A Versatile Method for Ice Particle Habit Classification Using Airborne Imaging Probe Data

Abstract

A versatile method to automatically classify ice particle habit from various airborne optical array probes is presented. The classification is achieved using a multinomial logistic regression model. For each airborne probe, the model determines the particle habit (among six classes) based on a large set of geometrical and textural descriptors extracted from the two-dimensional image of a particle. The technique is applied and evaluated using three probes with significantly different specifications: the high volume precipitation spectrometer, the two-dimensional stereo probe, and the cloud particle imager. Performance and robustness of the method are assessed using standard machine learning tools on the basis of thousands of images manually labeled for each of the considered probes. The three classifiers show good performance characterized by overall accuracies and Heidke skill scores above 90%. Depending on the application and user preferences, the classification scheme can be easily adapted. For a more precise output, intraclass subclassification can be achieved in a nested fashion, illustrated here with columnar crystals and aggregates. A comparative study of the classification output obtained with the three probes is presented for two aircraft flight periods selected when the three probes were operating together. Results are globally consistent in term of proportions of habit identified (once blurry and partial images have been automatically discarded). A perfect agreement is not expected as the three considered probes are sensitive to different particle size range.Plain Language Summary An automatic classification method to identify ice particle habit from images is proposed. The technique is applied and evaluated using three airborne probes mounted on research aircraft with significantly different specifications: the high volume precipitation spectrometer, the two-dimensional stereo probe, and the cloud particle imager. The method relies on thousand of images manually classified and advanced machine learning techniques to determine the snow crystal habit among six preset classes. High classification performance is achieved, with accuracies above 90% for each of the considered probes.

Classification of Ice Crystal Habits Observed From Airborne Cloud Particle Imager by Deep Transfer Learning

“…They sorted the images of ice crystals into nine habit categories (spheroids and small plates, long columns, thick plates/short columns, plates, rosettes, budding rosettes, bullet rosettes, irregulars, and complex crystals with side planes). Praz et al () used the method of multinomial logistic regression to classify the six ice crystal habits based on the geometrical and textural descriptors of ice crystals. Although the above studies analyzed the categories of ice crystal habits, a publicly available ice crystal database in China has still not been established.…”

Characterising optical array particle imaging probes: implications for small-ice-crystal observations

“…How-ever, it is possible that small particles could be misclassified as artefacts or vice versa, and as a result HALOHolo could either underestimate or overestimate the small ice concentration. For particles > 35 µm, it is estimated that the probe's detection rate is > 90 %, and previous work has shown excellent agreement with a CDP in liquid clouds (Schlenczek, 2017). However, HALOHolo PSDs should not be considered the true PSD but rather another piece of evidence that suggests for these cases OAPs overestimate small ice concentrations using current data-processing techniques.…”

“…The detection of small particles is limited by noise in the background image. Therefore, a minimum size threshold of 35 µm is applied, above which it is estimated that the probe's detection rate is greater than 90 % (Schlenczek, 2017). Shattered particles were minimised by removing all particles with interparticle distances less than 10 mm (Fugal and Shaw, 2009;O'Shea et al, 2016).…”

“…However, as described earlier, measurements of small particles from HALOHolo are limited by noise in the background image. For a complete description of the HALOHolo data-processing and qualitycontrol procedures, see Schlenczek (2017). HALOHolo uses supervised machine learning to discriminate real particles from artefacts due to noise in the background image.…”

Characterising optical array particle imaging probes: implications for small ice crystal observations

“…The shape of ice crystals is a key microphysical parameter impacting cloud radiative properties in a number of ways. A variety of automatic image recognition algorithms have been applied to OAP datasets to classify particles into different habits (Korolev & Sussman, 2000;Crosier et al, 2011;Praz et al, 2018). These algorithms typically rely on geometrical features extracted from OAP images that have characteristic values for specific habits.…”