FAAR abstract instructions and template
Please read carefully - the text contains instructions for abstract preparation
10 YEARS OF CLOUD DROPLET ACTIVATION DATA FROM PALLAS ATMOSPHEREECOSYSTEM SUPERSITE IN SUB-ARCTIC FINLAND
N. KIVEKÄS1, E. ASMI1, D. BRUS1, K. DOULGERIS1, E. O’CONNOR1, M. KOMPPULA2 and H.
LIHAVAINEN1
1
Finnish Meteorological institute, PO Box 503, FI-00101, Helsinki, Finland
2
Finnish Meteorological institute, PO Box 1627, FI-70211, Kuopio, Finland
Keywords: aerosol-cloud interactions, DMPS, long time series.
INTRODUCTION
The interactions between aerosol particles and clouds cause the largest uncertainty in our estimate of
present day radiative forcing compared to pre-industrial time (IPCC, 2013). The most important question
is cloud formation, specifically which particles activate into cloud droplets and which do not (McFiggans
et al., 2006). This has been studied a lot in simulation chambers and in ambient air during measurement
campaigns (eg. Sun et al., 2006). Continuous measurements with cloud condensing nuclei (ccn) counters
have allowed us to investigate the particles’ ccn-potential, i.e. which ambient particles would activate in a
given supersaturation of water in the air (eg. Burkart et al., 2011). This approach, however, does not tell
which particles really do activate in the formation process in a real cloud, as the supersaturation can vary
significantly as function of time and location even within a single cloud.
An indirect method to investigate the number and size distribution of activated particles is to measure the
particles inside a cloud simultaneously via two inlets, one allowing the cloud droplets to enter the
measurements and one blocking them. Once water is evaporated from both particle populations, they are
comparable and the difference between them refers to those particles that have been activated into cloud
droplets (Komppula et al., 2005).
Conducting this type of measurements at a mountain or hill top site (being frequently within cloud) allows
long time series of ambient cloud droplet activation data. This enables investigation of seasonal and interannual variation and trends in climate-relevant time scales. We have measured particle number size
distribution (PNSD) with this method at Sammaltunturi measurements site (24.12°, 67.97° N, 565 m asl),
located at a hill top at Pallas Atmosphere-Ecosystem Supersite (Lohila et al., 2015) in Finnish Lapland for
more than a decade. Our visibility data started at 1995, PNSD data without cloud droplets in 2000 and
with cloud droplets in 2005. All these measurement series are ongoing and continuous except for short
gaps.
METHODS
There are two parallel Differential Mobility Particle Sizers (DMPS) at the site, measuring the number
concentration and dry size distribution of atmospheric aerosol particles. One DMPS is connected to gas
line inlet, which prevents particles larger than about 5 µm from entering the sample line. The other DMPS
is connected to a total air inlet with a much larger, yet undefined, cut-off diameter. After each inlet the
particles are dried to evaporate any water in them. Subtracting the gas-line PNSD from the total air-line
PNSD gives the PNSD of those particles that have activated into cloud droplets. Aerosol particles with
very small diameter are not expected to activate into cloud droplets but can be scavenged by the larger
droplets or evaporated during the evaporation of water in the particles (Komppula et al., 2005). Therefore
only particles with diameter (Dp) larger than 50 nm are included in this analysis. We performed the data
analysis for hourly-averaged data. Periods when visibility or particle number-size-distribution were
changing significantly during the hour were excluded from the analysis.
FAAR abstract instructions and template
Please read carefully - the text contains instructions for abstract preparation
The measurement site is located on a hill top, about 300 m above the surrounding lowlands. The presence
of a cloud at the measurement site was based on visibility, and defined as horizontal visibility < 900 m.
This is roughly in line with the synoptic definition of fog (AMS, 2013), and is also supported by particle
activation measurements. Activated fraction of particles (Act%) was close to zero during higher visibility,
but increased sharply when visibility was below 900 m (Figure 1).
Figure 1. Mean activated fraction of three particle diameter ranges as function of visibility.
The presence of cloud at the top of Sammaltunturi was also estimated from Ceilometer measurements of
cloud height at Kenttärova, a low land site at 6 km distance from the Sammaltunturi site. Sammaltunturi
site was assumed to be inside cloud whenever the cloud height at Kenttärova was less than the altitude
difference between the sites.
RESULTS
The site was found to be inside cloud (in-cloud time, Tcloud) for 23 % of time during the entire time span of
the measurements. Annual mean Tcloud varied from year to year, ranging from 15 % in 2003 to 29 % in
2012 (Figure 2). Both methods of estimating Tcloud give similar numbers and similar interannual
variability, with high and low values of Tcloud occurring at same years. When only the time periods with
cloud but no precipitation were included, this variability pattern remained.
Figure 2. Annual fraction of time for which the measurement site was inside cloud, and annual mean
particle number concentration.
FAAR abstract instructions and template
Please read carefully - the text contains instructions for abstract preparation
There was also a clear seasonal pattern in Tcloud, the measurement site being inside cloud most often (Tcloud
= 46%) in November and least often (Tcloud = 9 %) in June (Figure 3). This pattern was similar for all the
measured years, and remained when precipitating clouds were removed. A more detailed analysis of the
Ceilometer data also revealed that the main difference between the summer and autumn clouds was the
cloud height, not the temporal coverage of clouds. In winter months there were also cases with very thin
low clouds, leaving the Sammaltunturi site above the cloud.
Figure 3. Mean monthly patterns of time for which the measurement site was inside cloud and particle
number concentration.
Aerosol particle number concentration, Np, (for particles with diameter between 50 nm and 500 nm)
showed a similar inter-annual pattern at the site when compared to Tcloud (Figure 2). No significant
correlation between these parameters at annual level were found. The seasonal patterns of Tcloud and Np
were, however, very different (Figure 3). Where Tcloud peaks strongly in autumn, Np peaks in summer. This
also means that even though the low clouds at the site are most frequent in fall, the more rare low summer
clouds have highest number of potential ccn. The high particle number concentration in summer clouds
leads to increase of D50 activation diameter (diameter at which 50% of particles with that diameter
activate), which is shown in Figure 4. Also the optically thickest clouds (in-cloud periods with lowest
visibility) were observed in summer.
Figure 4. Mean monthly pattern of D50 activation diameter during times when the measurement site was
inside cloud.
FAAR abstract instructions and template
Please read carefully - the text contains instructions for abstract preparation
CONCLUSIONS
This research demonstrates that visibility can be used as a proxy for the measurement site being inside a
cloud, and that 1000 m visibility is a good separation criterion between in-cloud and outside-cloud
periods. We observed signs of similar inter-annual patterns in cloudiness and particle number
concentration. The seasonal patterns, however, were very different from each other. Even though highest
in-cloud time fractions were observed in the fall months, the highest number concentrations of activated
particles and highest D50 activation diameters were observed in summer.
ACKNOWLEDGEMENTS
This work was supported by the EC FP7 project BACCHUS (grant 603445) and H2020 project ACTRIS-2
(grant 654109), by Academy of Finland (project 272041) and by Kone Foundation. We also want to
acknowledge all those who have made the data collection possible.
REFERENCES
AMS Glossary. Retrieved 16 March 2013.
Burkart, J., Steiner, G., Reischl, G. and Hitzenberger, R., 2011, Atmos. Environ 45(32), 5751-5759.
Hatakka, J. et al. (2003), , Boreal Environ. Res., 8, 365– 384.
IPCC, 2013: Summary for Policymakers. In: Climate Change 2013: The Physical Science Basis.
Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels,
Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA, 2013.
Komppula M., Lihavainen H. and Kerminen V.-M. 2005., J. Geophys. Res. 110(D0): 6204,
doi:10.1029/2004JD005200.
Lohila A., Penttilä T., Jortikka S., Aalto T., Anttila P., Asmi E., Aurela M., Hatakka J., Hellén H.,
Henttonen H., Hänninen P., Kilkki J., Kyllönen K., Laurila T., Lepistö A., Lihavainen H., Makkonen
U., Paatero J., Rask M., Sutinen R., Tuovinen J.-P., Vuorenmaa J. & Viisanen Y. 2015., Boreal Env.
Res. 20: 431–454.
McFiggans, G., Artaxo, P., Baltensperger, U., Coe, H., Facchini, M. C., Feingold, G., Fuzzi, S., Gysel, M.,
Laaksonen, A., Lohmann, U., Mentel, T. F., Murphy, D. M., O'Dowd, C. D., Snider, J. R., and
Weingartner, E. 2006, Atmos. Chem. Phys., 6, 2593-2649, doi:10.5194/acp-6-2593-2006.
Sun J, Ariya PA. 2006, Atmos Environ 40,795–820.