A case study of microphysical structures and hydrometeor phase in convection using radar Doppler spectra at Darwin, Australia

To understand the microphysical processes that impact diabatic heating and cloud lifetimes in convection, we need to characterize the spatial distribution of supercooled liquid water. To address this observational challenge, ground‐based vertically pointing active sensors at the Darwin Atmospheric Radiation Measurement site are used to classify cloud phase within a deep convective cloud. The cloud cannot be fully observed by a lidar due to signal attenuation. Therefore, we developed an objective method for identifying hydrometeor classes, including mixed‐phase conditions, using k‐means clustering on parameters that describe the shape of the Doppler spectra from vertically pointing Ka‐band cloud radar. This approach shows that multiple, overlapping mixed‐phase layers exist within the cloud, rather than a single region of supercooled liquid. Diffusional growth calculations show that the conditions for the Wegener‐Bergeron‐Findeisen process exist within one of these mixed‐phase microstructures. Key Points Convective cloud observed containing multiple mixed‐phase layers K‐means clustering of moments of radar spectra identifies layers Conditions favorable for WBF process observed in convective cloud Plain Language Summary We use radar measurements to identify regions of a thick tropical cloud with both liquid droplets and ice crystals at temperatures colder than 0°C. The coexistence of liquid and ice in the cloud changes how ice crystals grow. Because falling ice crystals melt into rain when they reach warmer air, observing where liquid and ice coexist can help us improve our understanding of when and how much clouds will rain.