Abstract

Formation of ice in mixed-phase clouds may proceed via primary nucleation and secondary ice production (SIP). Primary nucleation involves the presence of ice nucleating particles (INPs), while SIP follows primary nucleation, resulting in ice crystal number concentrations N_ice exceeding the number of INPs. The ice formation pathways in young congestus clouds are not well understood nor are the environmental conditions favorable for the onset of SIP. Coincident airborne measurements of INPs and N_ice collected during the Secondary Production of Ice in Cumulus Experiment (SPICULE) campaign over the U.S. central Great Plains are reported for young congestus clouds. Number concentrations of INPs and N_ice at temperatures between −10° and −20°C were used to categorize whether SIP was active in each analyzed cloud pass. Observational data suggested that fragmentation of freezing droplets (FFDs) was occurring and the most likely SIP mechanism for ice enhancement during the early stages of the sampled clouds, where higher cloud-base temperatures and stronger updrafts produced conditions favorable for the onset of FFD. Simple model simulations of a congestus cloud were used to compare predictions of N_ice from newly implemented SIP mechanisms (including FFD) alongside those from the existing Hallett–Mossop (HM) mechanism. Although predicted N_ice, when all SIP mechanisms were active, was in good agreement with observations, the dominant SIP mechanism in the model was predicted to be HM. The rate of heterogeneous freezing in raindrops was likely inhibiting realistic FFD rates in the simulated cloud. These discrepancies between the observations and simulations underscore the need for extended laboratory studies to formulate improved model representations of SIP in convective clouds.