Cite abstract as Author(s) (2007), Title, European Aerosol Conference 2007, Salzburg, Abstract T12A009
Estimating cloud droplet number concentration with particle number-to-volume
concentration ratio
N. Kivekäs, V.-M. Kerminen, T. Anttila, H. Korhonen, M. Komppula and H. Lihavainen
Finnish Meteorological Institute, PO box 503, Helsinki, Finland
Keywords: Aerosol cloud interaction, Modelling, Aerosol size distribution.
N (d > d c )
,
R( d c ) =
Vtot
(1)
where N(d>dc) is the number concentration of
submicron particles with dry diameter larger than a
cut-off diameter dc and Vtot is the total volume
concentration of all submicron particles (Kivekäs et
al., 2007).
With an adiabatic air parcel model (Korhonen
et al., 2005, Anttila et al., 2002) we simulated the
activation of particles to cloud droplets in rising air.
We varied the particle number size distribution
parameters, the soluble fraction of particle mass (ε)
and the updraft velocity of the air parcel (vup). Then
we fitted the numerical parameters to make the
parameterization fit the simulated values as well as
possible. The resulting formula was

R(0.1) 
 × Vtot − a 2 ,
CDNC =  a1 ×
d c 

(2)
where
b5
.
d c = (b1 × lnVtot + b2 × ln R(0.1) + b3) × ε b4 × v up
(3)
In the formulas above the numerical constants are
a1=0.10, a2=6cm-3, b1=0.014, b2=0.008, b3=0.016,
b4=-0.19 and b5=-0.32. The units are cm-3 for CDNC,
µm-3 for R(dc) and R(0.1), µm for dc, µm3/cm3 for Vtot
and m/s for vup. ε is dimensionless. R(0.1) in the
formula stands for R with dc=0.1µm.
The parameterization was tested against
measured cloud droplet activation data from Pallas
GAW station in northern Finland (Komppula et al.,
2005). There were 33 clouds where particle
activation was measured during the years 2000-2002.
The soluble fraction of the particle mass was
estimated to be 50% based on campaign
measurements at the same site during falls 2004 and
2005. There was no updraft velocity data available,
so we decided to set vup equal to 0.43m/s for all the
cloud events.
The parameterized vs measured values of
cloud droplet number concentration are shown in
figure 1. The correlation coefficient between those
two data sets was 0.76 and the largest offsets
produced by the parameterization were within 50%
of the measured value.
-3
CDNC from parameterization (cm )
In atmospheric aerosol systems, the number of
particles >70-120 nm in diameter is related closely to
the number of particles able to act as cloud
condensation nuclei (CCN). Many large-scale
atmospheric models, however, have the particle mass
or volume concentration as the only prognostic
variable.
The aim of this study is to investigate how
accurate the estimation of cloud droplet number
concentration (CDNC) would be if an empirical
relation between the number concentration of CCNsize particles and particle volume (or mass)
concentration was available. The relation used here is
500
400
300
200
100
0
0
100
200
300
400
500
-3
CDNC from measurements (cm )
Figure 1. The parameterized vs measured cloud
droplet number concentrations at Pallas during 33
cloud events.
The concept of parameterizing cloud droplet
number concentrations (CDNC) with number-tovolume concentration ratios (R(dc)) works in Pallas,
which is a station representing continental
background air. If R(0.1) is known or can be
parameterized, this kind of a parameterization would
be a computationally efficient way to estimate the
aerosol impact on cloud formation in large scale
models. This demonstrates the benefit achievable
from investigating parameterizations for R(0.1) or for
some other R(dc).
Anttila, T. and Kerminen, V.-M. (2002). J. Geophys.
Res., 107(D22), 4662, doi:10.1029/2001JD
001482.
Kivekäs N., Kerminen V.-M., Engler C., Lihavainen
H., Komppula M., Viisanen Y. and Kulmala M.
(2007), J. Geophys. Res., accepted for publication
Komppula M., Lihavainen H., Hatakka J., Aalto P.P.,
Kulmala M., and Viisanen Y. (2003), J. Geophys.
Res., 108(D9), 4295, doi:10.1029/2002JD002939.
Korhonen, H. , Kerminen, V. -M., Lehtinen, K. E. J.
and Kulmala, M. (2005), Atmospheric Chemistry
and Physics, 5, 2561- 2570.