How does riming influence the observed spatial variability of ice water in mixed-phase clouds?

Abstract. Mixed-phase clouds (MPC) are a key component of the Earth's climate system. Observations show that ice water content (IWC) is not distributed homogeneously in MPC. Instead, high IWC tends to occur in clusters. However, it is not sufficiently understood, which ice crystal formation and growth processes play a dominant role in IWC clustering. One important ice growth process is riming, which occurs when liquid water droplets freeze onto ice crystals upon contact. Here, airborne measurements of MPC in mid- and high-latitudes are used to study spatial variability of ice clusters and investigate how this variability is linked to riming. We use data from the IMPACTS (mid-latitudes) and the HALO-(AC)³ (high-latitudes) aircraft campaigns, where closely spatially and temporally collocated cloud radar and in situ measurements were collected. Ice cluster scales and IWC variability are quantified using pair correlation functions. By comparing IWC calculations accounting for riming to IWC calculations neglecting riming, we single out the influence of riming. During all analyzed flight segments, riming is responsible for 66 % and 63 % of total IWC during IMPACTS and HALO-(AC)³, respectively. In mid-latitude MPC, riming does not significantly change IWC cluster scales, but increases the probability of clusters occurrence. This enhancement occurs at similar scales as liquid water content variability. In cold air outbreak MPC observed during HALO-(AC)³, riming impacts IWC clustering at two distinctive scales. First, riming enhances the probability of IWC clusters at spatial scales below 2 km, which corresponds to the wavelength of the roll cloud updraft and circulation features. Second, riming leads to additional IWC clustering at spatial scales of 3–5 km. We find that the presence of mesoscale updraft features leads to enhanced occurrences of riming and therefore additional IWC clustering. An increased liquid water path might increase the effect, but is not a necessary criterion. These results help to improve our understanding of how riming is linked to IWC variability and can be used to evaluate and constrain models of MPC.