Investigating Secondary Ice Processes in Winter Mixed-Phase Cloud over the Swiss Alps

Abstract:

The co-existence of ice crystals and supercooled liquid water droplets in clouds plays a significant role in the formation of precipitation in mid-latitudes. After primary production of ice through homogeneous and heterogeneous nucleation, secondary ice processes cause ice crystals to multiply rapidly enabling mixed-phase clouds to glaciate faster. An example of such a secondary ice process occurring between -3 and -8 ᵒC is rime splintering.

In the case where only heterogeneous nucleation is the source of ice crystals in a mixed-phased cloud (MPCs), it is expected that the ice nucleating particle concentration (INPC) would be equal to the ice crystal number concentration (ICNC).

However, field campaigns conducted at Jungfraujoch, in Switzerland, show that measurements of ICNCs exceed the INPC by several orders of magnitude. Also, many modelling studies show that the simulated ICNC during winter is often underestimated by an order of magnitude in MPCs as compared to observations. Because temperatures were at or below –15 ᵒC, rime splintering can be ruled out as a source of secondary ice. Yet, rime splintering is the only secondary ice formation scheme used in these models. We hypothesize that the discrepancy between the model results and the observations may be explained with other secondary ice formation processes such as droplet shattering upon freezing and breakup through collisions between ice hydrometeors.

Here, we present the results from a sensitivity study with additional simulated in-cloud secondary ice processes on precipitation formation and surface precipitation. We simulated the passage of a cold front on 7 March 2019 with COSMO as part of the RACLETS field campaign in Davos, Switzerland. The largest simulated differences came from the breakup simulations that showed 1-3 orders of magnitude larger ICNC than in the reference, rime splintering or droplet shattering simulations. Subsequently, the ICNC from the breakup simulations (at a formation rate of 10³ L^-1 s^-1 at 5 km in altitude) at Gotschnagrad was in better agreement with in-situ measurements and could also reproduce radar-derived precipitation rates more accurately.