Quantifying uncertainties in determining morphological variables of ice crystals

• Main Authors: Park, S., Um, J., Järvinen, E., Schnaiter, M., McFarquhar, G., Kim, J.
• Format: Conference Object
• Language: English
• Published: 2023
• Subjects: Arctic; Southern Ocean
• Online Access: https://gfzpublic.gfz-potsdam.de/pubman/item/item_5019025

Morphological variables (e.g., maximum dimension D max , projected area, perimeter, width, and roundness) of ice crystals are fundamental information to represent microphysical and radiative properties of ice clouds in numerical models, which can be derived via in-situ aircraft measurements. Conventional airborne cloud probes are limited to a single observation angle to measure morphological variables. Non-spherical ice crystals may have different measured morphological variables depending on the observation angle, which causes uncertainties in the calculation of the microphysical and radiative properties of ice clouds. Thus, it is necessary to quantify uncertainties that depend on observation angles for determining the morphological variables of ice crystals. In this study, differences in determining the morphological variables of the ice crystals depending on observation angles were quantified using measurements of Particle Habit Imaging and Polar Scattering (PHIPS) probe that were acquired during the Arctic CLoud Observation Using airborne measurements during polar Day field campaign and the Southern Ocean Clouds Radiation Aerosol Transport Experimental Study field campaign. The PHIPS is a probe that provides stereo images of the sampled ice crystal from two different observation angles that are 120° apart. Morphological variables and habits of ice crystals were determined based on PHIPS stereo images. The uncertainties in determining morphological variables of ice crystals were quantified by calculating relative differences in morphological variables of ice crystals depending on observation angles. The nonsphericity of ice crystals sampled during ACLOUD and SOCRATES caused an average 14.69% (22.75%; 14.57%; 16.67%; 18.76%) uncertainty in determining Dmax (projected area; perimeter; width; roundness).