Remote Sensing Letters

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/trsl20

Monitoring cloud parameters using a ground-

based airglow imager

R. N. Ghodpage, A. Taori, R. P. Patil, M.K. Patil, O. B. Gurav, S. Sripathi & A.P.

Dimri

To cite this article: R. N. Ghodpage, A. Taori, R. P. Patil, M.K. Patil, O. B. Gurav, S. Sripathi &
A.P. Dimri (2023) Monitoring cloud parameters using a ground- based airglow imager, Remote
Sensing Letters, 14:8, 854-866, DOI: 10.1080/2150704X.2023.2247520

To link to this article: https://doi.org/10.1080/2150704X.2023.2247520

Published online: 21 Aug 2023.

Submit your article to this journal

View related articles

View Crossmark data

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

https://www.tandfonline.com/action/journalInformation?journalCode=trsl20
REMOTE SENSING LETTERS
2023, VOL. 14, NO. 8, 854–866
https://doi.org/10.1080/2150704X.2023.2247520

Monitoring cloud parameters using a ground- based airglow

imager
R. N. Ghodpagea, A. Taorib, R. P. Patil c
, M.K. Patild, O. B. Gurave, S. Sripathif
and A.P. Dimrif
a
M. F. Radar, Indian Institute of Geomagnetism, Shivaji University Campus, Kolhapur, India; bEarth & Climate
Science Area, National Remote Sensing Centre, Hyderabad, India; cIndian Institute of Tropical Meteorology,
Pune, India; dSchool of Physical Sciences, Swami Ramanand Teerth Marathwada University, Nanded, India;
e
Department of Physics, Bharati Vidyapeeth (Deemed to Be University), Pune, India; fIndian Institute of
Geomagnetism, Navi Mumbai, India

ABSTRACT ARTICLE HISTORY

The current study investigates the application of the photogram­ Received 15 March 2023
metric technique, commonly used for satellite measurements, on Accepted 4 August 2023
the All Sky Imager (ASI), which is used for night airglow observa­ KEYWORDS
tions from Kolhapur (16.8° N, 74.2° E), India. Using image processing Airglow; cloud; motion-
methods, the current study spanning from the year 2016 to 2020 vector; Climate
(March – May months) estimates the cloud speeds to range from 10
m s−1 to 18 m s−1 during the period under consideration. The
slowest speed of 10 ± 3 m s−1 was evidenced in 2017, while it
takes value of 15 ± 3 m s−1 over other years. The clouds are found
to move in a south-westerly direction during the time period under
consideration. The cloud cover fraction varies in between ~ 0.178 to
0.594 from 2016 to 2020. Our analysis indicates systematic changes
in the pre-monsoon cloud fraction and direction of the cloud move­
ment. The observed trend indicates that the monsoon pattern is
somewhat changing

1. Introduction
Clouds play multiple roles in the earth’s atmosphere. They scatter the incoming solar
radiation in the atmosphere and act as a blanket to the earth’s outgoing long-wave
radiation (e.g., Taylor Patric 2012). Clouds have significant impact on the Earth’s climate,
which by direct and indirect processes is non-linear in its nature, such as the effect is
modulated by the spatio-temporal distribution and their height, thickness, size distribu­
tion, etc (e.g., Cesana and Storelvmo 2017). Therefore, high resolution observations using
space-borne and ground—based instruments are essentially called for. Satellite sensors
detect the cloud motion and under the assumption that clouds move with winds and
derive the Cloud Motion Vector, ‘CMV’. Values of CMV are very useful in understanding the
synoptic scale atmospheric dynamics and circulations (e.g., Menzel 2001). In the last few
decades, there has been significant progress in retrieval techniques from a simple cross-
correlation analysis (Hubert and Whitney 1971; Izawa and Fujita 1969; Leese, Novak, and

CONTACT R. N. Ghodpage [email protected] M. F. Radar, Indian Institute of Geomagnetism, Shivaji

University Campus, Kolhapur 416004, India
© 2023 Informa UK Limited, trading as Taylor & Francis Group
 REMOTE SENSING LETTERS 855

Clark 1971) to very advance photometry and satellite imagery that tracks the clouds and
calculate CMVs with more accuracy (Deb et al. 2015; Kaur et al. 2014; Kishtawal et al. 2009;
Velden et al. 1997).
Ground-based cloud observations used to be based on visual observations at weather
stations. Though visual observations have provided basic information for weather analysis
and climate studies, their use is restricted to limited resolution in time and prone to
observer errors. It is understood that ground-based instrumentation with a large field of
view would augment satellite cloud observations (Klebe, Blatherwick, and Morris 2014; Liu
et al. 2013; Smith and Toumi 2008; Souza-Echer et al. 2006). The sky-imaging devices
include Whole-sky imager, Total-sky imager, All-sky imager, achieve real-time hemispheric
sky images and cloud fraction (Kazantzidis et al. 2012; Long et al. 2006; Yang et al. 2012).
Over Indian sector, Nikumbh et al. (2019) used a Total Sky Imager (TSI) the for a year-
long data, which is operated at a high altitude station Mahabaleshwar (17.93° N, 73.66° E,
1350 m amsl) during daytime and compared their findings on the cloud fraction with data
acquired by the Moderate Resolution Imaging Spectro Radiometer (MODIS) on board the
satellites Terra and Aqua (Salomonson et al. 1989). The investigations on the high spatial
resolution data on the clouds using the night-time ground-based airglow image measure­
ments over Indian sector is very scanty till date. In this regard, Night airglow imagers may
be of some use. These imagers are designed primarily for monitoring the emission
received from mesosphere and lower thermosphere (MLT) region with high spatial and
temporal resolution. These imagers, under clear sky conditions, are utilized to study
waves, transient short scale ripples, plasma depletions (Batista et al. 2000; Patil et al.
2016; Taori, Jayaraman, and Kamalakar 2013; Taylor, Bishop, and Taylor 1995) and other
dynamical features. All sky imagers are generally operated near to the new moon periods.
However, at the same time, its capability to observe clouds on a continuous basis provides
an opportunity to derive the CMVs and cloud coverage. Present study aims to explore the
capability of an airglow imager for deriving the Cloud cover and CMVs, thereby extending
its utilization to the lower atmosphere.

2. Instrumentation
The cloud images used in the present study are obtained with the help of an All Sky
Imager over Kolhapur (16.8° N, 74.2° E), which monitors night airglow OI 557.7 nm
emission. The selection of wavelength is done using interference filters (transmission
0.6, full width at half maxima 10 nm). The integration time for the selected wavelength is
2 minutes with cadence time ~6 min. The front-end optics collects the incoming radiation
in 180° field of view with a (f/4 Mamiya RB67) fish-eye lens that has a focal length of 24
mm and F-ratio of 4. A high-resolution PIXIS (1024 × 1024 pixels) charge-coupled device
(CCD) cooled to −80°C captures the image of the sky. To enhance the signal-to-noise ratio,
we binned the images for 2 × 2 pixel on the chip, making an effective 512 × 512 super
pixel image. We restrict the image full field of view to 100° to avoid non-linearity arising
due to curvature effects (e.g., Sivakandan et al. 2015). More details of the experimental
setup have been described in Ghodpage et al. (2014). In general, airglow imagers are
operated during the new moon period when sky is not over casted. Therefore, the clouds
detected in imager are mid- or high-altitude non-cumulus clouds. The present work
consists of i) detection of cloud, ii) tracking of the cloud for CMV computation and iii)
856 R. N. GHODPAGE ET AL.

estimating the cloud cover percentage. After these estimations, we analyse how the cloud
cover and CMV vary from year-to-year over Kolhapur.

3. Results
3.1. Detection of cloud
In general, airglow imagers detect the processes occurring at mesospheric and thermo­
spheric altitudes during the clear sky conditions. The cloud appearance in the sky diffuses
the optical signals coming from higher above. Therefore, whenever dark clouds are noted,
the ASI is not operated. However, when mid-level clouds appear in the sky, the brightness
of the stars reduces drastically, but imagers are operated due to the transience nature of
these passing clouds. Owing to these criteria, mid- and high-altitude non-cumulus clouds
are detected in the night airglow images. As the cloud-free sky conditions over India occur
during January–May (e.g., Taori1 et al. 2012), with monsoon transition during March-April
months, the data pertaining to the present study corresponds to pre-monsoon months,
viz., March, April and May.
We utilize the edge detection method for the cloud detection, which is an image processing
technique for finding the boundaries of objects within the images (e.g., Canny 1986). It detects
the discontinuities in the brightness. The Canny edge detection algorithm is composed of five
steps, viz., Noise reduction, Gradient calculation, Non-maximum suppression, double thresh­
old and edge tracking by hysteresis. This is shown in Figure 1. In this method, the target is
identified as the cloud that has the highest gradient value and number of pixels (Borde and
García-Pereda 2014). At the same time, the target should be isolated so that it persist for some
time (at least till the next image). Further, priority is given to the cloud (if more than one cloud
satisfies the above criteria), which is at the centre of the image. The centre of the image in
geometric coordinates roughly corresponds to the location of Kolhapur. Based on earlier
studies, we consider the height of the clouds noted in the airglow imager is ~ 9 ± 2 km and
computes a pixel width is 0.0858 km ±0.0190 km per pixel (Satheesh Kumar, Narayana Rao, and
Taori 2015). This is used for the estimation of spatial extent per pixel.

3.2. Tracking of target cloud and CMV computation

The target cloud tracing consists of three factors. One is search window, second is
temporal resolution between images and third one is identification of same cloud in
successive images. These are crucial in tracking the target cloud and define the
accuracy of CMV estimation (Figure 2). To reduce the computational load and to
avoid noisy structures entering into the area of interest, we track the target cloud in
a smaller search domain in successive images (e.g., Bresky et al. 2012).
Conventionally, wind guess (WG) information supplements this exercise and sets
the coordinates of smaller windows in the latter image before matching (e.g.,
Bresky et al. 2012; Velden et al. 1997). The spatial search window is fixed based on
the maximum wind speed obtained from wind climatology around 9–12 km altitude
range and the time interval between the successive images. For instance, the cloud
would move 2.25 km in the presence of 15.04 m s−1 wind speed in 2.5 min. The time
interval between successive images has a significant impact on the quality of the
 REMOTE SENSING LETTERS 857

Figure 1. Image processing steps involved in identifying the target cloud. (a) The original image size
512 × 512 pixels, (b) the cropped image having 256 × 256 pixels, (c) the image enhanced using the
grey-level histogram method, (d) the contours of high intensity identified by using the Canny method
of edge detection, (e) the cleaned cloud image after the removal of stars and noisy structures and (f)
the target cloud image. X and Y axis shows number of pixels.

derived CMVs. Although satellites generally use 30 min intervals to derive the CMVs,
studies have shown that a temporal gap of 5 min for 1 km pixel size would produce
numerous valid motion vectors (Garcia-Pereda and Borde 2014). Therefore, in the
present study, time interval between the successive images is 2.5 min, which is
nearly equal to the optimum time gap suggested by Garcia-Pereda and Borde
(2014). Identification of the same target cloud in the successive images is very
important. Since the airglow imager is a zenith looking system with 100° FoV,
there is a possibility that it identifies two different clouds in successive images (for
example, a low cloud can suddenly mask a high cloud). It is, therefore, required to
use a pattern recognition method to estimate the cloud motion. In the present
study, 2D cross correlation method is employed for this purpose. The images were
discarded if the correlation coefficient obtained from the cross correlation of two
successive images is < 0.5.
Figure 3 shows an example of identified target cloud in six successive images on
10–11 May 2016. The target cloud images are cross-correlated to obtain the cloud
lag or lead at the maximum correlation. The distance travelled by the cloud in
x and y directions are estimated from the number of pixels displaced from 0 and
the pixel width calculated from Equation 1. With the known time interval of
successive images recording (2.5 min) and the distance travelled by cloud in the
time; the velocity of cloud (CMV) are estimated. We calculated the CMV for this day
to be varied from 13.87 m s−1 to 17.85 m s−1. Some examples of CMV calculation
are shown in Table 1.
858 R. N. GHODPAGE ET AL.

Figure 2. Flow chart of image processing and cloud motion vector (CMV) detection.

3.3. Cloud cover computation

As elaborated in section 3.1, the detection of clouds in a given airglow image is possible.
This covers 256 × 256 pixels regions with Kolhapur in the centre of the image. The cloud
cover fraction over Kolhapur region is defined as the ratio of the number of the pixels
where cloud has been detected to the total number of pixels in the image. For
a meaningful comparison, the imager data are read into an array variable using Matlab
to obtain the pixel-wise grey scale values ranging from 0 to 1. A histogram is made for
obtaining the distribution of pixels into 10 bins having a width of 0.1. The peak
 REMOTE SENSING LETTERS 859

Figure 3. Sequential images of the detected target cloud on 11 May 2016.

Table 1. Depicted some examples of CMV calculation.

CMV Cloud Cloud
Cloud travel distance (m angle
10-May 2016 (cloud observation images) x_y_1 x_y_2 (km) s−1) (°) direction
KOP_20160510_201435_5577.png 47.21 −108.61 0.00 0.00 184 SW
KOP_20160510_201706_5577.png 41.14 −86.63 1.95 13.04 185 SW
KOP_20160510_201936_5577.png 30.92 −61.61 2.32 15.46 179 SE
KOP_20160510_202206_5577.png 22.94 −35.07 2.37 15.84 173 SE
KOP_20160510_202437_5577.png 19.15 −9.67 2.20 14.69 153 SE
KOP_20160510_202707_5577.png 12.76 10.81 1.84 12.28 65 NE
KOP_20160510_202937_5577.png 7.39 33.93 2.03 13.56 28 NE
KOP_20160510_203208_5577.png 3.98 53.27 1.68 11.23 21 NE
KOP_20160510_203438_5577.png 0.95 71.09 1.55 10.33 16 NE

distribution is evaluated to obtain a rough estimate of the threshold value for applying
a cut-off to screen the pixel either into cloud/non-cloud pixel. Using the approximate
threshold, the image is converted to a grey scale and read into the image. Each image is
checked visually to see if the applied threshold is matching the distribution of the clouds
and is free of any contamination due to stray light. Corrections to the threshold are made
to remove any error due to either high contrast/light contamination. Thus, using the final
threshold, obtained images are converted to black and white image, and the percentage
of white pixels is calculated to obtain the final percentage of cloud in each image.
Figure 4 compares the cloud coverage noted in the ASI with the one noted in INSAT-3D
data on 30 March 2019 and 29 April 2019. On 30 March 2019, INSAT-3D shows isolated
860 R. N. GHODPAGE ET AL.

Figure 4. Sample comparison between INSAT 3-D data and ground based airglow imager data. Colour
bar shows the intensity variation in INSAT-3D and airglow image data.

thick occurrences of clouds in north direction, which are consistent with the scene taken
from the imager. On 29 April 2019, it is seen from the airglow images that there is
a consistent shroud of cloudiness throughout the field of view. The same is being
reflected in the INSAT image wherein the entire scene is flagged as cloud. The cloud
cover fraction noted by the airglow imager on these days was 0.41 and 0.62, while the
INSAT-3D data show these to be 0.35 and 0.43, respectively, indicating a reasonable
agreement.

4. Discussion
It is pertinent to state that ASI has very high spatial resolution, which we are comparing
with INSAT data of 10 km resolution. Figure 5 shows comparison of 5 days of cloud cover
percentage data estimated using ASI and INSAT. The ground-based data primarily repre­
sent the bottom-side features of a cloud, while the space-borne INSAT depicts the topside
of a cloud. The resolution of the two datasets differs, as the ASI data are of very high
resolution compared to the INSAT, later may average out finer features. As a result, when
 REMOTE SENSING LETTERS 861

Figure 5. Comparison of cloud coverage estimates deduced from the INSAT 3-D data and Ground
based Airglow imager data and R2 is depicted as coefficient of determination.

the sky is completely clear or cloudy, we see very strong agreement. In spite of the
limitations on the viewing conditions and resolution differences, a correlation coefficient
of ~ 0.7 suggests that during these months days, possibly multi-layer and short spatial
scale cloud structures were not dominant.
Using the data collected over the period of 2016 to 2020 during March, April and May
months of airglow monitoring, we calculate the cloud motion vector, cloud coverage percen­
tage and direction of cloud movement (Figure 6). The top panel in Figure 6 describes the
variation of the wind speed. It is noteworthy that cloud speed varies from 10 m s–1 to 18 m s−1
during the years 2016–2020. The lowest speed is noted in the year 2017 (10 ± 3 m s−1) while on
other years the speed is noted to be 15 ± 3 m s−1. The range of wind speeds are in agreement
with the earlier reports by Joseph (2012) and Satheesh Kumar et al. (2015). The propagation
angle of the clouds during the period under consideration is shown in the centre panel of
Figure 6. It is noteworthy that during the years 2016–2020, during the March, April, May
seasons, the cloud movement is found to be in the South-West direction. This is in agreement
with the understanding that during pre-monsoon, direction of winds is from the North-West
and West in the North-Western India and from the South-West in the Arabian Sea and
adjoining coasts (e.g., Joseph 2012). The Kolhapur location being closer to the Arabian Sea,
the winds shall be in the South-West direction, and hence, the cloud motion noted in our data.
It is important to note that the direction of the cloud propagation is turning the Southward as
the year progresses. This is more evident in the March and April month data. This may have
a linkage to the large climatic changes as discussed by Kathayat et al (2017). A slow shift in the
wind pattern change may bring a large shift in the monsoon rainfall pattern which may be of
862 R. N. GHODPAGE ET AL.

Figure 6. Variation of the cloud parameters during 2016–2020.

high social impact. The bottom panel of Figure 6 plots the cloud fraction. It is noteworthy that
cloud fraction has decreased during March, April and May months over the Kolhapur region. In
a recent work by Saud et al. (2016), based on 13 year of climatology exhibited the cloud cover
fraction over India. It was noted that over the most region of India, the cloud coverage has
 REMOTE SENSING LETTERS 863

increased except for the regions nearby the Arabian Sea. Using the CERES (Clouds and Earth’s
Radiant Energy System) data Kant et al. (2019) showed that the cloud fractions over the
Kolhapur and Pune region witnesses decreasing trend. In an earlier report, Marchand (2013)
reported trends in clouds of multiple types based on their optical depths over India. They
reported increasing trend for the clouds with optical depth of 3.6 to 23 and a decreasing trend
in the clouds having optical depth more than 23. This means that high-altitude clouds show an
increasing trend. However, in their plots, they note multiple oscillatory features of the time
periods varying from 2 to 5 years. Based on these reports, it is apparent that the decrease in the
cloud fraction may depend on the phase of quasi biennial oscillation or higher period
oscillations related to the southern oscillations as discussed by Maharana et al. (2021). We
compared our results with NCEP re-analysis and MODIS satellite data (observing times sig­
nificantly different) from the year 2010 to 2021 (not shown here). We noted that although the
resolution (NECP grids of 2°) and MODIS (1°) are coarser than the resolution of ASI observation,
they exhibit nearly identical patterns in both the cloud percentage (CF) and CMV for March–
May months indicating relevance of our results from climatological perspectives. Overall, our
data suggest that there is a change in the pre-monsoon cloud fraction and direction of its
propagation. Note that in the absence of cloud base height, our data may have errors as high
as 15%; however, the trends noted in the analysis are significant. Bearing this, if this remain
consistent, this may indicate a shift occurring in the monsoon pattern which may get
imprinted into the rainfall behaviour as well.

Acknowledgments
Indian Institute of Geomagnetism (IIG), Navi Mumbai, conducts the nightglow all sky imager
campaign observations at Shivaji University, Campus Kolhapur, India. The directors of the IIG in
Mumbai and the NRSC, ISRO in Hyderabad are thanked by the authors for their support of this
collaborative investigation. On the RG’s request, the provided data can be made available. The
production of Figure 4 was assisted by Madhav Haridas of the URSC, Bengaluru. The NCEP reanalysis
data is downloaded from https://climate.copernicus.eu/climate-reanalysis, https://psl.noaa.gov/
data/gridded/data.ncep.reanalysis.html and for resolution and its data https://neo.gsfc.nasa.gov/
view.php?datasetId=MYDAL2_D_CLD_FR&year=2019#.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The work was supported by the Department of Science and Technology, Govt. of India .

ORCID
R. P. Patil http://orcid.org/0000-0002-2829-0711

Data availability statement

The data presented in this paper will be made available upon request to the RG.
864 R. N. GHODPAGE ET AL.

References
Batista, P. P., H. Takahashi, D. Gobbi, and A. F. Medeiros. 2000. “First Airglow All Sky Images at 23° S.”
Advances in Space Research 26 (6): 925–928. https://doi.org/10.1016/S0273-1177(00)00031-4.
Borde, R., and J. García-Pereda. 2014. “Impact of Wind Guess on the Tracking of Atmospheric Motion
Vectors.” Journal of Atmospheric and Oceanic Technology 31 (2): 458–467. https://doi.org/10.1175/
JTECH-D-13-00105.1.
Bresky, W. C., J. M. Daniels, A. A. Bailey, and S. T. Wanzong. 2012. “New Methods Toward Minimizing
the Slow Speed Bias Associated with Atmospheric Motion Vectors.” Journal of Applied
Meteorology & Climatology 51 (12): 2137–2151. https://doi.org/10.1175/JAMC-D-11-0234.1.
Canny, J. 1986. “A Computational Approach to Edge Detection.” IEEE Transactions on Pattern Analysis
& Machine Intelligence 8 (6): 679–698. https://doi.org/10.1109/TPAMI.1986.4767851.
Cesana, G., and T. Storelvmo. 2017. “Improving Climate Projections by Understanding How Cloud
Phase Affects Radiation.” Journal of Geophysical Research Atmospheres 122 (8): 4594–4599. https://
doi.org/10.1002/2017JD026927.
Deb, S. K., C. M. Kishtawal, I. Kaur, P. K. Pal, and A. S. Kiran Kumar. 2015. “Assessment of a New Quality
Control Technique in the Retrieval of Atmospheric Motion Vectors.” Meteorological Applications
22 (2): 178–186. https://doi.org/10.1002/met.1433.
Garcia-Pereda, J., and R. Borde. 2014. “Impact of the Tracer Size and the Temporal Gap Between
Images in the Extraction of Atmospheric Motion Vectors.” Journal of Atmospheric and Oceanic
Technology 30:1171–1179. https://doi.org/10.1175/JTECH-D-13-00235.1.
Ghodpage, R. N., A. Taori, P. T. Patil, S. Gurubaran, S. Sripathi, S. Banola, and A. K. Sharma. 2014.
“Simultaneous Optical Measurements of Equatorial Plasma Bubble (EPB) from Kolhapur (16.8°N,
74.2°E) and Gadanki (13.5°N, 79.2°E).” Journal of Atmospheric and Solar-Terrestrial Physics
121:196–205. https://doi.org/10.1016/j.jastp.2014.05.008i.
Hubert, L., and L. F. Whitney. 1971. “Wind Estimation from Geostationary-Satellite Pictures.” Monthly
Weather Review 99 (9): 665–672. https://doi.org/10.1175/1520-0493(1971)099<0665:WEFGP>2.3.CO;2.
Izawa, T., and T. Fujita. 1969. “Relationship Between Observed Winds and Cloud Velocities
Determined from Pictures Obtained by the ESSA 3, ESSA 5 and ATS-I Satellites.” In Space
Research, 571–579. Vol. IX. Amsterdam: North-Holland Publ. Co.
Joseph, P. V. 2012. Synoptic Systems During Monsoon Season - Monsoon Monograph, edited by A. Tyagi,
G. C. Asnani, U. S. De, H. R. Hatwar, and A. B. Mazumdar, Vol. 2, 1–34. Pune, India: Meteorological
Department.
Kant, S., J. Panda, and R. Gautam. 2019. “A Seasonal Analysis of Aerosol-Cloud-Radiation Interaction
Over Indian Region During 2000–2017.” Atmospheric Environment 201:212–222. https://doi.org/
10.1016/j.atmosenv.2018.12.044.
Kathayat, G., H. Cheng, A. Sinha, L. Yi, X. Li, H. Zhang, H. Li, Y. Ning, and R. L. Edwards. 2017. “The
Indian Monsoon Variability and Civilization Changes in the Indian Subcontinent.” Science
Advances 3 (12): e1701296. https://doi.org/10.1126/sciadv.1701296.
Kaur, I., S. K. Deb, C. M. Kishtawal, P. K. Pal, and R. Kumar. 2014. “Atmospheric Motion Vector Retrieval
Using Improved Tracer Selection Algorithm.” Theoretical and Applied Climatology 119 (1–2):
299–312. https://doi.org/10.1007/s00704-014-1115-1.
Kazantzidis, A., P. Tzoumanikas, A. F. Bais, S. Fotopoulos, and G. Economou. 2012. “Cloud Detection
and Classification with the Use F Whole-Sky Ground-Based Images.” Atmospheric Research
113:80–88. https://doi.org/10.1016/j.atmosres.2012.05.005.
Kishtawal, C. M., S. K. Deb, P. K. Pal, and P. C. Joshi. 2009. “Estimation of Atmospheric Motion Vectors
from Kalpana-1 Imagers.” Journal of Applied Meteorology & Climatology 48 (11): 2410–2421.
https://doi.org/10.1175/2009JAMC2159.1.
Klebe, D. I., R. D. Blatherwick, and V. R. Morris. 2014. “Ground-Based All-Sky Mid-Infrared and Visible
Imagery for Purposes of Characterizing Cloud Properties.” Atmospheric Measurement Techniques
7 (2): 637–645. https://doi.org/10.5194/amt-7-637-2014.
Leese, J. A., C. S. Novak, and B. B. Clark. 1971. “An Automated Technique for Obtaining Cloud Motion
from Geosynchronous Satellite Data Using Cross Correlation.” Journal of Applied Meteorology
10 (1): 118–132. https://doi.org/10.1175/1520-0450(1971)010<0118:AATFOC>2.0.CO;2.
 REMOTE SENSING LETTERS 865

Liu, L., X. J. Sun, T. C. Gao, and S. J. Zhao. 2013. “Comparison of Cloud Properties from Ground-Based
Infrared Cloud Measurement and Visual Observations.” Journal of Atmospheric and Oceanic
Technology 30 (6): 1171–1179. https://doi.org/10.1175/JTECH-D-12-00157.1.
Long, J. M., J. Sabburg, J. Calbo, and D. Pages. 2006. “Retrieving Cloud Characteristics from
Ground-Based Daytime Color All-Sky Images.” Journal of Atmospheric and Oceanic Technology
23 (5): 633–652. https://doi.org/10.1175/JTECH1875.1.
Maharana, P., R. Agnihotri, and A. P. Dimri. 2021. “Changing Indian Monsoon Rainfall Patterns Under
the Recent Warming Period 2001–2018.” Climate Dynamics 57 (9–10): 2581–2593. https://doi.org/
10.1007/s00382-021-05823-8.
Marchand, R. 2013. “Trends in ISCCP, MISR, and MODIS Cloud-Top-Height and Optical-Depth
Histograms.” Journal of Geophysical Research: Atmospheres 118 (4): 1941–1949. https://doi.org/
10.1002/jgrd.50207.
Menzel, W. P. 2001. “Cloud Tracking with Satellite Imagery: From the Pioneering Work of Ted Fujita
to the Present.” Bulletin of the American Meteorological Society 82 (1): 33–47. https://doi.org/10.
1175/1520-0477(2001)082<0033:CTWSIF>2.3.CO;2.
Nikumbh, A., B. Padmakumari, and S. Sunil. 2019. “Cloud Fraction Retrieval and Its Variability During
Daytime from Ground-Based Sky Imagery Over a Tropical Station in India.” Journal of Atmospheric
and Solar-Terrestrial Physics 190:74–83. https://doi.org/10.1016/j.jastp.2019.05.002.
Patil, P. T., R. N. Ghodpage, A. Taori, R. P. Patil, S. Gurubaran, S. N. Nikte, D. P. Nade, et al. 2016. “The
Study of Equatorial Plasma Bubble During January to April 2012over Kolhapur (India).” Annals of
Geophysics 59 (2): a0214. https://doi.org/10.4401/ag-6868.
Salomonson, V. V., W. L. Barnes, P. W. Maymon, H. E. Montgomery, and H. Ostrow. 1989. “MODIS:
Advanced Facility Instrument for Studies of the Earth as a System.” IEEE Transactions on
Geoscience and Remote Sensing: A Publication of the IEEE Geoscience and Remote Sensing Society
27 (2): 145–153. https://doi.org/10.1109/36.20292.
Satheesh Kumar, S., T. Narayana Rao, and A. Taori. 2015. “A Novel Approach for the Extraction of
Cloud Motion Vectors Using Airglow Imager Measurements.” Atmospheric Measurement
Techniques 8 (9): 3893–3901. https://doi.org/10.5194/amt-8-3893-2015.
Saud, T., S. Dey, S. Das, and S. Dutta. 2016. “A Satellite-Based 13-Year Climatology of Net Cloud
Radiative Forcing Over the Indian Monsoon Region.” Atmospheric Research 182:76–86. https://doi.
org/10.1016/j.atmosres.2016.07.017.
Sivakandan, M., A. Taori, S. Satish Kumar, and A. Jayaraman. 2015. “Multi-Instrument Investigation of
a Mesospheric Gravity Wave Event Absorbed into Background.” Journal of Geophysical Research:
Space Physics 33 (4): 3150–3159. https://doi.org/10.1002/2014JA020896.
Smith, S., and R. Toumi. 2008. “Measuring Cloud Cover and Brightness Temperature with a
Ground-Based Thermal Infrared Camera.” Journal of Applied Meteorology & Climatology 47 (2):
683–693. https://doi.org/10.1175/2007JAMC1615.1.
Souza-Echer, M. P., E. B. Pereira, L. S. Bins, and M. A. R. Andrade. 2006. “A Simple Method for the
Assessment of the Cloud Cover State in High-Latitude Regions by a Ground-Based Digital
Camera.” Journal of Atmospheric and Oceanic Technology 23 (3): 437–447. https://doi.org/10.
1175/JTECH1833.1.
Taori, A., A. Jayaraman, and V. Kamalakar. 2013. “Imaging of Mesosphere–Thermosphere Airglow
Emissions Over Gadanki (13.5°N, 79.2°E) — First Results.” Journal of Atmospheric and Solar-
Terrestrial Physics 93:21–28. https://doi.org/10.1016/j.jastp.2012.11.007.
Taori, A., V. Kamalakar, L. M. Joshi1, S. Sripathi, and A. K. Patra. 2012. “On the Linkage Between
Mesospheric Gravity Waves and Occurrence of Equatorial Plasma Bubble Observed During the
Low Solar Activity.” Indian Journal of Radio & Space Physics 41 (April): 264–270.
Taylor, M. J., M. B. Bishop, and V. Taylor. 1995. “All-Sky Measurements of Short Period Waves Imaged
in the OI(557.7 Nm), Na(589.2 Nm) and Near Infrared OH and O 2 (0,1) Nightglow Emissions
During the ALOHA-93 Campaign.” Geophysical Research Letters 22 (20): 2833–2836. https://doi.
org/10.1029/95GL02946.
Taylor Patric, C. 2012. “The Role of Clouds: An Introduction and Rapporteur Report, Surv.” Surveys in
Geophysics 33 (3–4): 609–617. https://doi.org/10.1007/s10712-012-9182-2.
866 R. N. GHODPAGE ET AL.

Velden, C. S., C. M. Hayden, S. J. Nieman, W. P. Menzel, S. Wanzong, and J. S. Goerss. 1997. “Upper-
Tropospheric Winds Derived from Geostationary Satellite Water Vapor Observations.” Bulletin of
the American Meteorological Society 78 (2): 173–195. https://doi.org/10.1175/1520-0477(1997)
078<0173:UTWDFG>2.0.CO;2.
Yang, J., L. Weitao, M. Ying, and Y. Wen. 2012. “An Automated Cirrus Cloud Detection Method for a
Ground-Based Cloud Image.” Journal of Atmospheric and Oceanic Technology 29 (4): 527–537.
https://doi.org/10.1175/JTECH-D-11-00002.1.