Mesospheric Zonal Mean Winds Derived from Consecutive Orbits of AIM Cips Images

Abstract:

In order to infer mesospheric wind velocities, polar mesospheric cloud (PMC) pattern variations are investigated using images from consecutive orbits taken by the Cloud Imaging and Particle Size instrument (CIPS) aboard the AIM satellite. CIPS measurements are analyzed to detect patterns that repeat from one orbit to the next, but are displaced in location; the displacement provides a measure of the wind velocity. Pattern matching is achieved by re-sampling the CIPS data to a standard geographic grid with a horizontal resolution of 0.2° longitude × 0.05° latitude (~25 km²), and correlating patterns within geographic frames of size 24° longitude × 3.6° latitude. Such a frame size is arbitrarily chosen, but it covers a hierarchy of cloud structures including scales as large as several hundred kilometers. A relatively larger frame is required because after ~90 minutes, the time of one orbit, the smaller scale features are no longer conserved. Several thousand pairs, taken from 10-14 July 2007, are matched to derive the statistics. These pairs are mostly evenly distributed at longitudes and latitudes north of 70⁰N for each given day. The results suggest that the zonal velocity probability distribution during this 5-day period was peaked at around -40m/s with a 1-σ scatter of ~35m/s. The meridional velocity distribution peaked at 0 m/s with a 1-σ scatter of ~25m/s. These prevailing velocities can be determined with high precision because the corresponding patterns are shifted by at least half of the frame size from one orbit to the next.

The CIPS cloud albedo on consecutive orbits is also examined for variations at fixed locations. The statistical results suggest that the mean cloud albedo within a given frame will most likely be weakened or strengthened by < 30% on consecutive orbits, although larger variations can occur with lower probability. Such a conclusion applies to both bright and dim clouds. This indicates that within 90 minutes the cloud brightness for a given location changes moderately rather than drastically in the next orbit. However, it is also found that the correlation of the two frames at a fixed location is generally poor. These findings suggest that cloud patterns are subject to wind advection, but the bright patch or (dim) cloud voids are more extended in size than the movement that occurs within 90 minutes.