Interacting effects of aerosols and ice formation processes on mixed-phase cold-air outbreak clouds

Mixed-phase clouds associated with cold-air outbreak (CAO) events are natural laboratories to study mixed-phase cloud processes which are important for our estimation of cloud-phase feedback. These CAO clouds have also been shown vital to the radiative bias over the Southern Ocean. Recent studies show that CAO clouds are sensitive to aerosols including cloud condensation nuclei (CCN) and ice-nucleating particles (INPs), as well as secondary ice production (SIP). Therefore, it is vital to understand how these processes affect the radiative properties of CAO clouds and their roles in the cloud-phase feedback mechanism. However, many modelling studies have investigated the effects of these processes by perturbing model parameters individually, limiting the investigation of joint effects from multiple processes on cloud properties.  

Here we investigated how six cloud microphysics parameters jointly affect CAO cloud properties by building model emulators trained on output from perturbed parameter ensembles (PPEs) of a high-resolution regional model. The selected CAO case was on 24 October 2022 over the Labrador Sea, which coincided with the M-Phase aircraft campaign. The parameters are cloud droplet number concentration (N_d), ice-nucleating particle concentration (N_INP), efficiencies of three SIP processes including the rime-splintering, ice-ice collisional breakup and droplet shattering, as well as the mixed-phase overlap factor (mpof) which controls the spatial overlap between liquid and ice clouds within model grid cells. The perturbed ranges of these parameters either match the observed ranges when available or were chosen based on uncertainty ranges suggested by previous studies.  

For the CAO case studied, N_d and N_INP most strongly control the cloud radiative properties in the stratocumulus region; whereas in the cumulus region, N_d and mpof are the most important parameters. Variations of SIP efficiencies have stronger effects in the cumulus region compared to their effects in the stratocumulus region, but their effects on radiative properties are generally weaker compared to the other three parameters (N_d, N_INP and mpof).  

Our results show that these parameters have non-linear joint effects such that the magnitude and even sign of cloud responses to a parameter are highly dependent on the values of other parameters. For example, the sensitivity of cloud albedo to increases in N_INP varies between near zero and strongly negative across the sampled parameter space. Therefore, perturbing parameters individually is an inadequate method for determining the cloud responses to model parameters and can potentially lead to misleading conclusions.  

This work illustrates the power of using PPEs and model emulation for systematically quantifying the sensitivities of CAO cloud properties to important cloud microphysics parameters and identifying the interactions among model parameters. Exploration of the entire parameter space is compulsory to fully understand the influences on CAO clouds from these parameters, and to constrain the uncertainties from mixed-phase cloud processes on cloud-phase feedback mechanism.