The effective density of small ice particles obtained from in situ aircraft observations of mid-latitude cirrus

The effective ice-particle density, parametrized through a mass-dimension relation, is widely used in ice microphysical schemes for weather and climate models. In this study, we use aircraft-based observations in mid-latitude cirrus taken during the Constrain field programme in 2010. The low temperatures and a humidity often close to ice saturation meant that the typical ice particles observed were small (maximum dimension 20-800 mu m) and ice water contents were low (0.001-0.05 g m(-3)). Two new instruments are included in this study: the Small Ice Detector Mark-2 (SID-2) and the deep-cone Nevzorov Total Water Content probe. SID-2 is a new single-particle light-scattering instrument and was used to identify and size small ice particles (10-150 mu m). The deep-cone Nevzorov probe is shown to be able to collect small ice masses with sufficient sensitivity. The focus of this article is on the effective density of small ice particles (both pristine ice crystals and small aggregates up to 600 mu m maximum dimension). Due to instrument limitations in previous studies, the effective density of small ice particles is questionable. Aircraft flights in six cirrus cases provided ice-particle measurements throughout the depth of the cirrus. The particle size distribution (PSD) was mostly bimodal. The smaller ice-crystal mode dominated the PSD near cloud top and the larger ice-aggregate mode dominated near cloud base. A mass-dimension relation valid for both ice crystals and aggregates was found that provided a best fit to the observations. For small ice particles (less than 70 mu m diameter) the density is constant (700 kg m(-3)), while for larger ice crystals or aggregates the mass-dimension relation is m(D) = 0.0257D(2.0). These measurements allow testing of the diagnostic split between ice crystals and aggregates used in the Met Office Unified Model.