In situ balloon-borne ice particle imaging in high-latitude cirrus

• Published in: Pure and Applied Geophysics
• Other Authors: Kuhn, Thomas (author), Heymsfield, Andrew J. (author)
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2016
• Subjects: Kiruna; Northern Sweden
• Online Access: https://doi.org/10.1007/s00024-016-1324-x

Cirrus clouds reflect incoming solar radiation, creating a cooling effect. At the same time, these clouds absorb the infrared radiation from the Earth, creating a greenhouse effect. The net effect, crucial for radiative transfer, depends on the cirrus microphysical properties, such as particle size distributions and particle shapes. Knowledge of these cloud properties is also needed for calibrating and validating passive and active remote sensors. Ice particles of sizes below 100 A mu m are inherently difficult to measure with aircraft-mounted probes due to issues with resolution, sizing, and size-dependent sampling volume. Furthermore, artefacts are produced by shattering of particles on the leading surfaces of the aircraft probes when particles several hundred microns or larger are present. Here, we report on a series of balloon-borne in situ measurements that were carried out at a high-latitude location, Kiruna in northern Sweden (68N 21E). The method used here avoids these issues experienced with the aircraft probes. Furthermore, with a balloon-borne instrument, data are collected as vertical profiles, more useful for calibrating or evaluating remote sensing measurements than data collected along horizontal traverses. Particles are collected on an oil-coated film at a sampling speed given directly by the ascending rate of the balloon, 4 m s(-1). The collecting film is advanced uniformly inside the instrument so that an always unused section of the film is exposed to ice particles, which are measured by imaging shortly after sampling. The high optical resolution of about 4 A mu m together with a pixel resolution of 1.65 A mu m allows particle detection at sizes of 10 A mu m and larger. For particles that are 20 A mu m (12 pixel) in size or larger, the shape can be recognized. The sampling volume, 130 cm(3) s(-1), is well defined and independent of particle size. With the encountered number concentrations of between 4 and 400 L-1, this required about 90- to 4-s sampling times to determine particle size ...