A model for ice nucleation in the AIDA cloud simulation chamber. Part 1: Observations and model description, using key measurements to constrain the model

The AIDA (Aerosol Interactions and Dynamics in the Atmosphere) aerosol and cloud chamber of Forschungszentrum Karlsruhe can be used to test the ice forming ability of aerosols. The AIDA chamber is instrumented with temperature, humidity and particle counters. Expansion cooling using mechanical pumps leads to ice supersaturation conditions and possible ice formation. The aerosol concentration and size distribution, and the ambient temperature, pressure, humidity and rate of cooling can be controlled and made similar to upper troposphere conditions. Particle concentrations and mean particle size are measured using a laser scattering device, the small ice detector (SID). The SID can count and size particles above one micron diameter and discriminate between liquid water drops and ice particles.

In order to describe the evolving chamber conditions during an expansion, a detailed microphysics size-resolving parcel model was modified to account for diabatic heat and moisture interactions with the chamber walls. The wall vapour flux is derived using the measured total water content which is now accurately measured when there are no large particles present.

Model results are shown for a series of expansions using desert dust as ice forming nuclei, over an initial chamber temperature which ranged from -20 to -60C. During each expansion, the initial formation of ice particles was clearly observed. For the colder expansions there were two clear ice nucleation regimes.

Firstly, ice particles were added to the model as a function of time so as to reproduce the observations of SID ice crystal concentration. The agreement between the model and the observations of chamber relative humidity and temperature, and of the ice particle concentration and average diameter improves the confidence in these measurements, and in the estimated wall flux of heat and water vapour. The time interval and chamber conditions over which ice nucleation occurs is therefore accurately known. By establishing the ability of the model to describe the evolving chamber conditions, it can then be used as a test bed for different representations of heterogeneous ice nucleation.