Ice Crystal Habits and Growth Processes in Stratiform Clouds with Embedded ConvectionExamined through Aircraft Observation in Northern China

SHICHAO ZHU

Key Laboratory for Cloud Physics, Chinese Academy of Meteorological Sciences, Beijing, and Key Laboratory for

Aerosol-Cloud-Precipitation of China Meteorological Administration, Nanjing University of Information Science

and Technology, Nanjing, and Anhui Weather Modification Office, Hefei, China

XUELIANG GUO, GUANGXIAN LU, AND LIJUN GUO

Key Laboratory for Cloud Physics, Chinese Academy of Meteorological Sciences, Beijing, China

(Manuscript received 8 July 2014, in final form 4 January 2015)

ABSTRACT

Ice crystal habits and growth processes in two cases of stratiform clouds with embedded convection are in-

vestigated using data observed simultaneously from three aircraft on 18 April 2009 and 1May 2009 as part of the

BeijingCloudExperiment (BCE). The results show that themajority of ice crystal habits found in the two cases at

temperatures between 08 and2168C included platelike, needle column, capped column, dendrite, and irregular.

Amixture of several ice crystal habitswas identified in all of the clouds studied.However, the ice crystals recorded

in the embedded convection regions containedmore dendrites and possessed heavier riming degrees, and the ice

crystals identified in the stratiform clouds contained more hexagonal plate crystals. Both riming and aggregation

processes played central roles in the broadening of particle size distributions (PSDs), and these processes were

more active in embedded convection regions than in stratiform regions. However, riming was more prevalent in

the 18 April case than aggregation, though aggregates were evident. In contrast, the 1 May case had a more

dominant aggregation processes, but also riming.With the decrease in height, PSDs broadened in both embedded

convection regions and stratiform regions, but the broadening rates between 4.8 km (T ’ 211.68C) and 4.2 km

(T ’ 288C) were larger than those between 4.2 km (T ’ 288C) and 3.6 km (T ’ 258C). In addition, the

broadening rates of PSDs in the embedded convection regions were larger than those in the stratiform clouds, as

the aggregation and riming processes of ice particles in embedded convection regions were active. High super-

cooled water content is critical to enhancing riming and aggregation processes in embedded convection regions.

1. Introduction

Stratiform clouds with embedded convection consti-

tute an important precipitation system that typically has

a long lifetime and that may bring either continuous or

intermittent precipitation to a large region (Hobbs and

Locatelli 1978; Matejka et al. 1980; Hobbs et al. 1980;

Herzegh and Hobbs 1981; Evans et al. 2005). Because of

the presence of high ice crystal concentrations and su-

percooledwater content in embedded convection regions

(Evans et al. 2005; Hobbs and Rangno 1990; Matejka

et al. 1980), these systems can improve the precipitation

efficiency of stratiform clouds by up to 20%–35%

(Herzegh and Hobbs 1980; Houze et al. 1981; Rutledge

and Hobbs 1983). Although ice crystal habit is central to

understanding the generation and growth processes of

precipitation particles in these clouds, it is difficult to

obtain accurate characteristics of crystals and their con-

tinuous growth processes.

At present, the airborne observation is an important

method used to obtain information on ice crystal habits

and growth mechanisms in natural clouds. In deep

stratiform clouds, the formation and growth processes of

ice particles are complex. Such processes include riming

(Ono 1969), the freezing of large drizzle drops (Korolev

et al. 2004), and the aggregation of ice crystals (Takahashi

and Fukuta 1988). However, the growth processes of ice

crystals in stratiform clouds with embedded convection

are more complex.

Corresponding author address:XueliangGuo, ChineseAcademy

of Meteorological Sciences, No.46, Zhongguancun Street, Haidian

District, Beijing 100081, China.

E-mail: guoxl@mail.iap.ac.cn

MAY 2015 ZHU ET AL . 2011

DOI: 10.1175/JAS-D-14-0194.1

� 2015 American Meteorological Society

A wide variety of particle habits were observed in both

convection and stratiform regions of clouds at tempera-

tures from 08C to;2208C (Stith et al. 2002). McFarquhar

and Black (2004) analyzed data recorded by Particle

Measuring Systems (PMS) two-dimensional cloud (2DC)

probe placed in tropical cyclones and found that the

composition (i.e., graupel or snow), number, and size of ice

particles can vary substantially in convection and strati-

form regions. In stratiform regions, small ice crystals, col-

umns, and medium-sized graupel were observed, while in

convection regions, medium to large graupel, aggregates,

and raindrops were identified. Stark et al. (2013) collected

solid precipitation habits at ground during two winter

storms in coastal, extratropical cyclones and showed that

convective seeder cells resulted in relatively cold

(,2158C) ice crystal characteristics (side planes, bullets,

and dendrites). Needlelike crystals were prevalent during

the pre-band period when the maximum vertical motion

occurred in the 258 to 2108C layer. Moderately rimed

dendritic crystals were observed at snowband maturity.

Ice crystal habits heavily depend on the cloud tem-

peratures at which the ice crystals form. Hogan et al.

(2003) identified pristine planar crystals in aircraft ob-

servations of a multilayered altocumulus cloud with

a cloud-top temperature (CTT) of2158C and found that

at a CTT of 2248C, crystals formed as complex poly-

crystals. Carey et al. (2008) sampled several altocumulus

cloud layers at temperatures between 2128 and 2268Cand found that pristine planar crystals developed close to

the top of the cloud. Westbrook and Heymsfield (2011)

summarized previous studies and concluded that single

pristine crystals are common in thin, mixed-phase clouds

and that the critical temperature for polycrystal particles

lies in the range of approximately 2208 to 2268C, de-pending on the characteristics of the individual cloud.

The findings of laboratory (Fukuta and Takahashi

1999; Heymsfield et al. 2010; Magono and Iwabuchi

1972; Pruppacher and Klett 1997) and field (Evans et al.

2005; Heymsfield et al. 2002; Stith et al. 2002; Woods

et al. 2008) studies indicate that ice crystals formed

between 08 and 248C are predominantly plates, while

those formed between248 and288C are predominantly

needle columns. Bailey and Hallett (2009) produced

a new ice crystal habit diagram based on field and lab-

oratory datasets while retaining descriptions of ice

crystal habits in temperatures above 2188C from older

diagrams. The authors proposed that the ice crystal

form changes from plate (from 08 to 248C) to column

(from 248 to 288C) and then to platelike (plate or

dendrite, from 288 to 2228C).A wide variety of particle growth processes have also

been observed in convection and stratiform regions of

clouds. Analyzing PMS probe data in midlatitude

cyclones, Herzegh and Hobbs (1980) found that riming

growth was dominant in embedded convection regions

where updrafts reached 60cms21, while deposition and

aggregation growth dominated at lower layers where up-

draft velocities were less than 15cms21. Analyzing 2DC

and two-dimensional precipitation (2DP) probe data from

winter storms on the coast of Newfoundland, Canada,

Lawson et al. (1993, 1998) showed that ice crystals grow

rapidly through aggregation processes in embedded con-

vection regions, and similar ice crystal growth character-

istics were found in Arctic clouds during the summer

(Lawson and Zuidema 2009). Studying 2DC and cloud

imaging particle (CIP) probe data from a strong cyclonic

storm over the northeastern Pacific Ocean, Evans et al.

(2005) showed that above the melting layer, vapor de-

position is the dominant growth process in the rainband.

Ice particle growth via riming was found to be negligible,

while significant ice particle aggregation occurred in the

region just above the melting layer. Therefore, dominant

ice crystal growth processes found in different cloud types

were not consistent with previous studies.

The Beijing Cloud Experiment (BCE), a component

of the National Science and Technology Pillar Program

(2006–11) Key Project, was conducted from April to

May 2009 in the Zhangjiakou area, which is located in

the upstream area of Beijing in northern China.

Throughout the experiment, three aircraft observed

aerosol and cloud features simultaneously at different

cloud levels, generating valuable data on aerosol–cloud

interactions in this area (Lu and Guo 2012). Northern

China is located in the middle latitude and it has

a complex precipitating cloud system including both

convective clouds (e.g., Fu andGuo 2012) and stratiform

clouds with embedded convection. The cloud system in

this region is also heavily affected by aerosols and ur-

banization processes owing to rapid industrialization

and population growth (Guo et al. 2006, 2014).

Ice crystal habits and growth processes in two cases of

stratiform clouds with embedded convection were in-

vestigated on 18 April 2009 and 1 May 2009 through

simultaneous observation from three aircraft during the

BCE. The aircraft tracks were layered with radar echoes

to reveal ice crystal distribution and growth processes in

different cloud regions (convection and stratiform) and

at different temperatures and cloud levels. The broad-

ening rate of vertical particle size distribution (PSD) in

different cloud regions was quantitatively calculated

using data collected through simultaneous observations

from three aircraft at different levels of clouds.

A brief description of the instruments and field ex-

periment are provided in section 2. Section 3 introduces

the synoptic weather conditions and the properties of

the cloud system observed. The ice crystal habits

2012 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

observed from the three aircraft are listed in section 4,

and the distribution and growth processes of ice crystal

habits are presented in section 5. Section 6 provides the

calculations and discusses the broadening rate of PSD.

A conclusion and discussion are provided in section 7.

2. Instruments and field experiment

Throughout the BCE field observations, simultaneous

measurements were conducted through ground-based

meteorological stations, radar, and aircraft observation.

A schematic diagram of the study area and the obser-

vational facilities used for the BCE are shown in Fig. 1.

The aircraft employed in this experiment included the

Cheyenne III-A from the Shijiazhuang Weather Modi-

ficationOffice (number 3625), theDatongY-12 from the

ShanxiWeatherModificationOffice (number 3817), and

the Beijing Y-12 from the BeijingWeatherModification

Office (number 3830). The three aircraft departed from

airports in Zhangjiakou, Shijiazhuang, and Taiyuan,

respectively, and simultaneously collected data at dif-

ferent cloud levels within the study area.

During the experiment, two cold-frontal systems

passed through the study area on 18 April 2009 and 1

May 2009, respectively, and the dominant cloud type

over the study area was stratiform clouds with em-

bedded convection. The three aircraft simultaneously

observed aerosol, cloud condensation nuclei (CCN),

and cloud microphysics processes at different cloud

levels and collected valuable data on cloud micro-

physical properties in the area. Table 1 provides a list

of the instruments used aboard the three research

aircraft. All instruments were calibrated by Droplet

Measuring Technologies (DMT) in the United States

before the experiment was conducted (Lu and Guo

2012).

FIG. 1. Schematic of the observational region and facilities (the colors indicate the altitudes of terrain).

MAY 2015 ZHU ET AL . 2013

3. Synoptic weather conditions and properties ofthe cloud system observed

On 18 April 2009, a weak cold-frontal system passed

through the study area and produced stratiform clouds

with embedded convection regions over the study area.

The total rainfall recorded at the Zhangjiakou meteo-

rological station was 6.7mm. The synoptic weather

conditions at 0800 BT (Beijing time) on 18 April 2009

are shown in Fig. 2, and the black box shown in the

image denotes the study area. Figure 2 shows that on 18

April 2009, a weak trough at 500 hPa approached the

study area, and a high pressure center in the East China

Sea blocked the movement of the weak trough to the

southeastern direction. The wind at 500 hPa was domi-

nated by southwesterly winds. Figure 2b shows infrared

cloud images from satellite FY-2C at 1900 BT on 18

April 2009. The figure shows that the cloud band was

dispersed throughout the northeast–southwestern re-

gion, and the main cloud band was located in the

northern area of the Zhangjiakou region. Clouds with

nonuniform top developed over the study area.

The radar composite reflectivity shown in Fig. 3

demonstrates that the strong echo band was located

along the front boundary of the cloud system, which was

produced through low-level airflow convergence. A

stratiform cloud with a weak echo formed behind the

convection region. The entire cloud bandwas positioned

in the northeast–southwestern area and moved from

northwest to southeast. Some obvious strong echo re-

gions were embedded in the cloud band, and the maxi-

mum radar reflectivity of the embedded convection

region was approximately 50 dBZ, which was approxi-

mately 10–20 dBZ higher than reflectivities recorded in

surrounding stratiform clouds. Both aircraft 3817 and

3830 recorded observations from the front edge of the

cloud system, and the flight route of aircraft 3625 fol-

lowed a north–south loop through the cloud band.

On 1 May 2009, a stronger frontal system passed

through the study area and generated 6–10mm of total

rainfall in the study area. The synoptic weather condi-

tions at 0800 BT on 1 May 2009 are shown in Fig. 4. The

figure shows that on 1 May 2009, the 500-hPa level over

the study area was controlled by a trough (Fig. 4a).

Comparing these conditions with those of 18 April 2009,

the trough was more obvious and was not blocked by the

high pressure center located in the East China Sea.

Winds at 500 hPa were dominated by southwesterly

winds. Figure 4b provides a visible cloud image taken

from satellite FY-2C at 0900 BT on 1 May 2009 and

shows that themain cloud system,moving in a northwest

to southeast direction, had already moved out of the

study area. The cloud observed from the aircraft was

mainly located in the postfrontal region.

Figure 5 provides a superimposed image of radar

reflectivity at 0930 BT with flight tracks on 1 May 2009.

The figure shows that the entire cloud band assumed a

northeast–southwest formation and moved from north-

west to southeast. Similar to the case on 18 April 2009,

some obvious strong echo regions were embedded in the

cloud band, and the highest radar reflectivity in the em-

bedded convection region was approximately 40dBZ,

which is approximately 5–10 dBZ higher than the re-

flectivities in the surrounding stratiform cloud. The flight

route of aircraft 3625 was not recorded because the GPS

system in the planemalfunctioned during the experiment.

4. Ice crystal habits observed from three aircraft

Ice crystal habits were observed using theDMTCloud

Imaging Probe (CIP) and PMS 2DC, and larger ice

particles were observed via the DMT Precipitation Im-

aging Probe (PIP) and PMS 2DP. CIP and PIP are ba-

sically modified 2DC and 2DP, respectively (Field et al.

2006). A detailed description of the probe features can

be found in Table 1.

TABLE 1. Instruments and measurements collected aboard three BCE aircraft.

Instrument Variable Size range Resolution Aircraft

PCASP-100X Aerosol particles 15 bins, 0.1–3mm Changes in particle size 3625

SPP-200 Aerosol particles 30 bins, 0.1–3mm Changes in particle size 3830

CCN Counter CCN concentration 0.75–10mm Changes in particle size 3525, 3817, 3830

CAS Aerosol and cloud particles 30 bins, 0.6–50mm Changes in particle size 3830

CDP Cloud particles 30 bins, 2–50mm Changes in particle size 3817

FSSP-ER Cloud particles 15 bins, 2–47mm 3mm 3625

OAP-2D-GA2 Cloud and precipitation particles 62 bins, 25–1550mm 25mm 3625

CIP Cloud and precipitation particles 62 bins, 25–1550mm 25mm 3817, 3830

OAP-2D-GB2 Precipitation particles 62 bins,100–6200mm 100mm 3625

PIP Precipitation particles 62 bins,100–6200mm 100mm 3817, 3830

King-LWC Liquid water content (LWC) 0–5 g/m3 — 3625

Hotwire-LWC Liquid water content 0–5 g/m3 3817, 3830

AIMMS-20 Meteorological parameters — — 3817, 3830

2014 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

FIG. 2. Synoptic weather background on 18 Apr 2009: (a) geopotential height (gpm) and wind fields at 500 hPa and 0800 BT, (b) FY-2C

infrared cloud image at 1900 BT. The black square indicates the study area.

MAY 2015 ZHU ET AL . 2015

Typical ice crystal habits identified on 18April 2009 are

shown in Fig. 6. The ice crystal images were classified into

the stratiform clouds and embedded convection regions

according to radar reflectivity and LWC in order to iden-

tify differences of ice crystal habits between two regions.

Theplatelike ice habits in this study included both plate

(Bailey andHallett 2009; Kajikawa andHeymsfield 1989;

Woods et al. 2005) and small spheroid (Connolly et al.

2005). The identification of riming and aggregation

levels of ice crystals may roughly depend on their size

and density. The ice particles with large size and high

density are generally identified as heavily rimed ice

crystals, and those with large size and low density (with

clear structure) are classified as aggregate.

Ice crystal habits identified by three aircraft between

228 and 29.58C were predominantly platelike, needle

column, dendrite, and irregular. But needles and plates

could be identified at all temperatures and the only actual

difference is that of size. It can be seen from Fig. 6 that

most of the ice crystals identified in the stratiform clouds

were platelike and needle column and some dendritic

crystals could be also identified. However, in the em-

bedded convection regions, in addition to platelike and

needle-column crystals, more dendritic crystals existed

between 268 and 29.58C and more aggregate existed be-

tween248 and258C. This phenomenon suggests that the

aggregation process in the embedded convection regions

was active. In addition, it also can be seen from Fig. 6 that

the ice crystal images recorded in embedded convection

regions had higher density than those recorded in the

stratiform regions, suggesting that rimingwere also evident

and larger ice particles such as graupel might occur.

Ice crystal habits observed in this case are generally

consistent with previous laboratory research (Fukuta and

Takahashi 1999; Heymsfield et al. 2010; Magono and

Iwabuchi 1972; Pruppacher and Klett 1997) and field obser-

vations (Evans et al. 2005; Heymsfield et al. 2002; Stith et al.

2002; Woods et al. 2008) as well as with the new ice crystal

habit diagram developed by Bailey and Hallett (2009).

Another case of stratiform clouds with embedded

convection was identified in the study area on 1 May

2009 (Fig. 4). Figure 7 shows the typical ice crystal habits

recorded by three aircraft on 1 May 2009. The temper-

ature ranges identified in the cloud from aircraft 3625,

3817, and 3830 were from 2118 to –168, from 288 to–11.68, and from 08 to –118C, respectively.The ice crystal images recorded by aircraft 3625 at

temperatures between2118 and2168C in Fig. 7a indicate

FIG. 3. Superimposed image of radar reflectivity (shading) at 1900 BT with flight tracks recorded on 18 Apr 2009 (green line represents

3817 path, red line represents 3830 path, blue line represents 3625 path).

2016 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

FIG. 4. Synoptic weather background at 0800 BT on 1 May 2009: (a) geopotential height (gpm) and wind fields of 500 hPa, The black

square indicates the study area. (b) FY-2C visible cloud image at 0900 BT. The red square indicates the study area, the pink line indicates

the land boundary of China, and the blue line represents the Yellow River.

MAY 2015 ZHU ET AL . 2017

that the ice crystal habits were predominantly needle col-

umn, platelike, capped column, dendrite, and irregular. The

ice crystal images recorded by aircraft 3817 at temperatures

between 288 and 211.68C shown in Fig. 7b demonstrate

that ice habits in these conditions were predominantly

platelike, needle column, capped column, and dendrite. Ice

crystal images recorded by aircraft 3830 at temperatures

between 08 and 2118C indicate that the majority of ice

crystals were platelike, dendrite, and irregular. Heavily

rimmed ice crystals were common at these layers.

Differences of ice crystal habits recorded between the

stratiform clouds and embedded convection regions

were basically similar to those recorded on 18 April

2009. However, there were still some differences be-

tween two cases. The 18 April case had more heavily

rimed ice particles and 1 May case had more aggregate.

This will be discussed in detail in the following sections.

The ice crystal habits recorded in the embedded con-

vection regions contained more dendritic crystals and

possessed a heavier riming degree. In addition, the ice

crystals recorded in the stratiform clouds contained

more hexagonal plate crystals (shown in Figs. 7a,b).

In summary, the ice crystal habits sampled in the two

cases with stratiform clouds containing embedded con-

vection were complex; the ice crystal habits between

08 and 2168C were predominantly platelike, needle

column, capped column, dendrite, and irregular. This

indicates that although temperature may have an effect

on crystal habit, it may not be the primary factor.

It can be concluded from this comparison of ice crystal

habits in different clouds that a mixture of several ice

crystal habits can always be identified. However, the ice

crystal habits recorded in the embedded convection re-

gions contained more dendritic crystals and heavier

riming degree, and the ice crystals recorded in the strat-

iform clouds contained more hexagonal plate crystals.

Previous studies suggest that the characteristics of ice

crystal habits in different stratiform clouds with embed-

ded convection are different because of different cloud

conditions (e.g., temperatures, cloud layers, cloud tops,

LWC, etc.) (McFarquhar and Black 2004; Evans et al.

2005; Stark et al. 2013). The deep clouds with high cloud

top and low temperature could produce more complex

ice crystal habits such as bullet, dendrite, and aggregate.

5. Distribution and growth processes of ice crystalhabits

a. The 18 April 2009 case

Table 2 lists observation data recorded from the air-

craft on 18 April 2009. The three aircraft flew at differ-

ent cloud levels, and the maximum LWC recorded from

FIG. 5. Superimposed image of radar reflectivity (shading) at 0930BTwith flight tracks on 1May 2009 (green line represents 3817 path, red

line represents 3830 path).

2018 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

aircraft 3817 and 3830 both exceeded 1 gm23, suggesting

high LWC in the clouds. To investigate the distribution

and growth processes of ice crystal habits in the clouds,

a detailed description of the procedure through which

the three aircraft penetrated the clouds is discussed in

this section.

A cross section of radar reflectivity in the flight path

and relevant observed data recorded from aircraft 3625

on 18 April 2009 are shown in Fig. 8. Aircraft 3625 pri-

marily flew horizontally at 4.8 km (T ’ 268C) in the

early stage and then at 5.1 km (T ’ 29.58C) in the later

stage, penetrating several embedded convection re-

gions. Pristine ice crystal habits identified from the air-

craft were primarily platelike, needle column, and

dendrite between 268 and 29.58C. The cloud particle

spectra derived from 2DC and 2DP shown in Fig. 8b and

Fig. 8c indicate that high concentrations of large ice

particles were present in the embedded convection re-

gion. The formation and growth of these large particles

are difficult to document, but the properties of the ice

habits sampled from aircraft 3625 demonstrate that both

riming and aggregation processes were obvious in the

embedded convection regions. Figure 8d shows that

LWC was not uniformly distributed in the clouds and

high LWC corresponded well with embedded convec-

tion regions. The ice crystal riming degree was heavy in

the embedded convection region with high LWC, such

as that between 1715 and 1728 BT shown in Fig. 8a,

suggesting that riming growth was an important growth

process in the embedded convection regions. At the

same time, it can be seen from the 2DP images show in

Fig. 8a (such as 1727 BT) that the apparent aggregates

appeared between 268 and 29.58C, suggesting that ag-

gregation growth was also an important growth process

within the embedded convection region. But it can be

seen from the ice particle images shown in Fig. 8a that

riming was more prevalent in this case than aggregation,

though aggregates were evident. In contrast, the 1 May

case discussed later had a more dominant aggregation

processes, but also riming.

FIG. 6. Typical ice crystal images recorded by the three aircraft on 18 Apr 2009 (ice crystal

habits at 5.1 and 4.8 km were recorded by 2DC aboard 3625, those at 4.2 km were recorded by

CIP aboard 3817, those at 3.6 km were recorded by CIP aboard 3830).

MAY 2015 ZHU ET AL . 2019

Aircraft 3817 penetrated the clouds at lower layers in

248 to approximately 258C conditions on 18 April 2009.

As the hotwire-LWC equipment used on aircraft 3817 was

in poor condition, no LWC data are available. Figure 9

shows the radar reflectivity cross section in the flight path

and relevant spectra of cloud particles sampled through

CIP and PIP. Figure 9a shows that aircraft 3817 penetrated

both weak and strong embedded convection regions as

well as the stratiform region. The pristine ice crystals from

248 to approximately 258C were predominantly needle

column; however, a wide variety of ice crystal habits, such

as heavily rimed dendrite and ice aggregate, were also

sampled at these layers. From 1743 to 1745 BT, a weak

embedded convection region with a thickness of less than

2km and a cloud top at less than 5km were sampled. The

corresponding CTT was warmer than 288C and the 08–88C layers are suitable for plate and needle-column

crystal formation (Takahashi et al. 1991, 1986). In addi-

tion, most of the cloud particles were single crystals

similar to those recorded by the PIP probe at 1744 BT,

suggesting that the ice crystal aggregation process was

weak and that the growth process should be dominated

by deposition growth.

At 1748 BT, when the CTT was colder than 2138C,aircraft 3817 sampled lightly rimed dendritic crystals in

the cloud. Laboratory studies (Takahashi et al. 1991,

1986) and field (Bailey andHallett 2009; Heymsfield and

Kajikawa 1987) have shown that the optimal tempera-

ture for dendritic crystal formation is approximately

2158C. However, at 4.2 km, the temperature recorded

from aircraft 3817 was approximately 258C. Therefore,lightly rimed dendritic crystals likely fell from higher,

colder cloud layers.

From 1748 to 1750 BT, the heavily rimed and aggre-

gated crystals (such as the images recorded via CIP

probing at 1749 and 1750 BT) were sampled in the

strong embedded convection region, suggesting that

both riming and aggregation processes were important

in the strong embedded convection region.

Since the maximum reflectivity of this cloud system

reached about 50dBZ and with the cloud base tempera-

ture of 128C, cloud-top temperature of2288C (estimated

from radar data in Fig. 9), and LWC of 1gm23, it is

possible for occurrence of graupel in the lower regions of

the embedded convection. Graupel formation has been

found in several other studies with embedded convection.

Figure 10 shows the penetration data recorded by

aircraft 3830 on 18April 2009. From 1800 to 1805 BT, ice

crystal habits recorded at a layer of 3.6 km (T ’ 228C)was dominated by rimed and aggregated plates and

FIG. 7. Typical ice crystal images recorded by the three aircraft

on 1 May 2009: (a) images recorded by 2DC aboard 3625, (b) im-

ages recorded by CIP aboard 3817, (c) images recorded by CIP

aboard 3830.

TABLE 2. Observed results from aircraft taken on 18 Apr 2009.

Number Time (BT)

Altitude

(m) T (8C)LWCmv

(gm23)

3625 1614–1900 4800–5100 From 268 to –9.58 1.5

3817 1655–1904 4200 258 Data error

3830 1705–1855 2700–3600 From 58 to –38 1.1

2020 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

FIG. 8. The penetration data recorded by aircraft 3625 on 18 Apr 2009: (a)

cross-section of radar reflectivity for flight path, ice crystal images recorded

with 2DC (panel top) and 2DP (panel bottom), and T (black line); (b) 2DC

instantaneous spectrum; (c) 2DP instantaneous spectrum; and (d) LWC.

MAY 2015 ZHU ET AL . 2021

FIG. 9. The penetration data recorded by aircraft 3817 on 18 Apr 2009: (a) cross section of radar

reflectivity for flight path, ice crystal images recorded through CIP (panel top)and PIP (panel

bottom), and T (black line); (b) CIP instantaneous spectrum; and (c) PIP instantaneous

spectrum.

2022 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

FIG. 10. The cloud penetration data recorded by aircraft 3830 on 18

Apr 2009: (a) cross section of radar reflectivity for flight path and ice

crystal habits recorded through CIP, where black solid line represents

flight track, red dashed line represents the 08C layer; (b) CIP in-

stantaneous spectrum; and (c) PIP instantaneous spectrum, (d) temper-

ature (black) and LWC (blue).

MAY 2015 ZHU ET AL . 2023

needle columns in the embedded convection region

(Fig. 10a). Previous laboratorial studies (Pruppacher and

Klett 1997; Bailey and Hallett 2009) have shown that

temperatures at this layer are suitable for plate formation.

Thus, the aggregated plates recorded at this layer may

have formed locally because of the longevity of the

stratiform region. In addition, Fig. 10a shows that crystals

identified at 1803BT included somepristine columns. This

is an interesting finding given that most previous studies

(Bailey and Hallett 2009; Crosier et al. 2011) show that

this crystal habit typically forms at temperatures lower

than228C. Ice formation at this temperature is very rare.

A more plausible explanation could be a secondary ice

formation event driven by Hallett–Mossop processes.

Both aggregation and riming processes became active

when the aircraft passed from the shallow stratiform

region (1801 BT) to the relatively deep embedded

convection region (1804 BT) with an increase in super-

cooled cloud water (Fig. 10d). The corresponding

spectra recorded by the CIP and PIP probes also

broadened rapidly (Figs. 10b,c). From 1805 to 1811 BT,

the aircraft flew downward and passed through the

melting layer as it descended. A small number of melt-

ing ice crystals were recorded by the CIP probewhile the

aircraft passed through the melting layer. With the

melting of ice particles, the cloud particle spectra grad-

ually narrowed because of changes in particle diameters

in the transition from snow to raindrop form. From 1811

to 1820 BT, the aircraft flew at 2.7 km (T ’ 58C), andparticles at this layer were dominated by cloud and rain

drops. From 1813 to 1815 BT, PSD broadened again

owing likely to larger particles aloft melting into larger

raindrops with the increase of cloud top. Upon traveling

from 1815 to 1817 BT, the aircraft passed through a re-

gion of strong radar echo (.30 dBZ). However, at this

moment, the CIP and PIP probes malfunctioned, and

thus, the PSDs recorded in this region were incorrect,

and the particle images recorded at 1816 BT were

dominated by larger raindrops.

b. The 1 May 2009 case

Table 3 lists observations from the three aircraft re-

corded on 1 May 2009. The three aircraft flew at dif-

ferent cloud levels, and the maximum LWC identified in

this case were lower than those recorded on 18 April

2009, suggesting that the ice crystal riming process was

relatively weaker than the process recorded on 18 April

2009. Therefore, pristine ice crystal habits recorded on 1

May 2009 were much clearer. The flight route of aircraft

3625 was not recorded because the GPS system mal-

functioned during the experiment. Therefore, only the

cloud penetration data recorded by aircraft 3817 and

3830 will be discussed in this section.

Figure 11 shows the cloud penetration data recorded

by aircraft 3817 for 1 May 2009. It indicates that aircraft

3817 penetrated both stratiform and embedded con-

vection regions with temperatures ranging from 278 to211.68C throughout the flight. The CIP images show

that the ice crystal habits sampled in the stratiform cloud

region were predominantly plate and needle column

(Fig. 11a, 0925 BT, 1017 BT), and those observed in the

embedded convection regions were predominantly

dendrites (Fig. 11a, 0937 BT, 1026 BT). The ice crystal

images sampled by CIP and PIP show that both riming

and aggregation processes existed in the embedded

convection regions, but riming was not as strong

a growth process as aggregation.

The particle images recorded by PIP probe (Fig. 11a)

show that aggregates were very common in the em-

bedded convection region over a height range of 4.2 km

(T ’ 288C) to 4.9 km (T ’ 2128C), suggesting that ag-

gregation was also a primary growth process in the em-

bedded convection regions. At the same time, the regions

of stronger reflectivity (.20dBZ) contained larger ag-

gregates. For example, the aggregates recorded at 0937,

0942, and 1005 BT were identifiably larger than those

recorded at 0926, 0948, and 1017 BT, indicating that the

aggregation process was more active in the embedded

convection region than in the stratiform regions.

Riming growth at this layer also showed a nonuniform

distribution. The riming degrees of the ice crystals

identified at 0937 and 1000 BT were clearly heavier than

those recorded at 0934 BT. The LWC detected at 0937

BT were higher than 0.1 gm23, but the LWC at 0934 BT

were lower than 0.02 gm23, suggesting that the non-

uniform riming growth distributions found in the clouds

were caused by variations of LWC in the stratiform and

embedded convection clouds. These results echo those

of the 18 April 2009 case. After 0957 BT was reached,

the Hotwire-LWC malfunctioned, so there are no

available data on cloud LWC. However, the particle

images show that riming growth was still nonuniform,

which is consistent with the results of Stark et al.

(2013), who found that the riming degree of ice crystals

is influenced by cloud updrafts in extratropical cy-

clones. The authors also noted that heavy riming in

high updraft regions was caused by high supercooled

LWC in these areas.

TABLE 3. Observed results from aircraft taken on 1 May 2009.

Number Time (BT) Altitude (m) T (8C)LWCmv

(gm23)

3625 0827–1120 4800–5400 From 2118 to –168 0.16

3817 0846–1105 4200–4800 From 288 to –11.68 0.21

3830 0831–1134 1700–4900 From 48 to –128 1.0

2024 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

In embedded convection regions (such as at 0938 BT),

both riming and aggregation processes were active

(shown in Fig. 11a), and the PSDs recorded by PIP

probe broadened in the embedded convection regions

(Fig. 11c). Although both riming and aggregation pro-

cesses contributed to the PSD broadening, the particle

images recorded by PIP probes show that the aggrega-

tion process was more critical to the broadening of PSDs

at this layer. Therefore, comparing with 18 April case,

the 1 May case had a more dominant aggregation pro-

cesses but also riming.

The ice crystal habits at this layer were also strongly

affected by CTT, which varied from 288C at the lowest

cloud top of approximately 4.5 km (such as 0925 BT,

1017 BT) to approximately 227.38C at the highest

cloud top of approximately 7 km (such as 0935 BT,

1008 BT). In addition to plate, needle-column, and den-

dritic crystals, capped-column crystals also appeared as

the CTT was colder than2188C (Fig. 11a, 0934 BT). The

initial columns likely formed over a temperature range

of 2188 to 2258C, and then capped columns formed

as these initial columns fell through the planar-crystal

growth region where temperatures ranged from 2128 to2188C (Heymsfield et al. 2002).

The particle images recorded by PIP probe also show

that the large aggregates typically consisted of dendritic

crystals (Fig. 11a). Higher cloud tops (.6km) with

lower temperatures (2208–278C) are suitable for gen-

erating dendritic and capped-column crystals. Dendritic

aggregate more easily than other crystal habits and can

easily capture supercooled cloud droplets (Hashino and

Tripoli 2011; Hobbs et al. 1974). In addition, Fig. 11b

shows that the embedded convection regions with higher

cloud tops (such as 0933 BT, 1009 BT) often contained

more ice crystals than other regions, and high concen-

trations of ice crystals can facilitate the aggregation pro-

cess (Cooper and Lawson 1984; Dye et al. 1976; Holroyd

and Jiusto 1971). Because the embedded convection re-

gions with higher cloud tops contained more ice crystals,

especially dendritic crystals, PSD broadening should be

apparent and precipitation efficiencies should also be

higher. Hobbs et al. (1980) found that precipitation effi-

ciency in wide, cold-frontal rainbands with generating

cells reaches levels of at least 80%, while levels in warm-

sector and narrow, cold-frontal rainbands reach only

40%–50% and 30%–50%, respectively.

Figure 12 shows the cloud penetration data recorded

by aircraft 3830 on 1 May 2009. Although the crystal

habits recorded by CIP probe varied greatly as the flight

altitude of aircraft 3830 changed considerably, ice crys-

tals in this region were dominated by heavily rimed

dendritic and irregular particles. More dendritic crystals

were found in embedded convection regions (such as

FIG. 11. As in Fig. 9, but on 1May 2009 and with (d) Hotwire-LWC

liquid water content.

MAY 2015 ZHU ET AL . 2025

0938 BT, 0958 BT), which is consistent with the ob-

served results of aircraft 3817.

The particle images recorded via CIP probing show

that most of the ice crystals were heavily rimed, in-

dicating that riming is an important growth process for

ice crystals in clouds in the range of 2.3 km (T’ 08C) to3.6 km (T ’ 258C). Compared to the results observed

from aircraft 3817, the riming process was more active in

the lower cloud layer because of the high supercooled

LWC in this layer. Figure 12d, which shows the LWC

FIG. 12. As in Fig. 11, but by aircraft 3830.

2026 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

distribution detected from aircraft 3830, demonstrates

that the maximum LWC in the clouds was more than

1 gm23, and most of regions observed from aircraft 3830

exhibited higher LWC values than the upper layer ob-

served from aircraft 3817. The particle images recorded

by PIP probe (Fig. 12a) suggest that the aggregation was

an important growth process in clouds in the range of

2.3 km (T’ 08C) to 4.9 km (T’2128C) and in stronger

embedded convection regions especially.

In summary, the distribution and growth processes of

ice crystal habits in two cases of stratiform clouds with

embedded convection show that the distribution of ice

crystal habits varied greatly between stratiform cloud

regions and embedded convection regions. Higher

concentrations of large ice particles, such as dendritic

crystals, and higher supercooled water content were

found in the embedded convection regions. Both rim-

ing and aggregation processes were more active in the

embedded convection regions than in the stratiform

clouds.

6. The broadening rate of PSDs

To investigate how the PSDs changed with vertical

distance, the three aircraft flew horizontally at different

cloud levels while maintaining constant longitude and

latitude locations (beginning at 0930 BT, 40.8078N,

114.9828E, and ending at 0940 BT, 40.9848N, 115.2928E).The flight levels of aircraft 3625, 3817, and 3830 were

4.8 km (T ’ 211.68C), 4.2 km (T ’ 288C), and 3.6 km

(T’258C), respectively. Although GPS data were lost,

the flight route of aircraft 3625 was recovered for the

first 10min from the pilot’s flight log.

Figure 13a shows the flight routes of the three aircraft

superimposedwith radar echo distributions from 0930 to

0940 BT. Figure 13a shows that, from 0935 to 0940 BT,

the three aircraft simultaneously passed through an em-

bedded, strong echo region. In this region, the LWC de-

tected from the three aircraft reached its peak value

(Figs. 13b,c,d), and the LWC detected from aircraft 3830

was higher than 1gm23.

The averaged spectra between 0935 and 0940 BT at

convection areas are shown in Fig. 13f. In the embedded

convection region, the Marshall–Palmer (MP) fit of the

PSDs shows that the PSDs broadened from 4.8km (T ’211.68C) to 3.6km (T’258C) but most noticeably from

4.8km (T ’ 211.68C) to 4.2km (T ’ 288C). Figure 13e

provides a comparison of particle mean volume di-

ameters. Figure 13e shows that the particle mean volume

diameters at 4.8 km (T ’ 211.68C), 4.2 km (T ’ 288C),and 3.6km (T ’ 258C) were 1.2mm (Dmv4.8 5 1.2mm),

3mm (Dmv4.2 5 3mm), and 3.2mm (Dmv3.6 5 3.2mm),

respectively. The PSD broadening rate from 4.8 km

(T ’ 211.68C) to 4.2 km (T ’ 288C) can be calculated

as follows:

V4:2-4:8km 5Dmv4:22Dmv4:8

0:6 km5

3mm2 1:2mm

0:6 km

5 3mmkm21 . (1)

From 4.2 km (T ’ 288C) to 3.6 km (T ’ 258C), thePSD broadening rate is

V3:6-4:2km5Dmv3:62Dmv4:2

0:6 km5

3:2mm23mm

0:6 km

’ 0:33mmkm21 . (2)

The three aircraft passed through stratiform clouds

from 0930 to 0935 BT (Fig. 13a), and the averaged

spectra between 0930 and 0935 BT at stratiform region

are shown in Fig. 13g. In the stratiform region, PSDs

broadened from 4.8 km (T ’ 211.68C) to 3.6 km (T ’258C), but most noticeably from 4.8km (T ’ 211.68C)to 4.2 km (T ’ 288C). Figure 13e shows that the par-

ticle mean volume diameters at 4.8 km (T ’ 211.68C),4.2 km (T’288C), and 3.6 km (T’258C) were 1.3mm

(Dmv4.85 1.3mm), 2.8mm (Dmv4.25 2.8mm), and 2.9mm

(Dmv3.6 5 2.9mm), respectively. The PSD broadening

rate from 4.8 km (T ’ 211.68C) to 4.2 km (T ’ 288C)can be calculated as follows:

V4:2-4:8km5Dmv4:22Dmv4:8

0:6 km5

2:8mm2 1:3mm

0:6 km

5 2:5mmkm21 . (3)

From 4.2 km (T ’ 288C) to 3.6 km (T ’ 258C), thePSD broadening rate is

V3:6-4:2km5Dmv3:62Dmv4:2

0:6 km5

2:9mm2 2:8mm

0:6 km

’ 0:17mmkm21 . (4)

The PSD broadening processes in the convection and

stratiform regions demonstrate that with the decrease of

height, the PSDs broadened in both regions. However,

the broadening rates between 4.8km (T’ 211.68C) and4.2km (T ’ 288C) were larger than those between

4.2km (T ’ 288C) and 3.6km (T ’ 258C). This phe-nomenon may be because the temperature range be-

tween 4.8km (T’211.68C) and 4.2km (T’288C) wasmore suitable for dendritic crystal formation. Furthermore,

once aggregates are formed, riming is more effective as

well (Bailey and Hallett 2009; Hashino and Tripoli 2011).

The PSD broadening rate in the convection region

was larger than that in the stratiform region. At the same

MAY 2015 ZHU ET AL . 2027

time, Figs. 13f and 13g show that concentrations of ice

particles of sizes less than 1.2mm (convection region)

and 1mm (stratiform region) decreased with the de-

crease in height owing to collision and collection forces

on larger ice particles. Heymsfield et al. (2002) showed

from TRMMfield campaigns that in deep tropical cirrus

and stratiform precipitating cloud, PSDs broaden from

the cloud top toward the cloud base. The broadening

rate was found to be 1–3mmkm21, and concentrations

of particles less than 1mm in size decreased with de-

creasing height. Our case showed similar characteristics

and results on the vertical evolution of PSDs. Particle

images recorded by PIP probe show that the aggregation

process was the main cause of the broadening phe-

nomenon (Fig. 13a).

The PSD broadening rate in the embedded convec-

tion region was found to be higher in our study. Lawson

and Zuidema (2009) analyzed 2DP data from stratiform

clouds with embedded convection regions at high lati-

tudes and found that in embedded convection regions

(CTT’2278C), the maximum diameter of ice particles

at 4 km (T ’ 2138C) was 15mm, and at 2.2 km (T ’08C), the value was approximately 30mm. The PSD

broadening rate was also approximately 6.25mmkm21.

While the broadening rate was higher in that study, their

results were similar to the results of this study, suggest-

ing that PSD broadening rates are higher in embedded

convection regions.

PSD broadening rates were higher in the embedded

convection than in the stratiform clouds. The reason is

too complicated to clearly clarify by the observations of

this study. Studies have shown that riming does enhance

aggregation (Hallgren and Hosler 1960), but when high

concentrations of dendritic crystals exist, aggregation

occurs rapidly (Hashino and Tripoli 2011; Hobbs et al.

1974), creating larger particles favored for riming

growth. The observations in this paper do not consis-

tently support one over the other. Some of the images

show large aggregates with little riming while others

show heavily rimed particles. The importance of one

process over another probably varies, but both are ac-

tive and lead to PSD broadening. Therefore, in the

higher LWC regions of embedded convection, both

particle growth processes are important and feedback

on one another.

7. Conclusions and discussion

In this study, we investigated ice crystal habits, dis-

tribution, and growth processes in two cases of strati-

form clouds with embedded convection on 18April 2009

and 1 May 2009 in northern China. Three aircraft were

employed to observe clouds simultaneously at different

FIG. 13. (a) The flight tracks of three aircraft 3625 (red line), 3817

(black line), and 3830 (green line) superimposed with the cross

section of radar reflectivity on 1May 2009; (b) RH (green line, %),

LWC (black line, gm23), and temperature (blue line, 8C) recordedfrom aircraft 3625; (c) as in (b), but from aircraft 3630; (d) as in (b),

but from 3630 aircraft; (e) comparison of particle mean volume

diameter; (f), average spectrum between 0935 and 0940 BT in

convection region; and (g) as in (f), but between 0930 and 0935 BT

in stratiform region.

2028 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

levels of the clouds while maintaining a constant longi-

tude and latitude position. This flight pattern facilitated

an investigation of PSD change with height in the clouds.

This combination of airborne observations and radar

echo distributions allows us to conduct an accurate

evaluation of ice particle distributions and main growth

processes in stratiformcloudswith embedded convection.

The 18 April 2009 case observed prefrontal clouds,

and the convection embedded in stratiform clouds had

a stronger radar echo with a maximum reflectivity of

FIG. 13. (Continued)

MAY 2015 ZHU ET AL . 2029

approximately 50 dBZ, which was stronger than re-

flectivities in the surrounding stratiform clouds of ap-

proximately 10–20 dBZ. The 1 May 2009 case observed

postfrontal clouds with a maximum reflectivity of ap-

proximately 40 dBZ in the embedded convection region,

which was stronger than reflectivities in the surrounding

stratiform clouds of approximately 5–10 dBZ. Two cases

studied here comprised typical stratiform clouds with

embedded convection.

The ice crystal habits sampled in the two cases of

stratiform clouds with embedded convection were com-

plex, and the ice crystal habits present between 08 and2168C were predominantly platelike, needle column,

capped column, dendrite, and irregular. Aggregates and

heavily rimed ice crystals were common in the clouds.

The ice crystal habits recorded from the aircraft were

strongly affected by CTT and cloud type. First, the plate

and needle-column crystals were predominant in clouds

with CTT warmer than 288C, and dendritic and capped-

column crystals were found only in CTT conditions colder

than 2138 and 2188C, respectively. Second, comparisons

of ice crystal habits in different clouds show that a mixture

of several ice crystal habits can always be found. However,

ice crystal habits identified in embedded convection re-

gions contained more dendrites and possessed a heavier

riming degree, and the ice crystals recorded in stratiform

clouds contained more hexagonal plate crystals.

Our case showed similar results to those of Heymsfield

et al. (2002), who identified ice crystal habits recorded

with a cloud particle imager (CPI) probe in deep tropical

cirrus and stratiform precipitating clouds, and concluded

that small, pristine particles are observable in clouds

formed by more transient convective clouds, with CTT

controlling the habits. Most of larger particles were

aggregates and rimed ice particles that were sampled

near deep convective clouds with sustained convection.

The dominant growth process is different for different

clouds. In shallow stratiform clouds, the growth process

is dominated by deposition growth, but in deep strati-

form clouds or embedded convection, both riming and

aggregation are important. The two cases in this study

show that riming was more prevalent in 18 April case

than aggregation, though aggregates were evident. In

contrast, the 1 May case had a more dominant aggrega-

tion processes but also riming. These results are generally

in agreement with previous studies conducted in other

regions, such asWashington, United States (Herzegh and

Hobbs 1980); Newfoundland, Canada (Lawson et al.

1998); Kwajalein, United States (Heymsfield et al. 2002);

and the Arctic (Lawson and Zuidema 2009; Lawson et al.

1998, 1993).

In addition, with changes of height in deep stratiform

clouds or embedded convection, ice crystal growth

processes were different. Both aggregation and riming

processes constituted main growth processes between

08 and 211.68C, but riming processes were more active

in lower cloud layers because of high supercooled LWC

in the clouds. Hou et al. (2010) analyzed 2DC and 2DP

data of a stratiform cloud with embedded convection

regions in northeastern China and found that ice parti-

cles primarily form and grow at layers between 4.0 and

5.5 km (from 218 to –88C), with depositional growth

dominating this process in addition to aggregation. In

contrast, our study shows that riming was also an im-

portant growth process in embedded clouds owing to

high supercooled LWC in the clouds.

The PSD broadening processes in convection and

stratiform regions show that with a decrease in height,

PSD broadened in both regions, but the broadening

rates between 4.8 km (T ’ 211.68C) and 4.2 km (T ’288C) were larger than those between 4.2 km (T ’288C) and 3.6 km (T ’ 258C). In addition, the PSD

broadening rates in embedded convection regions were

larger than those in stratiform clouds, as the aggregation

and riming processes of ice particles in embedded con-

vection regions were active. The existence and extent of

supercooled liquidwater is critical to enhancing riming and

aggregation processes in embedded convection region.

It should be noted that the quality of the microphys-

ical data due to the relatively low resolution of airborne

instruments and the limited data collections in this study

lead to many difficulties and uncertainties for analyzing

classification of particle habits and their growth pro-

cesses. In many cases, we cannot separate plates from

irregular shapes unless the images appear to be larger

size and identify riming and aggregation processes that

come first. On the one hand, riming comes first and then

enhances aggregation. On the other hand, when high

concentrations of dendritic crystals exist, aggregation

occurs rapidly, creating larger particles favored for

riming growth. The observations in this paper do not

consistently support one over the other. Some of the

images show large aggregates with little riming, and

others show heavily rimed particles. The importance of

one process over another probably varies, but both are

active and lead to PSD broadening. Therefore, in the

higher LWC regions of embedded convection, both

particle growth processes are important and feedback

on one another.

The ice particle shattering problems induced by the

older-technology cloud probes (Field et al. 2006; Korolev

et al. 2013) used in this study are not considered in this

paper, which was likely taking place during observations

in these clouds and influencing the data quality. All

these uncertainties need to be further clarified in future

studies.

2030 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72

Acknowledgments. The authors are grateful for criti-

cal and constructive comments from three anonymous

reviewers and editor, which greatly improved the quality

of this paper. This research was supported by the Key

Project of National Science and Technology Pillar

Program (Grant 2006BAC12B03), the Research and

Development Special Fund for Public Welfare In-

dustry (Meteorology) (Grants GYHY200806001 and

GYHY201306040), the Third Tibetan Plateau Scientific

Experiment: Observations for Boundary Layer and

Troposphere (GYHY201406001), the Key Project of

Basic Scientific Research and Operation fund of Chinese

Academy of Meteorological Sciences (Grant 2013Z009),

and theOpen Funding fromKeyLaboratory forAerosol-

Cloud-Precipitation of China Meteorological Adminis-

tration (KDW1402).

REFERENCES

Bailey, M. P., and J. Hallett, 2009: A comprehensive habit

diagram for atmospheric ice crystals: Confirmation from the

laboratory, AIRS II, and other field studies. J. Atmos. Sci.,

66, 2888–2899, doi:10.1175/2009JAS2883.1.

Carey, L. D., J. Niu, P. Yang, J. A. Kankiewicz, V. E. Larson, and

T. H. Vonder Haar, 2008: The vertical profile of liquid and

ice water content in midlatitude mixed-phase altocumulus

clouds. J. Appl. Meteor. Climatol., 47, 2487–2495, doi:10.1175/

2008JAMC1885.1.

Connolly, P. J., C. P. R. Saunders, M. W. Gallagher, K. N. Bower,

M. J. Flynn, T.W. Choularton, J.Whiteway, andR. P. Lawson,

2005: Aircraft observations of the influence of electric fields on

the aggregation of ice crystals.Quart. J. Roy.Meteor. Soc., 131,1695–1712, doi:10.1256/qj.03.217.

Cooper, W. A., and R. P. Lawson, 1984: Physical interpretation of

results from the HIPLEX-1 experiment. J. Appl. Meteor. Cli-

matol., 23, 523–540, doi:10.1175/1520-0450(1984)023,0523:

PIORFT.2.0.CO;2.

Crosier, J., and Coauthors, 2011: Observations of ice multiplication

in a weakly convective cell embedded in supercooled mid-

level stratus. Atmos. Chem. Phys., 11, 257–273, doi:10.5194/

acp-11-257-2011.

Dye, J. E., G. Langer, V. Toutenhoofd, T. W. Cannon, and C. A.

Knight, 1976: Use of a sailplane to measure microphysical

effects of silver iodide seeding in cumulus clouds. J. Appl.

Meteor., 15, 264–274, doi:10.1175/1520-0450(1976)015,0264:

UOASTM.2.0.CO;2.

Evans, A. G., J. D. Locatelli, M. T. Stoelinga, and P. V. Hobbs,

2005: The IMPROVE-1 storm of 1–2 February 2001. Part II:

Cloud structures and the growth of precipitation. J. Atmos.

Sci., 62, 3456–3473, doi:10.1175/JAS3547.1.

Field, P. R., A. J. Heymsfield, and A. Bansemer, 2006: Shat-

tering and particle interarrival times measured by optical

array probes in ice clouds. J. Atmos. Oceanic Technol., 23,

1357–1371, doi:10.1175/JTECH1922.1.

Fu, D., and X. Guo, 2012: A cloud-resolving simulation study on

the merging processes and effects of topography and envi-

ronmental winds. J. Atmos. Sci., 69, 1232–1249, doi:10.1175/JAS-D-11-049.1.

Fukuta, N., and T. Takahashi, 1999: The growth of atmospheric

ice crystals: A summary of findings in vertical supercooled

cloud tunnel studies. J. Atmos. Sci., 56, 1963–1979, doi:10.1175/

1520-0469(1999)056,1963:TGOAIC.2.0.CO;2.

Guo, X., D. Fu, and J. Wang, 2006: Mesoscale convective

precipitation system modified by urbanization in Beijing City.

Atmos. Res., 82, 112–126, doi:10.1016/j.atmosres.2005.12.007.

——, ——, X. Guo, and C. Zhang, 2014: A case study of aerosol

impacts on summer convective clouds and precipitation over

northern China. Atmos. Res., 142, 142–157, doi:10.1016/

j.atmosres.2013.10.006.

Hallgren, R. E., and C. L. Hosler, 1960: Preliminary results on the

aggregation of ice crystals. Physics of Precipitation, Geophys.

Monogr., Vol. 5, Amer. Geophys. Union, 257–263.

Hashino, T., and G. J. Tripoli, 2011: The spectral ice habit

prediction system (SHIPS). Part IV: Box model simulations

of the habit-dependent aggregation process. J. Atmos. Sci.,

68, 1142–1161, doi:10.1175/2011JAS3667.1.

Herzegh, P. H., and P. V. Hobbs, 1980: The mesoscale and mi-

croscale structure and organization of clouds and precipitation

in midlatitude cyclones. II: Warm-frontal clouds. J. Atmos.

Sci., 37, 597–611, doi:10.1175/1520-0469(1980)037,0597:

TMAMSA.2.0.CO;2.

——, and ——, 1981: The mesoscale and microscale structure and

organization of clouds and precipitation in midlatitude cy-

clones. IV: Vertical air motions and microphysical structures

of prefrontal surge clouds and cold-frontal clouds. J. Atmos.

Sci., 38, 1771–1784, doi:10.1175/1520-0469(1981)038,1771:

TMAMSA.2.0.CO;2.

Heymsfield, A. J., and M. Kajikawa, 1987: An improved ap-

proach to calculating terminal velocities of plate-like crys-

tals and graupel. J. Atmos. Sci., 44, 1088–1099, doi:10.1175/

1520-0469(1987)044,1088:AIATCT.2.0.CO;2.

——, A. Bansemer, P. R. Field, S. L. Durden, J. L. Stith, J. E. Dye,

W. Hall, and C. A. Grainger, 2002: Observations and param-

eterizations of particle size distributions in deep tropical cirrus

and stratiform precipitating clouds: Results from in situ ob-

servations in TRMM field campaigns. J. Atmos. Sci., 59, 3457–

3491, doi:10.1175/1520-0469(2002)059,3457:OAPOPS.2.0.

CO;2.

——, C. Schmitt, A. Bansemer, and C. H. Twohy, 2010: Improved

representation of ice particle masses based on observations in

natural clouds. J. Atmos. Sci., 67, 3303–3318, doi:10.1175/

2010JAS3507.1.

Hobbs, P. V., and J. D. Locatelli, 1978: Rainbands, precipitation

cores and generating cells in a cyclonic storm. J. Atmos.

Sci., 35, 230–241, doi:10.1175/1520-0469(1978)035,0230:

RPCAGC.2.0.CO;2.

——, and A. L. Rangno, 1990: Rapid development of high ice

particle concentrations in small polar maritime cumuliform

clouds. J. Atmos. Sci., 47, 2710–2722, doi:10.1175/1520-0469

(1990)047,2710:RDOHIP.2.0.CO;2.

——, S. Chang, and J. D. Locatelli, 1974: The dimensions and ag-

gregation of ice crystals in natural clouds. J. Geophys. Res., 79,

2199–2206, doi:10.1029/JC079i015p02199.

——, T. J.Matejka, P. H. Herzegh, J. D. Locatelli, and R.A. Houze,

1980: The mesoscale and microscale structure and organization

of clouds and precipitation in midlatitude cyclones. I: A case

study of a cold front. J. Atmos. Sci., 37, 568–596, doi:10.1175/

1520-0469(1980)037,0568:TMAMSA.2.0.CO;2.

Hogan,R. J., P. N. Francis, H. Flentje, A. J. Illingworth,M.Quante,

and J. Pelon, 2003: Characteristics of mixed-phase clouds. I:

Lidar, radar and aircraft observations from CLARE’98.

Quart. J. Roy. Meteor. Soc., 129, 2089–2116, doi:10.1256/

rj.01.208.

MAY 2015 ZHU ET AL . 2031

Holroyd, E. W., and J. E. Jiusto, 1971: Snowfall from a heavily

seeded cloud. J. Appl. Meteor., 10, 266–269, doi:10.1175/

1520-0450(1971)010,0266:SFAHSC.2.0.CO;2.

Hou, T., H. Lei, and Z. Hu, 2010: A comparative study of the mi-

crostructure and precipitation mechanisms for two stratiform

clouds in China. Atmos. Res., 96 (2–3), 447–460, doi:10.1016/

j.atmosres.2010.02.004.

Houze, R. A., S. A. Rutledge, T. J. Matejka, and P. V. Hobbs, 1981:

The mesoscale and microscale structure and organization of

clouds and precipitation inmidlatitude cyclones. III:Airmotions

and precipitation growth in a warm-frontal rainband. J. Atmos.

Sci., 38, 639–649, doi:10.1175/1520-0469(1981)038,0639:

TMAMSA.2.0.CO;2.

Kajikawa, M., and A. J. Heymsfield, 1989: Aggregation of ice

crystals in cirrus. J. Atmos. Sci., 46, 3108–3121, doi:10.1175/1520-0469(1989)046,3108:AOICIC.2.0.CO;2.

Korolev, A. V., M. P. Bailey, J. Hallett, and G. A. Isaac, 2004:

Laboratory and in situ observation of deposition growth of

frozen drops. J. Appl. Meteor., 43, 612–622, doi:10.1175/

1520-0450(2004)043,0612:LAISOO.2.0.CO;2.

——, E. F. Emery, J. W. Strapp, S. G. Cober, andG. A. Isaac, 2013:

Quantification of the effects of shattering on airborne ice

particle measurements. J. Atmos. Oceanic Technol., 30,2527–2553, doi:10.1175/JTECH-D-13-00115.1.

Lawson, R. P., and P. Zuidema, 2009: Aircraft microphysical and

surface-based radar observations of summertime arctic

clouds. J. Atmos. Sci., 66, 3505–3529, doi:10.1175/

2009JAS3177.1.

——, R. E. Stewart, J. W. Strapp, and G. A. Isaac, 1993: Aircraft

observations of the origin and growth of very large snowflakes.

Geophys. Res. Lett., 20, 53–56, doi:10.1029/92GL02917.

——, ——, and L. J. Angus, 1998: Observations and numerical

simulations of the origin and development of very large

snowflakes. J. Atmos. Sci., 55, 3209–3229, doi:10.1175/

1520-0469(1998)055,3209:OANSOT.2.0.CO;2.

Lu, G., and X. Guo, 2012: Distribution and origin of aerosol and

its transform relationship with CCN derived from the spring

multi-aircraft measurements of Beijing Cloud Experi-

ment (BCE). Chin. Sci. Bull., 57, 2460–2469, doi:10.1007/

s11434-012-5136-9.

Magono, C., and T. Iwabuchi, 1972: A laboratory experiment

on the electrification of ice crystals. Arch. Meteor., Geophys.

Bioklimatol., 21A, 287–298.

Matejka, T. J., R. A. Houze, and P. V. Hobbs, 1980: Microphysics

and dynamics of clouds associatedwithmesoscale rainbands in

extratropical cyclones.Quart. J. Roy. Meteor. Soc., 106, 29–56,

doi:10.1002/qj.49710644704.

McFarquhar, G. M., and R. A. Black, 2004: Observations of

particle size and phase in tropical cyclones: Implications for

mesoscale modeling of microphysical processes. J. Atmos.

Sci., 61, 422–439, doi:10.1175/1520-0469(2004)061,0422:

OOPSAP.2.0.CO;2.

Ono, A., 1969: The shape and riming properties of ice crystals in

natural clouds. J. Atmos. Sci., 26, 138–147, doi:10.1175/

1520-0469(1969)026,0138:TSARPO.2.0.CO;2.

Pruppacher, H. R., and J. D. Klett, 1997: Microphysics of Clouds

and Precipitation. Kluwer Academic Publications, 44 pp.

Rutledge, S. A., and P. Hobbs, 1983: The mesoscale and

microscale structure and organization of clouds and

precipitation in midlatitude cyclones. VIII: A model for the

‘‘seeder-feeder’’ process in warm-frontal rainbands. J. Atmos.

Sci., 40, 1185–1206, doi:10.1175/1520-0469(1983)040,1185:

TMAMSA.2.0.CO;2.

Stark, D., B. A. Colle, and S. E. Yuter, 2013: Observed micro-

physical evolution for two east coast winter storms and the

associated snow bands. Mon. Wea. Rev., 141, 2037–2057,

doi:10.1175/MWR-D-12-00276.1.

Stith, J. L., J. E.Dye,A. Bansemer,A. J.Heymsfield, C.A.Grainger,

W.A. Petersen, andR.Cifelli, 2002:Microphysical observations

of tropical clouds. J. Appl. Meteor., 41, 97–117, doi:10.1175/1520-0450(2002)041,0097:MOOTC.2.0.CO;2.

Takahashi, T., and N. Fukuta, 1988: Super cooled cloud tunnel

studies on the growth of snow crystals between -48C and -208C.J. Meteor. Soc. Japan, 66, 841–855.

——, C. Inoue, Y. Furukawa, T. Endoh, and R. Naruse, 1986: A

vertical wind tunnel for snow process studies. J.Atmos.Oceanic

Technol., 3, 182–185, doi:10.1175/1520-0426(1986)003,0182:

AVWTFS.2.0.CO;2.

——, T. Endoh, G. Wakahama, and N. Fukuta, 1991: Vapor dif-

fusional growth of free-falling snow crystals between -38C and

-238C. J. Meteor. Soc. Japan, 69, 15–30.Westbrook, C. D., and A. J. Heymsfield, 2011: Ice crystals

growing from vapor in supercooled clouds between 22.58and2228C: Testing current parameterization methods using

laboratory data. J. Atmos. Sci., 68, 2416–2429, doi:10.1175/

JAS-D-11-017.1.

Woods, C. P., M. T. Stoelinga, J. D. Locatelli, and P. V. Hobbs,

2005: Microphysical processes and synergistic interaction

between frontal and orographic forcing of precipitation during

the 13 December 2001 IMPROVE-2 event over the Oregon

Cascades. J. Atmos. Sci., 62, 3493–3519, doi:10.1175/JAS3550.1.

——,——, and——, 2008: Size spectra of snow particles measured

in wintertime precipitation in the Pacific Northwest. J. Atmos.

Sci., 65, 189–205, doi:10.1175/2007JAS2243.1.

2032 JOURNAL OF THE ATMOSPHER IC SC IENCES VOLUME 72