International Journal of Remote Sensing

ISSN: 0143-1161 (Print) 1366-5901 (Online) Journal homepage: https://www.tandfonline.com/loi/tres20

Vertical characteristics of radar reflectivity and

DSD parameters in intense convective clouds over
South East South Asia during the Indian Summer
monsoon: GPM observations

Shailendra Kumar & Yamina Silva

To cite this article: Shailendra Kumar & Yamina Silva (2019): Vertical characteristics of radar
reflectivity and DSD parameters in intense convective clouds over South East South Asia during
the Indian Summer monsoon: GPM observations, International Journal of Remote Sensing, DOI:
10.1080/01431161.2019.1633705

To link to this article: https://doi.org/10.1080/01431161.2019.1633705

View supplementary material

Published online: 02 Jul 2019.

Submit your article to this journal

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tres20
INTERNATIONAL JOURNAL OF REMOTE SENSING
https://doi.org/10.1080/01431161.2019.1633705

Vertical characteristics of radar reflectivity and DSD parameters

in intense convective clouds over South East South Asia during
the Indian Summer monsoon: GPM observations
Shailendra Kumar and Yamina Silva
Institute of Geophysics Peru, Badajoz, Peru

ABSTRACT ARTICLE HISTORY

Global precipitation measurement (GPM) launched in Received 23 August 2018
February 2014 as a legacy of Tropical Rainfall Measuring Mission Accepted 20 April 2019
(TRMM). Both satellites carry precipitation radar (PR), which mea-
sures the three-dimension structure of precipitation from space.
Compared to TRMM PR, GPM has dual-polarized radar (DPR) and
provides the raindrop size distribution (DSD) including mass-
weighted mean diameter (Dm, in mm) and normalized DSD scaling
parameter for concentration (Nw, in mm–1 m–3) (DSD parameters),
along with radar reflectivity factor (Ze) from the surface to
21.875 km. Here we investigated the regional differences in
intense convective clouds over South East & South Asia (SESA)
and explored the differences in the East and West coast of India,
by selecting the various areas. We defined two types of clouds,
namely Cumulonimbus towers (CbTs) and intense convective
clouds (ICCs) based on the Ze and height thresholds. CbTs must
consist of 20 dBZ at 12 km, with echo base height less than 3 km,
where ICCs are classified based on Ze threshold at 8 km (ICC8) and
3 km (ICC3). The average vertical profiles of CbTs indicate a strong
west to east gradient, as the west side/coast of India has intense
CbTs, with higher hydrometeors size, and decreases at east side/
coast of India. The results reveal that the western side of India
(Western Himalaya Foothills and Western Ghats) consists of fewer
CbTs, but they are more intense. ICC3s are distributed nearly
uniformly over the SESA but show the regional differences in Ze
and DSD parameters in the vertical profiles. Despite various height
and Ze thresholds used in the present study, WHF has the stron-
gest vertical profile for all type of cloud cells, and indicate the
importance of specific features and orographic modulated flow.
The hydrometeors size is highest over Western Himalaya Foothills
and least over the Bay of Bengal. Hydrometeors concentration
shows the north-south gradient and higher over oceanic areas.
Two coastal areas, Western Ghats and Myanmar show the different
characteristics. Western Ghats CbTs are more intense, with higher
hydrometeors size, whereas Myanmar has weaker convection and
consists of small-sized hydrometeors. Hydrometeors size and con-
centration show the opposite characteristics, as hydrometeors size
is higher for intense average Ze profiles, whereas hydrometeors
concentration is less for intense average Ze profiles.

CONTACT Shailendra Kumar [email protected]; [email protected] Institute of Geophysics Peru,

Badajoz 15012, Lima
Supplementary material can be accessed here.
© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 S. KUMAR AND Y. SILVA

1. Introduction
Cumulonimbus (Cb) clouds play an important role in the Earth’s atmosphere. They
transport energy and momentum vertically from Earth’s surface to the upper atmosphere
(Riehl and Malkus 1958) and known as hot towers (Zipser 2003). They also act as the
primary source of hydrometeors to anvil regions of the MCS (mesoscale convective
system, Houze 1993). In the last two decades, Tropical Rainfall Measuring Mission pre-
cipitation radar (TRMM PR) with other onboard sensors was used to explore the various
properties of precipitating cloud systems (Takayabu 2002; Hirose and Nakamura 2005;
Nesbitt, Cifelli, and Rutledge 2006; Zipser et al. 2006; Houze, Wilton, and Smull 2007; Liu
et al. 2008; Romatschke, Medina, and Houze 2010; Xu and Zipser 2012; Wu, Qie, and Yuan
2013; Kumar 2016). It is also used to investigate the vertical profiles of intense convective
clouds at the individual pixel’s scale of PR (Petersen and Rutledge 2001; Yuan and Qie
2008; Liu et al. 2012; Qie et al. 2014; Bhat and Kumar 2015 (BK15 henceforth); Kumar and
Bhat (2016) (KB16 henceforth); Kumar and Bhat (2017); Kumar 2017a,b; Kumar 2018).
Petersen and Rutledge (2001) reported that oceanic areas consist of relatively weak
intense vertical reflectivity structure. Yuan and Qie (2008) used the vertical profile of
radar reflectivity (VPRR) technique to explore the relationship between lightning activity
and precipitation in China sea. They found that lightning activity increased whenever
there is higher reflectivity in mixed phase altitude. Liu et al. (2012) also extended the
relationship between lightning activity and radar reflectivity in tropics and subtropics. Xu
and Zipser (2012) selected the maximum radar reflectivity as a function of temperature
from the precipitation features (PFs, Nesbitt, Zipser, and Cecil 2000) over tropics. They
observed that regional differences in the convective structure over continental, monsoon
and oceanic regions, and showed that regional differences depend on updraft speed in
mixed phase regimes. Qie et al. (2014) investigated the deep convective system (DCSs)
and intense deep convective systems (IDCSs) over the Tibetan Plateau and showed that
westernmost Himalaya has strongest DCSs, but few in numbers. On the other, Tibetan
Plateau has a higher frequency of DCSs, but they were small in size and weak in intense.
BK15 showed that the Western Ghats and Arabian sea have fewer Cb clouds, but they are
the most intense one. Kumar (2018) and Kumar and Bhat (2017) showed that Western
Ghats and Arabian sea consist of higher frequency/number of shallow precipitating
echoes during the Indian summer monsoon season.
LeMone and Zipser (1980) and Jorgensen and LeMone (1989) observed the updraft
width (approximately 5 km) in tropical cloud systems using the aircraft data. Guimond,
Heymsfield, and Turk (2010) and Heymsfield et al. (2010, H10 henceforth) used the data
from radar onboard on aircraft to investigate the updraft speed over the tropical land
and oceanic areas. They showed that updraft width is higher (approximately 10 km) over
tropical land areas compared to oceanic areas (approximately 6 km). In this study, we
use the data from Global precipitation measurement precipitation radar (GPM PR) to
investigate the vertical structure in intense convective clouds. The benefit of GPM PR
pixels is that its horizontal resolution nearly matches with the diameter of the Cb clouds
(Lucas, Zipser, and LeMone 1994) and updraft width observed in the aircraft measure-
ment (H10). Various attempts have been made to understand the monsoon by investi-
gating the rainfall, monsoon circulation pattern, intraseasonal variations, precipitation
and microphysical distribution and mesoscale convection over different regions of the
 INTERNATIONAL JOURNAL OF REMOTE SENSING 3

tropics (for e.g., Webster et al. 1998; Lee et al. 2011; Goswami et al. 2011; Sikka 2011;
Kumar et al. 2019a;b; Villalobos et al. 2019) and numerical modelling in recent (Moya-
Álvarez et al. 2018, 2019;b). The vertical structure of the monsoonal clouds has been
studied using ground-based measurement as well as satellite-based observation.
However, lack of raindrop size distribution (DSD) parameters left the uncertainty in the
distribution of rainfall rate estimation (Seto et al. 2013; Liuo et al. 2014). The main aim of
the present work is to investigate the distribution of hydrometeors in intense convective
clouds. The present study could be very useful to simulate the intense convective clouds
based on their hydrometeors size and concentration sensitivity, as they largely affect the
cloud microphysics (Gao et al. 2016). The understanding of the cloud vertical structure
during monsoon seasons is important due to their role in radiative effect (Li et al. 2016).

2. Background and definition of the sub-regions selected for the study

Figure 1 shows the spatial distribution of climatology of surface rainfall during the Indian
summer monsoon (June to September; JJAS months) for the year between 1998 to 2017
using TRMM 3B42 daily rainfall data (Huffman and Coauthors 2007). The regions are
selected based on past studies and spatial distribution of rainfall. Figure 1 also shows the
selected areas and discussed below. We selected two areas over the Western Himalaya

Figure 1. June to September (1998–2017) average surface rainfall from TRMM 3B42 daily rainfall
data. There are observed several precipitation maxims such as over Indo-Gangetic plain, Western
Ghats and Myanmar coast. The colour bar here is mm/day.
4 S. KUMAR AND Y. SILVA

Foothills (WHF, R-1 & R-2), where the orographic flow must overcome the capped moist
air and produces some extreme convection (Medina et al. 2010; Romatschke, Medina,
and Houze 2010). Indo-Gangetic plain (IGP, R-3) consists of Himalaya mountain in the
northeast and known as main monsoon zone. IGP is characterized by monsoon lows and
depression with heavy rainfall during JJAS months. West coast of India e.g., Western
Ghats (WG, R-6) also receive a heavy amount of rainfall during Indian summer monsoon
and affected by the orographic lifting (Rao 1976; Grossman and Durran 1984) but at the
same time it also consists of highest shallow precipitating clouds (Kumar and Bhat 2017;
Kumar 2018). Head of Bay (R-7) consists of frequent large and deep convective clouds
throughout the monsoon season (Bhat et al. 2001) and selected here. Arabian sea and
Bay of Bengal (BOB, R-9) have different sea surface temperatures and thermal atmo-
spheric conditions (Bhat et al. 2001; Bhat 2006), but only BOB is selected here to
investigate the vertical characteristics of intense convective clouds. We did not select
the box over Arabian sea because of the fewer number of cloud cells during the study
periods (see Figure 3, discussed later). We also selected the area over central India (R-4),
Myanmar (R-5) and near Cardamom mountains (CM, R-8) to understand the vertical
structure of intense clouds to the east coast of South East & South Asia (SESA). Table 1
shows the selected areas in the present study with their boundary. In this study, a tall Cb
cloud cell is termed ‘cumulonimbus tower’ (CbT).

3. Data and methods

GPM is launched in mid-February 2014 and designed by Japan Aerospace Exploration
Agency (JAXA) and the National Institute of Communication Technology (NICT) and covers
nearly 80% of the globe (65°S to 65°N). The core observatory of GPM has dual-polarized
radar (DPR) and consists of Ku band precipitation radar (Ku PR, 35.5 GHz) and a Ka-band
precipitation radar (Ka PR, 13.5 GHz). It is a non-sun synchronous satellite and scans the
globe 16 times in a day, and the number of GPM DPR passes are higher at higher latitude.
The horizontal and vertical resolution of GPM is about 5 km × 5 km and 0.125 km,
respectively. GPM DPR works in three scanning modes, namely normal scan (NS), match
scan (MS) and high scan (HS) mode. The NS is done by Ku PR and consists of 49 PR beams
and 176 PR levels. The beam width is 0.71°; nearby beams are separated by 0.71°, giving
a maximum scan angle of 17°. The 176 vertical levels in PR measurements are not the
actual height, but they represent the distance along the PR beams (BK15). It needs to be
corrected. The correction method is fully described in BK15, and we also followed the
similar method to correct the GPM Ku PR vertical levels. GPM DPR provides the three-
dimensional-attenuated corrected radar factor (Ze) at each PR levels. The minimum
detectable Ze is nearly 16.7 dBZ (Hou et al. 2014; Hamada and Takayabu 2015), and due
to its high sensitivity, GPM DPR in NS is not able to measure the weak precipitation and
drizzle, compared to Ka -band. The measurement accuracies of these three scan modes are
within 1 dBZ (Hou et al. 2014), and the sampling technique is shown in Kotsuki, Terasaki,
and Miyoshi (2014). TRMM PR used radar reflectivity to estimate the rainfall rate by
calculating the raindrop size distribution (DSD) parameters. This is a major source of
uncertainty in estimating the rainfall rate. But the inclusion of DPR provides a more
accurate estimate of rainfall rate in precipitating clouds. The DPR estimates the DSD
parameters at each sampled bin range, and then calculate the rain rate (Seto, Iguchi,
 INTERNATIONAL JOURNAL OF REMOTE SENSING 5

Table 1. Column 1 shows the name of the selected regions, and second column is the boundary of
the selected regions. The number of each type of cloud cells for each type in mentioned in column
3–5 for maximum height of 40, 30 and 20 dBZ. The median height for each type is written in column
6–8 in km.
Median height (km)
Region Boundary Ze CbTs ICC8s ICC3s CbT ICC8 ICC3
R-1 72°E–77°E H40 575 1307 2018 14.125 12.250 10.50
(WHF) 31°N–35°N H30 611 1410 2018 11.750 10.375 8.500
H20 613 1410 2018 8.250 7.750 6.250
R-2 73°E–78°E H40 720 1109 1766 14.375 12.812 10.625
(WHF) 25°N–30°N H30 760 1132 1766 11.750 10.375 8.125
H20 768 1132 1766 7.125 7.250 5.875
R-3 82°E–95°E H40 3455 3786 11,696 14.375 13.50 10.375
(IGP) 22°N–28°N H30 3772 3855 11,696 10.875 10.25 7.625
H20 3801 3855 11,696 6.375 6.625 5.500
R-4 76°E–85°E H40 1919 2070 7324 14.375 13.375 10.000
(Central 16°N–22°N H30 2059 2096 7324 10.875 10.125 7.375
India) H20 2069 2096 7324 6.500 6.750 5.500
R-5 90°E–100°E H40 578 781 5862 14.250 13.125 9.375
(Myanmar) 18°N–22°N H30 624 789 5862 10.625 9.875 6.875
H20 624 789 5862 6.375 6.500 5.250
R-6 72°E–78°E H40 502 497 4620 14.250 13.25 9.375
(WG) 10°N–18°N H30 524 505 4620 10.875 9.875 6.875
Slant H20 526 505 4620 6.500 6.625 5.250
R-7 85°E–91°E H40 613 844 2774 14.125 13.125 9.500
(Head of Bay) 10°N–16°N H30 671 850 2774 10.625 9.875 7.000
H20 676 850 2774 6.375 6.500 5.250
R-8 95°E–100°E H40 266 310 3945 14.125 13.125 9.125
(Cardamom hills) 12°N–16°N H30 283 313 3945 10.500 9.875 6.750
H20 288 313 3945 6.375 6.500 5.125
R-9 83°E–95°E H40 431 577 6497 14.125 13.00 9.125
(BOB) 5°N–12°N H30 482 585 6497 10.500 9.750 6.625
H20 495 585 6497 6.375 6.500 5.125

and Oki 2013; Liao, Meneghini, and Tokay 2014) using single and double frequency
methods (Iguchi et al. 2010). The two DSD parameters used in the study, namely mass-
weighted mean diameter (Dm, in mm) and normalized DSD scaling parameter for con-
centration (Nw, in mm–1 m–3). All the past studies based on TRMM PR data only provided
the variation in Ze in intense convective clouds, but here we also showed the variation in
hydrometeors size and concentration in intense convective clouds along with Ze. We did
not consider the hydrometeors size below 1.0 mm due to its limitation and uncertainty
(Seto, Iguchi, and Oki 2013; Liao, Meneghini, and Tokay 2014). Here four years (2014–2017)
of GPM data, during JJAS months is used to investigate the intense convection. At the
same time TRMM precipitation product 3B42 at 0.25° × 0.25° resolution (Huffman and
Coauthors 2007) is used to observe a long time average surface rainfall during the Indian
summer monsoon (Figure 1). The reanalysis data from NCEP are used for temperature and
moisture at 2.5° × 2.5° resolution (Kalnay et al. 1996) for the atmospheric condition.
To define the cloud cell from the radar data, we must need to define a Ze threshold
(Roca, Fiolleau, and Bouniol 2017). As an example, Dixon and Wiener (1993) used 35 dBZ,
as a threshold to define the tropical storms, whereas Zipser and Lutz (1994) used 40 (35)
6 S. KUMAR AND Y. SILVA

dBZ at 4.4 (3.9) km altitude for defining the mid-latitude (tropical) cloud systems. H10
used Ze ≥ 20 dBZ at and above 12 km to define the CbTs using the radar data onboard
on aircraft. A convective core is defined using three-dimensional Ze from TRMM PR and
consists of convective echoes ≥40 dBZ at and more than 10 km (Houze, Wilton, and
Smull 2007; Romatschke, Medina, and Houze 2010; Rasmussen and Houze 2011).
Guimond, Heymsfield, and Turk (2010) showed that a convective cloud tilt vertically,
and for considering this issue, we used VPRR approach (Zipser and Lutz 1994;
Supplementary Figure 1) to define cloud cells. Villalobos et al. (2019) used the GPM
DPR data to explore the cloud properties over the complex topography of Andes using
the VPRR technique. Here, we adopted the nearly similar methodology to classify the
CbTs, as used in BK15 and KB16, but with a fixed Ze threshold for ICCs. The method is
briefly discussed in BK15 and here too. First for each GPM PR pass the maximum Ze
(Zemax) is observed at 12 km over the area of study. If Zemax exceeds more than 20 dBZ
for a PR beam, we constructed cloud cells around this beam (b) using the VPRR
approach. Basically, we selected the maximum Ze at each level (from ground to max-
imum altitude) from the 9 PR beams, containing the beam bmax in centre
(Supplementary Figure 1). Once the cloud cell is formed, all the pixels of nine PR
beams are set to zero or negative, and the same algorithm is applied again, and next
cloud cell is formed if Zemax is higher than 20 dBZ at 12 km. This method is followed until
there are no Zemax is left, higher than 20 dBZ at 12 km altitude. Note that, for a given
area of interest, at the same time there could be several CbTs, and we are able to
capture them all. To remove the stratiform and thick anvil clouds from our CbTs, we only
considered the CbTs, with echo base height <3 km (Li and Schumacher 2011). At the
same time, we also stored the GPM PR based mass-weighted mean diameter (hydro-
meteors size) and normalized DSD scaling parameter for concentration for each CbTs to
investigate their microphysical characteristics. We constructed two ICCs, namely ICC8
and ICC3, based on 8 and 3 km, respectively. GPM PR provides the number of cloudy
pixels (Ze > 17 dBZ; not shown) as a function of altitude, such as TRMM PR (see Figure 2
in BK15). BK15 considered the top 5% Ze club at 3 and 8 km altitude to construct ICC3
and ICC8 and in that case, Ze thresholds were always higher than 30 dBZ and 40 dBZ at
8 km and 3 km in all the selected areas (BK15). Here a fixed Ze threshold is used; 30 (40)
dBZ at 8 (3) km altitude, respectively. Zuluaga and Houze (2015) used the 30 dBZ at 8 km
to investigate the extreme convection over the tropical land and oceanic areas, as 30
dBZ at 8 km represents the most intense end of the convective spectrum. So considering
the 30 dBZ at 8 km will allow us to capture the vigorous and intense convection. Kumar
(2017a;b) used 40 dBZ at 3 km altitude to investigate the intense convective clouds and
here too. Once the thresholds are defined, again VPRR algorithm is applied to capture
the ICC3 and ICC8, as we used for CbTs.
Figure 2 shows a vertical variation of Ze for a cloud system observe over an area of 82°
E–82.5°E and 25.2°N to 26°N for 3 July 2016 over the Indo-Gangetic plain (inset Figure 2).
Inset figure shows the composite Ze for a cloud system observed from GPM PR. Here we
can see many GPM PR beams (1–25 PR beams), derived from VPRR approach. The
vertical structure of Ze shows that there are several PR beams, which are crossing the
12 km altitude and even reaching the 18 km and correspond to CbTs in the present
study. Also, at the same time, there are several PR beams, with Ze > 40 dBZ between 2
and 6 km altitude, but they are not able to cross the 12 km altitude and contribute to
 INTERNATIONAL JOURNAL OF REMOTE SENSING 7

Figure 2. An example of vertical variation in Ze in GPM PR beams for a cloud system (inset figure)
observed over Indo-Gangetic plain region on 3 July 2016. The vertical variation shows the variation
of radar reflectivity factor in ICC8 for GPM pass between the box of 82°E–82.4°E and 25.2°N to 26°N.
The colour bar in dBZ.

Figure 3. Spatial distribution of cloud cells in each 1° × 1° grid box. The colour bar shows the
number of each type of cloud cell in 1° × 1° grid box.

ICC8. Also, few PR beams are not able to cross the 8 km, but with a higher Ze > 40 dBZ
between 2 and 4 km altitude and belong to ICC3s in the present study.
Simpson et al. (1998) showed that a Cb cloud can penetrate the upper troposphere
and known as hot towers (Riehl and Malkus 1958), and the main aim of defining the
CbTs is to capture such tall and deeper clouds. Eight kilometres lie within the mixed
8 S. KUMAR AND Y. SILVA

phase altitude and thus the cloudy pixels in ICC8 may belong to both liquid and ice
phase. ICC3 cells threshold height (3 km) lies well below the freezing level (about 5 km
in the tropical areas) and so ICC3 pixels may come from the Cb and cumulus congestus.
Leary and Houze (1979) showed the four stages of development (growth, intensify,
mature and decay phase) and different stages have different vertical structure in Ze. In
the early/growth stage, higher vertical velocity lifts the hydrometeors to upper tropo-
sphere. The hydrometeors grew in size, while their downwards journey in the lower
troposphere Ze peaks in early mature phase (Williams, Weber, and Orville 1989). Below
the freezing level, microphysical processes such as collision-coalescence, phase change
(ice-liquid conversion), and higher downward velocity tend to increase the Ze in the
lower troposphere (Houze 1993; Fabry and Zawadzki 1995). We used three heights and
Ze thresholds, and so the cloud cells in the present study may come from various stages/
phases of the evolution of Cb clouds, and so results of different microphysical processes.
As an example, CbTs may come from growing and mature phase, whereas ICC3s may
come from the mature and later stage of Cb clouds.

4. Results
To examine the regional differences and characteristics of three types of cloud cells,
defined in section 3, we plotted the average vertical profiles as well as the echo top height
of the cloud cells. Table 1 shows the number of cloud cells used in the present study. Figure
3 shows the spatial distribution of CbTs and ICCs in 1° × 1° box, normalized by the number
of GPM PR pass. CbTs occurrences match with the TRMM PR observations (BK15). Northern-
East of India has a higher number of CbTs compared to Southern-West side of India. The
higher number of CbTs occur over IGP, followed by WHF and central India. WG and Arabian
sea along with the tropical ocean (areas between Equator and 9°N) consist of fewer number
of CbTs and ICC8s. ICC3s are distributed nearly uniformly over most of the regions of SESA
and do not show the regional differences, but north-east areas have higher number of
ICC3s. WHF is known as to produces the most intense convection and wide convective
echoes during Indian summer monsoon (Romatschke, Medina, and Houze 2010;
Romatschke and Houze 2011a,b; Zipser et al. 2006; KB16) and here too, WHF is showing
a higher number of CbTs and ICC8s after IGP and central India. The spatial distribution of
different type of cloud cells shows that different regions have different types of cloud cells
in maxims. Head of Bay, Myanmar, CM have the higher number of ICC3 (nearly 3 ICC3s per
GPM pass) followed by WG (nearly 2 ICC3 per GPM pass). Equatorial Indian Ocean does not
have much CbTs and ICC8s, but comparable ICC3s compared to land-dominated areas.

4.1. Average vertical profile

Figure 4 shows the average vertical profile of Ze in CbTs and ICCs over the selected
areas. Only those vertical profiles are considered here which contribute to 95% of data
points to each PR levels. Based on the shape of Ze profiles, the average vertical profiles
can be divided into three regimes, below freezing level (<4 km), within mixed phase
altitude (6–9 km) and beyond upper troposphere (>12 km). Below 4 km the average Ze is
either increases or remains constant in CbTs and ICCs, towards the surface, depends on
the areas. Overall the Ze decreases fastest with altitude in ICC3s within the mixed phase
 INTERNATIONAL JOURNAL OF REMOTE SENSING 9

Figure 4. Average vertical of radar reflectivity factor in cloud cells. Only those vertical levels are
considered here which contribute at least 95% profiles from the population of cloud cells. Data
below 1.5 km altitude is omitted because of ground contamination.

altitude (approximately 6 to 9 km) and discussed later. Regional differences are least
below the 4 km and mainly observed between 5 and 10 km, and then they merge into
CbTs but dispersed in ICC3s. The average vertical profiles in CbTs reflect the west to east
gradient, as west side/coast of India has stronger (intense) vertical profiles and tends to
weaken as we move towards the east side/coast of India over land. WHF (R-1 and R-2)
have the strongest vertical profiles (e.g., higher Ze at higher altitude) among all the
areas, followed by R-6 (WG), R-4 (Central India) and R-3 (IGP) in CbTs, ICC8s, and ICC3s,
respectively. BOB (R-9) has the weakest average Ze profile (lowest Ze at all the altitude) in
CbTs and ICC8s, whereas Myanmar (R-8) has weakest (lowest Ze at all the altitude)
average vertical profile in ICC3, above 6 km altitude. So, the Ze variation in different
cloud cells in selected areas depends on reference height. Land-dominated areas such as
R-3 (IGP), R-4 (Central India), and R-5 (Myanmar) consist of similarity in CbTs but show
the differences in ICC3s and ICC8s.
The average variation in Ze within different cells follows nearly similar trends as
observed in BK15 and Kumar (2017a). The inclusion of Myanmar (R-5), CM hills (R-8) and
fixed Ze thresholds for ICCs shows a little difference in the average vertical profiles
compared BK15. CbTs over Myanmar (R-5) is weaker than that over the IGP (R-4) and
WG (R-6), although the frequency of occurrences of CbTs is nearly equal over Myanmar
and WG. WG and Arabian sea have the strongest CbTs among the Indian regions (BK15)
and here WG has the strongest CbTs only after WHF. But, we did not select the box over
the Arabian sea, as the number of CbTs were very small over the Arabian sea (Figure 3).
This is also true for IGP (R-3), which also shows the most intense ICC3, after WHF (BK15).
Overall BOB (R-9) has weakest (less Ze at high altitude) vertical profiles and consistent with
the previous study of weak convection over the ocean (Lutz and Zipser 1994). Head of Bay
(R-7) matches with the land convection below 4 km altitude. Below 5 km, WG matches
with oceanic convection, whereas above 8 km, WG behaves like continental convection in
10 S. KUMAR AND Y. SILVA

ICC3 and discussed later in details. The average vertical profiles indicate that except for
WHF, other regimes are thresholds and height dependent. The fixed Ze threshold does not
affect the outcomes compared to BK15. Maybe 30 dBZ and 40 dBZ are very close to 5% Ze
club at 8 km and 3 km and able to capture the vigorous convection.

4.2. Cloud top height distribution and the maximum height of 30 and 40 dbz
altitude
The maximum height of 20 dBZ is used to define the cloud tops (Kumar 2017a) and
termed that echo top height (ETH). Figure 5 shows the distribution of ETH in all the
selected areas. In each box-plot, the lower and upper edge of the box indicate 25 and
75 percentiles of ETHs, whereas red line and green dots are showing the median and
average ETHs, respectively. Land and topographic areas such as WHF (R-2), IGP (R-3) and
central India (R-4) have a higher median and average ETH (nearly 14.500 km), whereas
ETH is least over BOB (R-9, nearly 13.250 km) in CbTs. West coast/side of India (WG) has
a higher average and median ETH in CbTs compared to Myanmar and east/side, but less
compared to IGP. ICC8s show the higher regional differences compared to CbTs, and IGP
(R-3) has the highest average ETH followed by R-4 (central India). ICC3s show the highest
regional differences, and west to east coast/side differences are high. As an example,
west side/coast of India has much higher average ETH compared to east coast/side of
India in ICC3. Table 1 shows the median of ETH (H20) over the selected areas.
Zipser et al. (2006) mentioned that a higher Ze in mid-troposphere indicates a strong
convection, and maximum height of 30 and 40 dBZ (MH30 & MH40) used as a proxy for
convection (e.g., Zipser et al. 2006, H10). Figure 6 shows the frequency distribution of
MH30 and MH40 over the selected areas. Cloud cells based on different height thresh-
olds show the different characteristics. As an example, MH30 and MH40 are showing the
double (two) modes distribution in CbTs, whereas in ICCs, MH30 and MH40 are showing
the single mode in frequency distribution. MH30 (MH40) shows the double mode
distribution in CbTs, namely between 6 and 8 km (6 km) and then between 12 and
14 km (8–9 km) altitude. The modal altitude for MH30 (MH40) is 9 km (6 km) and 6 km
(5 km) in ICC8s and ICC3s, respectively, for all the selected areas. The regional differences
are also observed as MH30 and MH40 are higher over WHF (R-1 and R-2) in CbTs,
whereas Myanmar (R-5), CM hills (R-8) and BOB (R-9) have the least MH30 and MH40.
Nearly 8% CbTs are crossing the 15 km altitude (generally higher than tropical

Figure 5. Maximum height of 20 dBZ (CTH), considered as the cloud top height. Red colour shows
the median value whereas the lower and upper boundary of boxes is showing the 75 and
25 percentiles of maximum height. Green dots are showing the mean of maximum height of 20 dBZ.
 INTERNATIONAL JOURNAL OF REMOTE SENSING 11

Figure 6. Frequency distribution of maximum height of 30 (upper row) and 40 dBZ (bottom row)
distribution. The x-axis is height and the frequency is calculated at each 0.125 km interval.

tropopause layer (TTL), nearly 14 km near the tropics) over western side of India
including WHF (R-1 & R-2) and WG (R-6), whereas only 5% of CbTs are crossing 15 km
altitude over R-3 and R-4. Bay of Bengal (R-9) has a higher fraction of shallow echoes
compared to other areas. The western side of India (WHF and WG) has a higher
frequency of deeper clouds in ICC8, but WG does not have much deeper ICC3s. The
Ze > 40 dBZ above 10 km indicates the most intense convection (Houze, Wilton, and
Smull 2007; Kumar 2017a). Topographic areas such as WHF, WG, and IGP have a higher
number/frequency of CbTs and ICC8s, which are crossing the 10 km altitude, whereas
less than 8% of cloud cells are crossing the 10 km altitude in ICC3s. Table 1 summarized
the median height of MH30 and MH40 over the selected areas with the number of cloud
cells used to calculate them, and supplementary Figure 2 shows the cumulative fre-
quency distribution (CDF) of MH30 and MH40 in the selected areas.

4.3. Distribution of DSD parameters (drop size and concentration)

Figure 7 shows the spatial mean of mass-weighted mean diameter (mm) and normalized
scaling parameters for hydrometeors concentration (mm–1 m–3) at 3 and 8 km altitudes,
respectively, for the different type of cloud cells in 1° × 1° box. The differences are higher
at 8 km altitude, compared to 3 km, whereas within the different type of clouds, ICC3s
have the highest regional differences. The hydrometeors size distribution shows a strong
west-east gradient and decreases from west to east at both the altitudes. Mean hydro-
meteors size is highest over the Western side (WHF, R-1 and R-2, >3 mm) and decreases as
we move towards eastern side (<2 mm). Central India also consists of higher hydrome-
teors size (nearly 2.5 mm), but small compared to WHF. Few grid boxes (areas) over
oceanic regions also consist of higher average hydrometeors size, but overall, oceans have
small sized of hydrometeors compared to land-dominated areas. BOB (R-9) and Myanmar
(R-5) consist of the least average size of hydrometeors, along with equatorial Indian ocean.
12 S. KUMAR AND Y. SILVA

Figure 7. Spatial distribution of mean of mass-weighted hydrometeors diameters (a;b) and mean of
scaling parameter for concentration (c;d) for CbTs and ICCs. The hydrometers size is in mm whereas
concentration is in mm–1 m–3. (a) and (c) are at 3 km, whereas (b) and (d) are at 8 km.

At 8 km the west-east contrast/differences are higher and WHF has higher size of hydro-
meteors and decrease as we move towards the east of India. Rajeevan et al. (2013) used the
CloudSat data and plotted the mean effective radii (MER) over south Asia during Indian
summer monsoon seasons and showed the opposite characteristics. They showed that
oceanic areas have much higher MER, compared to land-dominated areas. That could arise
due to the sensitivity issue, as CloudSat has much less sensitivity (−29 dBZ) compared to
GPM PR. CloudSat is mainly designed to capture the small cloud particles compared to GPM
PR, which mainly sense precipitation-sized hydrometeors. Figure 7(c and d) show the
average of hydrometeors concentration at 3 and 8 km, respectively. It shows the opposite
characteristics compared to hydrometeors size, and oceanic regions have higher hydro-
meteors concentration (35–40 mm–1 m–3 and 30–35 mm–1 m–3 at 3 and 8 km, respectively)
compared to land-dominated areas (20–25 mm–1 m–3 and 10–15 mm–1 m–3 at 3 and 8 km,
respectively), and regional differences are higher at 8 km. Hydrometeors concentration
shows the north-south gradient and increases from north to south at 3 km, but at 8 km the
differences are much clearer. The westerly advected cloud condensation nuclei (CCN) play
a vital role in the formation of cloud droplets and their concentration, which grows and then
precipitation-sized hydrometeors are visible. As an example, Konwar et al. (2012) showed
the relationship between the drop size, concentration and cloud top in monsoonal clouds
using aircraft measurement over various locations over India. They explained that DSDs
mostly depends on meteorological conditions. They found higher drop size over WG
compared to IGP, because of the giant CCN (GCCN) and higher number of CCN over WG.
The westerly flow has higher CCN at the cloud base compared to easterly flow and causes
a higher drop size at the western side of India. They showed the effective drop radius over
IGP grows slowly but deeper into the atmosphere, where the precipitation-sized particle
could be visible. Rajeeven et al. (2012) used the CloudSat data and westerly advected
 INTERNATIONAL JOURNAL OF REMOTE SENSING 13

aerosols decrease the MER over the monsoon trough zone e.g., central and east side of India.
The higher hydrometeors concentration over the oceanic areas could be related to the
weaker updraft speed (Zipser and Lutz 1994). The higher updraft speed over the land areas
carries the hydrometeors at higher altitude whereas weaker updraft speed leaves the higher
hydrometeors over oceanic areas, which also leads to lower ETH.
To investigate the microphysical structure in CbTs and ICCs, we plotted the vertical
profiles of DSD parameters. The diameter of solid particles (e.g., hail or graupel) gen-
erally about factor of 2–5 (sometimes factor of 10) times larger than that of the liquid
drops. So, the contribution of the solid particle to the Ze, above the melting level, is
significantly higher than that of the water drops. Since we do not know the phase of the
hydrometeors, thus we restricted the DSD parameters up to 8 km, because above 8 km
the temperature could be very less below the freezing level and only ice could be
expected in that regimes. Figure 8 (9) shows the vertical profile for 90 percentiles
(50 percentile) of hydrometeors size and hydrometeors concentration over selected
areas. In general, intense Ze profiles have higher hydrometeors size and less/fewer
hydrometeors concentration. This indicates that deposition and riming above the freez-
ing level and collision-coalescence mechanism below the freezing level are effective in
intense convective clouds, which tends to increase the hydrometeors size (Houze 1993).
The hydrometeors size is highest (nearly 4.3 mm) over WHF at all PR levels, whereas
BOB has the least (nearly 2.5 mm) in CbTs. The hydrometeors size is nearly twice over
WHF compared to other areas and indicates the role of GCCN and CCN as discussed
earlier (Fig. 8a,9a). The hydrometeors size lies between nearly 4–4.5 mm and nearly
2–2.5 mm over WHF, followed by WG (>3 mm and 2.5 mm) for 90 and 50 percentiles,
respectively. Overall western side of India (WHF and WG) has higher hydrometeors size
(>3.5 mm and >2.5 mm for 90 and 50 percentile) compared to east side (<3 mm for
90 percentiles and <2.5 mm for 50 percentiles), and reflect more intense rain events for

Figure 8. Vertical profile of 90% percentile of mass-weighted hydrometeors diameters (in mm, upper
panels) and scaling parameter for droplet concentration (in mm–1 m–3, lower panels) for each type
of cloud cells.
14 S. KUMAR AND Y. SILVA

Figure 9. Vertical profile of 50% percentile of mass-weighted hydrometeors diameters (in mm, upper
panels) and scaling parameter for droplet concentration (in mm–1 m–3, lower panels) for each type
of cloud cells.

less deeper cloud cells (Figure 5). Myanmar, which consists of higher number of shallow
echoes (Figure 5), consists of least sized of hydrometeors. Longtime TRMM observations
reveal that WG and Myanmar have higher fraction of shallow precipitation during Indian
summer monsoon (Kumar, Hazra, and Goswami 2014; Kumar and Bhat 2017; Kumar
2018), but they are showing the large difference in hydrometeors size, as WG has larger
hydrometeors (>3 mm) compared to Myanmar (<2.5 mm). This indicates the rapid
intense precipitation over WG compared to Myanmar (Konwar et al. 2012). The hydro-
meteors concentration shows the opposite/reverse characteristics compared to hydro-
meteors size, and the most intense cloud cells have the least concentration. As an
example, WHF (R-1 and R-2) have the least hydrometeors concentration, whereas CM
hills (R-8) have the highest hydrometeors concentration. Oceanic areas such as Arabian
sea, BOB and Equatorial Indian Ocean have higher hydrometeors concentration com-
pared to land-dominated areas. This is consistent with the argument that weak updraft
over the oceanic areas leaves higher hydrometeors to lower altitude, which could
remove out quickly by shallow precipitation (Lucas, Zipser, and LeMone 1994). The
vertical distribution of hydrometeors size shows the peak near freezing height in 50%
percentile and could result due to melting of hydrometeors. Although, this needs more
investigation through numerical modelling.
To see the detailed structure of hydrometeors size, we plotted the contoured fre-
quency by altitude diagram (Figure 10–12, CFAD, Yuter and Houze 1995) for hydrome-
teors size. CbTs and ICC8s have higher width at all the PR levels compared to ICC3s and
show the two maxima, one between 8 and 12 km and other at 1.5–3 km altitude. WHF
(R-1, R-2) and IGP (R-3) have highest CFAD width in CbTs and ICC8s, whereas BOB (R-9)
has the least width. The major regional differences are observed in ICC3s, whereas WHF
(R-1) and IGP (R-3) has the maximum probability of higher sized hydrometeors at all PR
levels compared to other areas and correspond to higher Ze in these regions. The higher
 INTERNATIONAL JOURNAL OF REMOTE SENSING 15

Figure 10. Contoured frequency by altitude diagram (CFAD) for hydrometeors size (in mm) in CbTs.
The colour bar shows the relative occurrences of the hydrometeors size at each vertical level.

Figure 11. Contoured frequency by altitude diagram (CFAD) for hydrometeors size (in mm) in ICC8s.
The colour bar shows the relative occurrences of the hydrometeors size at each vertical level.
16 S. KUMAR AND Y. SILVA

Figure 12. Contoured frequency by altitude diagram (CFAD) for hydrometeors size (in mm) in ICC3s.
The colour bar shows the relative occurrences of the hydrometeors size at each vertical level.

width at higher altitude is consistent with the argument that in intense cloud cells
higher updraft speed takes the large size hydrometeors at high heights (H10). Even in
ICC3s CFAD, topographic areas such as WHF (R-1) and IGP (R-3) have maximum hydro-
meteors size at all the altitude and reveals the higher updraft speed in these areas
(Romatschke, Medina, and Houze 2010; H10).

5. Discussion
The radar reflectivity is used as a proxy for hydrometeors size and concentration and
directly proportional to sixth to the power of the diameters of the drop size (Houze
1993). There are several other factors such as the strength of low-level convergence,
vertical velocity (updraft and downdraft) inside the cloud, horizontal and vertical wind
shear influence the vertical distribution of hydrometeors (Lasher Trapp et al. 2018). The
average Ze profiles show the maximum Ze between 3 and 5 km and then decreases
towards its tops, and the rate of decreasing in Ze depends on the type of cloud cells. In
the upper troposphere, the average Ze reaches its detection limit and has lower Ze in the
upper atmosphere. The shape in Ze and DSD profiles is associated with humidity
structure as well as the cloud microphysics, which varies with the altitude (Liu and
Zipser 2013). These observations could be connected to various microphysical processes
and updraft speed (Zipser and Lutz 1994; H10), as the variations in Ze, are sensitive to
updraft speed in mixed phase height (H10). The higher updraft speed provides fewer
time to hydrometeors to grown up and take away the condensate water very quickly to
upper atmosphere (H10). In the mature stage of the life cycle, the updraft speed
weakens and the hydrometeors which grown up in upward journey start falling down
 INTERNATIONAL JOURNAL OF REMOTE SENSING 17

and show the variation in Ze in mixed phase regimes. Once the hydrometeors start
falling down, deposition, accretion, and aggregation enhance the size of falling hydro-
meteors. These falling hydrometeors have higher Ze in mixed phase regimes, as Ze is
directly proportional to the 6th powers of the hydrometeors size (Houze 1993). Near the
melting level, the ice turns into water and the dielectric effect of the water causes the
increase in Ze (Houze 1993). Higher Ze above the freezing level indicates the strong
updraft speed and large ice particles over the land areas (Zipser and Lutz 1994; Zipser
et al. 2006). We calculated the absolute slope in average vertical profiles between 5 and
7 km altitude (e.g., within the mixed phase regions) and showed in Figure 13, and it is
consistent with the TRMM PR observations (BK15). ICC3s have the highest slope and lies
between 4 and 6.5 dBZ km–1, and nearly 2 to 2.5 times higher than CbTs and ICC8s. WHF
has the least slope in all the cases, whereas Myanmar (R-5) and CM (R-8) have the largest
slope and cloud cells decay rapidly in the mentioned area. Oceanic slope (R-7 and R-9)
lies between the plain and topographic areas.
One of the main objectives of the present work is to explore the role of local factors on Ze
and DSD parameters. The local factors include the CCN, horizontal wind shear and moisture
availability, which could lead to a significant effect on Ze and DSD parameters within the
different selected areas. Sherwood, Minnis, and McGill (2004) explained that surrounding
moist atmosphere supports the development of deep convection and dry atmosphere
restrict them. Moist static energy (MSE) is a crucial parameter for moist convection and
relates the thermodynamic variable such as temperature and specific humidity in one by
using equation h = Cp × T+ Lv × q + g × Z. Here his MSE, Cp is the specific heat of air at
constant pressure, T the temperature, g acceleration due to gravity, Z the height above
mean sea level, Lv latent heat of evaporation of water and q the mass of water vapour per
unit mass of moist air, i.e. specific humidity. Supplementary Figure 3 shows the spatial
distribution of column water vapour and surface MSE for JJAS months during 2014 to 2017,
whereas supplementary Figure 4 shows the vertical distribution of MSE over the selected
areas during the study period. A less (small) value of MSE in mid-troposphere indicates the
drier air, because of less temperature gradient in the tropics (Sobel and Bretherton 2000).
WHF (R-1) and IGP (R-3) have highest MSE in the upper troposphere and correspond to
higher Ze and hydrometeors size in vertical profile and consist of the higher frequency of

Figure 13. Absolute slope estimated between 5 and 7 km altitude for different types of cloud cells.
18 S. KUMAR AND Y. SILVA

deeper ETH, but not true for all the areas. As WG (R-6) has less MSE in mid and upper
troposphere, but still contribute to deeper clouds (higher ETH) and intense vertical profile in
CbTs. BOB (R-9) has least MSE and corresponds to weakest vertical profiles, small-sized
hydrometeors and higher frequency of shallower echoes (less ETH). The Western coast/side
of India has relatively less MSE, drier mid-troposphere, and this may partially explain why
WG (R-6) has small Ze and small-sized hydrometeors in ICC3. Despite of various thresholds
used in the present study, WHF has the strongest vertical profile and indicate that the
intense convection over WHF could be results of its specific geographical features. Medina
et al. (2010) explained that the role of surface feedback flux and specific regional features
which produces the intense and deep convection. Over WHF a stable layer within the moist
boundary layer does not allow the instability to release outside the layer. Also, the unstable
flow moves along the slope of WHF up to saturation level to release the instability and
results in the deep convection. Land versus oceanic differences are not able to explain the
variation of Ze in CbTs and ICCs, as, at the same time land areas such as Myanmar (R-5) also
consists of weaker CbTs and ICCs compared to head of Bay (R-7). This could be possible
because of different height and Ze thresholds used in the present study (BK15). Convective
available potential energy (CAPE) is an indicator for updraft speed (Yuter and Houze 1995),
but again CAPE and updraft speed does not contain direct relationship with convective
intensity (e.g., Zipser 2003; H10).
WG shows the most interesting characteristics as average profiles in CbTs, matches
with most intense vertical profile over WHF. But in ICC3s, below 4 km altitude, WG has
a weaker average vertical profile but matches with WHF and IGP above 8 km altitude.
WG is one of the most explored areas through observational data (Rao 1976; Mukherjee
et al. 1978; Kumar, Hazra, and Goswami 2014), numerical and theoretical models (Sarker
1967; Grossman and Durran 1984; Oruga and Yoshizaki 1988). The intense convection
over WG could be orographically influenced. As an example, during Indian summer
monsoon seasons mean wind at low levels is normal to orography (Rao 1976;
Romatschke and Houze 2011a). Grossman and Durran (1984) explained that lifting of
near-surface air, higher vertical velocity at upstream of orographic slope produces the
extreme convection. Houze (2012) also showed the similar trends and explained that
a moisture loaded flow over the mountain enhances the clouds and precipitation
formation. However, at the same time, droplets could be generated at the faster rate
in supersaturation environment due to condensational growth, which leads to rapid and
local precipitation and turns to less ETH in the area (Konwar et al. 2014). A classic case of
land and ocean differences/contrast could be observed in the region based on R-3, R-4,
R-7, R-9, and Arabian sea. In general, the land dominated areas have intense vertical
profile, higher frequency (number) of deeper clouds, higher hydrometeors size and small
concentration compared to oceanic areas. As an example, Ze values over land-
dominated areas are approximately 6–8 dBZ higher in the different types of clouds
(Figure 4). Rosenfeld et al. (2008) showed the CCN concentration and moisture avail-
ability lead to these differences. As an example, over the west side of India, lower CCN
concentration (Pruppacher and Klett 1997) increases the warm rain frequency and then
leaves too fewer hydrometeors to carry them into upper atmosphere (Konwar et al.
2012). But higher CCN concentration delays the precipitation over land-dominated areas
(Rosenfeld et al. 2008; Konwar et al. 2012), and more condensate water could carry to
upper atmosphere/upper levels (e.g. Wang 2005). The saline sea water, salt particles and
 INTERNATIONAL JOURNAL OF REMOTE SENSING 19

GCCN affect the precipitation processes over the Arabian sea. Weak vertical velocity and
GCCN over the Arabian sea simulate the early precipitation, where collision-coalescence
mechanism enhances the warm rain processes (Konwar et al. 2012). Arabian sea has
small-sized hydrometeors and higher concentration, which also simulate the collision-
coalescence mechanism, and removed the condensed water nearer to the cloud base in
Arabian Sea clouds compared to Bay clouds (Bhat 2006).

6. Conclusions
We would like to mention that focus of the present study is to use the Global
Precipitation Measurement Precipitation Radar to investigate the variation in radar
reflectivity and DSD parameters in intense convective clouds at the PR pixels’ size.
Major conclusions from the present study are as follows:

(1) The average vertical profiles of CbTs indicate a strong west-east gradient, as west
side/coast of India has intense CbTs with higher hydrometeors size, and it decreases
as we move towards the east side/coast of India over land and topographic areas.
West side of India including Western Himalaya Foothills, and Western Ghats have
less number of CbTs, but they are the most intense one compared to the east side
of India. Convective clouds based on the 3 km altitude are uniformly distributed
over the Indian land and oceanic areas and show that ice microphysics is more
important to derive the cloud top height in South East South Asia.
(2) The average vertical profiles in the cloud cells depend on the height and reflectivity
thresholds except over the western Himalaya foothills. Despite the various height
and reflectivity thresholds used in the present study, WHF has the most intense
vertical profile for all type of cloud cells. The regional differences are least in
Cumulonimbus towers and highest in intense convective clouds based on the 3 km.
(3) Two coastal areas, Western Ghats and Myanmar show the different characteristics.
Both have the comparable number of intense convective clouds, but Western
Ghats have the most intense CbTs, whereas Myanmar contributes to weaker
convection. Western Ghats have higher hydrometeors size compared to
Myanmar and reflect the more intense and warm precipitation.
(4) Western side/coast of India has higher average hydrometeors size and decrease as we
move towards the east coast of India. Hydrometeors sizes are higher over the land
and topographic areas compared to oceanic areas in intense convective clouds.
Hydrometeors concentration shows the north-south gradient and higher over the
oceanic areas. It is related to CCN and updraft velocity over the different areas.
(5) Hydrometeors size and concentration show the opposite characteristics compared
to radar reflectivity, and hydrometeors size increases with altitude, whereas
hydrometeors concentration decreases with altitude in intense convective clouds.
The size of hydrometeors is highest over Western Himalaya Foothills and least
over the Bay of Bengal.

The manuscript presents the vertical distribution of hydrometeors, their size, and concen-
tration in intense convective clouds. We improved the analysis in the regions by including
the data from dual-polarized precipitation radar from GPM and included the analysis of
20 S. KUMAR AND Y. SILVA

hydrometeors size and concentration, which was missing in the past observational ana-
lysis. Gao et al. (2016) used Weather Research Forecasting (WRF) model to investigate the
sensitivity test based on the cloud droplets on rainfall over Tibetan Plateau. They showed
that doubling of the hydrometeors concentration and size affect the warm and cold rain
processes over Tibetan Plateau. It is important to improve the weather prediction/mon-
itoring over the region, and present study could be useful to simulate the intense
convection over SESA based on hydrometeors size and concentration.

Acknowledgements
Present study comes under the project “MAGNET-IGP: Strengthening the research line in physics and
micro physics of the atmosphere (Agreement Nº 010–2017–FONDECYT)”. I would like to thanks the
CONCYTEC, Peru, for financial support and Inter-American Institute for Cooperation on Agriculture
(IICA) for administrative support. GPM (https://mirador.gsfc.nasa.gov/cgi–bin/mirador/granlist.pl?
page=1&dataSet=GPM_2ADPR&version=05&allversion=05&keyword=GPM%20DPR&pointLocation=
(−90,-180),(90,180)&location=(−90,180),(90,180)&searchType=Location&event=&startTime=2014–03–
08&endTime=2019–0418%2023:59:59&search=&CGISESSID=6a18e607df380bd63841190cf
cef870f&nr=16&temporalres=1.5%20hours&prodpg=https://mirador.gsfc.nasa.gov/collections/GPM_
2ADPR__05.shtml&longname=GPM%20DPR%20Precipitation%20Profile%20L2A%201.5%20hours%
205%20km%20V05&granulePresentation=ungrouped) and 3B42 (http://mirador.gsfc.nasa.gov/cgi
bin/mirador/presentNavigation.pl?tree=project&dataset=3B42:%203Hour%200.25%20x%200.25%
20degree%20merged%20TRMM%20and%20other%20satellite%20estimates&project=
TRMM&dataGroup=Gridded&version=007) data are taken from NASA’s Earth-Sun System Division
website. The authors thank the anonymous reviewers for their constructive suggestions.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by the Inter-American Institute for Cooperation on Agriculture (IICA)
[010-2017-FONDECYT].

References
Bhat, G., and S. Kumar. 2015. “Vertical Structure of Cumulonimbus Towers and Intense Convective
Clouds over the South Asian Region during the Summer Monsoon Season.” Journal of
Geophysics Research Atmosphere 120 (5): 1710–1722. doi:10.1002/2014JD022552.
Bhat, G. S. 2006. “The Indian Drought of 2002, A Sub Seasonal Phenomenon?” Quarterly Journal of
the Royal Meteorological Society 132: 2583–2602. doi:10.1256/qj.05.13.
Bhat, G. S., S. Gadgil, P. V. H. Kumar, S. R. Kalsi, P. Madhusoodanan, V. S. N. Murty, C. V. K. Prasada
Rao, et al. 2001. “BOBMEX: The Bay of Bengal Monsoon Experiment”. Bulletin of the American
Meteorological Society 82: 2217–2243. 10.1175/1520-0477(2001)082<2217:BTBOBM>2.3.CO;2.
Dixon, M., and G. Wiener. 1993. “TITAN: Thunderstorm Identification, Tracking, Analysis, and
Nowcasting a Radar-Based Methodology.” Journal of Atmospheric Oceanic Technology 10:
785–797. doi:10.1175/1520-0426(1993)010<0785:TTITAA>2.0.CO;2.
Fabry, F., and I. Zawadzki. 1995. “Long-Term Radar Observations of the Melting Layer of
Precipitation and Their Interpretation.” Journal of Atmospheric Science 52: 838–851.
doi:10.1175/1520-0469(1995)052<0838:LTROOT>2.0.CO;2.
 INTERNATIONAL JOURNAL OF REMOTE SENSING 21

Gao, W., C.-H. Sui, J. Fan, Z. Hu, and L. Zhong. 2016. “A Study of Cloud Microphysics and
Precipitation over the Tibetan Plateau by Radar Observations and Cloud-Resolving Model
Simulations.” Journal of Geophysics Research Atmosphere 121: 13,735–13,752. doi:10.1002/
2015JD024196.
Goswami, B. N., M. C. Wheeler, J. C. Gottschalck, and D. E. Waliser. 2011. “Intraseasonal Variability
and Forecasting: A Review of Recent Research.” In The Global Monsoon System: Research and
Forecast. 2nd ed., 389–407. World Scientific Publishing Co. Pte.
Grossman, R. L., and D. R. Durran. 1984. “Interaction of Low-Level Flow with the Western Ghat
Mountains and Offshore Convection in the Summer Monsoon.” Monthly Weather Review 112:
652–672. doi:10.1175/1520-0493(1984)112<0652:IOLLFW>2.0.CO;2.
Guimond, S. R., G. M. Heymsfield, and F. J. Turk. 2010. “Multiscale Observations of Hurricane Dennis
2005.: The Effect of Hot Towers on Rapid Intensification.” Journal Atmospheric Science 67:
633–654. doi:10.1175/2009JAS3119.1.
Hamada, A., and Y. N. Takayabu. 2015. “Improvements in Detection of Light Precipitation with the
Global Precipitation Measurement Dual-Frequency Precipitation Radar GPM DPR.” Journal of
Atmospheric Oceanic Technology 33: 653–667. doi:10.1175/JTECH-D-15-0097.1.
Heymsfield, G. M., L. Tian, A. J. Heymsfield, L. Li, and S. Guimond. 2010. “Characteristics of Deep
Tropical and Subtropical Convection from Nadir-Viewing High-Altitude Airborne Doppler
Radar.” Journal of Atmospheric Science 67: 285–308. doi:10.1175/2009JAS3132.1.
Hirose, M., and K. Nakamura. 2005. “Spatial and Diurnal Variation of Precipitation Systems over Asia
Observed by the TRMM Precipitation Radar.” Journal of Geophysics Research Atmosphere 110:
D05106.
Hou, A. Y., R. K. Kakar, S. Neeck, A. A. Azarbarzin, C. D. Kummerow, M. Kojima, R. Oki, K. Nakamura,
and T. Iguchi. 2014. “The Global Precipitation Measurement Mission.” Bulletin of the American
Meteorological Society 95: 701–722. doi:10.1175/BAMS-D-13-00164.1.
Houze, R. A., Jr. 1993. 496. Cloud Dynamics. San Diego, CA: Academic Press.
Houze, R. A. 2012. “Orographic Effects on Precipitating Clouds.” Reviews of Geophysics 50: RG1001.
doi:10.1029/2011RG000365.
Houze, R. A., D. C. Wilton, and F. B. Smull. 2007. “Monsoon Convection in the Himalayan Region as
Seen by the TRMM Precipitation Radar.” Quarterly Journal of the Royal Meteorological Society
133: 1389–1411.
Huffman, G., and J. Coauthors. 2007. “The TRMM Multisatellite Precipitation Analysis TMPA:
Quasi-Global, Multilayear, Combined-Sensor Precipitation Estimates at Fine Scales.” Journal of
Hydrometeorology 8: 38–55. doi:10.1175/JHM560.1.
Iguchi, T., S. Seto, R. Meneghini, N. Yoshida, J. Awaka, and T. Kubota. 2010. “GPM/DPR Level#2
Algorithm Theoretical Basis Document. ” Technical Report, JAXA–NASA. http: //pps.gsfc.nasa.
gov/atbd.html.
Jorgensen, D. P., and M. A. LeMone. 1989. “Vertical Velocity Characteristics of Oceanic Convection.”
Journal of Atmospheric Science 46: 621–640. doi:10.1175/1520-0469(1989)046<0621:
VVCOOC>2.0.CO;2.
Kalnay, E., M. Kanamitsu, R. Kistler, W. Collins, D. Deaven, L. Gandin, M. Iredell, et al. 1996. “The
NCEP/NCAR 40 Year Reanalysis Project.” Bulletin of the American Meteorological Society 77: 437–
471. doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Konwar, M., R. S. Maheskumar, J. R. Kulkarni, E. Freud, B. N. Goswami, and D. Rosenfeld. 2012.
“Aerosol Control on Depth of Warm Rain in Convective Clouds.” Journal of Geophysical Research:
Atmospheres 117 (D13). doi:10.1029/2012JD017585.
Konwar, M., S. K. Das, S. M. Deshpande, K. Chakravarty, and B. N. Goswami. 2014. “Microphysics of
Clouds and Rain over the Western Ghat.” Journal of Geophysical Research: Atmospheres 119 (10):
6140–6159.
Kotsuki, S., K. Terasaki, and T. Miyoshi. 2014. “GPM/DPR Precipitation Compared with a 3.5-Km-
Resolution NICAM Simulation.” Sola 10: 204–209. doi:10.2151/sola.2014-043.
Kumar, S. 2016. “Three Dimensional Characteristics of Precipitating Cloud Systems Observed
during Indian Summer Monsoon.” Advances in Space Research 58 (6): 1017–1032. doi:10.1016/
j.asr.2016.05.052.
22 S. KUMAR AND Y. SILVA

Kumar, S. 2017a. “A 10-Year Climatology of Vertical Properties of Most Active Convective Clouds
over the Indian Regions Using TRMM PR.” Theoretical and Applied Climatology 127 (1–2):
429–440. doi:10.1007/s00704-015-1641-5.
Kumar, S. 2017b. “Vertical Characteristics of Reflectivity in Intense Convective Clouds Using TRMM
PR Data.” Environment and Natural Resources Research 72: 58. doi:10.5539/enrr.v7n2p58.
Kumar, S. 2018. “Vertical Structure of Precipitating Shallow Echoes Observed from TRMM during
Indian Summer Monsoon.” Theoretical and Applied Climatology 133 (3–4): 1051–1059.
doi:10.1007/s00704-017-2238-y.
Kumar, S., A. Hazra, and B. N. Goswami. 2014. “Role of Interaction between Dynamics,
Thermodynamics and Cloud Microphysics on Summer Monsoon Precipitating Clouds over the
Myanmar Coast and the Western Ghats.” Climate Dynamics 43: 911. doi:10.1007/s00382-013-
1909-3.
Kumar, S., and G. Bhat. 2016. “Vertical Profiles of Radar Reflectivity Factor in Intense Convective
Clouds in the Tropics.” Journal of Applied Meteorology and Climatology 555: 1277–1286.
doi:10.1175/JAMC-D-15-0110.1.
Kumar, S., and G. Bhat. 2017. “Vertical Structure of Orographic Precipitating Clouds Observed over
South Asia during Summer Monsoon Season.” Journal of Earth System Science 126 (8): 114.
doi:10.1007/s12040-017-0897-9.
Kumar, S., Y. Silva-Vidal, A. S. Moya-Álvarez, and D. Martínez-Castro. 2019a. “Effect of the Surface
Wind Flow and Topography on Precipitating Cloud Systems over the Andes and Associated
Amazon Basin: GPM Observations.” Atmospheric Research 225 (193–208). doi:10.1016/j.
atmosres.2019.03.027.
Kumar, S., Y. Silva-Vidal, A. S. Moya-Álvarez, and D. Martínez-Castro. 2019b. “Seasonal and Regional
Differences in Extreme Rainfall Events and Their Contribution to the World’s Precipitation: GPM
Observations: GPM Observations.” Advances in Meteorology. doi:10.1155/2019/4631609.
Lasher-Trapp, S., S. Kumar, D. H. Moser, A. M. Blyth, J. R. French, R. C. Jackson, D. C. Leon, and
D. M. Plummer. 2018. “On Different Microphysical Pathways to Convective Rainfall.” Journal of
Applied Meteorology and Climatology 57 (10): 2399–2417. doi:10.1175/JAMC-D-18-0041.1.
Leary, C. A., and R. A. Houze. 1979. “The Structure and Evolution of Convection in a Tropical Cloud
Cluster.” Journal of Atmospheric Science 36: 437–457. doi:10.1175/1520-0469(1979)036<0437:
TSAEOC>2.0.CO;2.
Lee, D. I., S. M. Jang, C. H. You, K. J. Seo, D. S. Kim, M. Y. Kang, H. Uyeda, M. Maki, and B. J. D. Jou.
2011. “The Characteristics of Summer Monsoon Rainfall at the Southwestern Ocean Area of
Korea: A Study of the 2007 Season.” In The Global Monsoon System: Research and Forecast,
319–338. World Scientific Publishing Co. Pte.
LeMone, M. A., and E. J. Zipser. 1980. “Cumulonimbus Vertical Velocity Events in GATE: Part
I Diameter, Intensity, and Mass Flux.” Journal of Atmospheric Science 37: 2444–2457.
doi:10.1175/1520-0469(1980)037<2444:CVVEIG>2.0.CO;2.
Li, J., D. E. Waliser, G. Stephens, and S. W. Lee. 2016. “Charactersing and Understanding Cloud Ice
and Radiation Budgets in Global Climate Models and Reanalysis.” Meteorological Monographs
56: 13.1–13.20. doi:10.1175/AMSMONOGRAPHS-D-15-0007.1.
Li, W., and C. Schumacher. 2011. “Thick Anvils as Viewed by the TRMM Precipitation Radar.” Journal
of Climate 24: 1718–1735. doi:10.1175/2010JCLI3793.1.
Liao, L., R. Meneghini, and A. Tokay. 2014. “Uncertainties of GPM DPR Rain Estimates Caused by
DSD Parameterizations.” Journal of Applied Meteorology and Climatology 53: 2524–2537.
doi:10.1175/JAMC-D-14-0003.1.
Liu, C., D. J. Cecil, E. J. Zipser, K. Kronfeld, and R. Robertson. 2012. “Relationships between
Lightning Flash Rates and Radar Reflectivity Vertical Structures in Thunderstorms over the
Tropics and Subtropics.” Journal of Geophysics Research Atmosphere 117: D06212.
Liu, C., and E. J. Zipser. 2013. “Why Does Radar Reflectivity Tend to Increase Downward toward the
Ocean Surface, but Decrease Downward toward the Land Surface?” Journal of Geophysics
Research Atmosphere 118: 135–148. doi:10.1029/2012JD018134.
 INTERNATIONAL JOURNAL OF REMOTE SENSING 23

Liu, C., E. J. Zipser, D. J. Cecil, S. W. Nesbitt, and S. Sherwood. 2008. “A Cloud and Precipitation
Feature Database from Nine Years of TRMM Observations.” Journal of Applied Meteorology and
Climatology 4710: 2712–2728. doi:10.1175/2008JAMC1890.1.
Lucas, C., E. J. Zipser, and M. A. LeMone. 1994. “Vertical Velocity in Oceanic Convection off Tropical
Australia.” Journal of Atmospheric Science 51: 3183–3193. doi:10.1175/1520-0469(1994)
051<3183:VVIOCO>2.0.CO;2.
Medina, S., R. A. Houze, A. Kumar, and D. Niyogi. 2010. “Summer Monsoon Convection in the
Himalayan Region: Terrain and Land Cover Effects.” Quarterly Journal of the Royal Meteorological
Society 136: 593–616.
Moya-Álvarez, A., J. Gálvez, A. Holguín, R. Estevan, S. Kumar, E. Villalobos, D. Martínez-Castro, and
Y. Silva. 2018. “Extreme Rainfall Forecast with the WRF-ARW Model in the Central Andes of
Peru.” Atmosphere 9 (9): 362. doi:10.3390/atmos9090362.
Moya-Álvarez, A. S., D. Martínez-Castro, S. Kumar, R. Estevan, Y. Silva. 2019. “Response of the WRF
Model to Different Resolutions in the Rainfall Forecast over the Complex Peruvian Orography.”
Theoretical and Applied Climatology. doi:10.1007/s00704-019-02782-3.
Mukherjee, A. K., M. K. Rao, and K. C. Shah. 1978. “Vortices Embedded in the Trough of Low
Pressure off Maharashtra–Goa Coasts during the Month of July.” Indian Journal of Meteorology,
Hydrology & Geophysics 29: 61–65.
Nesbitt, S. W., E. Zipser, and C. Cecil. 2000. “A Census of Precipitation Features in the Tropics Using
TRMM Radar, Ice Scattering, and Lightning Observations.” Journal of Climate 13: 4087–4106.
doi:10.1175/1520-0442(2000)013<4087:ACOPFI>2.0.CO;2.
Nesbitt, S. W., R. Cifelli, and S. A. Rutledge. 2006. “Storm Morphology and Rainfall Characteristics of
TRMM Precipitation Features.” Monthly Weather Review 134: 2702–2721. doi:10.1175/
MWR3200.1.
Oruga, Y., and M. Yoshizaki. 1988. “Numerical Study of Orographic-Convective Precipitation over
Eastern Arabian Sea and the Ghat Mountains during the Summer Monsoon.” Journal of the
Atmospheric Sciences 45: 2097–2121. doi:10.1175/1520-0469(1988)045<2097:NSOOCP>2.0.CO;2.
Petersen, W. A., and S. A. Rutledge. 2001. “Regional Variability in Tropical Convection: Observation
from TRMM.” Journal of Climate 13: 4087–4106.
Pruppacher, H. R., and J. D. Klett. 1997. Microphysics of Clouds and Precipitation, 961. Dordrecht:
Kluwer Academic Publishers.
Qie, X. S., X. K. Wu, T. Yuan, J. C. Bian, and D. R. Lu. 2014. “Comprehensive Pattern of Deep
Convective Systems over the Tibetan Plateau-South Asian Monsoon Region Based on TRMM
Data.” Journal of Climate 27: 6612–6626. doi:10.1175/JCLI-D-14-00076.1.
Rajeevan, M., P. Rohini, K. N. Kumar, J. Srinivasan, and C. K. Unnikrishnan. 2013. “A Study of Vertical
Cloud Structure of the Indian Summer Monsoon Using CloudSat Data.” Climate Dynamics 40 (3–
4): 637–650. doi:10.1007/s00382-012-1374-4.
Rao, Y. P., 1976. “Southwest Monsoon, India Meteorological Department.” Accessed http://
wwwimdgovin/section/nhac/dynamic/Monsoon_framehtm
Rasmussen, K. L., and R. A. Houze, Jr. 2011. “Orogenic Convection in Subtropical South America as
Seen by the TRMM Satellite.” Monthly Weather Review 139: 2399–2420. doi:10.1175/MWR-D-10-
05006.1
Riehl, H., and J. Malkus. 1958. “On the Heat Balance in the Equatorial Trough Zone.” Geophysica 6:
503–538.
Roca, R., T. Fiolleau, and D. Bouniol. 2017. “A Simple Model of the Life Cycle of Mesoscale
Convective Systems Cloud Shield in the Tropics.” Journal of Climate 30: 4283–4298.
doi:10.1175/JCLI-D-16-0556.1.
Romatschke, U., and R. A. Houze. 2011a. “Characteristics of Precipitating Convective Systems in the
Premonsoon Season of South Asia.” Journal of Hydrometeorology 12: 3–26. doi:10.1175/
2010JHM1289.1.
Romatschke, U., and R. A. Houze. 2011b. “Characteristics of Precipitating Convective Systems in the
South Asian Monsoon.” Journal of Hydrometeorology 12: 157–180. doi:10.1175/2010JHM1311.1.
24 S. KUMAR AND Y. SILVA

Romatschke, U., S. Medina, and R. A. Houze. 2010. “Regional, Seasonal, and Diurnal Variations of
Extreme Convection in South Asian Region.” Journal of Climate 23: 419–439. doi:10.1175/
2009JCLI3140.1.
Rosenfeld, D., U. Lohmann, G. B. Raga, C. D. O’Dowd, M. Kulmala, S. Fuzzi, A. Reissell, and
M. O. Andreae. 2008. “Flood or Drought: How Do Aerosols Affect Precipitation?” Science 321:
1309–1313. doi:10.1126/science.1160606.
Sarker, R. P. 1967. “Some Modification in a Dynamical Model of Orographic Rainfall.” Monthly
Weather Review 95: 673–684. doi:10.1175/1520-0493(1967)095<0673:SMIADM>2.3.CO;2.
Seto, S., T. Iguchi, and T. Oki. 2013. “The Basic Performance of a Precipitation Retrieval Algorithm
for the Global Precipitation Measurement Mission’s Single/Dual-Frequency Radar
Measurements, IEEE Trans.” Geoscience Remote Sensing 51: 5239–5251. doi:10.1109/
TGRS.2012.2231686.
Sherwood, S. C., P. Minnis, and M. McGill. 2004. “Deep Convective Cloud-Top Heights and Their
Thermodynamic Control during Crystal-Face.” Journal of Geophysics Research Atmosphere 109.
doi:10.1029/2004JD004811.
Sikka, D. R. 2011. “Synoptic and Meso-Scale Weather Disturbances over South Asia during the
Southwest Summer Monsoon Season.” In The Global Monsoon System: Research and Forecast,
183–204. World Scientific Publishing Co. Pte.
Simpson, J., C. Kummerow, W. Tao, and R. Adler. 1998. “Eyeing the Eye: Exciting Early Stage Science
Results from TRMM.” Bulletin of the American Meteorological Society 79: 1711. doi:10.1175/1520-
0477-79.8.1711.
Sobel, A., and C. S. Bretherton. 2000. “Modeling Tropical Precipitation in a Single Column.” Journal
of Climate 13: 4378–4392. doi:10.1175/1520-0442(2000)013<4378:MTPIAS>2.0.CO;2.
Takayabu, Y. N. 2002. “Spectral Representation of Rain Profile and Diurnal Variations Observed with
TRMM PR over the Equatorial Area.” Geophysics Research Letter 29 (12): 25–31. doi:10.1029/
2001GL014113.
Villalobos, E. E., D. Martinez-Castro, S. Kumar, Y. Silva, and O. Fashe. 2019. “Estudio De Tormentas
Convectivas Sobre Los Andes Centrales Del Perú Usando Los Radares PR-TRMM Y KuPR-GPM.”
Revista Cubana De Meteorología 25 (1): 59–75.
Wang, C. 2005. A modeling study of the response of tropical deep con-vection to the increase of
cloud condensation nuclei concentration: 1. Dynamics and microphysics. Journal of Geophysical
Research: 110: doi:10.1029/2004JD005720.
Webster, P. J., V. O. Magaña, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T. Yasunari. 1998.
“Monsoons: Processes, Predictability, and the Prospects for Prediction.” Journal Geophysical
Research Atmosphere 14451–14510. doi:10.1029/97JC02719.
Williams, E. R., M. E. Weber, and R. E. Orville. 1989. “The Relationship between Lighting Type and
Convective State of Thunderstorms.” Journal of Geophysics Research Atmosphere 94:
13213–13220. doi:10.1029/JD094iD11p13213.
Wu, X. K., X. S. Qie, and T. Yuan. 2013. “Regional Distribution and Diurnal Variation of Deep
Convective Systems over the Asian Monsoon Region.” Science China Earth Sciences 56: 843–854.
doi:10.1007/s11430-012-4551-8.
Xu, W., and E. J. Zipser. 2012. “Properties of Deep Convection in Tropical Continental,
Monsoon, and Oceanic Rainfall Regimes.” Geophysics Research Letter 39 (7). doi:10.1029/
2012GL051242.
Yuan, T., and X. Qie. 2008. “Study on Lightning Activity and Precipitation Characteristics before and
after the Onset of the South China Sea Summer Monsoon.” Journal Geophysical Research
Atmosphere 113: D14101. doi:10.1029/2007JD009382.
Yuter, S. E., and R. A. Houze. 1995. “Three-Dimensional Kinematic and Microphysical Evolution of
Florida Cumulonimbus Part II: Frequency Distribution of Vertical Velocity, Reflectivity, and
Differential Reflectivity.” Monthly Weather Review 123: 1941–1963. doi:10.1175/1520-0493-
(1995)123<1941:TDKAME>2.0.CO;2.
Zipser, E. J. 2003. “Some View on “Hot Towers” after 50 Years of Tropical Field Programs and Two
Years of TRMM Data.“ In Cloud Systems, Hurricanes, and the Tropical Rainfall Measuring Mission
TRMM, 49–58. Boston, MA: American Meteorological Society.
 INTERNATIONAL JOURNAL OF REMOTE SENSING 25

Zipser, E. J., D. J. Cecil, C. Liu, S. W. Nesbitt, and D. P. Yorty. 2006. “Where are the Most Intense
Thunderstorms on Earth?” Bulletin of the American Meteorological Society 87: 1057–1071.
doi:10.1175/BAMS-87-8-1057.
Zipser, E. J., and K. Lutz. 1994. “The Vertical Profile of Radar Reflectivity of Convective Cells a Strong
Indicator of Storm Intensity and Lightning Probability.” Monthly Weather Review 122:
1751–1759. doi:10.1175/1520-0493(1994)122<1751:TVPORR>2.0.CO;2.
Zuluaga, M. D., and R. A. Houze. 2015. “Extreme Convection of the Near-Equatorial Americas,
Africa, and Adjoining Oceans as Seen by TRMM.” Monthly Weather Review 143: 298–316.
doi:10.1175/MWR-D-14-00109.1.