JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 111, D06209, doi:10.1029/2005JD006208, 2006
Effect of cirrus clouds on the diurnal cycle of tropical
deep convective clouds
Gang Hong,1,2 Georg Heygster,1 and Carlos Augusto Morales Rodriguez3
Received 11 May 2005; revised 20 October 2005; accepted 23 November 2005; published 29 March 2006.
[1] The cirrus clouds tightly connected with tropical deep convective clouds can extend
and persist for some hours after the deep convective clouds themselves dissipate. This
can result in time lags of the diurnal cycle of deep convective clouds detected from
infrared satellite measurements with different brightness temperature thresholds because
different amounts of cirrus clouds contaminate the measurement. The diurnal cycles of
rain from the Tropical Ocean–Global Atmosphere (TOGA) radar during the Tropical
Rainfall Measuring Mission (TRMM) Wet Season Atmospheric Mesoscale Campaign
(WETAMC) Large-Scale Biosphere Atmosphere (LBA) Experiment in Amazonia and the
diurnal cycles of deep convective clouds and high cold clouds from the Precipitation
Radar (PR), Visible and Infrared Scanner (VIRS) onboard the TRMM satellite over the
tropics (30S–30N) from November 1998 to April 1999 are investigated to study the
influence of cirrus clouds on the observed diurnal cycle of tropical deep convective
clouds. A 2-hour time lag of the diurnal cycle of deep convective clouds from the VIRS
with respect to that from the PR is found over land. Over ocean the cirrus clouds generated
by deep convective clouds enhance the diurnal cycle of the deep convective clouds from
the VIRS, and a time lag similar to that over land also occurs. The influence of cirrus
clouds leads the diurnal cycle of the deep convective clouds from the VIRS to depend
strongly on the selected IR threshold and to be very different from that of the PR over the
maritime continent. Moreover, over ocean and the maritime continent, from late afternoon
to midnight the strong increase of the deep convective clouds from the VIRS is mainly due
to the developing cirrus clouds near and above the tropical tropopause layer.
Citation: Hong, G., G. Heygster, and C. A. M. Rodriguez (2006), Effect of cirrus clouds on the diurnal cycle of tropical deep
convective clouds, J. Geophys. Res., 111, D06209, doi:10.1029/2005JD006208.
1. Introduction
[2] Deep convective cloud is a more vertically developed
and localized convective event that is characterized by
rapid injection of boundary layer air near or through the
tropopause [e.g., Jiang et al., 2004]. It plays a major role in
the Earth’s climate by transporting heat, moisture, and
momentum from the lower to the upper troposphere.
Convective clouds penetrating into the tropical tropopause
layer (TTL, about 14– 18 km) contribute to the exchange of
air between the troposphere and the stratosphere, and hereby
influence the physical and chemical processes occurring
in the TTL and the stratosphere [e.g., Dessler, 2002;
Gettelman et al., 2002; Notholt et al., 2003].
[3] One of the main characteristics of tropical deep
convective cloud is its strong reaction to diurnal forcing
1
Institute of Environmental Physics, University of Bremen, Bremen,
Germany.
2
Now at Department of Atmospheric Sciences, Texas A&M University,
College Station, Texas, USA.
3
Instituto de Astronomia, Geofisica e Ciências Atmosféricas,
Universidade de São Paulo, São Paulo, Brazil.
Copyright 2006 by the American Geophysical Union.
0148-0227/06/2005JD006208$09.00
[e.g., Yang and Slingo, 2001; Machado et al., 2002; Tian et
al., 2004]. Diurnal changes in deep convective clouds affect
incoming shortwave and outgoing longwave radiation,
affecting the Earth’s radiation balance [e.g., Hendon and
Woodberry, 1993; Hall and Vonder Haar, 1999]. Knowledge of the diurnal cycle of tropical deep convective clouds
is important to evaluate the diurnal cycle described by
general circulation models [e.g., Hall and Vonder Haar,
1999; Yang and Slingo, 2001; Machado et al., 2002; Slingo
et al., 2004; Tian et al., 2004].
[4] The most frequently used technique for objectively
extracting cloud information from satellite images is the
infrared threshold technique [e.g., Machado et al., 1998;
Hall and Vonder Haar, 1999; Soden, 2000; Machado et al.,
2002; Machado and Laurent, 2004; Tian et al., 2004]
applied to geostationary satellite data providing high temporal sampling with wide spatial coverage. However, satellite infrared sensors only provide an indirect estimate of the
properties of deep convective clouds because they do not
sense directly radiation from their interior. They only
observe cloud top temperatures which frequently are similar
for deep convective clouds and cirrus clouds [Liu et al.,
1995; Sui et al., 1997; Hall and Vonder Haar, 1999]. Hong
et al. [2005a] found that the deep convective cloud fraction
(cloud/rain fraction in units of percent is defined as the ratio
D06209
1 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
Figure 1. (a) Regions used to estimate the diurnal cycles
of deep convective clouds, high cold clouds, and rain from
the TRMM PR and VIRS measurements, GOES 8 images,
and TOGA radar in the Amazon region and the WETAMC
LBA area. The square region shows the location of the
WETAMC LBA area. (b) Location of the 150 km range of
the TOGA radar (circle) and the WETAMC LBA area
(square).
of total of cloud/rain flagged pixels over the total number of
samples) detected with infrared methods by the brightness
temperature threshold of 215 K, is over three times larger
than that detected by the method of Hong et al. [2005b]
from passive microwave sensors sensitive to frozen hydrometeors in the interior of deep convective clouds. The
authors explained this discrepancy by the fact that infrared
threshold methods cannot distinguish between deep convective clouds and thick high cirrus clouds.
[5] The characteristics of the life cycle of deep convective
clouds are developing, mature, and dissipating stage. The
dominant cloud type characteristic at each stage is cumulus,
cumulonimbus, and anvil cloud (cirrus) [Inoue, 2003],
respectively. These cirrus clouds, tightly connected with
deep convective clouds [Garreaud and Wallace, 1997; Chou
and Neelin, 1999; Dessler and Yang, 2003; Comstock and
Jakob, 2004; Hong et al., 2005a], can extend and persist for
some hours after the deep convective clouds themselves
dissipate [Gray and Jacobson, 1977; Weickmann et al.,
1977]. It has been speculated that the extension and persistence of cirrus clouds may lead to some hours time lag for
the maximum of the deep convective cloud fraction detected
by infrared satellite observations [Slingo et al., 2004; Hong
et al., 2005a].
D06209
[6] Time lags of the diurnal cycle peak of high cold clouds
(clouds with high-altitude cloud top and low cloud top
temperature) with respect to that of deep convective clouds
have already been found using infrared geostationary satellite observations [e.g., Mapes and Houze, 1993; Janowiak
et al., 1994; Chen and Houze, 1997; Yang and Slingo, 2001;
Machado et al., 2002; Tian et al., 2004; Soden, 2004].
However, the deep convective clouds detected from the
infrared thresholds are still contaminated by cirrus clouds
generated from them [Liu et al., 1995; Sui et al., 1997; Hall
and Vonder Haar, 1999; Hong et al., 2005a]. How these
cirrus clouds impact the observationally based interpretation
of diurnal cycle of deep convective clouds observed with the
infrared thresholds was not investigated. The goal of this
study is to analyze the effect of cirrus clouds on the diurnal
cycle of deep convective clouds observed with infrared
threshold techniques. The Tropical Rainfall Measuring
Mission (TRMM) Precipitation Radar (PR) provides direct
three dimensional observations of precipitating cloud structure [e.g., Kummerow et al., 1998]. The PR is sensitive to
reflectivities above the 17 dBZ threshold [Anagnostou et al.,
2001] associated with large precipitating ice particle [Alcala
and Dessler, 2002], therefore not sensitive to cirrus clouds.
The TRMM satellite also carries the Visible and Infrared
Scanner (VIRS). The diurnal cycles of deep convective
clouds from the PR and VIRS over the tropics (30S–
30N), and the diurnal cycles of rainfall observed from
Tropical Ocean – Global Atmosphere (TOGA) radar during
the TRMM Wet Season Atmospheric Mesoscale Campaign
(WETAMC) Large-Scale Biosphere Atmosphere (LBA)
Experiment in Amazonia are studied here. In section 2, the
satellite data, the surface radar data, and the methods to
estimate the diurnal cycles of deep convective clouds and
high cold clouds are described. Section 3 first presents the
diurnal cycles of deep convective clouds and high cold
clouds from infrared measurements in the Amazon region
and the WETAMC LBA region. Then, the effect of cirrus
clouds generated by deep convective clouds is investigated
over the Amazon region and the whole tropics. Conclusions
are summarized in Section 4.
2. Data and Methodology
[7] This study uses four sources of data: the TRMM PR
rainfall rate and profile product 2A25 (Version 6), the
TRMM VIRS calibrated radiance product 1B01 (Version
6), the GOES 8 infrared brightness temperature images, and
the 3 km Constant Altitude Plan Position Indicator (CAPPI)
from the ground-based TOGA radar [Anagnostou and
Morales, 2002]. Figure 1 shows the regions used to estimate
the diurnal cycles of cloud and rain from the TRMM PR,
VIRS, GOES 8, and TOGA radar in the Amazon region and
the WETAMC LBA area. In order to match the period of the
WETAMC LBA campaign from 9 January to 27 February
1999, the TRMM data in January and February are used in
the Amazon region.
[ 8 ] The TRMM satellite was launched in 1997. It
provides nearly continuous observations of tropical precipitation (between 35S and 35N) [Kummerow et al., 1998].
The TRMM satellite takes about 46 days to return to a given
position at a given local time [Nesbitt and Zipser, 2003].
The 13.8 GHz PR onboard the TRMM satellite is the first
2 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
spaceborne precipitation radar. It scans a 215 km swath
every 0.6 s, measuring profiles of the radar reflectivity
factor. The PR 2A25 data product used in this study has a
horizontal resolution of 4.3 4.3 km2 and a vertical
resolution of 250 m at nadir. Lin et al. [2000] found that
at least three months of PR data must be combined
to adequately sample a 4 5 (longitude, latitude) grid
at 1-hour resolution. In order to obtain enough samples,
here a 20 10 area (Figure 1a) in the Amazon region is
chosen to sample the diurnal cycle of deep convective
clouds during the WETAMC LBA. Moreover, half a year
of data from the PR and VIRS from November 1998 to
April 1999 over the whole tropics are used to study the
relation between the diurnal cycles of cirrus cloud and
tropical deep convective cloud. Negri et al. [2002] suggested that the optimum period of accumulation is 4 hours
to minimize the potentially large variations in sampling
error. Thus, in our study, hourly cloud and rain fractions
from the PR and VIRS are obtained by 4-hour running
means.
[9] The method to detect tropical deep convective clouds
from the PR is based on the procedure of Alcala and
Dessler [2002]. First, convective clouds are discriminated
using the rain type flag values of 200, 210, and 220
indicating convective rain type from the TRMM PR product
2A25 (version 6) [Iguchi et al., 2000]. Then, those pixels
containing a top layer above 10 km and with a thickness
of at least 1.5 km where all the PR reflectivities exceed
17 dBZ, which is the minimum sensitive reflectivity for
the PR, are referred to as deep convective clouds.
[10] The VIRS onboard the TRMM satellite measures
radiances at five different wavelengths. It scans a 720 km
swath with a horizontal resolution of 2.2 km at nadir. For
this study, the 10.8 mm radiances from the VIRS 1B01 data
product are converted to brightness temperatures (T11) using
lookup table (from the TRMM Science Data and Information System, NASA). Cloud fractions are calculated using
different brightness temperature thresholds, T11 < 235 K and
T11 < 210 K respectively, to detect high cold clouds and
deep convective clouds as suggested by Machado et al.
[2002]. In order to distinguish the deep convective clouds
detected from the PR and VIRS, we refer to the clouds
detected from the PR and T11 < 210 K as the PR deep
convective clouds and 210 K deep convective clouds,
respectively. The result of the T11 < 235 K threshold is
referred to as the 235 K high cold clouds. The thresholds of
235 and 210 K correspond to the altitudes of cloud top at 10
and 13.5 km in the tropics, respectively.
[11] The GOES 8 TIFF image data are obtained from
NASA Goddard Space Flight Center (GSFC). The channel
4 image data (11 mm) are converted to brightness temperatures (T11) using the scaling factor provided by the GSFC.
The horizontal resolution is 4 4 km2 at nadir. The images
were available every half hour (at each 15 and 45 min)
according to the NOAA-GOES scanning strategy during the
TRMM LBA experiment [Machado et al., 2002]. The 235 K
high cold cloud and 210 K deep convective cloud fractions
are calculated over the 20 10 area of Figure 1a in the
Amazon region and over the 2.3 2.3 area of the
WETAMC LBA region (61.0W to 63.3W, 12.1S to
9.8S, Figure 1b) [e.g., Machado et al., 2002; Silva Dias
et al., 2002; Machado and Laurent, 2004].
D06209
[12] The TOGA radar was located at 62.34W, 10.78S.
Its complete volume was sampled in 10 minute intervals
over a cycle of 150 km radius. The TOGA radar data have
been extensively used to study the properties of clouds and
rainfall in the Amazon region [e.g., Silva Dias et al., 2002;
Rickenbach et al., 2002; Machado et al., 2002; Rickenbach,
2004]. In this study, the 3 km CAPPI data with a horizontal
resolution of 2 2 km2 are used to sample the diurnal cycle
of rain fraction over the WETAMC LBA region. The
reflectivity (Z) threshold of Z > 10 dBZ is used to determine
all types of rain [Machado et al., 2002]. In order to compare
to the deep convective cloud fraction, the convective rain
fraction is calculated based on the methods of Steiner et al.
[1995] and Rickenbach and Rutledge [1998]. The convective rain center is satisfied wherever Z > 40 dBZ. If a local
reflectivity less than 40 dBZ exceeds the mean value of its
10 10 neighborhood reflectivities by 4.5 dBZ, it is also
identified as convective rain center. Then, a convective cell
with a radius (between 1 and 5 km) that increased with the
mean background reflectivity is placed around the convective rain center.
3. Results
3.1. Diurnal Cycle of Clouds From Infrared
Measurements
[13] Because of the long return time of the TRMM
satellite, we compare in Figure 2 the diurnal cycles of
clouds from the VIRS to those from the GOES 8 over the
chosen 20 10 area in the Amazon region and the
WETAMC LBA area averaged over the months of January
and February 1999. The diurnal cycles of 235 K high cold
clouds and 210 K deep convective clouds from the GOES
over the 20 10 area peak at the same local time and
have very similar shapes as those from the GOES over the
WETAMC LBA area (Figure 2a). This reveals that 235 K
high cold clouds and 210 K deep convective clouds over the
two regions have the same diurnal characteristics. Then, the
diurnal cycles of clouds over the 20 10 area can be used
to represent the properties of the diurnal cycles of cloud
over the WETAMC LBA area.
[14] The diurnal cycles of the 235 K high cold clouds
and 210 K deep convective clouds from the VIRS over
the 20 10 area are estimated and compared to those
from the GOES over the same area (Figure 2b). Although
the diurnal cycles of the 235 K high cold clouds and
210 K deep convective clouds from the VIRS and GOES
show some differences in the absolute values of observed
cloud fractions, the diurnal cycles for the same thresholds
have similar shapes and tendencies for the VIRS and
GOES. The maximum correlations between the diurnal
cycles from the VIRS and GOES are 0.98 and 0.97 for
210 K deep convective clouds and 235 K high cold
clouds with 0 hour time lag, respectively. The diurnal
cycles from the VIRS and GOES peak at the same local
time of 1600 LT consistent with or close to many earlier
studies [e.g., Soden, 2000; Yang and Slingo, 2001;
Machado et al., 2002; Tian et al., 2004; Nowicki and
Merchant, 2004] using satellite infrared measurements.
The differences in the absolute cloud fractions are most
probably due to the different sampling characteristics of
the VIRS and GOES.
3 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
D06209
VIRS measurements. We explain the time difference by the
fact that the infrared brightness temperatures are only
sensitive to the cloud top temperature [e.g., Liu et al.,
1995; Sui et al., 1997; Hall and Vonder Haar, 1999; Hong
et al., 2005a], so that deep convective clouds and cirrus
clouds with similar cloud top altitude cannot be distinguished. In most deep convective clouds, the convective
rain area has a maximum during its developing and mature
stage, therefore leading in time the cloud shield expansion
[Houze, 1977; Rickenbach, 1999]. As a consequence, the
influence of cirrus clouds generated by deep convective
clouds can result in the observed 2-hour time lag for the
peak of the diurnal cycle of the 210 K deep convective
clouds with respect to the PR deep convective clouds. The
‘‘shoulder’’ between 1600 and 1800 LT visible in both VIRS
cloud fractions, but not in the PR cloud fraction can be
explained as cirrus clouds to which the PR is not sensitive.
[17] In order to investigate the effect of cirrus clouds
generated by deep convective cloud on the 210 K deep
convective cloud fractions from the VIRS, the 210 K
deep convective cloud fractions are compared to the
PR deep convective cloud fractions in Figure 3b. The
210 K deep convective cloud fractions from the VIRS are
larger than those from the PR before 0700 LT and after
1300 LT. The maximum value of the 210 K deep convective
clouds (7.5%) is about twice that of the PR deep convective
Figure 2. The 235 K high cold cloud and 210 K deep
convective cloud fractions from (a) GOES over the 20 10 area in the Amazon region and the WETAMC LBA area
(Figure 1) and (b) VIRS and GOES over the 20 10 area.
All data are averages over the period January and February
1999. Cloud fraction is in percent.
3.2. Diurnal Cycle Over the Amazon Region
[15] The diurnal cycles of 235 K high cold clouds and
210 K deep convective clouds from the VIRS and the PR
deep convective clouds, and the differences between the
210 K deep convective cloud fractions and the PR deep
convective cloud fractions over the 20 10 area in the
Amazon region during January – February 1999 are shown
in Figure 3. All cloud fractions derived from satellite data
were calculated as the ratio of all pixels in the region of
interest showing the cloud type in question, summed over
all overflights, to the sum of all pixels in the region of
interest which are covered by each individual overflight.
[16] From Figure 3a we see that the deep convective
cloud fractions from the PR and VIRS have different values
and phase shifts, but similar shapes. The PR deep convective clouds increase sharply starting from 0800 LT, and
reach their maximum at 1400 LT, then begin to decrease
strongly. However, the 210 K deep convective clouds from
the VIRS show a strong increase from 1000 LT, and reach
their maximum at 1600 LT, and then begin to decrease
significantly. The diurnal cycle of the PR deep convective
clouds peaks about 2 hours earlier than that derived from the
Figure 3. (a) Diurnal cycles of the 235 K high cold
clouds, 210 K deep convective clouds, and PR deep
convective clouds and (b) differences between the 210 K
deep convective cloud fractions and the PR deep convective
clouds over the 20 10 area. The dotted line indicates
that the difference is 0. Cloud fraction is in percent.
4 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
D06209
of the 210 K deep convective clouds with respect to the PR
deep convective clouds (Figure 3a) results from cirrus
clouds generated by deep convective clouds.
Figure 4. Diurnal cycles of total rain and convective rain
fractions from the TOGA radar CAPPI over the WETAMC
LBA area during January – February 1999. Cloud and rain
fractions are in percent.
clouds (4%). During 0700 – 1300 LT, the 210 K deep
convective cloud fractions are less than the PR deep
convective cloud fractions. This is because the T11 threshold
at 210 K can only detect the clouds with cloud top above
about 13.5 km which actually are close to the bottom of the
TTL. However, the PR can detect all deep convective
clouds with cloud top above 10 km. After 1300 LT, the
differences increase and the largest differences appear at
1600 – 1800 LT after the 210 K deep convective clouds
peak. This is because cirrus clouds outflown from the deep
convective clouds at mature and dissipating stages can
extend to a large area and persist for some hours [Gray
and Jacobson, 1977; Weickmann et al., 1977].
[18] As shown in section 3.1, the diurnal cycle over the
20 10 area has the same properties as that over the
WETAMC LBA area. The TOGA radar has been used to
investigate the diurnal cycle of clouds and rainfall in
Amazon [e.g., Machado et al., 2002; Rickenbach et al.,
2002; Rickenbach, 2004]. Thus the diurnal cycles derived
from the independent TOGA radar are used to compare to
and evaluate those derived from satellite measurements in
Figure 4. The total rain fractions and convective rain
fractions strongly increase from 0900– 1000 LT and reach
their maxima at 1300 – 1400 LT, then sharply decrease until
2200 – 2300 LT. The tendencies and peaking time are in
agreement with those of the PR deep convective clouds
(Figure 3a). The peaking times of the PR deep convective
clouds and rains from the TOGA radar are consistent with
the time of maximum rainfall documented by Nesbitt and
Zipser [2003] using three years of the PR and TRMM
Microwave Imager (TMI) measurements over land between
36S and 36N. The difference in slope between the two
rain curves from the TOGA radar between 1800– 2200 LT
(Figure 4) suggests that the separate processes of convective
rain and stratiform rain are active during this time interval
[Rickenbach et al., 2002]. From the comparison of
Figures 3a and 4 it is also found that both the total rain
fraction and convective rain fraction peak 2 – 3 hours earlier
than the 210 K deep convective clouds from the VIRS. This
further confirms our interpretation that the 2-hour time lag
3.3. Diurnal Cycle Over the Tropics
[19] In order to obtain a sufficiently dense sampling by
the VIRS and PR, the diurnal cycles of high cold clouds and
deep convective clouds are investigated over 10 10
(longitude, latitude) boxes in the tropics (30S – 30N) from
November 1998 to April 1999. For each 10 10 grid cell
the cloud fractions are determined individually using the
same procedure as described in the last section for the
Amazon region. The distribution of the PR deep convective
cloud fractions in the tropics is shown in Figure 5a. The
geographic distribution of the deep convective cloud activity is consistent with previous results [e.g., Alcala and
Dessler, 2002; Jiang et al., 2004; Hong et al., 2005b].
During the time period from November 1998 to April 1999,
continental deep convective clouds usually occur over the
South Africa and tropical South America. Deep convective
clouds over ocean frequently appear in the Intertropical
Convergence Zone (ITCZ) across the Atlantic, the South
Pacific Convergence Zone (SPCZ), the Western Pacific, and
the Indian Ocean. The region over the maritime continent
combines the deep convective clouds over both land and
ocean.
[20] Considering the geographical variability of diurnal
cycles of deep convective clouds and precipitation [e.g.,
Yang and Slingo, 2001; Nesbitt and Zipser, 2003; Slingo et
al., 2004; Tian et al., 2004; Collier and Bowman, 2004], in
this study the diurnal cycles of 10 10 boxes are
combined over larger regions similar to the regions studied
by Nesbitt and Zipser [2003] and Collier and Bowman
[2004]. The chosen regions are shown in Figure 5b, together
with the phases and maxima of the peaks of the diurnal
cycles of deep convective clouds from the VIRS, PR and the
high cold clouds from the VIRS for a first overview. The
phases and maxima are represented by arrow directions of a
24-hour clock and by the lengths of the arrows, respectively.
The normalized cloud fractions defined by dividing the
cloud fraction by its daily mean over the chosen regions
are shown in Figure 6 in detail.
[21] Over the tropical continental areas of Africa and
Australia, the deep convective clouds and the high cold
clouds from the VIRS have a distinct 2-hour time lags with
respect to the PR deep convective clouds (Figure 5b). Note
that turn off of the PR over a partial area of the chosen
region in Australia due to intergovernmental agreement
potentially influences the diurnal cycle from the PR. Moreover, the 235 K high cold cloud fractions and 210 K deep
convective cloud fractions both peak at 1700 – 1800 LT over
the two land regions (Figure 5b). Although over the tropical
South America the PR deep convective clouds peak only
one hour earlier than the 210 K deep convective clouds, the
distinct 2-hour time lag between them still occurs if investigating the details of their diurnal cycles (Figure 6b).
[22] Over ocean, both the 210 K deep convective clouds
and PR deep convective clouds peak in early morning at
0100– 0700 LT (Figure 5b). Moreover, similar as over land,
distinct time lags between both signals are also found over
the tropical Indian, west Pacific, and tropical Atlantic.
However, unlike over the tropical Africa and the tropical
5 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
D06209
Figure 5. (a) Distribution of deep convective cloud fractions from PR in the tropics (30S – 30N)
averaged from November 1998 to April 1999. (b) Nine regions of interest together with the phase and
amplitude of the diurnal cycles of deep convective clouds from PR and VIRS (T11 < 210, 235 K). The
direction of the arrow indicates the local time of the peak of the diurnal cycle (phase) and the length of the
arrow indicates the cloud fractions at the peak time (amplitude).
South America, in general their maximum fractions have
only small differences at peak time (Figure 5b). The diurnal
variation ranges (difference between the maximum cloud
fraction and the minimum cloud fraction) of the normalized
deep convective cloud and high cold cloud fractions over
ocean are much weaker than those over land (Figure 6).
This is consistent with previous observational studies on
deep convective clouds [e.g., Hendon and Woodberry,
1993; Soden, 2000; Yang and Slingo, 2001; Tian et al.,
2004] and precipitation [e.g., Gray and Jacobson, 1977;
Dai, 2001; Nesbitt and Zipser, 2003; Collier and Bowman,
2004]. On the other hand, the normalized diurnal variation
ranges of the 210 K deep convective cloud fractions are
generally twice as large as those of the PR deep convective
cloud fractions varying in the range of 0.7– 1.4 (Figures 6d –
6h). In contrast to these findings over ocean, over land the
normalized diurnal variation ranges of the 210 K deep
convective cloud fractions are similar to those of the PR
deep convective cloud fractions. We conclude that over
ocean, the cirrus clouds generated by deep convective
clouds not only result in time lags relative to the PR deep
convective clouds, but also enhance the diurnal variations of
the 210 K deep convective clouds.
[23] The peaking times of the 235 K high cold cloud
fractions vary from 0800 to 1600 LT over ocean (Figure 5b).
Moreover, over the tropical Indian, the east Pacific, and the
tropical Atlantic (Figures 6d, 6g, and 6h), the 235 K high
cold cloud fractions also have a small secondary peak in the
6 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
D06209
Figure 6. Normalized diurnal cycles of high cold clouds and deep convective clouds from the VIRS and
deep convective clouds from the PR over the nine regions shown in Figure 5b.
early morning (0300– 0500 LT) when the deep convective
clouds from the PR and VIRS peak. Over the West Pacific
(Figure 6e), the secondary peaks of the 235 K high cold
cloud fractions are in the middle afternoon (1600 LT). The
feature of multiple diurnal peaks of the 235 K high cold
cloud fractions over ocean differing from those over land is
most probably due to the mechanisms of deep convective
clouds over ocean [e.g., Gray and Jacobson, 1977; Randall
et al., 1991; Chen and Houze, 1997]. Figure 7 shows the
averaged combined diurnal cycles over land, ocean, and the
maritime continent from Figure 6, respectively. The time lag
of 2 hours already observed over land in Figure 6 for the
maximum of the 210 K deep convective clouds and the
235 K high cold clouds with respect to the PR deep convective clouds, which is due to the influence of cirrus clouds
clearly shows up in the average of Figure 7a. The 210 K deep
convective clouds over ocean show stronger variations than
the PR deep convective clouds, and they peak at 0600 LT
2 hours later than the PR deep convective clouds. Again, this
indicates that over ocean the cirrus clouds generated by deep
convective clouds not only can result in a time lag of similar
as shown over land, but also they enhance the diurnal
variations of the 210 K deep convective clouds.
[24] Over the tropical oceans, previous studies found that
the time of maximum occurrence of cold clouds varies
substantially with the infrared brightness temperature
thresholds [e.g., Fu et al., 1990; Janowiak et al., 1994;
Chen and Houze, 1997; Soden, 2000; Yang and Slingo,
2001; Tian et al., 2004]. We observe a 7 – 9 hours delay of
the maximum of the 235 K high cold clouds relative to both
the 210 K deep convective clouds and the PR convective
clouds (Figure 7b). The maximum of deep convective
clouds is in the early morning (0400 –0600 LT). A possible
explanation of these observations is that after sunrise,
warming at the cloud top due to the solar absorption
increases the stability and therefore reduces convective
activities [e.g., Randall et al., 1991]. Consequently, deep
convective clouds strongly dissipate till the late afternoon
and therefore the 235 K high cold clouds strongly increase
till noon (1300 LT). The increase of the 235 K high cold
clouds and the decrease of the 210 K deep convective
clouds from the late morning to the early afternoon, also
can be interpreted that the warming at the cloud tops
depresses the developing of high cirrus clouds above
13.5 km corresponding to the 210 K infrared temperature.
From late afternoon to early morning, the 210 K deep
7 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
Figure 7. PR and 210 K deep convective cloud fractions
and the 235 K high cold cloud fractions over (a) ocean,
(b) land, and (c) maritime continent in the tropics. Cloud
fraction is in percent.
convective clouds generally increase, and just 2 hours after
they start increasing, the PR deep convective clouds also
increase, however, the 235 K high cold clouds generally
decrease until 2300 LT. This indicates that the increase of
the 210 K deep convective clouds is mainly due to the
development of cirrus clouds above 13.5 km, probably
caused by the destabilization resulting from infrared cooling
at cloud tops although it also includes some contributions
from the development of the deep convective clouds themselves. In this study the threshold of 235 K is used to
identify high cold clouds. It should be remarked that thresholds in the range of 230 – 240 K also have been used
extensively as an indicator of deep convective clouds over
both land and ocean [e.g., Arkin and Meisner, 1987;
Hendon and Woodberry, 1993; Yang and Slingo, 2001;
Ricciardulli and Sardeshmukh, 2002; Wilcox, 2003; Tian
D06209
et al., 2004]. From Figure 7b it is obvious that over ocean
the deep convective clouds identified by these brightness
temperature thresholds (e.g., 230 – 240 K) have a large
influence due to the cirrus clouds generated by deep
convective clouds.
[25] Over the maritime continent, the normalized diurnal
variation ranges of deep convective cloud and high cold
cloud fractions (Figure 6i) are much weaker than those over
land; however, the cloud fractions still have distinct diurnal
cycles. The PR deep convective cloud fractions peak at
1400 LT which is close to the time as over land, while the
210 K deep convective clouds peak at midnight (Figures 7a
and 7c). The PR deep convective cloud fractions also have a
secondary peak at 0300 LT close to the peaking time of the
PR deep convective cloud fractions over land. Moreover,
the 235 K high cold cloud fractions peak at 1500 LT
between the peaking times over ocean (1300 LT) and land
(1700 LT). These features can be explained by understanding the mechanisms of convective clouds over the maritime
continent as a combination of that over land and those over
ocean. From the late morning to early afternoon, the PR
deep convective clouds monotonically increase. This process is mainly attributed to a direct thermodynamic response
to the strong diurnal cycle of the land surface temperature,
which is the same as that over land. The high PR deep
convective cloud fraction lasts till the late afternoon. However, during this period the solar absorption warms cloud
top and depresses the developing of deep convective clouds
over ocean [Randall et al., 1991], so that the 210 K deep
convective clouds (cloud top above 13.5 km) have low
values although the 235 K high cold clouds (cloud top
above 10 km) have their maximum. From the early evening
to midnight, the increase of the deep convective clouds from
the PR and VIRS and the decrease of the 235 K high cold
clouds are similar to those over ocean. So during this period,
the increase of the 210 K deep convective clouds contain a
strong contribution from the developing cirrus clouds above
13.5 km. In the early morning, the PR deep convective
clouds reach their secondary maximum at 0300 LT, and
5 hours later (0800 LT), the 210 K deep convective clouds
reach their secondary maximum. The 210 K deep convective clouds increase strongly with a sharp decrease of the PR
deep convective clouds from 0500 to 0800 LT. This can be
explained as the effect of thick cirrus clouds generated by
the deep convective cloud. The dissipation of deep convective clouds extend their cloud shields and generate high
thick cirrus clouds with large areas, consequently resulting
in the time delay of the 210 K deep convective cloud
maximum fraction.
4. Conclusions and Discussions
[26] The diurnal cycle of tropical deep convective clouds
is an important key for understanding the earth radiation
budget. Knowledge of the diurnal cycle is very important to
validate the diurnal cycle produced by numerical simulations [e.g., Hall and Vonder Haar, 1999; Yang and Slingo,
2001; Machado et al., 2002; Slingo et al., 2004; Tian et al.,
2004]. The geostationary satellite infrared sensor is widely
used to study deep convective clouds since the investigator
can do a consistent analysis of large geographic regions
[e.g., Tian et al., 2004]. However, satellite infrared sensors
8 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
are only sensitive to the cloud top which acts as an umbrella
making deeper cloud properties invisible for this type of
sensors [e.g., Liu et al., 1995; Sui et al., 1997; Hall and
Vonder Haar, 1999; Hong et al., 2005a].
[27] Time lags between cloud fraction diurnal cycles
detected with high temperature thresholds and those with
low temperature thresholds have been found earlier [e.g.,
Mapes and Houze, 1993; Janowiak et al., 1994; Tian et al.,
2004; Soden, 2000; Chen and Houze, 1997; Yang and
Slingo, 2001], but the influence of cirrus clouds generated
by deep convective clouds on infrared observation of the
diurnal cycle of deep convective clouds has not been
investigated until now. In order to isolate the influence of
cirrus clouds on the diurnal cycle of infrared-observed
tropical deep convective clouds, in this study the PR
reflectivity profiles are used to determine deep convective
cloud fractions using the method of Alcala and Dessler
[2002]. The diurnal cycles of tropical deep convective
clouds from the PR are compared to those from the VIRS
onboard the same satellite TRMM using the brightness
temperature thresholds of 210 K for deep convective clouds
and 235 K for cold clouds respectively in the Amazon
region during the period of the WETAMC LBA experiment.
Comparison shows that the diurnal cycle of deep convective
clouds from the PR peaks at 1400 LT, which is in agreement
well with the diurnal cycles of rain from the TOGA radar.
Furthermore, cirrus clouds generated by deep convective
clouds can result in about a 2-hour time lag for deep
convective clouds from the VIRS.
[28] The effect of cirrus clouds on the diurnal cycles of
deep convective clouds over regions with frequent deep
convective clouds in the whole tropics (30S – 30N) is
investigated using six months (November – April 1998 –
1999) of the PR and VIRS data.
[29] Over the tropical Africa, tropical South America, and
Australia, consistently a distinct 2-hour time lag of the
210 K deep convective clouds with respect to the PR deep
convective clouds is found, while there is no distinct time
difference between the peaks of the 210 K deep convective
clouds and the 235 K high cold clouds (Figures 6a – 6c). It
seems that over land the peaking time of cold clouds is
independent of the temperature thresholds of 210 and 235 K,
in agreement will with the finding of Janowiak et al. [1994],
showing that the cirrus clouds over land are thick and have
high cloud tops. The deep convective cloud fractions from
the PR and VIRS are strongly different since the influence
of cirrus clouds and their normalized diurnal variation
ranges are strong over land, but their normalized diurnal
variation ranges are very similar (Figures 6a – 6c).
[30] Over ocean, the deep convective clouds from the PR
and VIRS both peak in the early morning (Figures 6d– 6h),
and a 2-hour time lag similar to that over land is found
(Figure 7b). Moreover, it is found that the cirrus clouds
generated by the deep convective clouds enhance the
diurnal cycles of the 210 K deep convective clouds. The
210 K deep convective clouds have stronger variations than
the PR deep convective clouds. We also found that the
maximum of the 235 K high cold clouds occurs 7 – 9 hours
later than the maximum of both PR and 210K deep
convective clouds (Figure 7b).
[31] Over the maritime continent, the diurnal cycle of the
210 K deep convective clouds is totally different from that
D06209
of the PR deep convective clouds. The 210 K deep
convective clouds peak at midnight and the PR deep
convective clouds peak in the afternoon close to the time
over land. The different characteristics of the 210 K deep
convective cloud fractions and PR deep convective clouds
in the evening and early morning are mainly due to the
developing cirrus clouds above 13.5 km.
[32] Inconsistencies between the actual timing of the
precipitation maximum and deep convective cloud maximum inferred from infrared brightness temperatures have
been noted by Yang and Slingo [2001] and Slingo et al.
[2004]. Janowiak et al. [1994] found that the maximum
precipitation observed from in situ measurements of rainfall
during the TOGA Coupled Ocean-Atmosphere Response
Experiment (TOGA COARE) generally occurred some
3 hours earlier than the maximum extent of 215 K deep
convective clouds. Similar results were also found by
Petersen et al. [1996], Chen and Houze [1997], and
Rickenbach [1999] in the same experiment. McGarry and
Reed [1978] noted that over West Africa the rainfall
maxima preceded the convective cloud maxima by up to
2 hours, which they associated with the spreading out of the
convective cloud tops after the peak in precipitation.
Machado et al. [2002] found that in the Amazon Basin
the intense rain fraction (Z > 35 dBZ) from the TOGA
radar peaks 2 – 3 hours earlier than the 210 K deep convective clouds from the GOES during the WETAMC LBA.
These time lags of deep convective clouds from infrared
measurements with respect to the rainfall maximum are
consistent with the 2-hour time lag of deep convective
clouds from infrared measurements with respect to the peak
of the deep convective clouds from the PR found in this
study over land and ocean. Moreover, the peaking times of
the PR deep convective clouds over land occurring at about
1500 LT and ocean occurring at about 0400 LT are in
agreement well with the precipitation maxima derived from
three years of the TRMM PR and TMI data [Nesbitt and
Zipser, 2003]. Thereby, understanding the contribution of
cirrus cloud generated by deep convective clouds on the
infrared-observed diurnal cycles is not only very important
to determine exacter diurnal cycles themselves, but also very
useful for understanding the mechanisms behind the evolution of tropical deep convective clouds, which in turn are
important for climate modeling.
[33] There have been relatively few studies on the diurnal
evolution of deep convective clouds in climate models.
Most of them compared the cycles in models to observations from infrared sensors [e.g., Yang and Slingo, 2001;
Wilcox, 2003; Slingo et al., 2004; Tian et al., 2004].
Although multi satellite sensor retrievals have been used
to analyze deep convective clouds [Roca et al., 2002], the
diurnal cycle of rainfall [Negri et al., 2002; Sorooshian et
al., 2002; Hong et al., 2005c], convective intensity [Nesbitt
and Zipser, 2003], and convective structure [Petersen et al.,
2002], only the recent study by Collier and Bowman [2004]
compared the diurnal cycle of tropical precipitation in a
general circulation model to observations from the TRMM
TMI and PR. The present study has investigated the effect
of cirrus clouds on the infrared-derived diurnal cycle of
deep convective clouds. Combining our findings with
observations from different satellite sensors including infrared [e.g., Tian et al., 2004], microwave [e.g., Hong et al.,
9 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
2005b], and precipitation radar [e.g., Alcala and Dessler,
2002] in order to separate quantitatively the diurnal cycles
of deep convective clouds and cirrus clouds will be a
helpful tool to evaluate the diurnal cycles of both cloud
types in climate models.
[34] Acknowledgments. We wish to thank L. A. T. Machado for
providing the TOGA radar data and the TRMM Ground Validation Office
for providing the radar Volume Scans in UF format. The GOES 8 data used
in this study were acquired from the TRMM LBA. The VIRS and PR data
used in this study were acquired as part of TRMM. The algorithms were
developed by the TRMM Science Team. The data were processed by the
TRMM Science Data and Information System (TSDIS) and the TRMM
Office; they are archived and distributed by the Goddard Distributed Active
Archive Center. TRMM is an international project jointly sponsored by the
Japan National Space Development Agency (NASDA) and the NASA
Office of Earth Sciences. We thank the three anonymous reviewers for
constructive comments and helpful suggestions.
References
Alcala, C. M., and A. E. Dessler (2002), Observations of deep convection
in the tropics using the Tropical Rainfall Measuring Mission (TRMM)
precipitation radar, J. Geophys. Res., 107(D24), 4792, doi:10.1029/
2002JD002457.
Anagnostou, E. N., and C. A. Morales (2002), Rainfall estimation from
TOGA radar observations during LBA field campaign, J. Geophys. Res.,
107(D20), 8068, doi:10.1029/2001JD000377.
Anagnostou, E. N., C. A. Morales, and T. Dinku (2001), The use of TRMM
precipitation radar observations in determining ground radar calibration
biases, J. Atmos. Oceanic Technol., 18, 616 – 628.
Arkin, P. A., and B. N. Meisner (1987), The relationship between largescale convective rainfall and cold cloud over the Western Hemisphere
during 1982 – 84, Mon. Weather Rev., 115, 51 – 74.
Chen, S. S., and R. A. Houze Jr. (1997), Diurnal variation and life-cycle of
deep convective systems over the tropical Pacific warm pool, Q. J. R.
Meteorol. Soc., 123, 357 – 388.
Chou, C., and J. D. Neelin (1999), Cirrus detrainment-temperature feedback, Geophys. Res. Lett., 26, 1295 – 1298.
Collier, J. C., and K. P. Bowman (2004), Diurnal cycle of tropical precipitation in a general circulation model, J. Geophys. Res., 109, D17105,
doi:10.1029/2004JD004818.
Comstock, J. M., and C. Jakob (2004), Evaluation of tropical cirrus cloud
properties derived from ECMWF model output and ground based measurements over Nauru Island, Geophys. Res. Lett., 31, L10106,
doi:10.1029/2004GL019539.
Dai, A. (2001), Global precipitation and thunderstorm frequencies. Part II.
Diurnal variations, J. Clim., 14, 1112 – 1128.
Dessler, A. E. (2002), The effect of deep tropical convection on the tropical
tropopause layer, J. Geophys. Res., 107(D3), 4033, doi:10.1029/
2001JD000511.
Dessler, A. E., and P. Yang (2003), The distribution of tropical thin cirrus
clouds inferred from Terra MODIS data, J. Clim., 16, 1241 – 1247.
Fu, R., A. D. Del Genio, and W. B. Rossow (1990), Behavior of deep
convective clouds in the tropical Pacific deduced from ISCCP radiances,
J. Clim., 3, 1129 – 1152.
Garreaud, R. D., and J. M. Wallace (1997), The diurnal march of convective
cloudiness over the Americas, Mon. Weather Rev., 125, 3157 – 3171.
Gettelman, A., M. L. Salby, and F. Sassi (2002), The distribution and
influence of convection in the tropical tropopause region, J. Geophys.
Res., 107(D10), 4080, doi:10.1029/2001JD001048.
Gray, W. M., and R. W. Jacobson Jr. (1977), Diurnal variation of deep
cumulus convection, Mon. Weather Rev., 105, 1171 – 1188.
Hall, T. J., and T. H. Vonder Haar (1999), The diurnal cycle of west Pacific
deep convection and its relation to the spatial and temporal variations of
tropical MCSs, J. Atmos. Sci., 56, 3401 – 3415.
Hendon, H. H., and K. Woodberry (1993), The diurnal cycle of tropical
convection, J. Geophys. Res., 98(D9), 16,623 – 16,637.
Hong, G., G. Heygster, and K. Kunzi (2005a), Intercomparison of deep
convective cloud fractions from passive infrared and microwave radiance
measurements, IEEE Geosci. Remote Sens. Lett., 2, 18 – 22.
Hong, G., G. Heygster, J. Miao, and K. Kunzi (2005b), Detection of tropical deep convective clouds from AMSU-B water vapor channels measurements, J. Geophys. Res., 110, D05205, doi:10.1029/2004JD004949.
Hong, Y., K.-L. Hsu, S. Sorooshian, and X. Gao (2005c), Improved representation of diurnal variability of rainfall retrieved from the Tropical
Rainfall Measurement Mission Microwave Imager adjusted Precipitation
Estimation From Remotely Sensed Information Using Artificial Neural
D06209
Networks (PERSIANN) system, J. Geophys. Res., 110, D06102,
doi:10.1029/2004JD005301.
Houze, R. A., Jr. (1977), Structure and dynamics of a tropical squall line
system, Mon. Weather Rev., 105, 1540 – 1567.
Iguchi, T., T. Kozu, R. Meneghini, J. Awaka, and K. Okamoto (2000),
Rain-profiling algorithm for the TRMM precipitation radar, J. Appl. Meteorol., 39, 2038 – 2052.
Inoue, T. (2003), Deep convection observed from split windows of GOES
and PR/TRMM, LIS/TRMM, paper presented at the International
Geoscience and Remote Sensing Symposium 2003, Geosci. Remote
Sens. Soc., Toulouse, France.
Janowiak, J. E., P. A. Arkin, and M. Morrissey (1994), An examination of
the diurnal cycle in oceanic tropical rainfall using satellite and in situ
data, Mon. Weather Rev., 122, 2296 – 2311.
Jiang, J. H., B. Wang, K. Goya, K. Hocke, S. D. Eckermann, J. Ma, D. L.
Wu, and W. G. Read (2004), Geographical distribution and interseasonal
variability of tropical deep convection: UARS MLS observations and
analyses, J. Geophys. Res., 109, D03111, doi:10.1029/2003JD003756.
Kummerow, C., W. Barnes, T. Kozu, J. Shiue, and J. Simpson (1998), The
Tropical Rainfall Measuring Mission (TRMM) sensor package, J. Atmos.
Oceanic Technol., 15, 809 – 817.
Lin, X., D. A. Randall, and L. D. Fowler (2000), Diurnal variability of the
hydrologic cycle and radiative fluxes: Comparisons between observations
and a GCM, J. Clim., 13, 4159 – 4179.
Liu, G., J. A. Curry, and R.-S. Sheu (1995), Classification of clouds over
the western equatorial Pacific Ocean using combined infrared and microwave satellite data, J. Geophys. Res., 100, 13,811 – 13,826.
Machado, L. A. T., and H. Laurent (2004), The convective system area
expansion over Amazonia and its relationships with convective system
life duration and high-level wind divergence, Mon. Weather Rev., 132,
714 – 725.
Machado, L. A. T., W. B. Rossow, R. L. Guedes, and A. W. Walker (1998),
Life cycle variations of mesoscale convective systems over the Americas,
Mon. Weather Rev., 126, 1630 – 1654.
Machado, L. A. T., H. Laurent, and A. A. Lima (2002), Diurnal march of
the convection observed during TRMM-WETAMC/LBA, J. Geophys.
Res., 107(D20), 8064, doi:10.1029/2001JD000338.
Mapes, B. E., and R. A. Houze Jr. (1993), Cloud clusters and superclusters
over the oceanic warm pool, Mon. Weather Rev., 121, 1398 – 1415.
McGarry, M. M., and R. J. Reed (1978), Diurnal variations in convective
activity and precipitation during phases II and III, Mon. Weather Rev.,
106, 101 – 113.
Negri, A. J., T. L. Bell, and L. Xu (2002), Sampling of the diurnal cycle of
precipitation using TRMM, J. Atmos. Oceanic Technol., 19, 1333 – 1344.
Nesbitt, S. W., and E. J. Zipser (2003), The diurnal cycle of rainfall and
convective intensity according to three years of TRMM measurements,
J. Clim., 16, 1456 – 1475.
Notholt, J., et al. (2003), Enhanced upper tropical tropospheric COS:
Impact on the stratospheric aerosol layer, Science, 300, 307 – 310.
Nowicki, S. M. J., and C. J. Merchant (2004), Observations of diurnal and
spatial variability of radiative forcing by equatorial deep convective
clouds, J. Geophys. Res., 109, D11202, doi:10.1029/2003JD004176.
Petersen, W. A., S. A. Rutledge, and R. E. Orville (1996), Cloud-to-ground
lightning observations from TOGA COARE: Selected results and lightning location algorithms, Mon. Weather Rev., 124, 602 – 620.
Petersen, W. A., S. W. Nesbitt, R. J. Blakeslee, R. Cifelli, P. Hein, and S. A.
Rutledge (2002), TRMM observations of intraseasonal variability in convective regimes over the Amazon, J. Clim., 15, 1278 – 1294.
Randall, D. A., Harshvardhan, and D. A. Dazlich (1991), Diurnal variability of the hydrologic cycle in a general circulation model, J. Atmos. Sci.,
48, 40 – 62.
Ricciardulli, L., and P. D. Sardeshmukh (2002), Local time- and space
scales of organized tropical deep convection, J. Clim., 15, 2775 – 2790.
Rickenbach, T. M. (1999), Cloud-top evolution of tropical oceanic squall
lines from radar reflectivity and infrared satellite data, Mon. Weather Rev.,
127, 2951 – 2976.
Rickenbach, T. M. (2004), Nocturnal cloud systems and the diurnal variations of clouds and rainfall in southwestern Amazonia, Mon. Weather
Rev., 132, 1201 – 1219.
Rickenbach, T. M., and S. A. Rutledge (1998), Convection in TOGA
COARE: Horizontal scale, morphology and rainfall production, J. Atmos.
Sci., 55, 2715 – 2729.
Rickenbach, T. M., R. N. Ferreira, J. B. Halverson, D. L. Herdies, and
M. A. F. Silva Dias (2002), Modulation of convection in the southwestern
Amazon basin by extratropical stationary fronts, J. Geophys. Res.,
107(D20), 8040, doi:10.1029/2000JD000263.
Roca, R., M. Viollier, L. Picon, and M. Desbois (2002), A multisatellite
analysis of deep convection and its moist environment over the Indian
Ocean during the winter monsoon, J. Geophys. Res., 107(D19), 8012,
doi:10.1029/2000JD000040.
10 of 11
D06209
HONG ET AL.: EFFECT OF CIRRUS CLOUDS ON DIURNAL CYCLE
Silva Dias, M. A. F., et al. (2002), Cloud and rain processes in a biosphereatmosphere interaction context in the Amazon region, J. Geophys. Res.,
107(D20), 8072, doi:10.1029/2001JD000335.
Slingo, A., K. I. Hodges, and G. J. Robinson (2004), Simulation of the
diurnal cycle in a climate model and its evaluation using data from
Meteosat 7, Q. J. R. Meteorol. Soc., 130, 1449 – 1467.
Soden, B. J. (2000), The diurnal cycle of convection, clouds, and water
vapor in the tropical upper troposphere, Geophys. Res. Lett., 27, 2173 –
2176.
Soden, B. J. (2004), The impact of tropical convection and cirrus on upper
tropospheric humidity: A Lagrangian analysis of satellite measurements,
Geophys. Res. Lett., 31, L20104, doi:10.1029/2004GL020980.
Sorooshian, S., X. Gao, K. Hsu, R. A. Maddox, Y. Hong, H. V. Gupta, and
B. Imam (2002), Diurnal variability of tropical rainfall retrieved from
combined GOES and TRMM satellite information, J. Clim., 15, 983 –
1001.
Steiner, M., R. A. Houze Jr., and S. E. Yuter (1995), Climatological characterization of three-dimensional storm structure from operational radar
and rain gauge data, J. Appl. Meteorol., 34, 1978 – 2007.
Sui, C.-H., K.-M. Lau, Y. N. Takayabu, and D. A. Short (1997), Diurnal
variations in tropical oceanic cumulus convection during TOGA
COARE, J. Atmos. Sci., 54, 639 – 655.
D06209
Tian, B., B. J. Soden, and X. Wu (2004), Diurnal cycle of convection,
clouds, and water vapor in the tropical upper troposphere: Satellites
versus a general circulation model, J. Geophys. Res., 109, D10101,
doi:10.1029/2003JD004117.
Weickmann, H. K., A. B. Long, and L. R. Hoxit (1977), Some examples of
rapidly growing oceanic cumulonimbus clouds, Mon. Weather Rev., 105,
469 – 476.
Wilcox, E. M. (2003), Spatial and temporal scales of precipitating tropical
cloud systems in satellite imagery and the NCAR CCM3, J. Clim., 22,
3545 – 3559.
Yang, G.-Y., and J. Slingo (2001), The diurnal cycle in the tropics, Mon.
Weather Rev., 129, 784 – 801.
G. Heygster, Institute of Environmental Physics, University of Bremen,
D-28334 Bremen, Germany.
G. Hong, Department of Atmospheric Sciences, Texas A&M University,
College Station, TX 77843, USA. ([email protected])
C. A. M. Rodriguez, Instituto de Astronomia, Geofisica e Ciências
Atmosféricas, Universidade de São Paulo, São Paulo, SP 05508-900,
Brazil.
11 of 11