ausblenden:
Schlagwörter:
-
Zusammenfassung:
Individual crystal imagery (order of magnitude: 100-10000 particles per second) with state-of-art binary optical array probes (OAP, e.g. 2DS and PIP) was continuously improved and operated with complementary cloud instruments to document size, fall speed, mass, and density of ice crystals, also co-existence of supercooled water droplets. However, the full potential of these images still has to be exploited in terms of quantitative ice particle morphological analysis.<pIn this study, we investigate the link between ice crystal morphological shape and associated growth regimes (vapor diffusion, aggregation growth, and riming) which are active in space and time during the cloud life cycle. The study utilizes the dataset from the HAIC-Cayenne 2015 flight campaign with corresponding satellite observations. First, we characterize the environmental conditions of sampled mesoscale convective systems (MCSs) to determine the convective, stratiform, and cirriform parts of MCSs. Then, number fraction of morphological classes during these flights are produced, thereby applying the recently developed Convolutional Neural Network tool for sophisticated automatic ice crystal classification (Jaffeux et al., 2022). The statistical analyses on crystal shapes should help in identifying which morphologies are the most abundant during the spatiotemporal evolution of cloud ice. This work should help to better quantify the three principal growth processes’ mass contributions to individual morphological classes.
