Investigations of the ice nucleation ability of natural soil dusts in artificial clouds

While the intense investigation of ice nucleation behaviour of model systems like pure minerals or specific biological matter is essential to provide a fundamental understanding of ice nucleation processes, there is also a profound interest for a description of ice formation in natural aerosol systems for use in climate models. Especially the ice nucleation abilities of natural soil dusts recently have come into focus of research due to their substantial contribution to the global dust burden through for example intense agricultural activities.

Here we present the investigation of aerosols generated by the dry dispersion of soil dusts, which have been sampled in southern and northern Germany, Mongolia and Argentina. The samples have been characterized for composition and biological activity in the bulk phase as well as for the percentage of bio-fluorescent particles in the suspended aerosol. The ice nucleation studies were conducted in the large cloud simulation chamber AIDA (Aerosol Interaction and Dynamics in the Atmosphere) at temperatures in the range from 230 K to 265 K. While in the immersion mode we find significantly enhanced ice nucleation abilities compared to those of desert dusts, we did observe just a slight enhancement in the deposition mode well below liquid water saturation.

For use in atmospheric models, we will give a parameterization of ice nucleation active surface site (INAS) densities describing our observations in the immersion mode. We will also correlate our findings with the preceding analysis of the soil dust samples. In addition, we present a comparison to comprehensive investigations of the ice nucleation behavior using the mobile ice nuclei counters FRIDGE, PINC and SPIN, which have been done during the IN19 campaign at the AIDA facility. Here, identical aerosol samples have been used, so that any differences in the measured nucleation efficiency have their origin in the specific instrumental design and detection.