Mineralogy Sensitive Immersion Freezing Parameterization in DREAM

Abstract Dust aerosols are abundant in the atmosphere and are very efficient ice nucleating particles at temperatures below −15°C. Depending on temperature, dust particles containing certain minerals (i.e., feldspar and quartz) are the most active as ice nuclei. A mineralogy‐sensitive immersion freezing parameterization for ice nucleating particle concentration (INPC) is implemented in Dust Regional Atmospheric Model (DREAM) for the first time. Additionally, four mineralogy‐indifferent parameterizations are implemented, two for immersion freezing and two for deposition nucleation. Dust concentration and its feldspar and quartz fractions are forecasted by DREAM for a dust episode in the Mediterranean in April 2016. DREAM results are compared with vertical profiles of cloud‐relevant dust concentrations and INPC from ground‐based lidar measurements in Potenza, Italy and Nicosia, Cyprus. INPC predictions are also compared with vertical profiles of ice crystal number concentration (ICNC) from satellite observations for two overpasses over the dust plume. The model successfully simulates the evolution and vertical extent of the dust plume. Mineralogy‐sensitive and mineralogy‐indifferent INPC parameterization results generally differ by about an order of magnitude. Forecasted INPC and observed ICNC values differ by an order of magnitude for all parameterizations. Feldspar fraction increase within a dust plume during transport can increase INPC by around 6% at −35°C, and up to 17% at −25°C, but sedimentation can reduce this effect. Over the Atlantic, mineralogy‐sensitive parameterization predicts horizontal distribution of clouds with higher probability of success, while in the Mediterranean; the results for different parameterizations show lower variability.