Comparing observations and model prediction of drop growth in near-adiabatic cumulus cores during RICO

Model predictions and observations of drop growth are compared for two RICO cumuli. The observations were obtained using the NCAR/NSF C130 aircraft and the model calculations were done using a kinematically simple, but microphysically very sophisticated condensation and Monte-Carlo gravitational coalescence cloud model.

Discrepancies between model predictions and observations of warm cloud microphysics have traditionally been found in many studies. In the present study we constrain the observations in the following 5 ways:

(1) Sub-cloud aerosol size distributions must cover the entire range from ultra-giant particles to aerosol particles so small that they do not activate in the cloud updraft. In the present case we use complete aerosol size distributions derived from the NCAR giant aerosol impactor, FSSP and PCASP instruments and the NCAR RDMA instrument. We assume a chemical composition, dependent on size, to arrive at the soluble mass size spectra.

(2) Thermodynamic measurements of the air below an observed cloud must be known; in the present case we sampled both aerosol size spectra and thermodynamic properties of the sub-cloud air.

(3) Entrainment must be characterized in terms of source of entrained air and amount of entrained air. Using aircraft observations of thermodynamic (energy and total water mixing ratio) and chemical tracers (ozone and DMS), we compare the properties between cloudy air from two penetrations to that of sub-cloud air. The result shows that cloudy air in each of the two penetrations contained significant cores of near-adiabatically transported sub-cloud air.

(4) The mean updraft at cloud base is inferred based on the aerosol spectrum and the concentration of cloud droplets observed in the near-adiabatic cores at low altitude. Sensitivity studies are performed to examine robustness of the inferred cloud base updraft.

(5) The 'age' of the cloudy air, that is the time since the air moved up through cloud base, is estimated here based on the measured updraft speed during cloud penetration and the altitude above cloud base.

Using the above constraints, we initialize the condensation and Monte-Carlo stochastic coalescence model, and calculate the droplet growth at the altitudes of the two observed near-adiabatic cloud penetrations. The results of the model runs, that is the predicted cloud droplet and rain drop spectra, will be compared against the C-130 observations using FSSP and optical array probes.