HR: 17:30h
AN: A12F-07    [PDF]
TI: Sensitivity Study of the Vertical Velocity Variation on Cloud Droplet Nucleation Process Using an Adiabatic Parcel Model
AU: * Peng, Y
EM: peng@mathstat.dal.ca
AF: Dalhousie University, Dept. of Physics and Atmospheric Science, Dalhousie University, HALIFAX, NS B3H 3J5 Canada
AU: Lohmann, U
EM: Ulrike.Lohmann@dal.ca
AF: Dalhousie University, Dept. of Physics and Atmospheric Science, Dalhousie University, HALIFAX, NS B3H 3J5 Canada
AU: Leaitch, R W
EM: Richard.Leaitch@ec.gc.ca
AF: Meteorological Service Canada, Meteorological Service Canada, 4905 Dufferin Street, Downsview, ON M3H 5T4 Canada
AB: Eleven profiles through liquid water cloud obtained during RACE (Radiation, Aerosol and Cloud Experiment) and NARE (North Atlantic Regional Experiment) are used to study the sensitivity of cloud droplet nucleation to the vertical gust velocity. Selected cloud microphysical data, size-distributed aerosol properties and particle chemistry are applied in an adiabatic parcel model to predict the activated cloud droplet number concentrations (N) using the frequency distribution of the measured in-cloud vertical velocities and again using a vertical velocity characteristic of observations. The simulated adiabatic value of N obtained using the standard deviation of the vertical velocity distribution agrees with the observed maximum N (the cloud droplet number in an adiabetic core) to within 5%. If the parameterization derived by Lin et al. [1997] is applied to obtain the cloud-average N from the maximum N, the average N agrees with the observed cloud-average N to within 20%. The simulated N obtained using the full probability density function of the vertical gust velocities is one approach that has been used to represent the cloud average N. This is based on the assumption that the average N is controlled by all variations in the updraft and not by the mixing process [Leaitch et al. 1996]. The value of N obtained in this manner is found to be higher than the observed average N by a factor of two. We believe that this result is because low vertical velocities do not contribute effectively to the cloud droplet nucleation. If we neglect the lowest 45% of all vertical velocities, then the difference between the simulated average N and the observed mean N is reduced to within 13%. These results suggest that it is appropriate to use a characteristic vertical velocity to predict the cloud droplet number concentration in climate models as done by Lohmann et al. [1999], where the subgrid variation of vertical velocity is diagnosed from the turbulent kinetic energy. The frequency distribution of vertical velocity is needed to represent the in-cloud vertical fluctuations. However, this work suggests that a modified frequency distribution of vertical velocity accounting for only the larger vertical velocities to predict the cloud droplet number concentration is more appropriate than using the full frequency distribution. To obtain a better understanding of this issue, it will be important to use both modeling and observations to investigate which updrafts in clouds contribute to cloud droplet nucleation. References: Lin, H., and Leaitch, W.R., Development of an in-cloud aerosol activation parameterization for climate modeling, in Proceedings of the WMO workshop on Measurement of Cloud Properties for Forecast of Weather, Air Quality and Climate, pp. 328-335, Geneva, World Meteorol. Organ, 1997. Leaitch, W.R., Banic, C.M., Isaac, G.A., Couture, M.D., Liu, P.S.K., Gultepe, I., Li, S.-M., Kleinman, L.I., Daum, P.H. and MacPherson, J.I., Physical and chemical observations in marine stratus during the 1993 NARE: Factors controlling cloud droplet number concentrations, J. Geophys. Res., 101, 29123-29135, 1996. Lohmann, U., Feicher, J., Chuang, C.C., and Penner, J.E., Predicting the number of cloud droplets in the ECHAM GCM, J. Geophys. Res., 104, 9169-9198, 1999.
DE: 0320 Cloud physics and chemistry
SC: Atmospheric Sciences [A]
MN: 2003 Fall Meeting