Quantifying riming from airborne data during the HALO-(AC) 3 campaign

• Published in: Atmospheric Measurement Techniques
• Main Authors: Maherndl, Nina (author), Moser, Manuel (author), Lucke, J.R. (author), Mech, Mario (author), Risse, Nils (author), Schirmacher, Imke (author), Maahn, Maximilian (author)
• Format: Article in Journal/Newspaper
• Language: English
• Published: 2024
• Subjects: Fram Strait; Svalbard
• Online Access: http://resolver.tudelft.nl/uuid:7e632475-60c7-4570-aae1-e258a9cdf3a9; https://doi.org/10.5194/amt-17-1475-2024

Riming is a key precipitation formation process in mixed-phase clouds which efficiently converts cloud liquid to ice water. Here, we present two methods to quantify riming of ice particles from airborne observations with the normalized rime mass, which is the ratio of rime mass to the mass of a size-equivalent spherical graupel particle. We use data obtained during the HALO-(AC)3 aircraft campaign, where two aircraft collected radar and in situ measurements that were closely spatially and temporally collocated over the Fram Strait west of Svalbard in spring 2022. The first method is based on an inverse optimal estimation algorithm for the retrieval of the normalized rime mass from a closure between cloud radar and in situ measurements during these collocated flight segments (combined method). The second method relies on in situ observations only, relating the normalized rime mass to optical particle shape measurements (in situ method). We find good agreement between both methods during collocated flight segments with median normalized rime masses of 0.024 and 0.021 (mean values of 0.035 and 0.033) for the combined and in situ method, respectively. Assuming that particles with a normalized rime mass smaller than 0.01 are unrimed, we obtain average rimed fractions of 88ĝ€¯% and 87ĝ€¯% over all collocated flight segments. Although in situ measurement volumes are in the range of a few cubic centimeters and are therefore much smaller than the radar volume (about 45ĝ€¯m footprint diameter at an altitude of 500ĝ€¯m above ground, with a vertical resolution of 5ĝ€¯m), we assume they are representative of the radar volume. When this assumption is not met due to less homogeneous conditions, discrepancies between the two methods result. We show the performance of the methods in a case study of a collocated segment of cold-air outbreak conditions and compare normalized rime mass results with meteorological and cloud parameters. We find that higher normalized rime masses correlate with streaks of higher radar reflectivity. ...