Comparison
of Measured and Predicted Cloud Condensation Nuclei in the Free Troposphere
over the Eastern Pacific Ocean during DYCOMS-II
Hiroshi Takagi, MS candidate,
University of Wyoming
![](../../usr/local/Office51/gallery/rulers/blurulr6.gif)
Atmospheric aerosols that nucleate cloud droplets are known as cloud
condensation nuclei or CCN. CCN modulate cloud droplet concentration and size
and thus cloud albedo and precipitation. These connections are most evident
within marine stratocumulus cloud systems; however budgets for marine boundary
layer CCN are poorly constrained. Two confounding factors are 1) the transport
of free-tropospheric CCN into the marine boundary layer, and 2) the prediction
of CCN from aerosol size spectra. This work focuses on both issues by analyzing
airborne measurements of CCN and aerosol obtained during 30-minute flight
segments conducted 100 to 400 m above marine stratocumulus cloud during
DYCOMS-II. Flight-to-flight variability of both CCN and aerosol is dependent on
back trajectories. Larger concentrations were documented for three study days
associated with back trajectories that passed over British
Columbia, Washington,
Oregon or California
before reaching the DYCOMS-II study area (~200 km west of San
Diego, California). In
contrast, two study days associated with oceanic back trajectories which did
not pass over North America exhibited substantially
lower concentrations. In a closure study, CCN concentrations measured at an
effective supersaturation of 0.8% were compared to concentrations predicted
using aerosol size spectra. Closure was achieved for the two study days
uninfluenced by pollution originating from North America.
This result could be used to either initialize or validate aerosol models which
treat the coupling of free troposphere and marine boundary layer. However, the
measured-to-predicted CCN ratios were lower than unity for the cases associated
with back trajectories which passed over North America
and for one of these cases the measured-to-predicted CCN ratio remained smaller
than unity after accounting for possible biases in the measurements and
calculations. Plausible reasons for this disparity are explored.