Global Importance of Secondary Ice Production
Transcript of Global Importance of Secondary Ice Production
Global Importance of Secondary Ice Production
Xi Zhao1 and Xiaohong Liu1
1Department of Atmospheric Sciences, Texas A&M University, College Station, Texas, 77840, USA
Corresponding author: Xiaohong Liu ([email protected])
Key Points:
Secondary ice production can enhance ice number concentrations by three orders
of magnitude in moderately cold clouds
Secondary ice production decreases the global annual average liquid water path
by 22% and increases ice water path by 23% in a global climate model
Secondary ice production changes the global annual average shortwave,
longwave, and net cloud forcing by 2.1, –1.0, and 1.1 W m–2, respectively
Acc
epte
d A
rtic
le
This article has been accepted for publication and undergone full peer review but has not been throughthe copyediting, typesetting, pagination and proofreading process, which may lead to differences betweenthis version and the Version of Record. Please cite this article as doi: 10.1029/2021GL092581.
This article is protected by copyright. All rights reserved.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Abstract
Measured ice crystal number concentrations are often orders of magnitude higher
than the number concentrations of ice nucleating particles, indicating the existence of
secondary ice production (SIP) in clouds. Here, we present the first study to examine the
global importance of SIP through the droplet shattering during freezing of rain, ice-ice
collision fragmentation, and rime splintering, using a global climate model. Our results
show that SIP happens quite uniformly in the two hemispheres and dominates the ice
formation in the moderately cold clouds with temperatures warmer than -15℃. SIP
decreases the global annual average liquid water path by –14.6 g m–2 (–22%), increases
the ice water path by 8.7 g m–2 (23%), improving the model agreement with observations.
SIP changes the global annual average shortwave, longwave, and net cloud forcing by
2.1, –1.0, and 1.1 W m–2, respectively, highlighting the importance of SIP on cloud
properties on the global scale.
Plain Language Summary
Global climate models (GCMs) show large cloud biases in terms of phase partitioning
due to their poor representations of ice microphysical processes. Ice formation in mixed-
phase clouds can remarkably change the cloud albedo, lifetime, and precipitation
efficiency. Despite strong evidence of SIP in moderately cold clouds, it is still poorly
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
represented in GCMs and its impacts are not quantified. Here, we present the first study
to examine the global importance of SIP using a GCM. Our study shows that SIP
decreases the global annual average liquid water path by –14.6 g m–2 (–22%) and
increases the ice water path by 8.7 g m–2 (23%), improving the model agreement with
observations. SIP changes the global annual average net cloud radiative forcing by 1.1 W
m–2, highlighting the importance of SIP on cloud properties on the global scale.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
1. Introduction
It has been frequently reported that measured ice crystal number concentrations
(Ni) are often orders of magnitude higher than number concentrations of ice nucleating
particles (INPs). Previous studies interpreted this observed discrepancy as evidence of
secondary ice production (SIP) in clouds (Mossop, 1985; Field et al., 2016; Korolev et
al., 2020). SIP is referred to as the ice production processes from preexisting ice crystals.
Ice nucleation (or primary ice production, hereafter referred to as PIP), is the ice
production processes involving aerosol particles via droplet freezing or deposition
nucleation (Vali et al., 2015). Currently, there are several proposed SIP mechanisms:
rime splintering (also known as the Hallett and Mossop (HM) process), ice-ice collision
fragmentation (IIC), droplet shattering during freezing (FR), fragmentation during the
sublimation of ice bridge, ice fragmentation during thermal shock, and activation of INPs
in a transient supersaturation (Field et al., 2016; Korolev and Leisner, 2020).
Observational studies have provided strong evidence of SIP in different types of
clouds and in different geographic regions. By analyzing the aircraft data, Hobbs and
Rangno (1985,1990,1998) attributed the occurrence of high Ni in maritime cumulus
clouds off the Washington coast to SIP, and suggested the importance of other SIP
mechanisms in addition to the HM. Based on in-situ measurements of the Arctic stratus
clouds, Rangno and Hobbs (2001) revealed that IIC and FR are the two important SIP
mechanisms besides the HM. Based on in-situ measurements by particle imaging probes,
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Lawson et al. (2015) and Heymsfield and Willis (2014) found that FR contributed
significantly to the high Ni in cumulus clouds in the Caribbean and Africa. Similar results
were reported by Taylor et al. (2016) in maritime cumulus over the southwest peninsula
of the United Kingdom. Using radar and aircraft measurements of convection in a warm-
frontal mixed-phase cloud, Hogan et al. (2002) attributed the formation of column ice
crystals in the convection to the SIP. Evidence of SIP was also reported by Crosier et al.
(2014) in their study of convection in a strong cold front.
Despite its potential importance, representations of SIP in global climate models
(GCMs) are still very crude. Most of the current GCMs only consider the HM. However,
this process operates at a relatively narrow temperature range of –3 to –8℃, and is not
efficient to reproduce observed Ni (Sullivan et al., 2018; Sotiropoulou et al., 2020).
Recent studies indicated that GCMs such as the DOE’s Energy Exascale Earth System
Model version 1 (E3SMv1) and the NCAR’s Community Earth System Model version 2
(CESM2) significantly underestimate observed ice water path (IWP) while overestimate
observed liquid water path (LWP) in mid- and high-latitude mixed-phase clouds
(D’Alessandro et al., 2019; Zhang et al., 2019; Zhang et al., 2020). The cloud phase
regulates the cloud radiative forcing (Shupe and Intrieri, 2004; Lawson and Gettelman,
2014; Korolev et al., 2017), and plays an important role in determining the mixed-phase
cloud feedbacks (Stephens, 2005; Gettelman and Sherwood, 2016; Ceppi et al., 2017;
Tan et al., 2019) and climate sensitivity (Tan et al., 2016; Bodas-Salcedo et al., 2019;
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Gettelman et al., 2019; Zelinka et al., 2020). Furthermore, cloud phase can also affect the
Arctic amplification (Tan and Storelvmo, 2019).
To bridge the gap between measurements and GCM representations of SIP, we
incorporated two new types of SIP, i.e., FR and IIC into CESM2, and improved the
agreement of modeled clouds with observations in the Arctic (Zhao et al., 2020). The
parameterization methods for the two new types of SIP were based on the latest
theoretical developments of Phillips et al. (2017, 2018). In this study, we attempt to
further quantify the global importance of SIP on cloud properties and cloud radiative
forcing on the global scale.
2. Method
In this study, we use the Community Atmosphere Model version 6 (CAM6), the
atmosphere component of CESM2 (Danabasoglu et al., 2020). The treatment of ice
nucleation in cirrus clouds includes the homogeneous freezing of sulfate and
heterogeneous immersion freezing on dust, and the competition between these two
mechanisms at temperatures below –37℃ (Liu and Penner, 2005). In mixed-phase clouds
with temperatures between –37℃ and 0℃, the classical nucleation theory (CNT) is used to
describe the three mechanisms of heterogeneous ice nucleation, i.e., immersion freezing,
contact freezing, and deposition freezing on mineral dust and black carbon (Hoose et al.,
2010; Wang et al., 2014). The aerosol processes are represented by the 4‐ mode version of
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
the Modal Aerosol Module (Liu et al., 2016, Liu et al., 2012). The cloud microphysics in
stratiform clouds is described by a two-moment bulk scheme, called the MG2 scheme
(Gettelman et al., 2015). MG2 includes the prognostic precipitation and considers only one
type of SIP, the HM. As mentioned before, this process only operates at the temperature
range from –8 to –3℃.
We implemented two new types of SIP into CAM6, the fragmentation of freezing
rain droplets (FR) and ice-ice collision breakup (IIC), following Phillips et al. (2017,
2018). The FR contains two modes (Phillips et al., 2017). The first mode describes the
fragmentation of a freezing rain droplet triggered by contacting with an INP or a smaller
ice crystal. In this process, the rain droplet is more massive. In the second mode, a rain
droplet collides with a more massive ice particle so that the rain particle freezes and
breaks up. The IIC includes snow-snow, snow-cloud ice, snow-graupel, cloud ice-
graupel, and graupel-graupel collisions (Phillips et al., 2018). A approach based on Zhao
et al. (2017) was applied to the MG2 scheme to diagnose the graupel mass and number
mixing ratios for the IIC calculation. Since these processes happen at the particle scale,
an emulated bin framework was developed for implementing the two new types of SIP in
the MG2 cloud microphysics (Zhao et al., 2020).
We conducted four model experiments in this study, as listed in Table S1. The
control (CTL) experiment uses the default CAM6 model in which only the HM is
considered. The SIP_CNT experiment is the same as CTL but considers all the three
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
types of SIP, i.e., HM, FR, and IIC. To examine the impacts of primary ice nucleation on
SIP, we conducted two sensitivity experiments: one with the DeMott et al. (2015) ice
nucleation scheme to replace CNT in SIP_CNT, hereafter the experiment as SIP_D15,
and one with the Niemand et al. (2012) ice nucleation scheme to replace CNT in
SIP_CNT, hereafter the experiment as SIP_N12. All simulations were performed for 6
years with a horizontal resolution of 0.9° (latitudes) by 1.25 (longitude) and 32 vertical
levels from the surface to 3 hPa. The model is initialized with prescribed greenhouse gas
concentrations and the climatological sea ice and sea surface temperature. The present-
day (year 2000) precursor gas and aerosol emissions are used to drive the model. The last
five-year simulations are used in the analysis with the first-year simulation used as the
model spin-up.
3. Results
3.1 Relative importance of SIP and PIP
Figure 1 shows the annual zonal mean ice production rates of SIP and PIP from the
SIP_CNT experiment. The SIP rate includes contributions from the FR, IIC, and HM.
The SIP production rate has a relatively homogeneous distribution pattern between the
two hemispheres: the rate over the Southern Hemisphere (SH) is in the same magnitudes
as that over the Northern hemisphere (NH) (Figure 1a). In contrast, the PIP production
rate shows an obvious hemispheric asymmetry, with the rate over NH two orders of
magnitude larger than that over SH (Figure 1b). PIP dominates the ice production over
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
the NH, which is mainly contributed from the immersion freezing on dust aerosol (Figure
1c and Figure S1). The PIP rate in the NH is around 10 to 50 kg–1 s–1 with the maximum
rate at ~400 hPa over 45° N. As a comparison, the PIP rate in the SH is around 1 to 5 kg–1
s–1, and is much smaller than that in the NH where most of the dust INPs are originated
from. On a global perspective, the SIP rate is 1 kg–1 s–1 with two peaks. The first peak
locates at ~60° S over the Southern Ocean (SO), extended from surface to 700 hPa with
the maximum value of 5 kg–1 s–1 at –10℃. The FR is the main contributor to the ice
production there (Figure S1). The second peak is at 400 hPa over the tropics, with the
most contribution from the FR and IIC. The slightly larger SIP rate in the SH than that in
the NH is related to the larger SH cloud and precipitation mass mixing ratios and number
concentrations (Figure S2).
Figure 1b also shows that the maximum PIP rate locates at about –25 ℃ isotherm
altitude because of the inverse dependence of immersion freezing rate on temperature as
well as its positive dependence on dust concentration. However, SIP has maximum rates
at lower levels because the FR and HM rates are dependent on liquid water amount,
which is higher at lower levels. The seasonal variation of SIP rate shows that maximum
SIP rates at about 7 kg–1 s–1 occur below 700 hPa altitude over 60°S in the austral winter
(Figure S3). As a result, the SIP rate is larger than the PIP rate in the temperature range
from –25 to 0℃ in the tropics and SH. The largest ice multiplication (i.e., SIP/PIP) by 3
orders of magnitude occurs in these regions at temperatures warmer than –10℃ (Figure
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
1c). In fact, the ice multiplication depends on PIP. The PIP rate is much larger in NH than
in SH, while the SIP rate has similar magnitudes in NH as in SH. Therefore, the ice
multiplication is lower in NH. The dominant role of SIP in the tropics and SH at warmer
temperatures than –10℃ is related to the lower PIP rate there.
3.2 SIP influence on clouds and radiation
Figure 2 shows annual zonal‐ mean liquid water path (LWP), ice water path
(IWP), shortwave cloud forcing (SWCF), longwave cloud forcing (LWCF), and cloud
fractions from CTL, SIP_CNT, and observations. SWCF and LWCF observation data are
from the Cloud and the Earth’s Radiant Energy System Energy Balanced and Filled
(CERES-EBAF) (Loeb et al., 2009). Six LWP observation datasets are used in the
comparison with modeled LWP. Since some of the LWP datasets are only available over
the ocean and there are large uncertainties with observed LWP over the land, we only
include LWP over the ocean in the comparison. IWP observation data are from the
CloudSat version 5.1 (Waliser et al., 2009). The IWP retrievals from CloudSat may be
underestimated since the satellite is unable to penetrate through the liquid-top cloud
layer, while a large amount of ice particles exist near cloud base. Since cloud ice is not
separated from snow in the observation, we include the snow component in modeled IWP
to be consistent with the observation. More details about the observation datasets and
their uncertainties can be found in the supplementary.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
The CTL experiment substantially overestimates the LWP in mid- to high
latitudes, i.e., 40° N north and 40° S south, which was also shown in previous studies
with CESM (e.g., D’Alessandro et al., 2019). The maximum bias occurs at 60° S, with
the LWP values of 140 and 58 g m–2 in CTL and observation, respectively. This positive
bias is largely reduced in the SIP_CNT experiment, which gives a LWP value of 60 g m–2
at 60° S, in an excellent agreement with the observation. LWP is still underestimated in
the tropical regions by the model compared with the observation. This calls for the
improvement of convective cloud parameterizations in the model, as our implementation
of SIP is only for the stratiform clouds which are more important in mid- and high
latitudes than in the tropics.
Meanwhile, IWP is systematically underestimated by 10 to 30 g m–2 in all
latitudes in CTL. After considering SIP in the model, this low bias is much reduced in the
SIP_CNT experiment. The maximum IWP in SH is 85, 105, and 115 g m–2 in CTL,
SIP_CNT, and observation, respectively. We also notice the increase of IWP in the
tropics in SIP_CNT compared with CTL, despite no obvious change in LWP there.
Although the model agreement with the observation is improved in SIP_CNT, modeled
IWP is still lower than the observation in the tropics and NH mid- to high latitudes.
Patnaude et al. (2021) showed that CAM6 underestimates ice water content (IWC) for the
cirrus cloud regime (T ≤ –40℃), and the underestimation of IWC in the NH midlatitudes
is stronger than that in the SH midlatitudes. This underestimation of IWC in the cirrus
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
clouds could contribute to the overall model low biases of IWP, particularly in the NH
midlatitudes.
Reduced LWP leads to weaker SWCF in mid- and high-latitudes, with a
maximum change of 5 W m–2 at 50° S (Fig. 2c). SWCF from SIP_CNT is closer to the
observation from 40 to 55° S and from 50 to 90° N compared to CTL. Meanwhile,
LWCF decreases slightly in SIP_CNT compared to CTL, especially over the SH mid-
latitudes. Need to mention that SWCF is more sensitive to the low-cloud amount, cloud
droplet size, and LWP, while LWCF is more sensitive to the high-cloud amount, ice
crystal size, and IWP. The increase of IWC mainly happens in the mixed-phase clouds
below 400 hPa altitude (Figure S2), which has a small influence on LWCF. However,
IWC and Ni are slightly reduced in SIP_CNT at around 300 hPa in high clouds (Figure
S2), due to less cloud droplets available for the homogeneous freezing to form ice after
the introduction of SIP. As a result, LWCF is slightly reduced in SIP_CNT compared to
CTL. The model does poorly in simulating LWP and IWP in the tropics, while still
captures the SWCF and LWCF there compared to the observation due to the model
compensating errors. Our implementation of SIP influences the microphysics in
stratiform mixed-phase clouds but does not directly change the properties of convective
clouds, which dominate the cloud forcing in the tropics.
Table 1 shows the global annual mean cloud water path and cloud forcing from
CTL and SIP_CNT as well as absolute and relative differences between these two
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
experiments. Global annual mean LWP decreases by 14.6 g m–2 (-22%) with the
introduction of SIP, while IWP increases by 8.7 g m–2 (23%). Global annual mean LWCF
is reduced by 1.0 W m–2, while SWCF is weakened by 2.1 W m–2. As a result, the net
cloud forcing (sum of LWCF and SWCF) is increased (positive) by 1.1 W m–2. Global
annual mean low-, mid-, high-, and total cloud fractions are all slightly reduced. Global
spatial distributions of LWP, IWP, SWCF, and LWCF changes between CTL and
SIP_CTL are shown in Figure S4. In summary, the CESM2 model with SIP improves
simulated LWP and IWP compared with observations, leading to weaker cloud forcings,
in terms of SWCF, LWCF, and net cloud forcing.
3.3 Ice budget analysis
As shown in Figure 2, SIP has significant impacts on the global LWP and IWP.
However, the SIP rate is smaller than the PIP rate, particularly in the NH and at cold
temperatures (Figure 1). Furthermore, although the SIP rate is 1-3 orders of magnitude
smaller than the PIP rate in the NH mid- and high latitudes, we still see significant
changes in LWP and IWP there caused by SIP (Figure 2). Below, we examine the
physical mechanisms behind these changes.
Figure 3 shows the global annual mean vertical profiles of rates of several key ice
microphysical processes from the CTL and SIP_CNT experiments, with a complete list
of ice budgets shown in Figure S5. We notice that both SIP and PIP are very small terms
in the ice mass budgets compared with other terms such as water vapor deposition,
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Wegener–Bergeron–Findeisen (WBF) process, ice melting, ice sedimentation, and
accretion of cloud droplets by snow (Figure S5). However, when introducing the SIP into
the model, Ni increases. This leads to a larger surface area of ice for water vapor
deposition (Figure 3a). Consequently, the cloud ice mixing ratio increases. This further
accelerates the autoconversion of cloud ice to snow (i.e., more negative rate in Figure 3b
as this is a loss term for cloud ice), because this process is proportional to the cloud ice
mixing ratio. The snow accretion of cloud ice only shows slight changes (Figure 3c).
Higher rates of these processes give rise to higher snow mixing ratio, by which the WBF
process of cloud ice and snow increases substantially (Figure 3d). This results in a strong
reduction in LWP and a noticeable increase in IWP from CTL to SIP_CNT. The ice-
related processes largely determine the liquid and ice partitioning in mixed-phase clouds
(Solomon et al., 2015; Tan & Storelvmo, 2016; Kalesse et al., 2016). Although SIP is not
the direct contributor to the changes of these process rates but it triggers a chain of
interactions among these processes.
4. Comparing SIP versus PIP
We further examine the sensitivity of model results to different ice nucleation
parameterizations, as shown in Figure 4. The SIP_N12 experiment shows the largest PIP
rate compared with the SIP_CNT and SIP_D15 experiments. The SIP rate is quite similar
among the three PIP parameterizations. Consequently, the SIP is more important in
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
SIP_CNT than SIP_N12. However, general patterns for the relative importance between
SIP and PIP are similar among these experiments. Figure S6 shows that the LWP, IWP,
SWCF, and LWCF from SIP_CNT, SIP_N12, and SIP_D15 experiments are close to
each other. This suggests that the global importance of SIP is not significantly dependent
on the primary ice nucleation parameterization. We also give the global annual means of
LWP, IWP, cloud forcings, and cloud fractions from these experiments in Table S2.
Differences among these experiments are much smaller than the differences between
these experiments and the CTL experiment.
It is worth noting that the modeled primary ice nucleation currently does not include
INPs from marine organic aerosol, which is identified as the dominant INPs over the
remote Southern Ocean (Burrows et al., 2013; Wilson et al., 2015; Zhao et al., 2021). The
model also does not include the ice nucleation on terrestrial biogenic INPs (O’Sullivan et
al., 2018; Kanji et al., 2017). Future work will focus on improving the representations of
primary ice nucleation on these INPs as well as representations of other SIP mechanisms
(e.g., fragmentation during the sublimation of ice bridge). The modeled ice number
concentration will be compared against recent observations over the Southern Ocean
(e.g., from the Southern Ocean Clouds, Radiation, Aerosol Transport Experimental Study
(SOCRATES)).
5. Conclusions
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Physical representations of the two new SIP mechanisms (i.e., FR and IIC) were
introduced to CAM6 and evaluated for simulating the Arctic clouds against the ARM
observations in our earlier study (Zhao et al., 2020). In this study, we further examine the
impacts of SIP on cloud properties and cloud radiative forcing on the global scale. Our
results show that the SIP occurs pretty uniformly in the two hemispheres with a
maximum process rate of 5 kg–1 s–1 at –10℃. SIP dominates the ice formation in
moderately-cold clouds in SH and in the tropics, where the ice multiplication reaches up
to three orders of magnitude at temperatures warmer than –10℃. In contrast, PIP
dominates the ice formation in the relatively-cold clouds with temperatures lower than –
25℃ in all latitudes.
Comparing with the CTL experiment, the model with the two new SIP
mechanisms decreases the LWP by 40 to 60 g m–2 at mid- to high latitudes of both
hemispheres, improving the model agreement with the observation. IWP increases at all
latitudes with a globally averaged increase of 8.7 g m–2 (23%). Finally, SIP results in
weaker cloud forcings in terms of SWCF, LWCF, and net cloud forcing by 2.1, –1.0, and
1.1 W m–2, respectively.
Through ice budget analysis, we found that SIP triggers a series of changes of ice-
related processes, which ultimately change the liquid-ice partitioning and cloud radiative
forcing. These findings highlight the importance of incorporating physical representations
of SIP in GCMs.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Competing interests: The authors declare that they have no conflict of interest.
Data availability: The model code is available at https://github.com/CESM-
Development. The observation data of LWCF/SWCF is available at
https://ceres.larc.nasa.gov/data/. The IWP data is available at
https://cloudsat.atmos.colostate.edu/data. The LWP data is available at:
https://climatedataguide.ucar.edu/climate-data/water-vapor-nvap.
Acknowledgment: This research was supported by the DOE Atmospheric System
Research (ASR) Program (grant DE-SC0020510). The authors would also like to
acknowledge the use of computational resources for conducting the model simulations
(ark:/85065/d7wd3xhc) at the NCAR-Wyoming Supercomputing Center provided by the
NSF and the State of Wyoming and supported by NCAR’s Computational and
Information Systems Laboratory.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
References
Bodas-Salcedo, A., J. P. Mulcahy, T. Andrews, K. D. Williams, M. A. Ringer, P. R.
Field, and G. S. Elsaesser (2019), Strong Dependence of Atmospheric Feedbacks on
Mixed-Phase Microphysics and Aerosol-Cloud Interactions in HadGEM3, Journal
of Advances in Modeling Earth Systems, 11, 1735-1758, doi:
10.1029/2019MS001688.
Burrows, S. M., C. Hoose, U. Pöschl, and M. G. Lawrence (2013), Ice nuclei in marine
air: Biogenic particles or dust?, Atmos Chem Phys, 13, 245-267, doi: 10.5194/acp-
13-245-2013.
Ceppi, P., Brient, F., Zelinka, M. D., & Hartmann, D. L. (2017). Cloud feedback
mechanisms and their representation in global climate models. Wiley
Interdisciplinary Reviews: Climate Change, 8(4), e465.
https://doi.org/10.1002/wcc.465.
Crosier, J., T. W. Choularton, C. D. Westbrook, A. M. Blyth, K. N. Bower, P. J.
Connolly, C. Dearden, M. W. Gallagher, Z. Cui, and J. C. Nicol (2014),
Microphysical properties of cold frontal rainbands†, Q J Roy Meteor Soc, 140,
1257-1268, doi: 10.1002/qj.2206.
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, Journal of Climate,
32, 2781-2805, doi: 10.1175/JCLI-D-18-0232.1.
Danabasoglu, G., et al. (2020), The Community Earth System Model Version 2
(CESM2), Journal of Advances in Modeling Earth Systems, 12, doi:
10.1029/2019MS001916.
DeMott, P. J., et al. (2015), Integrating laboratory and field data to quantify the
immersion freezing ice nucleation activity of mineral dust particles, Atmos Chem
Phys, 15, 393-409.
Field, P. R., et al. (2016), Chapter 7. Secondary Ice Production - current state of the
science and recommendations for the future, Meteorological Monographs, 58, 7.1-
7.20, doi: 10.1175/amsmonographs-d-16-0014.1.
Gettelman, A., et al. (2019), High Climate Sensitivity in the Community Earth System
Model Version 2 (CESM2), Geophys Res Lett, 46, 8329-8337, doi:
10.1029/2019GL083978.
Gettelman, A., and H. Morrison (2015), Advanced two-moment bulk microphysics for
global models. Part I: Off-line tests and comparison with other schemes, Journal of
Climate, 28, 1268-1287, doi: 10.1175/JCLI-D-14-00102.1.
Gettelman, A., and S. C. Sherwood (2016), Processes Responsible for Cloud Feedback,
in Current Climate Change Reports, edited, pp. 179-189, Springer.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Heymsfield, A., and P. Willis (2014), Cloud conditions favoring secondary ice particle
production in tropical maritime convection, Journal of the Atmospheric Sciences,
71, 4500-4526, doi: 10.1175/JAS-D-14-0093.1.
Hobbs, P. V., and A. L. Rangno (1985), Ice Particle Concentrations in Clouds, Journal of
the Atmospheric Sciences, 42, 2523-2549, doi: 10.1175/1520-
0469(1985)042<2523:IPCIC>2.0.CO;2.
Hobbs, P. V., and A. L. Rangno (1990), Rapid Development of High Ice Particle
Concentrations in Small Polar Maritime Cumuliform Clouds, Journal of the
Atmospheric Sciences, 47, 2710-2722, doi: 10.1175/1520-
0469(1990)047<2710:RDOHIP>2.0.CO;2.
Hobbs, P. V., and A. L. Rangno (1998), Microstructures of low and middle-level clouds
over the Beaufort Sea, Q J Roy Meteor Soc, 124, 2035-2071, doi:
10.1002/qj.49712455012.
Hogan, R. J., P. R. Field, A. J. Illingworth, R. J. Cotton, and T. W. Choularton (2002),
Properties of embedded convection in warm-frontal mixed-phase cloud from aircraft
and polarimetric radar, Q J Roy Meteor Soc, 128, 451-476, doi:
10.1256/003590002321042054.
Hoose, C., J. E. Kristjánsson, J. P. 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, Journal of the Atmospheric Sciences,
67, 2483-2503, doi: 10.1175/2010JAS3425.1.
Kalesse, H., G. de Boer, A. Solomon, M. Oue, M. Ahlgrimm, D. Zhang, M. D. Shupe, E.
Luke, and A. Protat (2016), Understanding rapid changes in phase partitioning
between cloud liquid and ice in stratiform mixed-phase clouds: An arctic case study,
Mon Weather Rev, 144, 4805-4826, doi: 10.1175/MWR-D-16-0155.1.
Kanji, Z. A., L. A. Ladino, H. Wex, Y. Boose, M. Burkert-Kohn, D. J. Cziczo, and M.
Krämer (2017), Overview of Ice Nucleating Particles, Meteorological Monographs,
58, 1.1-1.33, doi: 10.1175/amsmonographs-d-16-0006.1.
Korolev, A., and T. Leisner (2020), Review of experimental studies of secondary ice
production, Atmos Chem Phys, 20, 11767-11797, doi: 10.5194/acp-20-11767-2020.
Korolev, A., et al. (2017), Mixed-Phase Clouds: Progress and Challenges, Meteorological
Monographs, 58, 5.1-5.50, doi: 10.1175/amsmonographs-d-17-0001.1.
Lawson, R. P., and A. Gettelman (2014), Impact of Antarctic mixed-phase clouds on
climate, P Natl Acad Sci USA, 111, 18156-18161, doi: 10.1073/pnas.1418197111.
Lawson, R. P., S. Woods, and H. Morrison (2015), The microphysics of ice and
precipitation development in tropical cumulus clouds, Journal of the Atmospheric
Sciences, 72, 2429-2445, doi: 10.1175/JAS-D-14-0274.1.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Liu, X., et al. (2012), Toward a minimal representation of aerosols in climate models:
description and evaluation in the Community Atmosphere Model CAM5, Geosci
Model Dev, 5, 709-739.
Liu, X., P. L. Ma, H. Wang, S. Tilmes, B. Singh, R. C. Easter, S. J. Ghan, and P. J. Rasch
(2016), Description and evaluation of a new four-mode version of the Modal
Aerosol Module (MAM4) within version 5.3 of the Community Atmosphere Model,
Geosci Model Dev, 9, 505-522.
Liu, X., and J. E. Penner (2005), Ice nucleation parameterization for global models,
Meteorologische Zeitschrift, 499-514, doi: 10.1127/0941-2948/2005/0059.
Loeb, N. G., B. A. Wielicki, D. R. Doelling, G. L. Smith, D. F. Keyes, S. Kato, N.
Manalo-Smith, and T. Wong (2009), Toward optimal closure of the Earth's top-of-
atmosphere radiation budget, Journal of Climate, 22, 748-766, doi:
10.1175/2008JCLI2637.1.
Mossop, S. C. (1985), Secondary ice particle production during rime growth: The effect
of drop size distribution and rimer velocity, Q J Roy Meteor Soc, 111, 1113-1124,
doi: 10.1002/qj.49711147012.
Niemand, M., et al. (2012), A Particle-Surface-Area-Based Parameterization of
Immersion Freezing on Desert Dust Particles, Journal of the Atmospheric Sciences,
69, 3077-3092, doi: 10.1175/JAS-D-11-0249.1.
O’Sullivan, D., et al. (2018), Contributions of biogenic material to the atmospheric ice-
nucleating particle population in North Western Europe, Sci Rep-Uk, 8, 13821, doi:
10.1038/s41598-018-31981-7.
Patnaude, R., M. Diao, X. Liu, and S. Chu (2021), Effects of thermodynamics, dynamics
and aerosols on cirrus clouds based on in situ observations and NCAR CAM6,
Atmos Chem Phys, 21, 1835-1859, doi: 10.5194/acp-21-1835-2021.
Phillips, V. T. J., S. Patade, J. Gutierrez, and A. Bansemer (2018), Secondary ice
production by fragmentation of freezing drops: Formulation and theory, Journal of
the Atmospheric Sciences, 75, 3031-3070, doi: 10.1175/JAS-D-17-0190.1.
Phillips, V. T. J., J. I. Yano, and A. Khain (2017), Ice multiplication by breakup in ice-ice
collisions. Part I: Theoretical formulation, Journal of the Atmospheric Sciences, 74,
1705-1719, doi: 10.1175/JAS-D-16-0224.1.
Rangno, A. L., and P. V. Hobbs (2001), Ice particles in stratiform clouds in the Arctic
and possible mechanisms for the production of high ice concentrations, Journal of
Geophysical Research: Atmospheres, 106, 15065-15075, doi:
10.1029/2000JD900286.
Shupe, M. D., and J. M. Intrieri (2004), Cloud radiative forcing of the Arctic surface: The
influence of cloud properties, surface albedo, and solar zenith angle, Journal of
Climate, 17, 616-628, doi: 10.1175/1520-
0442(2004)017<0616:CRFOTA>2.0.CO;2.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
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, 10631-10643, doi: 10.5194/acp-15-10631-2015.
Sotiropoulou, G., S. Sullivan, J. Savre, G. Lloyd, T. Lachlan-Cope, A. M. L. Ekman, and
A. Nenes (2020), The impact of secondary ice production on Arctic stratocumulus,
Atmos Chem Phys, 20, 1301-1316, doi: 10.5194/acp-20-1301-2020.
Stephens, G. L. (2005), Cloud feedbacks in the climate system: A critical review, in
Journal of Climate, edited, pp. 237-273, American Meteorological Society.
Sullivan, S. C., C. Hoose, A. Kiselev, T. Leisner, and A. Nenes (2018), Initiation of
secondary ice production in clouds, Atmos Chem Phys, 18, 1593-1610, doi:
10.5194/acp-18-1593-2018.
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, doi: 10.1029/2018GL081590.
Tan, I., T. Storelvmo, and M. D. Zelinka (2016), Observational constraints on mixed-
phase clouds imply higher climate sensitivity, Science, 352, 224-227, doi:
10.1126/science.aad5300.
Taylor, J. W., et al. (2016), Observations of cloud microphysics and ice formation during
COPE, Atmos Chem Phys, 16, 799-826, doi: 10.5194/acp-16-799-2016.
Vali, G., DeMott, P. J., Möhler, O., & Whale, T. F. (2015). Technical Note: A proposal
for ice nucleation terminology. Atmospheric Chemistry and Physics, 15(18), 10263–
10270. https://doi.org/10.5194/acp-15-10263-2015
Waliser, D. E., et al. (2009), Cloud ice: A climate model challenge with signs and
expectations of progress, Journal of Geophysical Research Atmospheres, 114, doi:
10.1029/2008JD010015.
Wang, Y., X. Liu, C. Hoose, and B. Wang (2014), Different contact angle distributions
for heterogeneous ice nucleation in the Community Atmospheric Model version 5,
Atmos Chem Phys, 14, 10411-10430.
Wilson, T. W., et al. (2015), A marine biogenic source of atmospheric ice-nucleating
particles, Nature, 525, 234-238, doi: 10.1038/nature14986.
Zelinka, M. D., T. A. Myers, D. T. McCoy, S. Po-Chedley, P. M. Caldwell, P. Ceppi, S.
A. Klein, and K. E. Taylor (2020), Causes of Higher Climate Sensitivity in CMIP6
Models, Geophys Res Lett, 47, doi: 10.1029/2019GL085782.
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 and Space Science, 7, e2020EA001125, doi:
10.1029/2020EA001125.
Zhang, Y., S. Xie, W. Lin, S. A. Klein, M. Zelinka, P. L. Ma, P. J. Rasch, Y. Qian, Q.
Tang, and H. Y. Ma (2019), Evaluation of Clouds in Version 1 of the E3SM
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Atmosphere Model With Satellite Simulators, Journal of Advances in Modeling
Earth Systems, 11, 1253-1268, doi: 10.1029/2018MS001562.
Zhao, X., Y. Lin, Y. Peng, B. Wang, H. Morrison, and A. Gettelman (2017), A single ice
approach using varying ice particle properties in global climate model microphysics,
Journal of Advances in Modeling Earth Systems, 9, 2138-2157, doi:
10.1002/2017MS000952.
Zhao, X., Liu, X., Phillips, V. T. J., and Patade, S. (2020): Impacts of Secondary Ice
Production on Arctic Mixed-Phase Clouds based on ARM Observations and
CESM2, Atmos. Chem. Phys. Discuss. [preprint], https://doi.org/10.5194/acp-2020-
1276, accepted.
Zhao, X., Liu, X., Burrows, S. M., & Shi, Y. (2021). Effects of marine organic aerosols
as sources of immersion-mode ice-nucleating particles on high-latitude mixed-phase
clouds. Atmospheric Chemistry and Physics, 21(4), 2305–2327.
https://doi.org/10.5194/acp-21-2305-2021
Reference for Supporting Information
Amenu, G. G., and P. Kumar (2005), NVAP and reanalysis-2 global precipitable water
products: Intercomparison and variability studies, B Am Meteorol Soc, 86, 245-256,
doi: 10.1175/BAMS-86-2-245.
Greenwald, T. J., R. Bennartz, M. Lebsock, and J. Teixeira (2018), An Uncertainty Data
Set for Passive Microwave Satellite Observations of Warm Cloud Liquid Water
Path, Journal of Geophysical Research: Atmospheres, 123, 3668-3687, doi:
10.1002/2017JD027638.
Khanal, S., and Z. Wang (2018), Uncertainties in MODIS-Based Cloud Liquid Water
Path Retrievals at High Latitudes Due to Mixed-Phase Clouds and Cloud Top
Height Inhomogeneity, Journal of Geophysical Research: Atmospheres, 123,
11,154-111,172, doi: 10.1029/2018JD028558.
Khanal, S., Z. Wang, and J. R. French (2020), Improving middle and high latitude cloud
liquid water path measurements from MODIS, Atmospheric Research, 243, 105033,
doi: 10.1016/j.atmosres.2020.105033.
Lebsock, M., and H. Su (2014), Application of active spaceborne remote sensing for
understanding biases between passive cloud water path retrievals, Journal of
Geophysical Research: Atmospheres, 119, 8962-8979, doi: 10.1002/2014JD021568.
Loeb, N. G., B. A. Wielicki, D. R. Doelling, G. L. Smith, D. F. Keyes, S. Kato, N.
Manalo-Smith, and T. Wong (2009), Toward Optimal Closure of the Earth's Top-of-
Atmosphere Radiation Budget, Journal of Climate, 22, 748-766, doi:
10.1175/2008jcli2637.1.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
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, Journal of Climate, 21, 1721-1739, doi: 10.1175/2007JCLI1958.1.
Platnick, S., M. D. King, S. A. Ackerman, W. P. Menzel, B. A. Baum, J. C. Riédi, and R.
A. Frey (2003), The MODIS cloud products: Algorithms and examples from terra,
Ieee T Geosci Remote, 41, 459-472, doi: 10.1109/TGRS.2002.808301.
Randel, D. L., T. H. Vonder Haar, M. A. Ringerud, G. L. Stephens, T. J. Greenwald, and
C. L. Combs (1996), A New Global Water Vapor Dataset, B Am Meteorol Soc, 77,
1233-1246, doi: 10.1175/1520-0477(1996)077<1233:ANGWVD>2.0.CO;2.
Waliser, D. E., et al. (2009), Cloud ice: A climate model challenge with signs and
expectations of progress, J Geophys Res-Atmos, 114, doi: Artn
D00a2110.1029/2008jd010015.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Figure 1. Annual zonal mean pressure‐ latitude cross sections of (a) secondary ice
production rate (in unit of kg–1s–1), (b) primary ice production rate (in unit of kg–1s–1), and
(c) ice multiplication denoted as Log10(secondary ice production rate over primary ice
production rate) from the SIP_CNT experiment. Black solid lines show the –40℃, –25℃
and –10℃ isotherms, respectively.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Figure 2. Annual zonal‐ mean distributions of (a) liquid water path over ocean (in unit of
g m–2), (b) ice water path (in unit of g m–2), (c) shortwave cloud forcing (in unit of W m–
2), and (d) longwave cloud forcing (in unit of W m–2), (e) total cloud fraction (in unit
of %), (f) high cloud fraction (in unit of %), (g) middle-level cloud fraction (in unit of %),
and (h) low cloud fraction (in unit of %) from CTL (orange) and SIP_CNT (green)
experiments, and observations (OBS, shown as gray dashed lines). Modeled IWP
includes the snow component to be consistent with the observation, since cloud ice is not
separated from snow in the observation. In order to fairly compare modeled cloud
properties against satellite observations, the MODIS cloud simulator is used to calculate
the modeled cloud fraction.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Figure 3. Global annual mean vertical profiles of rates (in unit of 10-10 kg kg–1s–1) of (a)
deposition of water vapor on cloud ice and snow, (b) autoconversion of cloud ice to
snow, (c) accretion of cloud ice by snow, and (d) Wegener–Bergeron–Findeisen (WBF)
process of cloud ice and snow from CTL (orange) and SIP_CNT (green) experiments.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
Figure 4. Annual zonal mean pressure‐ latitude cross sections of primary ice production
rate (first row, in unit of kg–1s–1), secondary ice production rate (second row, in unit of
kg–1s–1), and ice multiplication denoted as Log10(secondary ice production rate over
primary ice production rate) (third row) from SIP_CNT (first column), SIP_D15 (second
column), and SIP_N12 (third column) experiments. Black solid lines show the –40, –25,
and -10℃ isotherms.
Acc
epte
d A
rtic
le
This article is protected by copyright. All rights reserved.
29
Table 1. Global annual means of cloud liquid water path (LWP), ice water path (IWP),
shortwave, longwave and net cloud forcings (SWCF, LWCF, and net CF), and low-,
mid-, high-, and total cloud fractions from CTL and SIP_CNT experiments as well as
absolute and relative differences (in parenthesis, %) between these two experiments.
CTL SIP_CNT SIP_CNT–CTL
LWP (g m–2) 67.02 52.39 -14.6 (-22%)
IWP (g m–2) 37.86 46.59 8.7 (23%)
SWCF (W m–2) -47.84 -45.75 2.1 (-4.4%)
LWCF (W m–2) 23.90 22.86 -1.0 (4.3%)
Net CF (W m–2) -23.94 -22.89 1.1 (4.4%)
Low cloud (%) 40.88 40.37 -0.5 (-1.2%)
Mid-level cloud (%) 30.02 29.19 -0.8 (-2.8%)
High cloud (%) 43.57 43.27 -0.3 (-0.67%)
Total cloud (%) 69.08 68.48 -0.6 (-0.86%)