The Arctic region, where clouds have a great impact on the radiation budget, is sensitive to climate change (Curry et al. 1996). In the Arctic, liquid-containing clouds have stronger longwave surface radiative forcing than ice clouds (Shupe and Intrieri 2004). Arctic stratiform cloud (ASC), a kind of liquid-containing clouds in the Arctic, often exhibited liquid top and mixed- or ice-phase below, which is called the phase inversion (Shupe et al. 2008), and occurred 42.3% of the time annually (Qiu et al. 2015) with relatively long lifetime more than 12 hours (Shupe et al. 2006). Therefore, ASC and their phase inversion structure are important to surface radiation budget in the Arctic and should be well understood to improve simulations of climate change.
The phase inversion structure in ASC, often coincident with temperature and specific humidity inversions within or above cloud (Qiu et al. 2015; Sedlar et al. 2012), is suggested a result of persistent liquid layer generation aloft and gravitational ice precipitation below (Morrison et al. 2012). Firstly, supercooled liquid cloud layer forms and ice nucleates within the cloud afterward (de Boer et al. 2011). After nucleation, ice particles grow via the Wegener–Bergeron–Findeisen (WBF) process and precipitate out by gravity, forming an ice-dominant cloud layer below the liquid-dominant cloud layer.
The depletion of liquid by the WBF process is compensated by the production due to updrafts of radiative-cooling-induced turbulence mixing (Morrison et al. 2012), or entrainment of water vapor from moist inversion above the cloud (Solomon et al. 2011).
Consequently, the magnitude of the WBF process plays a key role in the maintenance of the liquid layer. Hence, ice number concentration, affecting the WBF process, is critical to phase inversion in ASC.
Ice nucleation itself may also lead to phase inversion in ASC if the environment is not suitable for the liquid layer to nucleate ice. Ice nucleation in ASC is dominated by heterogeneous nucleation process because ASC are found to be most common in the ‐25 to ‐10⁰C environment, which is much higher than the temperature for significant homogeneous freezing rate (de Boer et al. 2011). Heterogeneous nucleation process requires ice nuclei (IN) to assist nucleation, including four ways: deposition nucleation, immersion freezing, condensation freezing, and contact freezing. Deposition nucleation is the direct deposition of water vapor on IN; immersion freezing is the freezing of water droplets with immersed IN; condensation freezing is the process of condensation of water vapor on IN quickly followed by freezing of the condensed water; contact freezing is the nucleation occurring at the site of the collision between IN and water droplets. Among all kinds of heterogeneous processes, immersion freezing is thought to be the main process in ASC (de Boer et al. 2010; Prenni et al. 2009), which meets the observation of first liquid formation (de Boer et al. 2011). However, the true role of different heterogeneous processes is not well understood yet because of the poor knowledge of IN species and the usage of empirical parameterizations in the model.
Empirical parameterizations mostly do not account for IN species and often relate ice nucleation to a single environmental factor simply regardless of the actual variation of IN. For example, immersion freezing is often parameterized as a function of temperature only (Bigg 1953; Diehl and Wurzler 2004), while deposition nucleation is often associated with supersaturation only (Meyers et al. 1992; Prenni et al. 2007).
Moreover, the ice nucleation rate is often calculated by the difference between the number concentration of IN and that of existing ice particles (de Boer et al. 2010; Morrison et al.
2009) regardless of possible removal by precipitation scavenging. These methods lack
Paukert and Hoose (2014) fixed these problems by using a more complex ice nucleation parameterization by considering temperature, saturation and IN species to simulate ASC, but they accounted for immersion freezing only. As immersion freezing and deposition nucleation of dust, and immersion freezing and contact freezing of soot are dominant in the Arctic (Hoose et al. 2010), the role of IN combining with nonlinear effect between different environmental factors on different ice nucleation processes in ASC was not fully investigated before. Ice nucleation processes are affected by temperature, saturation, and IN species and number concentration. Unfavorable conditions in any of these factors can cause a lower ice nucleation rate in the liquid layer and thus maintain the phase inversion structure in ASC, and these factors can be investigated only when a prognostic treatment of IN and the classical nucleation theory are considered at the same time.
The classical nucleation for heterogeneous nucleation derived in Chen et al. (2008) accounts for deposition nucleation and immersion freezing, while condensation freezing and contact freezing may be linked to immersion freezing if the preceding process (i.e, condensation and contact) can be accounted for. The nucleation rate can be expressed as
J = A’ ∙ rN2∙ √f ∙ exp(−ΔgkT#−Δgg), ( 1 ) where k is the Boltzmann constant, T is the temperature, A’ depends on ambient parameters, rN is the radius of IN, f is a geometric factor to account for the curvatures of the IN and ice germ, Δg# represents the desorption energy for deposition nucleation and the activation energy for immersion freezing, and Δgg is the germ formation energy.
The curvature of the ice germ is associated with the contact angle θ . The effects of curvature and solute on freezing is considered by including the water activity term in Δgg. This formula, which would be used in this study, can calculate different nucleation processes for different IN species in a simple way and can discover the roles of aerosol
and ice nucleation in phase inversion.
This study aims to further clarify the mechanism of phase inversion and find out the actual role of specific IN species and the associated ice nucleation process in ASC by model simulation. We seek to advance our knowledge of ASC and the role of aerosol particles in such systems.