1. Introduction
The Arctic is warming at more than twice the pace of the rest of the globe according to observations of the past century (Serreze et al. 2009; Serreze and Barry 2011). Global climate model (GCM) projections of the surface temperature warming suggest that by the end of the twenty-first century the Arctic will continue to warm at the ongoing rate (Barnes and Polvani 2015; Boeke and Taylor 2018). Most GCMs simulate a higher degree of warming in the Arctic relative to the rest of the globe; however, the spread in the amount of projected warming is also largest in the Arctic. Clouds play a critical role in the Arctic climate system through their impacts on both the surface (Kay and Gettelman 2009; Cesana et al. 2012; Bennartz et al. 2013) and top-of-the-atmosphere (TOA) radiation budget (Curry and Ebert 1992; Beesley 2000; Xie et al. 2013; Kay and L’Ecuyer 2013; English et al. 2014), which also includes a nonlinear interaction with Arctic sea ice. GCMs exhibit a particularly large spread in the representation of boundary layer Arctic clouds (Karlsson and Svensson 2011; Boeke and Taylor 2016), many of which are a combination of supercooled liquid droplets and ice crystals. These mixed-phase clouds are ubiquitous in the Arctic boundary layer as revealed by in situ aircraft and remote sensing observations from field campaigns as well as ground and satellite remote sensing instruments (Curry et al. 1996; Pinto 1998; Korolev et al. 2003; Chylek and Borel 2004; Zuidema et al. 2005; Verlinde et al. 2007; Shupe et al. 2011; Mioche et al. 2015; Ehrlich et al. 2019). These clouds have been observed to persist for days to weeks due to a complex interaction of microphysical, radiative, dynamical, and turbulent processes (Morrison et al. 2012). In particular, the dynamics of supercooled liquid clouds, ice nucleation, and the ice mass and number flux at cloud base have been shown to contribute to the resilience of mixed-phase clouds (Ovchinnikov et al. 2014).
The large spread in the model representation of various cloud properties including cloud fraction, ice water path (IWP), and liquid water path (LWP) has been linked to uncertainty in the parameterization of ice cloud microphysical processes (Klein et al. 2009; Taylor et al. 2019). The disagreement in the representation of the annual cycle of Arctic low-level clouds in phase 5 of the Coupled Model Intercomparison Project (CMIP5) has also been linked to the parameterization of ice cloud microphysical processes (Taylor et al. 2019). Furthermore, a number of single-column models and cloud-resolving models were shown to underestimate Arctic cloud LWP relative to in situ aircraft and ground-based remote sensing observations due to the interaction between liquid and ice cloud microphysics (Klein et al. 2009). The underestimate of LWP relative to the total mixed-phase cloud condensate is an issue that is not solely relevant to the Arctic, but rather affects all regions represented by a multitude of GCMs (Komurcu et al. 2014; Cesana et al. 2015; McCoy et al. 2016). Both the horizontal and vertical variability in clouds have been identified as reasons for the lack of supercooled liquid in mixed-phase clouds. GCMs do not represent horizontal subgrid variability of supercooled liquid and ice that comprise mixed-phase clouds for which there is evidence in nature (Korolev et al. 2003). This may lead to an unrealistically efficient Wegener–Bergeron–Findeisen (WBF) process (Wegener 1911; Bergeron 1935; Findeisen 1938) and in turn may lead to underestimates in supercooled liquid water within mixed-phase clouds (Tan and Storelvmo 2016) and so impact their microphysical properties (Zhang et al. 2020). Moreover, Arctic mixed-phase clouds are particularly challenging to represent in GCMs due to their unique vertical structure consisting of a thin supercooled liquid top with a mean thickness of approximately 0.5 km and precipitating ice virga that often reaches the surface (de Boer et al. 2009; Mioche et al. 2017; Silber et al. 2021). This unique vertical structure of mixed-phase clouds with their supercooled liquid droplets found at temperatures colder than its ice crystals implies that the temperature-dependent phase partitioning schemes that are used in some GCMs (Cesana et al. 2015; Taylor et al. 2019) are unrealistic for the representation of Arctic mixed-phase clouds, which call for more physically based prognostic schemes (Liu et al. 2007). The coarse vertical resolution of GCMs was determined to be a key reason for the lack of supercooled liquid in midlatitude altocumulus mixed-phase clouds due to biases in ice microphysical process rates near cloud top (Barrett et al. 2017). In addition to horizontal and vertical variability, the more physically based two-moment cloud microphysics that predict the mass and number concentration of cloud liquid and ice were shown to be capable of representing the unique vertical structure of mixed-phase cloud; however, they showed larger underestimates in supercooled liquid (Barrett et al. 2017). Two-moment microphysical schemes were indeed found to perform better in terms of the overall simulated cloud properties in comparison to various observations (Klein et al. 2009; Taylor et al. 2019).
A central component of two-moment cloud microphysics schemes is the primary production of ice via nucleation. Since temperatures in the Arctic boundary layer are typically insufficiently low for the homogeneous freezing of cloud droplets, which typically occurs at approximately −38°C (Pruppacher and Klett 2010), ice formation must be mediated by ice-nucleating particles (INPs). Several studies have shown that Arctic cloud properties and radiation are sensitive to ice nucleation in GCMs (Prenni et al. 2007; Liu et al. 2011; Fan et al. 2012; Fan 2013; Xie et al. 2013; English et al. 2014; Wang et al. 2018) and large-eddy simulations (Jiang et al. 2000; Fridlind et al. 2007, 2012; Solomon et al. 2015). These studies have shown that ice nucleation strongly impacts the vertical structure, coverage, and longevity of Arctic stratiform mixed-phase clouds. Depending on the parameterization, ice nucleation can either improve or worsen different aspects of clouds and radiation when compared to satellite and ground-based measurements.
Heterogeneous ice nucleation is commonly thought to occur via four different modes: deposition, condensation, contact and immersion (Rogers and Yau 1989; Vali et al. 2015). Out of these four modes, immersion freezing, where ice is initiated by an INP immersed in a supercooled liquid droplet, is likely the most common ice nucleation mode that occurs in the atmosphere (Prenni et al. 2009; de Boer et al. 2010; Wiacek et al. 2010; Murray et al. 2012). This is supported by several lines of evidence. First, multilayer water coverage is easily built on the surface of dust and soot particles even if they lack water-soluble materials (Laaksonen et al. 2020). Hence it is unlikely that INPs with partial water coverage, a condition necessary for deposition and condensation ice nucleation (Welti et al. 2014), would represent a substantial fraction of the aerosol population in the upper troposphere. Also, it has been shown that cloud parcels are commonly exposed to water saturation before reaching the upper troposphere, triggering cloud droplet activation before ice nucleation occurs (Wiacek et al. 2010). In the Arctic, ice formation is typically modulated by the presence of liquid water (de Boer et al. 2010; Lance et al. 2011; Morrison et al. 2012), suggesting an immersion freezing mechanism. A dominant immersion freezing regime may also be responsible for the persistence and structure (i.e., supercooled liquid on top of ice) characteristic of Arctic mixed-phase clouds, since ice crystals formed at the top of the cloud layer where temperatures are the lowest would rapidly fall, leaving a supercooled layer of liquid behind (de Boer et al. 2010).
Here, the impact of four different ice nucleation parameterizations of immersion freezing on Arctic cloud properties and radiation was studied using the fifth version of the National Aeronautics and Space Administration (NASA) Goddard Earth Observing System (GEOS-5) model (Molod et al. 2020). The Arctic is defined as the region between 60° and 80°N due to the lack of quantity and quality of the satellite instruments used in this study, and to be consistent with previous studies (Xie et al. 2013; English et al. 2014). The parameterizations were selected to represent the different levels of sophistication currently used in GCMs, the diversity in their method of derivation and to produce a wide range of INP concentrations and vertical structures. They include a classical nucleation theory (CNT)-based scheme (Barahona and Nenes 2009, hereafter BN09), a scheme based on field observations in Wyoming in the United States (Meyers et al. 1992, hereafter M92), a scheme based on laboratory measurements (Ullrich et al. 2017, hereafter U17), and a scheme based on field observations from around the globe (DeMott et al. 2010, hereafter D10). The primary intention of this manuscript is to understand the characteristics of the various distinctive features of the diverse ice nucleation parameterizations and their interactions with other cloud microphysical processes, and to determine their influence on clouds and radiation in GEOS-5. In each parameterization, immersion freezing refers to freezing of activated droplets, whether in cloud droplets or rain drops. Approximate comparisons are also made between the simulated cloud properties with observations from NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) and NASA’s Clouds and the Earth’s Radiant Energy System (CERES).
Section 2 provides a description of the model, including the four immersion freezing parameterizations, as well as the MODIS and CERES EBAF datasets. This is followed by section 3, which presents a comparison of the cloud microphysical and macrophysical properties as well as TOA and surface cloud radiative effects (CREs) between the model and the observations. Finally, the results are discussed in section 4 and a summary is provided in section 5.
2. Method
a. Description of the GEOS-5 model
GEOS-5 is a GCM that consists of a set of components that numerically represent different aspects of the Earth system (atmosphere, ocean, land, sea ice, and chemistry), coupled following the Earth System Modeling Framework (https://gmao.gsfc.nasa.gov/GEOS_systems/). GEOS-5 is the base model supporting the version 2 of the Modern-Era Retrospective Analysis for Research and Applications (MERRA-2) and it is commonly used to conduct operational weather and seasonal forecast (Molod et al. 2020). For this work, the atmospheric GCM configuration of GEOS-5 was used where sea ice fraction and sea surface temperature (SST) are prescribed as time-dependent boundary conditions (Reynolds et al. 2002; Rienecker et al. 2008). However, to reduce the effect of interannual meteorological variability that may influence ice nucleation, sea ice fraction and SSTs were prescribed instead from an annual climatology (1979–2019) calculated from MERRA-2. GEOS-5 has been shown to reproduce the global distribution of clouds, radiation, and precipitation in agreement with satellite retrievals and in situ observations (Barahona et al. 2014; Molod et al. 2020; Breen et al. 2021).
The Goddard Chemistry Aerosol and Radiation model (GOCART) computes the evolution of aerosols and gaseous tracers (Colarco et al. 2010), which interactively calculates the transport and evolution of dust, black carbon, organic material, sea salt, and SO2. Dust and sea salt emissions are prognostic whereas surface emissions of biomass burning, anthropogenic SO2, black carbon, and organic carbon are obtained from the MERRA-2 dataset (Randles et al. 2017), and maintained at 2005 levels, again to reduce the effect of interannual variability. Local aerosol emissions from volcanoes at high latitudes have been shown to impact the TOA and surface radiation budgets (Young et al. 2012) and long-range aerosol transport into the Arctic (Bian et al. 2013) and warm and moist air intrusions (Cullather et al. 2016) potentially laden with aerosols, both represented in GEOS-5 may also have a similar effect. Therefore the year 2005 was selected as it is associated with low volcanic activity (Carn et al. 2016). GOCART treats each aerosol type as externally mixed. INPs are predicted and interactively evolve with GOCART at each model time step; they are therefore subject to aerosol scavenging and transport processes. As it is based on a single-moment scheme, GOCART does not account explicitly for ice nucleation processes but rather uses a bulk scheme that removes the mass of different aerosol species at each time step using the total precipitation flux. The distributions of dust and sea salt are represented using five lognormal bins ranging from 0.1 to 1.0 μm, from 1.0 to 1.8 μm, from 1.8 to 3.0 μm, from 3.0 to 6.0 μm, and from 6.0 to 10.0 μm; single lognormal modes are used for the other aerosol components (Chin et al. 2009; Colarco et al. 2010).
Cirrus formation at temperatures T < 235 K is parameterized according to Barahona and Nenes (2009) where the ice crystal concentration is determined by the maximum supersaturation achieved during a parcel ascent. The scheme includes homogeneous and heterogeneous ice nucleation, the latter in the deposition and immersion modes (Ullrich et al. 2017), and it is forced using an annual climatology of the standard deviation in vertical velocity σw (Barahona et al. 2017). To allow for supersaturation in cirrus levels the saturation threshold required for new cloud formation is calculated as a function of the ice nucleation rate (Barahona et al. 2014). Ice formation in convective clouds is assumed to proceed by immersion in activated droplets and contact ice nucleation, the latter of which is described by the Ullrich et al. (2017) spectrum. Contact ice nucleation is described by shifting the immersion freezing efficiency (calculated using the U17 spectrum in all experiments) by 3 K and scaled by the particle–droplet collision rate (Hoose et al. 2010; Ladino Moreno et al. 2013). Deposition, condensation, and immersion within solution droplets are not explicitly represented at mixed-phase temperatures.
The model results presented herein represent averages in the last 30 years of a total of 40 years of simulations using a c90 cubed-sphere horizontal grid (nominally about 1° in latitude and longitude) and 48 vertical levels. Near the surface, the pressure levels were spaced 25 hPa apart until 700 hPa, above where the vertical resolution decreased to 50 hPa until 100 hPa. The topmost level extends to 0.01 hPa (∼85 km). The time step of the simulations was 900 s. The first 10 simulated years were omitted from the analysis as part of the model spinup. Simulations are free-running except for the prescribed SSTs and sea ice; nudging and data assimilation are not applied.
b. Immersion freezing schemes
Ice nucleation in the stratiform mixed-phase regime (235 K < T < 273 K) is assumed to always occur in the presence of activated droplets and rainwater but not aerosol solution, and is primarily driven by immersion and contact freezing. Immersion ice nucleation in interstitial aerosol is neglected since concentrated solution droplets provide an unfavorable environment for ice formation. The four immersion ice nucleation schemes used in this work are either empirical or semi-empirical, as described below and summarized in Table 1. Only dust and soot aerosol species are considered INPs except in D10, where all aerosol species with diameters greater than 0.5 μm are considered INPs.
List of freezing modes, model simulations, and their description.
1) The BN09 scheme
Despite being formulated in terms of Si, nINP is effectively a function of T and the aerosol number concentration in BN09. Since the INPs are immersed within a droplet, their local environment is saturated with respect to water. Active INP concentrations are thus always calculated at Si at water saturation (i.e., Si,1 = pw/pi, where pw and pi are the saturation vapor pressures for liquid and ice, respectively; Murphy and Koop 2005). This is a common assumption in GCMs, even when nucleation modes other than immersion freezing are considered. However, it would be incorrect to apply the same assumption to deposition and condensation ice nucleation without accounting for the dynamic feedback between ice crystal growth and supersaturation.
The BN09 scheme is an example of the single-θ classical nucleation theory (CNT) approach to immersion freezing. It is known that such a scheme overpredicts the ice crystal concentration because it does not take into account the distribution of active sites in the INP population (Ickes et al. 2017) and because CNT does not account for kinetic limitations inherent to immersion ice nucleation (Barahona 2018). Here BN09 is thus used to represent a limiting behavior where a large fraction of dust and soot act as INPs in mixed-phase clouds.
2) The M92 scheme
The M92 scheme is a widely used parameterization in several GCMs (e.g., Morrison and Gettelman 2008; Kawai et al. 2019); however, it was noted by Prenni et al. (2009) that it overestimates INP concentrations in the Arctic by several orders of magnitude compared to the Mixed-Phase Arctic Cloud Experiment (M-PACE) field campaign (Verlinde et al. 2007), which can in turn reduce the longevity of mixed-phase clouds and influence the Arctic surface radiation budget. However, the M92 scheme was included in this work to enable a comparison of its influence on Arctic clouds and radiation with previous studies (Xie et al. 2013; English et al. 2014). Note that the M92 parameterization does not depend on aerosol concentration and hence it is active regardless of the presence of INPs.
3) The U17 scheme
4) The D10 scheme
c. Observations
Model simulations of clouds and radiation are compared against those derived from three different satellite remote sensing products. Cloud fraction, water path, and optical thickness in the NASA Moderate Resolution Imaging Spectroradiometer (MODIS)/Aqua Collection 6.1 monthly 1° gridded dataset for the 17-yr time period extending from 2003 to 2019. The retrieved cloud fraction as opposed to the cloud mask fraction were compared, where the former refers to the cloud fraction calculated using the number of pixels for which cloud optical properties were successfully retrieved and the latter refers to the cloud fraction calculated by the number of pixels that were deemed to be either cloudy or probably cloudy (Pincus et al. 2012). Thus, the retrieved cloud fraction is always smaller than the cloud mask fraction. In addition to MODIS LWP, 1° gridded oceanic LWP derived from the Multisensor Advanced Climatology of Liquid Water Path (MAC-LWP) product based on a suite of 12 different spaceborne passive and microwave sensors (Elsaesser et al. 2017) for the 14-yr period from 2003 (the same year the MODIS/Aqua observations were available) to 2016 are also included for comparison. In contrast to the MODIS LWP product, the MAC-LWP product retrieves cloud properties based on brightness temperature rather than solar reflectance and therefore avoids sampling biases and solar zenith angle effects. However, the MAC-LWP product suffers from potential errors resulting from cross-talk, cloud-rain partitioning, cloud temperature and height, clear-sky biases and beam-filling effects, scattering by large ice particles, and cloud-rain partitioning (O’Dell et al. 2008). Finally, ground-based observations of LWP from Barrow, Alaska, are also included and are from the United States Department of Energy’s (DOE’s) Atmospheric Radiation Measurement (ARM) MicroWave Radiometer (MWR) Retrieval (MWRRET) value-added product (VAP) measured at the DOE ARM’s North Slope of Alaska (NSA) site from 1998 to 2018. Brightness temperature offsets at the 31.4-GHz channel were applied to reduce clear-sky biases in LWP in the MWRRET VAP (Gaustad and Turner 2007; Xie et al. 2010). Ground-based observations of IWP from the DOE’s ARM Cloud Retrieval Ensemble Dataset (ACRED) product at the NSA site measured from 1999 to 2008 were also included (Zhao 2011).
Monthly averaged radiative fluxes at the TOA and surface gridded at the 1° resolution were obtained by NASA’s Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) Level 3b Edition 4.1 satellite dataset. The monthly TOA fluxes were measured by the CERES instruments onboard the Terra, Aqua, Suomi National Polar-Orbiting Partnership (SNPP), and NOAA-20 satellites (Loeb et al. 2018) while the monthly surface fluxes were computed using satellite-derived cloud and aerosol properties and temperature reanalysis-derived specific humidity profiles (Kato et al. 2018). The surface fluxes are therefore more prone to error and TOA fluxes were used to better constrain the surface fluxes (Kato et al. 2018).
Monthly averaged air temperature data at the horizontal resolution of 0.5° latitude and 0.625° longitude and vertical resolution of 72 hybrid sigma-pressure levels from the MERRA-2 reanalysis product (Gelaro et al. 2017) were used to compute the lower tropospheric stability (LTS) defined as the difference in potential temperature at between 800 hPa and at 2 m above the surface. Model results are compared with satellite observations using the MODIS simulator part of the first version of the Cloud Feedback Model Intercomparison (CFMIP) Observation Simulator Package (COSP) (Bodas-Salcedo et al. 2011) implemented in GEOS-5. The COSP framework allows for “apples-to-apples” comparisons between the model and the satellite observations applying a subcolumn generator within the model grid cell to mimic pixel-level satellite retrievals, and uses these samples to construct statistics of cloud properties.
3. Results
a. Comparison of INP concentrations
Figure 1 shows the average nINP and
Arctic average vertical profiles of INP (left) concentration and (right) immersion freezing rates for the four different immersion freezing parameterizations in rows, for (a) winter (DJF), (b) spring (MAM), (c) summer (JJA), and (d) autumn (SON), and (e) annually averaged. The right axes show the average temperature from all simulations at each level.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
For the four schemes, nINP increases up to a maximum at around 240 K, in agreement with observations suggesting that more INPs become activated at lower temperatures (Murray et al. 2012). The step reduction in nINP for T < 240 K is likely the result of both a decrease in supercooled water droplets that subsequently froze and removed INPs and the formation of cirrus clouds instead of mixed-phase clouds at T > 235 K. On the other hand, the microphysical tendency,
Table 2 compares surface-level nINP from the four different parameterizations against measurements taken in the Canadian high Arctic in Alert (82.5°N, 62.5°W) during the spring season reported by Si et al. (2019). Due to the fact that SST and aerosol emissions were fixed to avoid confounding their influence with immersion freezing, these comparisons and subsequent comparisons with observations are meant as order-of-magnitude estimates to evaluate the credibility of the model. Table 2 also shows the integral of the immersion freezing rate over the model time step,
Mean spring (MAM) INP concentration (L−1) for the Canadian High Arctic (82.5°N, 62.5°W) calculated by direct application of the immersion freezing parameterizations at the surface (defined as p > 980 hPa) and by integration of the immersion freezing rates over the model time step at cloud level. The INP concentration includes the average within ±1°C at each isotherm. Also shown are daily values taken during the 2016 spring season reported by Si et al. (2019) and are reported in units of L−1.
As expected, nINP and nINP,int for BN09 are much larger than the observed nINP since the parameterization was implemented such that it does not take into account the distribution of freezing efficiencies on the INP population. M92 also produces nINP and nINP,int values much larger than observed since it is based on measurements in an INP-rich environment. U17 and D10 show more interesting behavior. The nINP in D10 is of the same order of magnitude of the observations at −25°C, but nINP,int tends to be underestimated at all isotherms where the observations are available. On the other hand, nINP in U17 is overestimated, but nINP,int is comparable to the field values. Parameterizations based on laboratory measurements such as U17 are expected to produce higher nINP when applied without modification because the effects of the availability of the liquid phase and the dynamics are not taken into account. However, when accounting for these factors in the form expressed in Eq. (3), U17 does result in INP concentrations comparable to observations. As D10 is obtained from in- and near-cloud measurements, it is likely that in situ effects are implicit in the parameterization and it is likely more representative of the cloud environment compared to the other three parameterizations. The strong effect of temperature on nINP and
b. Impact of immersion freezing on cloud properties and radiation
The monthly-averaged LWPs generated by the MODIS simulator from March to September in all four simulations are compared against the compatible MODIS (Pincus et al. 2012) product designed for comparison with the MODIS simulator in Figs. 2a–d. These comparisons are meant to roughly compare the statistical similarities between the model and retrievals, rather than matching specific retrievals during a coincident time period. It should also be noted that the MODIS LWP as well as cloud optical property retrievals are compromised at high latitudes due to large solar zenith angles (Seethala and Horvath 2010; Grosvenor and Wood 2014). The range of simulated average Arctic LWPs obtained from the MODIS simulator is narrow, spanning from ∼78 to ∼81 g m−2, and the spatial distribution of LWP is also similar among the simulations. However, the magnitude and spatial distribution of LWP in all of the simulations differ from the MODIS observations which averages ∼184 g m−2 over the Arctic during the same seasons. This discrepancy likely derives from differences in the partitioning between rain and cloud in the model and in the retrieval. To test this, the simulations were also compared against the MAC-LWP product (Elsaesser et al. 2017) (Fig. 2f). This comparison reveals that the magnitude and spatial distributions of LWP are within the range of the modeled values, with an average of 78 g m−2 and maxima occurring over the North Atlantic and Bering Sea. The partial annual cycle of LWP (Fig. 2g) is also in agreement with the MAC-LWP product, increasing steadily from the spring until a maximum is reached during the summer months.
(a)–(d) Arctic projections of LWP produced by the MODIS simulator, averaged from March to September from the GEOS-5 model using the four different immersion freezing parameterizations described in Table 1. Observations from the (e) MODIS simulator product and (f) MAC-LWP are shown. (g) Observed average LWP and MODIS simulator LWP from March to September. (h) Simulated LWP and ground-based microwave observations of LWP from Barrow, Alaska. Error bars represent one standard deviation.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
An evident feature upon comparing the magnitude and annual cycle of LWP is that the simulations can be roughly grouped into two categories, with BN09 and M92 forming a high nimm,int category, which will be referred to as HighNINP,int, and U17 and D10 forming a low nimm,int category, which will be referred to as LowNINP,int. The tendency for the four simulations to group into the two categories is also evident in several cloud properties. HighNINP,int simulations have higher average LWP values overall compared to the LowNINP,int simulations despite producing larger immersion freezing rates resulting from enhanced nINP,int at T > 250 K. This is at odds with previous studies where excess INPs were found to deplete cloud liquid via the WBF process (Liu et al. 2011; Xie et al. 2013; Tan and Storelvmo 2016). This counterintuitive result can be explained by the cloud microphysical tendencies (Fig. 3) and is attributed to a combination of the following three factors: the large INP concentrations in HighNINP,int, particularly in the lower troposphere near the melting line (Fig. 1); the weighted grid-mean vapor pressure following Fu and Hollars (2004) to account for the subgrid variability of liquid and ice pockets in mixed-phase clouds within a GCM grid cell; and the faster sedimentation and melting rates in HighNINP,int. In the presence of large INP concentrations, the large-scale condensation rate which is dictated by Eq. (2) in the macrophysics module of GEOS-5 is lowered so that excess condensate is produced. The rapid ice sedimentation and melting rates in HighNINP,int respectively sediment and melt large amounts of cloud ice that is present in the lower troposphere where p > ∼800 hPa (just below ∼2 km), thus leading to large LWPs particularly in the summer when atmospheric moisture content is relatively high (left column of Fig. 3). This emphasizes both the need to exercise caution when extrapolating empirical parameterizations outside their valid temperature ranges since it may lead to large nINP values at high temperatures (T >260 K), where they can easily glaciate the cloud and also the sensitivity of cloud macrophysical properties to nonlinear interactions of multiple cloud microphysical processes.
Cloud microphysical mass tendencies for (left) liquid and (right) ice for the JJA season. The source terms for cloud liquid include condensation (COND), convective detrainment (DCNVL), and melting (MELT). The sink terms for cloud liquid include ice nucleation (ICENUC), the WBF process for ice crystals and snow (WBF and WBFSNOW), autoconversion (AUT), accretion by snow (ACRLSNOW), and accretion by rain (ACRLRAIN). The source terms for cloud ice include ice nucleation (ICENUC), ice crystal growth via deposition (DEP), the WBF process for ice crystals (WBF), secondary ice production via the Hallett–Mossop process (HM), and convective detrainment (DCNVI). The sink terms for cloud ice include autoconversion (AUTICE), melting (MELT), and accretion by snow (ACRISNOW) and sedimentation (SDM).
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
The cloud ice water path (IWP) counterpart of Fig. 2 is displayed in Fig. 4. Note that the IWP includes contributions from cirrus clouds in addition to the ice water path from mixed-phase clouds, and represents measurements at cloud top. Here, the IWP in HighNINP,int simulations is larger than in LowNINP,int and is consistent with larger INP concentrations in the former. Figure 3 confirms the primary role of ice nucleation, with smaller contributions from the WBF process and deposition as the primary sources of the IWP with the exception of D10, which shows a larger contribution from the WBF process at most heights. The faster WBF process in D10 is discussed toward the end of this section. The spatial distribution of IWP, however is similar among the simulations, exhibiting maxima near the storm track regions. The magnitude, spatial distributions, and annual cycle of IWP compare relatively well to the MODIS retrieval. IWP values derived from the model without a satellite simulator are shown in Fig. 4g. These values are smaller than those obtained from the MODIS simulator, likely due to the fact that the model frequently produces cloud tops with medium optical thicknesses such that the MODIS simulator classifies the entire column as ice cloud despite the fact that liquid cloud and precipitation could exist below it. Conversely, if MODIS detects optically thin ice clouds at cloud top but optically thick liquid clouds exist below the cloud top, then MODIS will assume that the entire column consists of liquid clouds and precipitation. IWP also shows little variation between the simulations; only D10 results in a statistically significant difference from the other schemes based on a two-sided Student’s t test. Although counterintuitive since D10 results in lower nINP (Fig. 1), it also has the most efficient WBF process in the summer months (Fig. 3). This occurs because D10 also maintains a larger amount of supercooled water (discussed below), which favors the WBF process since more available liquid is available to convert to cloud ice. The slight underestimation between the ground-based data and the model likely originates from model sampling and from the fixed SSTs and aerosol emissions in the simulations. HighNINP,int generally produces higher IWP during summer, but it decreases below those of LowNINP,int in the nonsummer months. This “reversal” pattern in the simulations during the transition from spring to summer is explained by enhanced deposition (the solid phase counterpart of condensation) in HighNINP,int.
(a)–(d) Arctic projections of IWP produced by the MODIS simulator, averaged from March to September, from the GEOS-5 model using the four different immersion freezing parameterizations described in Table 1. (e) Observations from the MODIS simulator product. (f) Annual cycle of simulated and observed average Arctic IWP from March to September. (g) Annual cycle of simulated IWP and ground-based IWP from Barrow, Alaska. Error bars represent one standard deviation.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
LTS exerts a strong control over cloud properties, particularly low-cloud fraction. The annual cycle of LTS during polar day, is shown alongside the low-level, midlevel, and high cloud fractions generated by the MODIS simulator in Fig. 5. Here, the three cloud categories are defined according to the cloud-top pressure (CTP) thresholds of CTP > 680 hPa, 440 hPa < CTP < 680 hPa, and CTP < 440 hPa for low, midlevel, and high clouds, respectively. LTS from MERRA-2 is also displayed in Fig. 5, as are the combined Terra/Aqua MODIS low, middle, and high cloud fractions. The simulated low- and middle-level cloud fraction are within 10% of the MODIS observations during the summer months but overestimate the observations during spring. A shortcoming of the MODIS retrieval, however, is that it is difficult for the instrument to identify clouds over bright ice-covered surfaces, namely in the Arctic, due to the lack of contrast in reflectance between the clouds and the surface (King et al. 2004). This limitation is particularly relevant to thin cirrus clouds and may explain the overestimate in high cloud fraction in GEOS-5 compared to the MODIS observations. Broken cloud pixels and cloud edges within the MODIS instrument’s 1-km field of view also contribute to the discrepancy between the model and observations (King et al. 2004; Platnick et al. 2016).
Annual cycle of (a) LTS, (b) MODIS simulator low-cloud fraction, (c) MODIS simulator middle cloud fraction, and (d) MODIS simulator high cloud fraction from March to September. LTS computed from MERRA-2 reanalysis are included in (a) and combined Terra/Aqua MODIS data are included in (b)–(d). Error bars represent one standard deviation.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
Stronger LTS occurs in early spring before sea ice melt begins and corresponds to the relatively smaller values of low-cloud fraction. This negative correlation between LTS and low-cloud fraction arises from the decoupling of the surface to the cloud layer, which deprives low clouds from surface moisture fluxes, and is exhibited in all the simulations. Moisture inversions above cloud-top may supply moisture to the cloud layer (Solomon et al. 2011); however, the vertical resolution of GEOS-5 (about 10 hPa near the surface) may be insufficient for resolving these moisture inversions. As spring transitions to summer, sea ice melts and weakens LTS, which in turn supplies the cloud layer with a source of moisture from the underlying ocean. This situation is analogous to coupled boundary layers in the lower latitudes where strong LTS enhances low-cloud cover (Klein and Hartmann 1993). This strong control of LTS over low-cloud cover can explain the “reversal pattern” of low-cloud fraction observed during the transition from spring to summer; whereas LTS is weaker during spring in the HighNINP,int (BN09 and M92) than in the LowNINP,int (U17 and D10) simulations, the situation reverses during in the summer. The strong influence of LTS on low-cloud cover has been shown using satellite observations (Taylor et al. 2015) and models (Barton et al. 2014), including models participating in CMIP5 (Taylor et al. 2019).
Inspection of the microphysical tendency rates (Fig. 3) also elucidates the spring-to-summer reversal pattern of low-cloud fraction, IWP, and LWP. Due to the limited ice production in LowNINP,int simulations, and the strong T of dependency of nINP (Fig. 1), the relatively few newly formed ice crystals near the cloud top interact with the liquid-rich environment as they settle, efficiently producing snow and enhancing accretion and WBF rates (ACRLSNOW and WBF). Such an enhancement is however limited in HighNINP,int simulations since high nINP values limit ice autoconversion rates and deplete water, inhibiting accretion processes. Thus during summer, sinks of cloud ice and liquid are more efficient in LowNINP,int than in HighNINP,int simulations. During winter accretion is inhibited (not shown) and WBF is more efficient in LowNINP,int simulations since they maintain a higher amount of supercooled water. In other words, the low nINP and its stronger dependency on temperature in D10 and U17 lead to more processes that are sinks of cloud liquid and ice during the polar day than in BN09 and M92, reversing the pattern found in winter. For the same reason, LowNINP,int exhibit larger midlevel cloud fraction year-round than HighNINP,int simulations, since slow removal processes allow the cloud to build up during winter in the former. Warmer clouds are preferentially scavenged during summer hence reducing low-level cloud more efficiently than midlevel clouds. The relatively large midlevel cloud fraction in D10 is consistent with previous studies, and was attributed to complex interactions between cloud macrophysics, microphysics, and the large-scale environment (Xie et al. 2013).
An important cloud property specific to mixed-phase clouds is the ratio of the mass of supercooled liquid to the total mass of condensate, referred to as the supercooled liquid fraction (SLF), and has been shown to have important implications for mixed-phase cloud feedbacks (McCoy et al. 2014; Tan et al. 2016; Frey and Kay 2018; Tan et al. 2019). Arctic histograms of the frequency of occurrence of SLF as a function of air temperature are displayed for the four simulations in Fig. 6. A common metric used to approximate the degree of glaciation of mixed-phase clouds is T50, which represents the temperature at which half of all mixed-phase clouds glaciate (Naud and Genio 2006; McCoy et al. 2016; Coopman et al. 2020). Defining T50 as the temperature at which the maximum frequency of occurrence of SLF is equal to 0.5, the T50 values for BN09, M92, U17 and D10 are −8°, −8°, −22°, and −24°C, respectively. Although these T50 values were not generated by a satellite simulator and are thus not directly comparable to satellite observations, a previous study that took model and satellite differences into account used the CAM5 with MG08 microphysics to show that M92 scheme tends to underestimate SLFs relative to Arctic satellite observations (Tan and Storelvmo 2019). Using the same model and satellite comparison method, the D10 scheme was also shown to generally improve SLF when implemented in a number of different GCMs (Komurcu et al. 2014).
Histograms of Arctic monthly SLF and air temperature normalized by the maximum frequency of occurrence for the (a) BN09, (b) M92, (c) U17, and (d) D10 simulations. Each SLF bin represents a 0.05 increment while each air temperature bin represents a 1°C increment. Also shown are vertical profiles of Arctic SLF in (e) winter, (f) spring, (g) summer, and (h) fall.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
The plateau in the maximum frequency of SLF shown in all of histograms of Fig. 6 suggests a complex cloud structure for Arctic clouds. This feature occurs due to the appearance of a liquid-rich cloud layer at high levels, near cloud top, shown in Fig. 7. The modulation of ice production by the liquid phase in immersion freezing restricts ice production not only to low temperatures, but also to the presence of liquid. Primary ice production rates are thus higher in the midaltitude levels, around 550 hPa (corresponding to an altitude of ∼4 km), as shown in Fig. 3 (ICENUCL). After freezing, the new ice crystals fall quickly and scavenge droplets in the lower layers, as evidenced by high sedimentation and accretion rates (ACRLSNOW and SDM in Fig. 3). This however leaves behind a layer of supercooled liquid between 400 and 500 hPa (corresponding to an altitude of 4–5.5 km) evidenced in Fig. 7, in agreement with the average cloud-top height in the Arctic (Intrieri et al. 2002). As expected, such a liquid-on-top-of-ice structure is favored when nINP is low and highly dependent on temperature and hence more prominent in the LowNINP,int simulations (D10 and U17). This is a significant result: by allowing the liquid phase to modulate ice nucleation via immersion freezing, GEOS-5 is able to reproduce the unique vertical structure of Arctic clouds, in agreement with observations and detailed modeling studies (de Boer et al. 2010; Barrett et al. 2017).
Average supercooled liquid fraction (SLF), including snow and rain, for the JJA season for the simulations described in Table 1.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
The impact of immersion freezing on the average Arctic droplet effective radius is shown in Fig. 8a. LowNINP,int simulations tend to produce larger radii than HighNINP,int, and generally within a few micrometers of that observed by MODIS (Figs. 8a,b). The larger cloud droplets in LowNINP,int may result from a more efficient removal of small droplets near cloud base by snow and ice. Only D10 produces smaller ice crystals than the other simulations, although the difference is not statistically significant based on a two-sided Student’s t test (Fig. 8b). This results from high ice sedimentation and accretion rates that preferentially eliminate large ice crystals within the clouds.
Average Arctic MODIS simulator (a) droplet effective radius, (b) ice crystal effective radius, (c) LWP, (d) IWP, (e) cloud optical thickness (COT) for liquid clouds, (f) COT for ice clouds, and (g) total COT for the four simulations and MODIS observations from March to September. MAC-LWP average LWP is represented by the orange bar. Error bars represent one standard deviation.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
Overall, since cloud optical thickness is proportional to water path and inversely proportional to effective radius, the smaller droplet effective radii (Figs. 8a,b) and larger water paths (Figs. 8c,d) in HighNINP,int contribute to their larger cloud optical thicknesses compared to LowNINP,int (Figs. 8e–g). This result is expected based on the larger LWP and IWP observed in the HighNINP,int simulations in Figs. 2 and 4 and contrasts with the brighter midlevel and high clouds that D10 produced relative to M92 in the CAM5 model (Xie et al. 2013). Although both CAM5 and the version of GEOS-5 employed in this study implement MG08 microphysics, the differences in the WBF process can be attributed to the more efficient WBF process for ice crystals and scavenging process in D10 compared to M92 in GEOS-5, whereas ice crystal growth was attributed to a slowdown of the WBF process in CAM5 (Xie et al. 2013). In the version of GEOS-5 employed in this study, the accelerated WBF process in D10 is the result of enhanced ice crystal number concentration that is particularly prominent in autumn. The faster WBF process that enhanced ice growth at upper levels coupled with the faster ice sedimentation rate near the surface in D10 (Fig. 3) contributed to reduced cloud optical thickness in D10.
The enhanced cloud optical thickness in HighNINP,int simulations results in a more negative TOA shortwave cloud radiative effect (SWCRE) (Fig. 9a). In contrast, only D10 shows a statistically significant higher longwave cloud radiative effect (LWCRE) at the TOA than the other simulations. This is associated with the larger middle cloud fraction produced in D10 (Fig. 5c), which corroborates similar results with the CAM5 model (Xie et al. 2013). Overall, the LowNINP,int simulations resulted in less negative net CREs and were in better agreement with observations from CERES. At the surface, the optically thicker low-level clouds in HighNINP,int compared to LowNINP,int resulted in more negative surface SWCREs due to the enhanced reflection of sunlight during polar day (Fig. 8d). Surface SWCRE was calculated weighting by surface albedo, hence accounting for the impact of sea ice on the clear-sky radiative fluxes. The surface LWCRE tended to be more positive in HighNINP,int than in LowNINP,int simulations due to the optically thicker clouds in the former. A notable exception is the U17 parameterization that results in stronger surface LWCRE due to optically thicker midlevel cloud fraction. Overall, the negative surface SWCRE outweighed the positive surface LWCRE, and the HighNINP,int simulations produced excessive cooling at the Arctic surface during polar day as a result of its optically thick and extensive cloud cover.
Arctic average (a) TOA SWCRE, (b) TOA LWCRE, (c) TOA net CRE, (d) surface SWCRE, (e) surface LWCRE, and (f) surface net CRE for the simulations described in Table 1 and CERES observations. Note that the TOA radiative fluxes in all panels include data only from March to September in order to consistently compare the cloud radiative effects with the corresponding cloud properties in Fig. 7. Error bars represent one standard deviation.
Citation: Journal of Climate 35, 13; 10.1175/JCLI-D-21-0368.1
4. Discussion
Two different types of interesting behavior were exhibited in the four parameterizations. The first concerns the tendency for the four simulations to exhibit cloud properties that can broadly be grouped into two categories. The cloud properties including liquid and ice water path and droplet and ice crystal effective radii, as well as TOA and surface CREs produced by the CNT-based BN09 and the aerosol-independent M92 parameterizations that comprise HighNINP,int, are very similar to each other in terms of their magnitude and distribution despite the much larger INP concentration in BN09 relative to M92, although there are statistically significant differences in the cloud properties between the two parameterizations. Several factors contribute to this behavior. For p < 450 hPa (above ∼6.5 km), immersion freezing rates are limited by the availability of cloud droplets, so that
There are larger and statistically significant differences in the cloud properties produced by the laboratory-based U17 and the aircraft in situ-based D10 parameterizations that comprise LowNINP,int. The small nINP predicted by these parameterizations allows the cloud to maintain substantial SLF at almost all levels. As such, cloud properties are highly influenced by immersion freezing. This is particularly notable in the larger midlevel cloud fraction and smaller mean ice crystal effective radius in D10 than in U17. The former results from the slow rate of ice nucleation in D10 that limits removal processes extending the cloud lifetime during winter and early spring. The same low nINP in D10 however results in active accretion by snow, and large sedimentation rates during the summer season that preferentially remove large particles, and lead to small ice crystal radii. Larger LWP combined with smaller cloud droplet effective radii cause the low-cloud optical thickness to be larger in HighNINP,int than in LowNINP,int simulations. TOA and surface SWCRE are therefore more negative, and the surface LWCRE more positive, in the former.
The second interesting behavior exhibited in the four parameterizations is the “reversal” pattern in the relative magnitudes of water path and low-cloud fraction that occurred as the simulations transitioned from spring to summer. During winter and early spring LowNINP,int simulations exhibit higher LWP and IWP than HighNINP,int, whereas this was reversed during the summer season. This behavior results from the stronger T dependency of nINP in the LowNINP,int simulations, which limits ice nucleation to the top cloud layers during summer. The newly formed ice crystals grow quickly at the expense of the readily available supercooled water, triggering more efficient removal of cloud liquid and ice processes during summer in the LowNINP,int than in the HighNINP,int simulations.
The impact of the M92 and D10 parameterizations on cloud and radiation properties has been previously studied using the CAM5 model by Xie et al. (2013). The authors reported a smaller mean LWP during polar day when using the M92 scheme compared to the D10 scheme, which is at odds with the results shown in Fig. 8c. This discrepancy can be explained by enhanced condensation in the presence of high INP concentrations during summer in GEOS-5 in M92 due to the implementation of the Fu and Hollars (2004) scheme in the presence of high INP concentrations. Furthermore, accretion and sedimentation rates are enhanced in D10 compared to M92, reducing LWP in the former. Therefore, although the WBF process and ice nucleation undoubtedly exert important controls over cloud properties and radiation, the evolution of Arctic mixed-phase clouds can be only explained by holistically considering the complete set of microphysical processes that influence the clouds. Another significant difference between the study of Xie et al. (2013) and this study is that in CAM5 M92 and D10 represent the efficiency of INP in the condensation and deposition modes as opposed to the emphasis on immersion placed here. CAM5 does account for immersion freezing using a T-dependent parameterization, however using the Bigg (1953) scheme, which accounted for less than 10% of the total active INP concentration (English et al. 2014). Thus in CAM5, primary ice formation is not modulated by the liquid phase. Active INPs are thus readily available throughout the cloud, with their concentration increasing exponentially with decreasing temperature, which may lead to efficient cloud glaciation.
5. Summary
Ice nucleation is fundamental to the formation of Arctic mixed-phase clouds, yet there is a large variation in how they are parameterized in GCMs. The focus of this study is on the impact of ice nucleation via the immersion freezing mode, which is likely the most relevant mechanism of ice formation in mixed-phase clouds. In particular, the impacts of four different immersion ice nucleation parameterizations implemented in the NASA GEOS-5 model on cloud properties and radiation at the TOA and surface during polar day, from March to September, were investigated. The four parameterizations were selected to represent a wide range of variability in active INP concentrations. They were derived from a number of different methods, including classical nucleation theory (CNT), in situ field campaign measurements, and laboratory studies. The GEOS-5 model was run with SSTs and sea ice fixed to the 1979–2019 climatology of MERRA-2 and black carbon emissions and sulfate precursors were held constant at 2005 levels in order to isolate the impact of ice nucleation on clouds and radiation.
Simulations performed with the four parameterizations resulted in cloud properties broadly grouped by their spatial and temporal similarities into two categories, with BN09 and M92 comprising HighNINP,int and U17 and D10 comprising LowNINP,int. Immersion freezing rates were much larger in HighNINP,int, but came close to LowNINP,int simulations at p < 500 hPa (corresponding to ∼5 km in altitude), where they were limited by the availability of water, a necessary condition for immersion ice nucleation.
In terms of differences within the two categories, BN09 and M92 showed similar cloud properties overall despite large differences in INP concentrations and immersion freezing rates. At p < 450 hPa (above ∼6.5 km) this was the result of the limiting effect of cloud droplet number on immersion freezing rates. At lower levels it was attributed to the glaciation of the cloud by the abundant INP, which itself limits further freezing by immersion. This “saturation effect,” where the presence of more INP does not significantly affect cloud properties, seems to occur for INP concentrations roughly exceeding 5 L−1. Differences between cloud properties and radiation were larger within LowNINP,int. In particular, D10 resulted in a higher midlevel cloud fraction in early spring and smaller ice crystal effective radius during summer than the other simulations. This was attributed to the impact of low INP concentrations limiting accretion rates by snow during winter but enhancing them during summer. A high midlevel cloud fraction in simulations using D10 in GEOS-5 is consistent with previous studies using CAM5 (Xie et al. 2013; English et al. 2014).
LowNINP,int produced larger IWP, LWP, and low-level cloud fraction during winter and early spring than HighNINP,int simulations, whereas the opposite occurred during the summer season. LTS was strongly negatively correlated with low-cloud fraction in all simulations, but tended to be higher in early spring in HighNINP,int than in LowNINP,int simulations, but vice versa during summer. This explains in part the reversal pattern in the magnitude of the cloud properties between the two categories of simulations after May. These differences were caused by the effect of ice nucleation on ice production, liquid condensational growth, and accretion and WBF processes on snow and ice. The strong effect of temperature on INP concentrations in LowNINP,int simulations played a key role in the enhancement of ice cloud scavenging and melting processes during the summer season. The mean droplet effective radii in HighNINP,int was also lower than that in LowNINP,int, due to faster accretion rates by rain in the lower troposphere during summer. Collectively, the smaller average droplet effective radii, larger water paths and larger overall low-cloud fractions in HighNINP,int than in LowNINP,int simulations led to more negative TOA and surface SWCRE, as well as more positive LWCRE during polar day. The SWCRE outweighed the LWCRE contributions at the TOA and surface, resulting in net negative radiative fluxes that were larger in magnitude in the HighNINP,int simulations.
The SLF was also different between the two groups of simulations, with LowNINP,int maintaining a higher SLF at p < 800 hPa (just below ∼2 km). However all simulations resulted in a characteristic cloud structure, with a layer rich in supercooled water near the cloud top and ice and snow at lower levels. This liquid-on-top-of-ice structure is in agreement with field observations of Arctic clouds and results from the strong modulation of the ice formation by the liquid phase in immersion freezing, which forces immersion freezing rates to peak at midcloud levels, leaving a layer of supercooled water at cloud top.
To conclude, although previous studies have emphasized ice nucleation and the WBF process in determining mixed-phase cloud properties, this study emphasizes that the interaction between ice nucleation and a number of different cloud microphysical processes and the stability of the lower troposphere can also ultimately determine cloud properties and their impact on the Arctic’s radiation budget at the TOA and the surface. In particular, accretion and sedimentation in addition to the WBF process appear to play important roles in determining the influence of ice nucleation on clouds and radiation. The LowNINP,int simulations with relatively low INP concentrations produce a more realistic representation of Arctic clouds and radiation. Although the simulations in this study are not meant to exactly reproduce observations due to the fixed SSTs and aerosol emissions in the model, LowNINP,int simulations are also in better agreement with available satellite retrievals and in situ data. Importantly, the implementation of the various ice nucleation parameterizations as immersion freezing schemes allowed GEOS-5 to successfully reproduce the observed vertical cloud structure characteristic of Arctic mixed-phase clouds. Future efforts should be dedicated to understanding the role of the nonlinear interactions between cloud microphysics and ice nucleation, their role on the Arctic cloud feedback and the evolution of SSTs, and their ultimate impact on Earth’s climate.
Acknowledgments.
The effort of I. T. was supported by NASA Grant 80NSSC18K1599. D.B. was supported by the NASA Modeling and Analysis Program, Grant 16-MAP16-0085. Resources supporting this work were provided by the NASA High-End Computing (HEC) Program through the NASA Center for Climate Simulation (NCCS) at Goddard Space Flight Center. I. T. thanks Q. Coopman for his help with the analysis of in situ data for this work.
Data availability statement.
The GEOS-5 source code is available under the NASA Open Source Agreement at http://opensource.gsfc.nasa.gov/projects/GEOS-5/. All data generated in this work will be made publicly available through the NASA technical reports server (https://ntrs.nasa.gov) and PubSpace (https://www.nasa.gov/open/researchaccess/pubspace). Observations from MODIS are available for download at https://search.earthdata.nasa.gov/search?q=MCD06COSP. CERES observations were downloaded from https://ceres.larc.nasa.gov/data/ and the MERRA-2 Reanalysis dataset was downloaded from https://disc.gsfc.nasa.gov/. The MAC-LWP observations were downloaded from https://disc.gsfc.nasa.gov/datasets/MACLWP_mean_1/summary?keywords=mac-lwp. Ground-based observations of LWP and IWP were downloaded from https://www.arm.gov/data/.
REFERENCES
Abdul-Razzak, H., and S. Ghan, 2000: A parameterization of aerosol activation, 2. Multiple aerosol types. J. Geophys. Res., 105, 6837–6844, https://doi.org/10.1029/1999JD901161.
Bacmesiter, J. T., M. J. Suarez, and F. R. Robertson, 2006: Rain reevaporation, boundary layer–convection interactions, and Pacific rainfall patterns in an AGCM. J. Amer. Sci., 63, 3383–3403, https://doi.org/10.1175/JAS3791.1.
Barahona, D., 2018: On the thermodynamic and kinetic aspects of immersion ice nucleation. Atmos. Chem. Phys., 18, 17 119–17 141, https://doi.org/10.5194/acp-18-17119-2018.
Barahona, D., and A. Nenes, 2008: Parameterization of cirrus formation in large scale models: Homogeneous nucleation. J. Geophys. Res., 113, D11211, https://doi.org/10.1029/2007JD009355.
Barahona, D., and A. Nenes, 2009: Parameterizing the competition between homogeneous and heterogeneous freezing in cirrus cloud formation–monodisperse ice nuclei. Atmos. Chem. Phys., 9, 369–381, https://doi.org/10.5194/acp-9-369-2009.
Barahona, D., A. Molod, J. Bacmeister, A. Nenes, A. Gettelman, H. Morrison, V. Phillips, and A. Eichmann, 2014: Development of two-moment cloud microphysics for liquid and ice within the Nasa Goddard Earth Observing System Model (GEOS-5). Geosci. Model Dev., 7, 1733–1766, https://doi.org/10.5194/gmd-7-1733-2014.
Barahona, D., A. Molod, and H. Kalesse, 2017: Direct estimation of the global distribution of vertical velocity within cirrus clouds. Sci. Rep., 7, 6840, https://doi.org/10.1038/s41598-017-07038-6.
Barnes, E. A., and L. M. Polvani, 2015: CMIP5 projections of Arctic amplification, of the North American/North Atlantic circulation, and of their relationship. J. Climate, 28, 5254–5271, https://doi.org/10.1175/JCLI-D-14-00589.1.
Barrett, I., R. J. Hogan, and R. M. Forbes, 2017: Why are mixed-phase altocumulus clouds poorly predicted by large-scale models? Part 2. J. Geophys. Res. Atmos., 122, 9927–9944, https://doi.org/10.1002/2016JD026322.
Barton, N. P., S. A. Klein, and J. S. Boyle, 2014: On the contribution of longwave radiation to global climate model biases in Arctic lower tropospheric stability. J. Climate, 27, 7250–7269, https://doi.org/10.1175/JCLI-D-14-00126.1.
Beesley, J. A., 2000: Estimating the effect of clouds on the Arctic surface energy budget. J. Geophys. Res., 105, 10 103–10 117, https://doi.org/10.1029/2000JD900043.
Bennartz, R., and Coauthors, 2013: Greenland melt extent enhanced by low-level liquid clouds. Nature, 496, 83–86, https://doi.org/10.1038/nature12002.
Bergeron, T., 1935: On the physics of clouds and precipitation. Proces Verbaux de l’Association de Meteorologie. International Union of Geodesy and Geophysics, 156–178.
Bian, H., and Coauthors, 2013: Source attributions of pollution to the western Arctic during the NASA ARCTAS field campaign. Atmos. Chem. Phys., 13, 4707–4721, https://doi.org/10.5194/acp-13-4707-2013.
Bigg, E. K., 1953: The formation of atmospheric ice crystals by the freezing of droplets. Quart. J. Roy. Meteor. Soc., 79, 510–519, https://doi.org/10.1002/qj.49707934207.
Bodas-Salcedo, A., and Coauthors, 2011: COSP: Satellite simulation software for model assessment. Bull. Amer. Meteor. Soc., 92, 1023–1043, https://doi.org/10.1175/2011BAMS2856.1.
Boeke, R. C., and P. C. Taylor, 2016: Evaluation of the Arctic surface radiation budget in CMIP5 models. J. Geophys. Res. Atmos., 121, 8525–8548, https://doi.org/10.1002/2016JD025099.
Boeke, R. C., and P. C. Taylor, 2018: Seasonal energy exchange in sea ice retreat regions contributes to differences in projected Arctic warming. Nat. Commun., 9, 5017, https://doi.org/10.1038/s41467-018-07061-9.
Breen, K. H., D. Barahona, T. Yuan, H. Bian, and S. C. James, 2021: Effect of volcanic emissions on clouds during the 2008 and 2018 Kilauea degassing events. Atmos. Chem. Phys., 21, 7749–7771, https://doi.org/10.5194/acp-21-7749-2021.
Carn, S., L. Clarisse, and A. J. Prata, 2016: Multi-decadal satellite measurements of global volcanic degassing. J. Volcanol. Geotherm. Res., 311, 99–134, https://doi.org/10.1016/j.jvolgeores.2016.01.002.
Cesana, G., J. E. Kay, H. Chepfer, J. M. English, and G. de Boer, 2012: Ubiquitous low-level liquid-containing Arctic clouds: New observations and climate model constraints from CALIPSO-GOCCP. Geophys. Res. Lett., 39, 2012GL053385, https://doi.org/10.1029/2012GL053385.
Cesana, G., D. E. Waliser, X. Jiang, and J.-L.-F. Li, 2015: Multimodel evaluation of cloud phase transition using satellite and reanalysis data. J. Geophys. Res. Atmos., 120, 7871–7892, https://doi.org/10.1002/2014JD022932.
Chin, M., T. Diehl, O. Dubovik, T. F. Eck, B. N. Holben, A. Sinyuk, and D. G. Streets, 2009: Light absorption by pollution, dust, and biomass burning aerosols: A global model study and evaluation with AERONET measurements. Ann. Geophys., 27, 3439–3464 https://doi.org/10.5194/angeo-27-3439-2009.
Chylek, P., and C. Borel, 2004: Mixed phase cloud water/ice structure from high spatial resolution satellite data. Geophys. Res. Lett., 31, L14104, https://doi.org/10.1029/2004GL020428.
Colarco, P., A. da Silva, M. Chin, and T. Diehl, 2010: Online simulations of global aerosol distributions in the NASA GEOS-4 model and comparisons to satellite and ground-based aerosol optical depth. J. Geophys. Res., 115, D14207, https://doi.org/10.1029/2009JD012820.
Connolly, P. J., O. Möhler, P. R. Field, H. Saathoff, R. Burgess, T. Choularton, and M. Gallagher, 2009: Studies of heterogeneous freezing by three different desert dust samples. Atmos. Chem. Phys., 9, 2805–2824, https://doi.org/10.5194/acp-9-2805-2009.
Coopman, Q., J. Riedi, S. Zeng, and T. J. Garrett, 2020: Space-based analysis of the cloud thermodynamic phase transition for varying microphysical and meteorological regimes. Geophys. Res. Lett., 47, e2020GL087122, https://doi.org/10.1029/2020GL087122.
Cullather, R. I., Y. K. Lim, L. N. Boisvert, L. Brucker, J. N. Lee, and S. M. Nowicki, 2016: Analysis of the warmest Arctic winter, 2015–2016. Geophys. Res. Lett., 43, 10 808–10 816, https://doi.org/10.1002/2016GL071228.
Curry, J. A., and E. E. Ebert, 1992: Annual cycle of radiation fluxes over the Arctic Ocean: Sensitivity to cloud optical properties. J. Climate, 5, 1267–1280, https://doi.org/10.1175/1520-0442(1992)005<1267:ACORFO>2.0.CO;2.
Curry, J. A., W. B. Rossow, D. Randall, and J. L. Schramm, 1996: Overview of Arctic cloud and radiation characteristics. J. Climate, 9, 1731–1764, https://doi.org/10.1175/1520-0442(1996)009<1731:OOACAR>2.0.CO;2.
D’Alessandro, J. J., M. Diao, C. Wu, X. Liu, J. B. Jensen, and B. B. Stephens, 2019: Cloud phase and relative humidity distributions over the Southern Ocean in austral summer based on in situ observations and CAM5 simulations. J. Climate, 32, 2781–2805, https://doi.org/10.1175/JCLI-D-18-0232.1.
de Boer, G., E. W. Eloranta, and M. D. Shupe, 2009: Arctic mixed-phase stratiform cloud properties from multiple years of surface-based measurements at two high-latitude locations. J. Amer. Sci., 66, 2874–2887, https://doi.org/10.1175/2009JAS3029.1.
de Boer, G., T. Hashino, and G. J. Tripoli, 2010: Ice nucleation through immersion freezing in mixed-phase stratiform clouds: Theory and numerical simulations. Atmos. Res., 96, 315–324, https://doi.org/10.1016/j.atmosres.2009.09.012.
DeMott, P. J., and Coauthors, 2010: Predicting global atmospheric ice nuclei distributions and their impacts on climate. Proc. Natl. Acad. Sci. USA, 107, 11 217–11 222, https://doi.org/10.1073/pnas.0910818107.
Ehrlich, A., and Coauthors, 2019: A comprehensive in situ and remote sensing data set from the Arctic cloud observations using airborne measurements during polar day (ACLOUD) campaign. Earth Syst. Sci. Data, 11, 1853–1881, https://doi.org/10.5194/essd-11-1853-2019.
Elsaesser, G. S., C. W. O’Dell, M. D. Lebsock, R. Bennartz, T. J. Greenwald, and F. J. Wentz, 2017: The multisensor advanced climatology of liquid water path (MAC-LWP). J. Climate, 30, 10 193–10 210, https://doi.org/10.1175/JCLI-D-16-0902.1.
English, J., J. E. Kay, A. Gettelman, X. Liu, Y. Wang, Y. Zhang, and H. Chepfer, 2014: Contributions of clouds, surface albedos, and mixed-phase ice nucleation schemes to Arctic radiation biases in CAM5. J. Climate, 27, 5174–5197, https://doi.org/10.1175/JCLI-D-13-00608.1.
Fan, S.-M., 2013: Modeling of observed mineral dust aerosols in the Arctic and the impact on winter season low-level clouds. J. Geophys. Res. Atmos., 118, 11 161–11 174, https://doi.org/10.1002/jgrd.50842.
Fan, S.-M., and Coauthors, 2012: Inferring ice formation processes from global-scale black carbon profiles observed in the remote atmosphere and model simulations. J. Geophys. Res., 117, D23205, https:/doi.org/10.1029/2012JD018126.
Field, P. R., R. J. Hogan, P. R. A. Brown, A. J. Illingworth, T. W. Choularton, P. H. Kaye, E. Hirst, and R. Greenaway, 2004: Simultaneous radar and aircraft observations of mixed-phase cloud at the 100 m scale. Quart. J. Roy. Meteor. Soc., 130, 1877–1904, https://doi.org/10.1256/qj.03.102.
Findeisen, W., 1938: Kolloid-Meteorologische Vorgänge bei Neiderschlagsbildung. Meteor. Z., 55, 121–133.
Frey, W. R., and J. E. Kay, 2018: The influence of extratropical cloud phase and amount feedbacks on climate sensitivity. Climate Dyn., 50, 3097–3116, https://doi.org/10.1007/s00382-017-3796-5.
Fridlind, A. M., A. S. Ackerman, G. McFarquhar, G. Zhang, M. R. Poellot, P. J. DeMott, A. J. Prenni, and A. J. Heymsfield, 2007: Ice properties of single-layer stratocumulus during the Mixed-Phase Arctic Cloud Experiment: 2. Model results. J. Geophys. Res., 112, D24202, https://doi.org/10.1029/2007JD008646.
Fridlind, A. M., B. V. Diedenhoven, A. S. Ackerman, A. Avramov, A. Mrowiec, H. Morrison, P. Zuidema, and M. D. Shupe, 2012: A FIRE-ACE/SHEBA case study of mixed-phase Arctic boundary layer clouds: Entrainment rate limitations on rapid primary ice nucleation processes. J. Amer. Sci., 69, 365–389, https://doi.org/10.1175/JAS-D-11-052.1.
Fu, Q., and S. Hollars, 2004: Testing mixed-phase cloud water vapor parameterizations with SHEBA/FIRE–ACE observations. J. Amer. Sci., 61, 2083–2091, https://doi.org/10.1175/1520-0469(2004)061<2083:TMCWVP>2.0.CO;2.
Gaustad, K. L., and D. D. Turner, 2007: MWRRET value-added product: The retrieval of liquid water path and precipitable water vapor from microwave radiometer (MWR) datasets. Tech. Rep. DOE/SC-ARM/TR-081, 19 pp., https://doi.org/10.2172/948370.
Gelaro, R., and Coauthors, 2017: The Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2). J. Climate, 30, 5419–5454, https://doi.org/10.1175/JCLI-D-16-0758.1.
Gettelman, A., and Coauthors, 2010: Global simulations of ice nucleation and ice supersaturation with an improved cloud scheme in the Community Atmosphere Model. J. Geophys. Res., 115, D18216, https://doi.org/10.1029/2009JD013797.
Gettelman, A., H. Morrison, S. Santos, P. Bogenschutz, and P. M. Caldwell, 2015: Advanced two-moment bulk microphysics for global models. Part II: Global model solutions and aerosol–cloud interactions. J. Climate, 28, 1288–1307, https://doi.org/10.1175/JCLI-D-14-00103.1.
Grosvenor, D. P., and R. Wood, 2014: The effect of solar zenith angle on MODIS cloud optical and microphysical retrievals within marine liquid water clouds. Atmos. Chem. Phys., 14, 7291–7321, https://doi.org/10.5194/acp-14-7291-2014.
Hoose, C., and O. Möhler, 2012: Heterogeneous ice nucleation on atmospheric aerosols: A review of results from laboratory experiments. Atmos. Chem. Phys., 12, 9817–9854, https://doi.org/10.5194/acp-12-9817-2012.
Hoose, C., J. Kristjansson, J.-C. Chen, and A. Hazra, 2010: A classical-theory-based parameterization of heterogeneous ice nucleation by mineral dust, soot, and biological particles in a global climate model. J. Atmos. Sci., 67, 2483–2503, https://doi.org/10.1175/2010JAS3425.1.
Ickes, L., A. Welti, and U. Lohmann, 2017: Classical nucleation theory of immersion freezing: Sensitivity of contact angle schemes to thermodynamic and kinetic parameters. Atmos. Chem. Phys., 17, 1713–1739, https://doi.org/10.5194/acp-17-1713-2017.
Intrieri, J. M., M. D. Shupe, T. Uttal, and B. J. McCarty, 2002: An annual cycle of Arctic cloud characteristics observed by radar and lidar at SHEBA. J. Geophys. Res., 107, 8030, https://doi.org/10.1029/2000JC000423.
Jiang, H., W. R. Cotton, J. O. Pinto, J. A. Curry, and M. J. Weissbluth, 2000: Cloud resolving simulations of mixed-phase Arctic stratus observed during BASE: Sensitivity to concentration of ice crystals and large-scale heat and moisture advection. J. Amer. Sci., 57, 2105–2117, https://doi.org/10.1175/1520-0469(2000)057<2105:CRSOMP>2.0.CO;2.
Kärcher, B., 2003: Simulating gas–aerosol–cirrus interactions: Process-oriented microphysical model and applications. Atmos. Chem. Phys., 3, 1645–1664, https://doi.org/10.5194/acp-3-1645-2003.
Karlsson, J., and G. Svensson, 2011: The simulation of Arctic clouds and their influence on the winter surface temperature in present-day climate in the CMIP3 multi-model dataset. Climate Dyn., 36, 623–635, https://doi.org/10.1007/s00382-010-0758-6.
Kato, S., and Coauthors, 2018: Surface irradiances of edition 4.0 Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) data product. J. Climate, 31, 4501–4527, https://doi.org/10.1175/JCLI-D-17-0523.1.
Kawai, H., S. Yukimoto, T. Koshiro, N. Oshima, T. Tanaka, H. Yoshimura, and R. Nagasawa, 2019: Significant improvement of cloud representation in the global climate model MRI-ESM2. Geosci. Model Dev., 12, 2875–2897, https://doi.org/10.5194/gmd-12-2875-2019.
Kay, J. E., and A. Gettelman, 2009: Cloud influence on and response to seasonal Arctic sea ice loss. J. Geophys. Res., 114, D18204, https://doi.org/10.1029/2009JD011773.
Kay, J. E., and T. L’Ecuyer, 2013: Observational constraints on Arctic Ocean clouds and radiative fluxes during the early 21st century. J. Geophys. Res. Atmos., 118, 7219–7236, https://doi.org/10.1002/jgrd.50489.
Khairoutdinov, M., and Y. Kogan, 2000: A new cloud physics parameterization in a large-eddy simulation model of marine stratocumulus. Mon. Wea. Rev., 128, 229–243, https://doi.org/10.1175/1520-0493(2000)128<0229:ANCPPI>2.0.CO;2.
King, M. D., S. Platnick, P. Yang, G. T. Arnold, M. A. Gray, J. C. Riedi, S. A. Ackerman, and K.-N. Liou, 2004: Remote sensing of liquid water and ice cloud optical thickness and effective radius in the Arctic: Application of airborne multispectral MAS data. J. Atmos. Oceanic Technol., 21, 857–875, https://doi.org/10.1175/1520-0426(2004)021<0857:RSOLWA>2.0.CO;2.
Klein, S. A., and D. L. Hartmann, 1993: The seasonal cycle of low stratiform clouds. J. Climate, 6, 1587–1606, https://doi.org/10.1175/1520-0442(1993)006<1587:TSCOLS>2.0.CO;2.
Klein, S. A., and Coauthors, 2009: Intercomparison of model simulations of mixed-phase clouds observed during the ARM mixed-phase Arctic cloud experiment. I: Single-layer cloud. Quart. J. Roy. Meteor. Soc., 135, 979–1002, https://doi.org/10.1002/qj.416.
Knopf, D. A., and P. A. Alpert, 2013: A water activity based model of heterogeneous ice nucleation kinetics for freezing of water and aqueous solution droplets. Faraday Discuss., 165, 513–534, https://doi.org/10.1039/c3fd00035d.
Komurcu, M., and Coauthors, 2014: Intercomparison of the cloud water phase among global climate models. J. Geophys. Res. Atmos., 119, 3372–3400, https://doi.org/10.1002/2013JD021119.
Koop, T., B. Luo, A. Tsias, and T. Peter, 2000: Water activity as the determinant for homogeneous ice nucleation in aqueous solutions. Nature, 406, 611–614, https://doi.org/10.1038/35020537.
Korolev, A. V., G. A. Isaac, S. G. Cober, J. W. Strapp, and J. Hallett, 2003: Microphysical characterization of mixed-phase clouds. Quart. J. Roy. Meteor. Soc., 129, 39–65, https://doi.org/10.1256/qj.01.204.
Laaksonen, A., J. Malila, and A. Nenes, 2020: Heterogeneous nucleation of water vapor on different types of black carbon particles. Atmos. Chem. Phys., 20, 13 579–13 589, https://doi.org/10.5194/acp-20-13579-2020.
Ladino Moreno, L. A., O. Stetzer, and U. Lohmann, 2013: Contact freezing: A review of experimental studies. Atmos. Chem. Phys., 13, 9745–9769, https://doi.org/10.5194/acp-13-9745-2013.
Lance, S., and Coauthors, 2011: Cloud condensation nuclei as a modulator of ice processes in Arctic mixed-phase clouds. Atmos. Chem. Phys., 11, 8003–8015, https://doi.org/10.5194/acp-11-8003-2011.
Lin, Y.-L., R. D. Farley, and H. D. Orville, 1983: Bulk parameterization of the snow field in a cloud model. J. Appl. Meteor. Climatol., 22, 1065–1092, https://doi.org/10.1175/1520-0450(1983)022<1065:BPOTSF>2.0.CO;2.
Liu, X., S. Xie, and S. J. Ghan, 2007: Evaluation of a new mixed-phase cloud microphysics parameterization with CAM3 single-column model and M-PACE observations. Geophys. Res. Lett., 34, L23712, https://doi.org/10.1029/2007GL031446.
Liu, X., and Coauthors, 2011: Testing cloud microphysics parameterizations in NCAR CAM5 with ISDAC and M-PACE observations. J. Geophys. Res., 116, D00T11, https://doi.org/10.1029/2011JD015889.
Liu, Y., D. Daum, and S. Yum, 2006: Analytical expression for the relative dispersion of the cloud droplet size distribution. Geophys. Res. Lett., 33, L02810, https://doi.org/10.1029/2005GL024052.
Liu, Y., P. Daum, H. Guo, and Y. Peng, 2008: Dispersion bias, dispersion effect, and the aerosol–cloud conundrum. Environ. Res. Lett., 3, 045021, https://doi.org/10.1088/1748-9326/3/4/045021.
Loeb, N. G., and Coauthors, 2018: Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) top-of-atmosphere (TOA) edition-4.0 data product. J. Climate, 31, 895–918, https://doi.org/10.1175/JCLI-D-17-0208.1.
McCoy, D. T., D. L. Hartmann, and D. P. Grosvenor, 2014: Observed Southern Ocean cloud properties and shortwave reflection. Part II: Phase changes and low cloud feedback. J. Climate, 27, 8858–8868, https://doi.org/10.1175/JCLI-D-14-00288.1.
McCoy, D. T., I. Tan, D. Hartmann, M. D. Zelinka, and T. Storelvmo, 2016: On the relationships among cloud cover, mixed-phase partitioning, and planetary albedo in GCMs. J. Adv. Model. Earth Syst., 8, 650–668, https://doi.org/10.1002/2015MS000589.
Meyers, M. P., P. J. DeMott, and W. R. Cotton, 1992: New primary ice-nucleation parameterizations in an explicit cloud model. J. Appl. Meteor. Climatol., 31, 708–721, https://doi.org/10.1175/1520-0450(1992)031<0708:NPINPI>2.0.CO;2.
Mioche, G., and Coauthors, 2017: Vertical distribution of microphysical properties of Arctic springtime low-level mixed-phase clouds over the Greenland and Norwegian seas. Atmos. Chem. Phys., 17, 12 845–12 869, https://doi.org/10.5194/acp-17-12845-2017.
Mioche, G., O. Jourdan, M. Ceccaldi, and J. Delanoë, 2015: Variability of mixed-phase clouds in the Arctic with a focus on the Svalbard region: A study based on spaceborne active remote sensing. Atmos. Chem. Phys., 15, 2445–2461, https://doi.org/10.5194/acp-15-2445-2015.
Molod, A., L. Takacs, M. Suarez, J. Bacmeister, I. Song, and A. Eichmann, 2012: The GEOS-5 atmospheric general circulation model: Mean climate and development from MERRA to Fortuna. Tech. Rep. Series on Global Modeling and Data Assimilation, Vol. 28, NASA Goddard Space Flight Center, 115 pp.
Molod, A., and Coauthors, 2020: GEOS-S2S version 2: The GMAO high-resolution coupled model and assimilation system for seasonal prediction. J. Geophys. Res. Atmos., 125, https://doi.org/10.1029/2019JD031767.
Morrison, H., and A. Gettelman, 2008: A new two-moment bulk stratiform cloud microphysics scheme in the Community Atmosphere Model, version 3 (CAM3). Part I: Description and numerical tests. J. Climate, 21, 3642–3659, https://doi.org/10.1175/2008JCLI2105.1.
Morrison, H., G. D. Boer, G. Feingold, J. Harrington, M. D. Shupe, and K. Sulia, 2012: Resilience of persistent Arctic mixed-phase clouds. Nat. Geosci., 5, 11–17, https://doi.org/10.1038/ngeo1332.
Muhlbauer, A., T. P. Ackerman, J. M. Comstock, G. S. Diskin, S. M. Evans, R. P. Lawson, and R. T. Marchand, 2014: Impact of large-scale dynamics on the microphysical properties of midlatitude cirrus. J. Geophys. Res. Atmos., 119, 3976–3996, https://doi.org/10.1002/2013JD020035.
Murphy, D., and T. Koop, 2005: Review of the vapour pressures of ice and supercooled water for atmospheric applications. Quart. J. Roy. Meteor. Soc., 131, 1539–1565, https://doi.org/10.1256/qj.04.94.
Murray, B. J., S. L. Broadley, T. W. Wilson, J. D. Atkinson, and R. H. Wills, 2011: Heterogeneous freezing of water droplets containing kaolinite particles. Atmos. Chem. Phys., 11, 4191–4207, https://doi.org/10.5194/acp-11-4191-2011.
Murray, B. J., D. O’Sullivan, J. D. Atkinson, and M. E. Webb, 2012: Ice nucleation by particles immersed in supercooled cloud droplets. Chem. Soc. Rev., 41, 6519–6554, https://doi.org/10.1039/c2cs35200a.
Naud, C., and A. D. D. Genio, 2006: Observational constraints on the cloud thermodynamic phase in midlatitude storms. J. Climate, 19, 5273–5288, https://doi.org/10.1175/JCLI3919.1.
O’Dell, C. W., F. J. Wentz, and R. Bennartz, 2008: Cloud liquid water path from satellite-based passive microwave observations: A new climatology over the global oceans. J. Climate, 21, 1721–1739, https://doi.org/10.1175/2007JCLI1958.1.
Ovchinnikov, M., and Coauthors, 2014: Intercomparison of large-eddy simulations of Arctic mixed-phase clouds: Importance of ice size distribution assumptions. J. Adv. Model. Earth Syst., 6, 223–248, https://doi.org/10.1002/2013MS000282.
Pincus, R., S. Platnick, S. A. Ackerman, R. S. Hemler, and R. J. P. Hofmann, 2012: Reconciling simulated and observed views of clouds: MODIS, ISCCP, and the limits of instrument simulators. J. Climate, 25, 4699–4720, https://doi.org/10.1175/JCLI-D-11-00267.1.
Pinto, J. O., 1998: Autumnal mixed-phase cloudy boundary layers in the Arctic. J. Amos. Sci., 55, 2016–2038, https://doi.org/10.1175/1520-0469(1998)055<2016:AMPCBL>2.0.CO;2.
Platnick, S., and Coauthors, 2016: The MODIS cloud optical and microphysical products: Collection 6 updates and examples from Terra and Aqua. IEEE Trans. Geosci. Remote Sens., 55, 502–525, https://doi.org/10.1109/TGRS.2016.2610522.
Popovitcheva, O., E. Kireeva, N. Persiantseva, T. Khokhlova, N. Shonija, V. Tishkova, and B. Demirdjian, 2008: Effect of soot on immersion freezing of water and possible atmospheric implications. Atmos. Res., 90, 326–337, https://doi.org/10.1016/j.atmosres.2008.08.004.
Prenni, A. J., and Coauthors, 2007: Can ice-nucleating aerosols affect Arctic seasonal climate? Bull. Amer. Meteor. Soc., 88, 541–550, https://doi.org/10.1175/BAMS-88-4-541.
Prenni, A. J., P. J. DeMott, D. C. Rogers, S. M. Kreidenweis, G. McFarquhar, G. Zhang, and M. R. Poellot, 2009: Ice nuclei characteristics from M-PACE and their relation to ice formation in clouds. Tellus, 61, 436–448, https://doi.org/10.1111/j.1600-0889.2008.00415.x.
Pruppacher, H. R., and J. D. Klett, 2010: Microphysics of Atmospheric Clouds and Precipitation. Springer, 954 pp.
Randles, C. A., and Coauthors, 2017: The MERRA-2 aerosol reanalysis, 1980 onward. Part I: System description and data assimilation evaluation. J. Climate, 30, 6823–6850, https://doi.org/10.1175/JCLI-D-16-0609.1.
Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15, 1609–1625, https://doi.org/10.1175/1520-0442(2002)015<1609:AIISAS>2.0.CO;2.
Rienecker, M., and Coauthors, 2008: The GEOS-5 Data Assimilation System—Documentation of versions 5.0.1, 5.1.0, and 5.2.0. Technical Report Series on Global Modeling and Data Assimilation, Vol. 27, 97 pp.
Rogers, R. R., and M. K. Yau, 1989: A Short Course in Cloud Physics. Elsevier, 308 pp.
Schill, G. P., and Coauthors, 2020: The contribution of black carbon to global ice nucleating particle concentrations relevant to mixed-phase clouds. Proc. Natl. Acad. Sci. USA, 117, 22 705–22 711, https://doi.org/10.1073/pnas.2001674117.
Seethala, C., and A. Horvath, 2010: Global assessment of AMSR-E and MODIS cloud liquid water path retrievals in warm oceanic clouds. J. Geophys. Res., 115, D13202, https://doi.org/10.1029/2009JD012662.
Seinfeld, J. H., and S. N. Pandis, 1998: Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. John Wiley and Sons, 1326 pp.
Serreze, M. C., and R. G. Barry, 2011: Processes and impacts of Arctic amplification: A research synthesis. Global Planet. Change, 77, 85–96, https://doi.org/10.1016/j.gloplacha.2011.03.004.
Serreze, M. C., A. P. Barrett, J. C. Stroeve, D. N. Kindig, and M. M. Holland, 2009: The emergence of surface-based Arctic amplification. Cryosphere, 3, 11–19, https://doi.org/10.5194/tc-3-11-2009.
Shi, Y., and X. Liu, 2019: Dust radiative effects on climate by glaciating mixed-phase clouds. Geophys. Res. Lett., 46, 6128–6137, https://doi.org/10.1029/2019GL082504.
Shupe, M. D., V. P. Walden, E. Eloranta, T. Uttal, J. R. Campbell, S. M. Starkweather, and M. Shiobara, 2011: Clouds at Arctic atmospheric observatories. Part I: Occurrence and macrophysical properties. J. Appl. Meteor. Climatol., 50, 626–644, https://doi.org/10.1175/2010JAMC2467.1.
Si, M., E. Evoy, J. Yun, and Y. Xi, 2019: Concentrations, composition, and sources of ice-nucleating particles in the Canadian high Arctic during spring 2016. Atmos. Chem. Phys., 19, 3007–3024, https://doi.org/10.5194/acp-19-3007-2019.
Silber, I., A. M. Fridlind, A. S. Ackerman, G. V. Cesana, and D. A. Knopf, 2021: The prevalence of precipitation from polar supercooled clouds. Atmos. Chem. Phys., 21, 3949–3971, https://doi.org/10.5194/acp-21-3949-2021.
Solomon, A., M. D. Shupe, P. O. G. Persson, and H. Morrison, 2011: Moisture and dynamical interactions maintaining decoupled Arctic mixed-phase stratocumulus in the presence of a humidity inversion. Atmos. Chem. Phys., 11, 10 127–10 148, https://doi.org/10.5194/acp-11-10127-2011.
Solomon, A., G. Feingold, and M. D. Shupe, 2015: The role of ice nuclei recycling in the maintenance of cloud ice in Arctic mixed-phase stratocumulus. Atmos. Chem. Phys., 15, 10 631–10 643, https://doi.org/10.5194/acp-15-10631-2015.
Sullivan, S. C., D. Lee, L. Oreopoulos, and A. Nenes, 2016: Role of updraft velocity in temporal variability of global cloud hydrometeor number. Proc. Natl. Acad. Sci. USA, 113, 5791–5796, https://doi.org/10.1073/pnas.1514039113.
Szakáll, M., and Coauthors, 2021: Comparative study on immersion freezing utilizing single-droplet levitation methods. Atmos. Chem. Phys., 21, 3289–3316, https://doi.org/10.5194/acp-21-3289-2021.
Tan, I., and T. Storelvmo, 2016: Sensitivity study on the influence of cloud microphysical parameters on mixed-phase cloud thermodynamic phase partitioning in CAM5. J. Amer. Sci., 73, 709–728, https://doi.org/10.1175/JAS-D-15-0152.1.
Tan, I., and T. Storelvmo, 2019: Evidence of strong contributions from mixed-phase clouds to Arctic climate change. Geophys. Res. Lett., 46, 2894–2902, https://doi.org/10.1029/2018GL081871.
Tan, I., T. Storelvmo, and M. D. Zelinka, 2016: Observational constraints on mixed-phase clouds imply higher climate sensitivity. Science, 352, 224–227, https://doi.org/10.1126/science.aad5300.
Tan, I., L. Oreopoulos, and N. Cho, 2019: The role of thermodynamic phase shifts in cloud optical depth variations with temperature. Geophys. Res. Lett., 46, 4502–4511, https://doi.org/10.1029/2018GL081590.
Taszarek, M., N. Pilguj, J. T. Allen, V. Gensini, H. E. Brooks, and P. Szuster, 2021: Comparison of convective parameters derived from ERA5 and MERRA-2 with rawinsonde data over Europe and North America. J. Climate, 34, 3211–3237, https://doi.org/10.1175/JCLI-D-20-0484.1.
Taylor, P. C., S. Kato, K.-M. Xu, and M. Cai, 2015: Covariance between Arctic sea ice and clouds within atmospheric state regimes at the satellite footprint level. J. Geophys. Res. Atmos., 120, 12 656–12 678, https://doi.org/10.1002/2015JD023520.
Taylor, P. C., R. C. Boeke, Y. Li, and D. W. J. Thompson, 2019: Arctic cloud annual cycle biases in climate models. Atmos. Chem. Phys., 19, 8759–8782, https://doi.org/10.5194/acp-19-8759-2019.
Thompson, D. R., B. H. Kahn, R. O. Green, S. A. Chien, E. M. Middleton, and D. Q. Tran, 2018: Global spectroscopic survey of cloud thermodynamic phase at high spatial resolution, 2005–2015. Atmos. Meas. Tech., 11, 1019–1030, https://doi.org/10.5194/amt-11-1019-2018.
Tiedtke, M., 1993: Representation of clouds in large-scale models. Mon. Wea. Rev., 121, 3040, https://doi.org/10.1175/1520-0493(1993)121<3040:ROCILS> 2.0.CO;2.
Ullrich, R., and Coauthors, 2017: A new ice nucleation active site parameterization for desert dust and soot. J. Amer. Sci., 74, 699–717, https://doi.org/10.1175/JAS-D-16-0074.1.
Vali, G., P. J. DeMott, O. Möhler, and T. F. Whale, 2015: A proposal for ice nucleation terminology. Atmos. Chem. Phys., 15, 102263–102270, https://doi.org/10.5194/acp-15-10263-2015.
Vergara-Temprado, J., and Coauthors, 2018: Is black carbon an unimportant ice-nucleating particle in mixed-phase clouds. J. Geophys. Res. Atmos., 123, 4273–4283, https://doi.org/10.1002/2017JD027831.
Verlinde, J., and Coauthors, 2007: The Mixed-Phase Arctic Cloud Experiment. Bull. Amer. Meteor. Soc., 88, 205–222, https://doi.org/10.1175/BAMS-88-2-205.
Wang, Y., D. Zhang, X. Liu, and Z. Wang, 2018: Distinct contributions of ice nucleation, large-scale environment, and shallow cumulus detrainment to cloud phase partitioning with NCAR CAM5. J. Geophys. Res. Atmos., 123, 1132–1154, https://doi.org/10.1002/2017JD027213.
Wegener, A., 1911: Thermodynamik der Atmosphäre. J. A. Barth, Leipzig, 331 pp.
Welti, A., Z. A. Kanji, F. Lüönd, O. Stetzer, and U. Lohmann, 2014: Exploring the mechanisms of ice nucleation on kaolinite: From deposition nucleation to condensation freezing. J. Amer. Sci., 71, 16–36, https://doi.org/10.1175/JAS-D-12-0252.1.
Wiacek, A., T. Peter, and U. Lohmann, 2010: The potential influence of Asian and African mineral dust on ice, mixed-phase and liquid water clouds. Atmos. Chem. Phys., 10, 8649–8667, https://doi.org/10.5194/acp-10-8649-2010.
Xie, S., and Coauthors, 2010: Clouds and more: ARM climate modeling best estimate data: A new data product for climate studies. Bull. Amer. Meteor. Soc., 91, 13–20, https://doi.org/10.1175/2009BAMS2891.1.
Xie, S., X. Liu, C. Zhao, and Y. Zhang, 2013: Sensitivity of CAM5-simulated arctic clouds and radiation to ice nucleation parameterization. J. Climate, 26, 5981–5999, https://doi.org/10.1175/JCLI-D-12-00517.1.
Young, C. L., N. Sokolik, and J. Dufek, 2012: Regional radiative impact of volcanic aerosol from the 2009 eruption of Mt. Redoubt. Atmos. Chem. Phys., 12, 3699–3715, https://doi.org/10.5194/acp-12-3699-2012.
Zhang, M., S. Xie, X. Liu, W. Lin, K. Zhang, H.-Y. Ma, X. Zheng, and Y. Zhang, 2020: Toward understanding the simulated phase partitioning of Arctic single-layer mixed-phase clouds in E3SM. Earth Syst. Sci., 7, e2020EA001125, https://doi.org/10.1029/2020EA001125.
Zhao, C., 2011: ARM cloud retrieval ensemble data set (ACRED). DOE/SC-ARM-TR-99, 35 pp.
Zobrist, B., T. Koop, B. Luo, C. Marcolli, and T. Peter, 2007: Heterogeneous ice nucleation rate coefficient of water droplets coated by a nonadecanol monolayer. J. Phys. Chem. C, 111, 2149–2155, https://doi.org/10.1021/jp066080w.
Zuidema, P., and Coauthors, 2005: An Arctic springtime mixed-phase cloudy boundary layer observed during SHEBA. J. Amer. Sci., 62, 160–176, https://doi.org/10.1175/JAS-3368.1.
