Analysis of In situ Observations of Cloud Microphysics from M-PACE Final Report, DOE Grant Agreement No. DE-FG02-06ER64168 Page: 3 of 14

                                
from level in-cloud legs have been examined to characterized distributions on smaller
scales; other techniques have been used to derive larger scale cloud properties.
Two consecutive missions were flown on October 9 and 10 UTC (October 9 local
time) that sampled the same single cloud event. The cloud was forced by cold air flowing
off Arctic ice pack over open ocean water and cloud rolls developed in rolls parallel to
the flow. The overall depth of the cloud increased with distance from the ice pack. To
analyze these data the cloud samples were partitioned according to the vertical location in
cloud, i.e., lower third, middle third or upper third of the cloud layer. This yielded a total
of eight values in the lower two thirds of the cloud on the first flight, with six points in
the top third due to incomplete sampling, and five average sample points at each level on
the second flight. On average, liquid water content and mean droplet diameter increased
from 0.10 g m-3 and 12 microns in the lower third to 0.32 g m-3 and 20 microns in the
upper third of the layer, with droplet size about 2.5 microns larger on the second flight.
Droplet concentrations were constant throughout the layer on each flight but were
significantly lower on the second (55 cm-3) relative the first (75 cm-3).
Large-scale Variability
This analysis yields several interesting results. One is that the variability in LWC
and mean diameter with distance in both flights suggests microphysical variations on a
scale of 150-175 km and the larger number of samples on the first mission also reveals
variability on a scale about 1/3-1/2 of this value. A second feature is that the
microphysical properties of the cloud changed noticeably over the 4 hours that
encompassed both sample periods. Finally, the constant value of droplet concentration in
the vertical indicates that this parameter could be used as a measure of horizontal
variability in time series of ramp leg data.
Another approach in determining the horizontal distribution of liquid water was to
examine the variability of liquid water path (LWP) through the layer. The LWP was
computed for each ramp climb and descent that extended through the full cloud layer.
This was done for the October 9 and 10 missions and also for the single layer flights on
September 30 and October 1, although the profiles for the two additional flights occurred
relatively close to Oliktok Point. The LWP values ranged from < 50 g m-2 to over 200 g
m2, with the highest values occurring on September 30. The LWP maxima on October 9
were spaced approximately 200 km apart; the distribution was more uniform on the 10th
(but there were fewer points). Late in the flight on the 10th seven consecutive spirals
through the layer were conducted over Barrow. Over a 45-minute period, the LWP
increased by 60%, which at a mean layer advection speed of 12 m s corresponds to a
cloud layer spatial distance of about 30 km.
Small- to Mid-scale Variability
Analysis of the ice and liquid water data plus examination of high resolution
images of cloud particles indicated that regions of all ice and all liquid were separated at
times by only tens of meters. Times series of data from level legs were also analyzed to