set for GW studies. Figure 2 illustrates the sampling pattern for the limb-tracking mode, showing the MLS suborbital tracks on 28 December 1994. As detailed in the inset, the atmosphere is sampled every 2 seconds (or ~15km in distance), where the larger gap in the middle is caused by a 6-second instrument calibration. The limb-tracking mode was implemented since December 1994 and normally scheduled for every-third-day operation when MLS was on. Radiance measurements from channels 1-2 and 14-15 are not shown here because these channels are not fully saturated at high latitudes with the 18-km-limb-tracking mode and pointing variations may contaminate the radiance variances.

Shown in Figure 3 are examples of limb-tracking radiances taken on 28 December 1994, when there were large temperature gradients near the vortex edge (~40-50oN) seen in channels 3-5. The weaker latitudinal gradients in channel 6-8 radiances suggest that the vortex was weakening at altitudes above 50 km. The radiance gradient differs from orbit to orbit because the vortex was not zonally symmetric during that day. Coherent variations can be seen in the fluctuations at different channels (or altitudes) with scales from hundreds to thousands of kilometers. Figure 4 highlights short-scale fluctuations where large-scale (>1000 km) variations are removed. Some 0.5-3 K variations at 10o-30oS are evident over South America where tropospheric deep convection is activie during this period, while oscillations of 1-2K in the mid- and high-latitude Northern Hemisphere might be related to jetstream and mountain waves. Phase coherence and amplitude growth among the varianaces at different altitude are interesting, indicating wave-like nature of the perturbation. The 0-30oN region is relatively quiet where the fluctuations are mainly instrument noise.
 
 
Figure 2. UARS/MLS sampling tracks on 28 December 1994 marked by the first measurement of each major frame (65.5 seconds). The inset details the set of individual measurements in a single major frame with the short lines indicating the orientation of the temperature weighting functions (see text). On this day MLS sampling was biased towards the Northern Hemisphere.
map of UARS MLS sampling tracks
กก
Figure 3. MLS Channel 3-8 radiance measurements from the ascending part of orbit 1 on 28 December 1994.
plot of pressure as a function of latitude for ascending radiance measurements
กก
Figure 4. Radiance fluctuations derived from Figure 3 with large-scale (>1000 km) variations removed. Radiance fluctuations of each channel are displaced by 5 K with the channel number indicated at the left of each measurement series. There are more samples near the turning latitudes at 34oS and 80oN than elsewhere. In the middle portion of the orbit, a 10o latitude band at 20oN corresponds to a horizontal distance of ~1300 km.
plot of radiance perturbations as a function of latitude for radiance fluctuations
กก