Evaluation of an entraining droplet activation parameterization using in situ cloud data

This study investigates the ability of a droplet activation parameterization (which considers the effects of entrainment and mixing) to reproduce observed cloud droplet number concentration (CDNC) in ambient clouds. Predictions of the parameterization are compared against cloud averages of CDNC from ambient cumulus and stratocumulus clouds sampled during CRYSTAL-FACE (Key West, Florida, July 2002) and CSTRIPE (Monterey, California, July 2003), respectively. The entrainment parameters required by the parameterization are derived from the observed liquid water content profiles. For the cumulus clouds considered in the study, CDNC is overpredicted by 45% with the adiabatic parameterization. When entrainment is accounted for, the predicted CDNC agrees within 3.5%. Cloud-averaged CDNC for stratocumulus clouds is well captured when entrainment is not considered. In all cases considered, the entraining parameterization compared favorably against a statistical correlation developed from observations to treat entrainment effects on droplet number. These results suggest that including entrainment effects in the calculation of CDNC, as presented here, could address important overprediction biases associated with using adiabatic CDNC to represent cloud-scale average values.