From Aerosol to Cloud Droplet Number Concentration in VOCALS-REx

The dependence of cloud droplet number concentration (CDNC) on aerosol is central to investigations of the local cloud albedo, precipitation development and cloud morphology. However, prediction of the CDNC is complicated by its sensitivity to the cloud condensation nuclei (CCN) activation spectrum and to updraft velocities within the droplet activation region (cloud base to altitudes ~50 m higher). Here, we use data collected during VOCALS-REx to examine the predictability of CDNC based on airborne CCN and updraft measurements. The VOCALS-REx dataset is well-suited to this analysis because of the suite of instrumentation onboard the NSF/NCAR C-130 aircraft and the range of conditions - averaged CCN concentrations at supersaturation = 0.4% range from 500 per cubic centimeter, near the Chilean coast, to 50 per cubic centimeter 500 km to the west within the broken stratocumulus. The latter may be a situation where the CDNC is aerosol-limited, i.e., nearly all positive vertical velocities convert the totality of the aerosol population to cloud droplets. There are two implications: 1) within the broken stratocumulus region a measurement of the concentration of particles larger than ~0.01 micrometer could be a reliable predictor of the CDNC (assuming there are no ultrafine nuclei), and 2) the CDNC modeling in these regions is independent of updraft, and thus simplified. This work compares three observational estimates of the CDNC during VOCALS-REx. The first comes from in-cloud measurements made with the C-130 cloud droplet probe (CDP); the second is from sub-cloud values of the CCN (Wyoming CCN instrument, PCASP and aerosol chemistry), and measurements of the updraft probability distribution function (gust probe), input to a parcel model; and the third is from Wyoming Cloud Lidar via a CDNC retrieval algorithm. We refer to the CDNC values derived from these three approaches as the “measured”, the “modeled” and the “retrieved.” The three approaches are complimentary. The measured values are affected by two processes (entrainment/evaporation and precipitation scavenging) not accounted for in the modeled values; the measured/modeled comparison must acknowledge displacements in space (~100 km) and time (~15 min) between the in-cloud and sub-cloud flight legs; the retrieved values are available coincident with the modeled values; and the modeled values come with a vertical profile of the cloud droplet size spectrum which is useful for guiding the lidar-based CDNC retrieval. This presentation will discuss the sub-cloud and in-cloud vertical velocity probability distribution functions, the departure of cloud liquid water content and in-cloud temperature from adiabatic, the cloud droplet closure problem, and comparisons of the retrieved CDNC values to those derived via the two other approaches.