Convectively-coupled High-frequency Atmospheric waves triggered Kerala floods in 2018 and 2019
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Convectively-coupled High-frequency Atmospheric
waves triggered Kerala floods in 2018 and 2019
Kiran S. R.
Department of Civil Engineering
Central Polytechnic College
Thiruvananthapuram, Kerala
Email: kiransreekumarr@gmail.com
ORCID Id: 0000-0003-3200-8624
arXiv:2110.01944v1 [physics.ao-ph] 5 Oct 2021
Abstract—Floods have repeatedly battered the South Indian South China Sea and West Pacific [4]. Similarly, Ref. [5]
state, Kerala, as a result of the unprecedented heavy rainfall deduced that the Kerala floods of 2018 was triggered by the
during Boreal Summers, in recent years. The state witnessed formation of an intense and long-duration atmospheric river
large departures from normal rainfall in 2018 and 2019. Previous
studies have seldom adopted a systematic approach to understand with more than 60% of moisture supply from the Central-
the phenomenon responsible for the recurrent extreme events. Eastern Indian Ocean.
Hence, this study, based on spectral methods, identifies a char- Although, the above studies attempted to investigate the
acteristic propagation of high-frequency equatorial waves in the extreme precipitation in 2018 from the general wind field con-
atmosphere, which travelled from near tropical west Pacific to the
east coast of Africa. These waves stimulated intense convection ditions existed then, no conclusive evidences to support their
and ensured sufficient availability of moisture over the state, and theories exist hitherto, even after the extreme events repeated
are hence responsible for Kerala Floods. in 2019 and 2020 during the same season. This incited the
Index Terms—Kerala Floods, Indian Monsoon, Mid- necessity to investigate the climate dynamics responsible for
tropospheric convection, Moisture convergence, High-frequency the recurrent disaster.
Equatorial waves
I. I NTRODUCTION
II. DATA & M ETHODOLOGY
Kerala, the south-west coastal state of India, experienced
heavy rainfall and consequent floods in 2018 and 2019 during
Summer. The state received extremely heavy rainfall during Tropical Rainfall Measuring Mission (TRMM) gives daily
the South-West Monsoon in the two years – from 8th to rainfall data at a spatial resolution of 0.25o × 0.25o over the
17th August 2018 and from 3rd to 10th August 2019, which Indian Ocean region. In addition, the daily gridded rainfall
caused unprecedented floods resulting in massive loss of life data from raingauge observations over Kerala is obtained
and property. These recurrent extreme events in the state have from Indian Meteorological Department (IMD) from 2011 to
raised questions of changing climate in the southern part of 2019. Further, the ERA5 Reanalysis of the European Centre
Peninsular India. Hence, there arose a necessity to identify for Medium-Range Weather Forecasts (ECMWF) gives the
the physical phenomenon which derailed the hydrology of the specific humidity (kg/kg) and winds (zonal & meridional
State in recent years. winds in m/s and vertical velocity (or omega) in Pa/s) for
Vimal and Harsh (2019) observed no increase in mean and different vertical pressure levels from 1000 to 100hPa.
extreme precipitation in Kerala over the past six decades and In this study, the daily departures from the mean of
identified that the extreme rainfall event in August 2018 was atmospheric variables during 2018 and 2019 by deducting
driven by anomalous atmospheric conditions due to climate the average of daily data of the atmospheric variables from
variability rather than anthropogenic factors [1]. Hunt and 2011 to 2017 from the daily data for the years 2018 and
Menon (2020) concluded that the precipitation in Kerala may 2019 separately. Since, the two-day maximum precipitation
be 18% heavier for the pre-industrial period and 36% heavier of 2018 was abnormally high (220mm) and surpassed even
for a prospective climate in 2100 [2]. According to Ref. the 95th percentile of 70 years (150mm), a 7-year mean would
[3], the presence of a strong westerly jet along with the suffice to capture the anomaly [1]. Hovmöller diagrams, which
formation of the offshore vortex trough and the transport depict the temporal variation of atmospheric variables along
of mid-tropospheric moisture from various sources under the zonal/meridional direction, are used to identify the propagation
presence of conducive vertical shear of horizontal wind may of waves and its characteristics. Fast Fourier Transform (FFT)
have contributed to the extreme rainfall over Kerala in 2018. is performed on time series of atmospheric variables or its
Further, there are theories that the strong westerly flow, which anomalies for the ease of representation in frequency domain.
is a part of the cross equatorial Monsoon flow, was constantly The dominant harmonics, thus identified, are filtered out using
propelled by a couple of cyclonic systems over Bay of Bengal, Band-Pass Filters (BPF) for further analysis.Fig. 1. (first & third figure) TRMM rainfall on the days of peak rainfall in 2018 and 2019 respectively. (second & fourth figure) Corresponding values of
ERA5 Vertical velocity at 600hPa and 350hPa respectively.
Rainfall anomalies of 2018 = Rainfall of 2018 minus 7year mean daily rainfall Amplitude Spectrum of Precipitation over Kerala
200 5
North Kerala 33.2 days
Rainfall anomalies of 2018
Central Kerala
Amplitude Spectrum |P(f)|
4 7year Mean precipitation
Precipitation anomalies (mm)
150 South Kerala
16.6 days 7.8 days
5.5 days
3
100
2
50
1
0
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
-50 frequency (Hz)
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
Time
Fig. 4. Amplitude spectrum of IMD rainfall anomalies of 2018 averaged over
Fig. 2. IMD daily rainfall anomalies (in millimetres) of 2018 over North, the state of Kerala, in comparison to that of the 7 year mean precipitation.
Central and South Kerala, measured with respect to the 7 year mean
precipitation. Amplitude Spectrum of Precipitation Anomalies of 2019
6
26.1 days Rainfall anomalies of 2019
15.9 days
Amplitude Spectrum |P(f)|
Rainfall anomalies of 2019 = Rainfall of 2019 minus 7year mean daily rainfall 5 7year Mean precipitation
North Kerala 9.4 days
200 4
Central Kerala 5.1 days
Precipitation anomalies (mm)
South Kerala 6.4 days
150 3
2
100
1
50
0
0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
frequency (Hz)
-50
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
Time
Fig. 5. Amplitude spectrum of IMD rainfall anomalies of 2019 averaged over
the state of Kerala, in comparison to that of the 7 year mean precipitation.
Fig. 3. IMD daily rainfall anomalies (in millimetres) of 2019 over North,
Central and South Kerala, measured with respect to 7 year mean precipitation.
the state may be determined from ERA5 omega distribution.
At a geographical location in Kerala (76.5o E, 10o N), unusually
III. R ESULTS & D ISCUSSIONS high negative values of omega (rising air) were observed
at 600hPa in 2018 and at 350hPa in 2019 (Fig. 1 (second
A. Rainfall over Kerala in 2018 & 2019
& fourth figure), and are hence juxtaposed with rainfall for
Kerala experienced anomalous rainfall events in 2018 and comparison. It is interesting to conclude very high spatial
2019, with peak daily precipitation on 15th August and 8th correlation of precipitation with mid-tropospheric convection
August respectively [3,6]. Figure 1 (first & third figure) shows in 2018 and with upper-tropospheric convection in 2019, on
the spatial distribution of TRMM rainfall over Peninsular India the days of extreme events.
on these days. One observes heavy precipitation confined From IMD gridded rainfall point data, precipitation anoma-
only to the windward side of Western Ghats, especially its lies were determined for the North, Central and South Kerala,
southern region. Any propensities of localized convection over for the years 2018 and 2019 (Fig. 2 and Fig. 3). In 2018,
2Bandpass-filtered rainfall anomalies of 2018
0.15
5-8days
12-20days
0.1 5-8days bandpass filtered ERA5 Meridional Velocity in 2018 ×10 -3
Precipitation anomalies (mm)
20-60days
Pressure Levels (hPa)
200 3
0.05 400
0 600
2
800
-0.05
1000
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
-0.1 1
12-20days bandpass filtered ERA5 Meridional Velocity in 2018
Pressure Levels (hPa)
200
-0.15
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep 400
Time
600 0
Fig. 6. 5-8days, 12-20days and 20-60days bandpass-filtered IMD rainfall 800
anomalies of 2018 averaged over the state of Kerala. 1000
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep -1
20-60days bandpass filtered ERA5 Meridional Velocity in 2018
Pressure Levels (hPa)
Bandpass-filtered rainfall anomalies of 2019
200
0.15
5-8days
12-20days
400
-2
0.1
Precipitation anomalies (mm)
20-60days
600
0.05 800
1000 -3
0
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
Time
-0.05
-0.1
Fig. 8. 5-8days, 12-20days and 20-60days bandpass-filtered ERA5 Meridional
-0.15
velocities of 2018 averaged over the state of Kerala with altitude.
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
Time
Fig. 7. 5-8days, 12-20days and 20-60days bandpass-filtered IMD rainfall
anomalies of 2019 averaged over the state of Kerala. 5-8days bandpass filtered ERA5 Meridional Velocity in 2019 ×10 -3
200 3
Pressure Levels (hPa)
400
Kerala witnessed two distinct extreme events – on 8th August
600
(over 50mm) and on 15th August (over 150mm). Similarly, 2
two extreme events were observed in 2019 too – on 8th August 800
(over 150mm) and on 15th August (over 50mm). 1000
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
B. Dominant modes of variability 1
12-20days bandpass filtered ERA5 Meridional Velocity in 2019
Identification of the constituent harmonics, which con- 200
Pressure Levels (hPa)
tributed to the precipitation anomalies of 2018 and 2019, 400
shall expose the real phenomenon responsible for the extreme 0
600
event. Hence, Fast Fourier Transforms (FFT) of the spatially-
averaged rainfall anomalies were performed for both the years, 800
and the resultant amplitude spectrum are shown in Fig. 4 and 1000
Fig. 5. The spectrum corresponding to the 7 year mean daily 1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep -1
rainfall is superimposed for comparison. In 2018 and 2019, 200
20-60days bandpass filtered ERA5 Meridional Velocity in 2019
Pressure Levels (hPa)
the rainfall spectra are dominated by frequencies in the bands
of period 5-8 days, 12-20 days and 20-60 days. The semi- 400
-2
annual variability (180 days), a characteristic response of the 600
equatorial Indian Ocean region during Spring and Fall [7], can
800
be neglected for this study.
For further analysis, the time series of precipitation anoma- 1000 -3
1-Jun 15-Jun 1-Jul 15-Jul 1-Aug 15-Aug 1-Sep
lies for the concerned years were filtered separately for the Time
bands 5-8 days, 12-20 days and 20-60 days. Fig. 6 shows the
bandpass-filtered rainfall anomalies for the year 2018. The 5-8
days high frequency mode made an exceptional contribution to Fig. 9. 5-8days, 12-20days and 20-60days bandpass-filtered ERA5 Meridional
the extreme event in August, compared to the other two modes. velocities of 2019 averaged over the state of Kerala with altitude.
The high frequency mode rightly captured the two prominent
35-8days bandpassed V-wind 12-20days bandpassed V-wind 20-60days bandpassed V-wind
200hPa 700hPa 200hPa 200hPa ×10-3
1-Sep 1-Sep 1-Sep 1-Sep 3
2
21-Aug 21-Aug 21-Aug 21-Aug 1
Time
0
15-Aug 15-Aug 15-Aug 15-Aug
-1
7-Aug 7-Aug 7-Aug 7-Aug
-2
Cp = 12m/s
1-Aug 1-Aug 1-Aug 1-Aug -3
80 100 120 80 100 120 80 100 120 80 100 120
Longitude (degrees) Longitude (degrees) Longitude (degrees) Longitude (degrees)
Fig. 10. (first & second figure) 5-8 days bandpass-filtered ERA5 Meridional velocities of August 2018 meridionally averaged over the state of Kerala at
pressure levels 200hPa and 700hPa respectively. (third figure) 12-20 days bandpass-filtered Meridional velocities at 200hPa in 2018. (fourth figure) 20-60 days
bandpass-filtered Meridional velocities at 200hPa in 2018.
peaks in rainfall departures during the month. In 2019, the 13]. As a consequence, it can be inferred that the Quasi-
extreme precipitation which peaked on 8th August was the biweekly Oscillations (QBO) of 12-20 days period and the
unambiguous consequence of superposition of all the three Madden-Julian Oscillations (MJO) of 20-60 days period might
modes in-phase (Fig. 7). Yet, the 5-8 days periodicity leads have exacerbated the extreme event in August 2019 [12], but
to capture the second extreme event of the year. Hence, it is cannot drive the anomalous event by itself. Therefore, the
important to investigate the phenomenon which is responsible strong connection between the high-frequency mode with the
for these dominant modes of variability, especially the high convective tendencies at mid-troposphere in 2018 and at upper-
frequency mode, which has assumed great significance to the troposphere in 2019 stands established.
extreme rainfall events of 2018 and 2019.
ERA5 Reanalysis meridional winds over the state of Kerala, C. Westward-propagating High-frequency Waves
for different pressure levels in 2018 and 2019, are bandpass- In order to study the temporal evolution of the filtered
filtered for the dominant modes of variabilty (Fig. 8 and meridional winds at any pressure levels, Hovemöller diagrams
Fig. 9). In 2018, the 5-8 days mode is associated with are prepared for the month of August by meridional averaging
the intensification of meridional winds at mid-troposphere over the latitudinal extend of the state of Kerala. As per the
(800hPa - 500hPa). Similar anomalous wind patterns are above discussion, the middle and higher levels of troposphere
seldom observed for the other modes at mid-troposphere; resonated to the 5-8 days mode during the extreme event
although these had modulated the upper troposphere, but not of 2018, while only the upper atmosphere resonated to 12-
exclusive to period of extreme rainfall. Hence, the role of 20 days and 20-60 days periodicity during the same event.
the high frequency mode in causing the Kerala floods of Therefore, Hovemöller diagram corresponding to 700hPa is
2018 stands substantiated, as it conforms to the anomalous prepared only for the most dominant high frequency mode,
precipitation and convection observed in Fig. 1. while it is prepared at 200 hPa for all the three modes of
On the other hand, during the extreme event in August 2019, variability (Fig. 10). The 5-8 days mode at 700hPa exposes
a stark variability in meridional winds at upper atmosphere a interesting dynamical phenomenon, vide Fig. 10(second
(400hPa - 200hPa) is observed for the 5-8 days periodicity figure). During the extreme event of 2018, signals of strong
(Fig. 9). Other modes possessed relatively lower magnitudes anomalous meridional winds propagated from the West equa-
at any levels during this period, although all the three modes torial Pacific (100o E - 120o E) to the east coast of Africa, with
contributed comparably to the precipitation anomalies as per the phase speed (Cp ) of about 12 m/s. These westward prop-
Fig. 7. But, these modes can be observed to constitute the agating signals possess wave characteristics which resemble
zonal winds over Kerala (Figure not shown), which are long the equatorially trapped atmospheric waves of synoptic time
deemed classical Intra-seasonal Oscillation (ISO) modes as- scales [8, 9, 10]. The variability associated with meridional
sociated with the zonal component of the Monsoon Jet [11, winds corresponding to 12-20 days and 20-60 days at 200hPa,
45-8days bandpassed V-wind 12-20days bandpassed V-wind 20-60days bandpassed V-wind
200hPa 200hPa 200hPa ×10-3
1-Sep 1-Sep 1-Sep 3
2
21-Aug 21-Aug 21-Aug 1
Time
0
15-Aug 15-Aug 15-Aug
-1
7-Aug 7-Aug 7-Aug
-2
Cp=12m/s
1-Aug 1-Aug 1-Aug -3
80 100 120 80 100 120 80 100 120
Longitude (degrees) Longitude (degrees) Longitude (degrees)
Fig. 11. (first figure) 5-8 days bandpass-filtered ERA5 Meridional velocities of August 2019 averaged meridionally over the state of Kerala at 200hPa pressure
level. (second figure) 12-20 days bandpass-filtered Meridional velocities at 200hPa in 2019. (third figure) 20-60 days bandpass-filtered Meridional velocities
at 200hPa in 2019.
in Fig. 10(third figure) and Fig. 10(fourth figure) respectively, tropospheric convection in 2018 was the response to a low
do not reveal any wave characteristics in the wind field. pressure region created by the circulations in the field induced
Similarly, Hovemöller diagrams are prepared from merid- by waves. As a consequence, the waves dilated the wind
ional winds at 200hPa pressure level in 2019 for the three field to stimulate convection and further conveyed additional
variability modes (Fig. 11). As observed in 2018, westward moisture to cause extreme precipitation on the windward side
propagating waves of 5-8 days period, resonated the upper of Kerala in consecutive years.
troposphere for the second consecutive year (Fig. 11(first
IV. S UMMARY & C ONCLUSIONS
figure)). The high-frequency waves originated near the east
Indian Ocean (100o E) crossed over Kerala and finally reached Kerala experienced a spate of flood events in 2018 and 2019
Africa, with the same phase speed as in 2018. As antici- during Summer Monsoon. This study resolves to understand
pated, the other modes had not assumed significance during the dynamics of anomalous rainfall events in the state and
the extreme event (Fig. 11(second figure) and Fig. 11(third gives better insights from the available observations. The
figure)). Therefore, the westward-propagating high frequency summary of analysis and the conclusions thus drawn are
waves were simultaneously excited in the atmosphere with the discussed here.
anomalous convection over Kerala. 1) Extreme precipitation events recurred in Kerala in 2018
The ERA5 specific humidity in the atmosphere is vertically and 2019, which peaked on 15th and 8th August re-
integrated from near surface (1000 hPa) to the mid-troposphere spectively. The convective activity at mid-troposphere
(600 hPa) and bandpass-filtered for the high-frequency mode and upper-troposphere were strongly correlated with the
for the year 2018, with the objective of identifying the source spatial distribution of precipitation.
of moisture (Fig. 12). The figure is overlaid with the winds 2) Fourier Transforms performed on the precipitation
(5-8 days bandpassed) at 700 hPa. Alternate cyclonic and anti- anomalies exposed the prominent modes of variability
cyclonic circulations were observed to travel westward at mid- – 5 to 8 days, 12 to 16 days and 20 to 60 days. Out of
troposphere from Equatorial Pacific to Arabian Sea, which these, the high frequency mode (5 to 8 days) is observed
is nothing but the manifestation of the westward-propagating to dominate over the other two in 2018 during the
high-frequency wave train. The waves appear to possess zonal extreme event, although all the three seemed to possess
wave length of about 6000 km as they propagate with a comparable magnitudes as they fell in-phase in 2019.
phase speed of 12m/s and are trapped to the equator. One 3) The high frequency mode of period 5-8 days domi-
observes the transport of moisture along the wave trajectory, nated mid-troposphere in 2018 as well as at the upper-
especially with the cyclonic vortices, which infers that the mid- troposphere in 2019, all particular to the period of ex-
5moisture as it propagated. The waves not only stimu-
14-Aug-2018 ×10 -5 lated convection along its trajectory, but also ensured
15N 3
sufficient moisture availability. Therefore, the convective
7N activities which intensified in the mid-troposphere and
0N upper-atmosphere in 2018 and 2019 respectively, were
7S the direct consequence of the equatorially-trapped high
2 frequency waves, which played a significant role in
15S driving the recurrent anomalous precipitation in the
80E 90E 100E 110E 120E
South Indian state.
15-Aug-2018
15N ACKNOWLEDGMENT
7N 1 The author thanks Indian Meteorological Department (IMD)
0N and Asia-Pacific Data Research Center (APDRC) services for
7S granting free online access to the observations and Reanalysis
data for all users.
15S
80E 90E 100E 110E 120E AUTHOR C ONTRIBUTIONS
0
16-Aug-2018 K.S.R. performed the analysis, drawn conclusions and pre-
15N
pared the manuscript.
7N
C OMPETING I NTERESTS
0N
7S
-1 The author declares that there exist no competing interest
with any person or agency in matters related to this research,
15S and has not received any funds or grants in any form for the
80E 90E 100E 110E 120E
same.
17-Aug-2018
15N R EFERENCES
-2
7N [1] Vimal Mishra and Harsh L. Shah, “Hydroclimatological perspective of
the Kerala flood of 2018,” J. Geo. Soc. India, Vol. 92, 2019, pp. 645–
0N 650.
7S [2] M. R. Hunt and Arathy Menon. “The 2018 kerala floods: a climate
change perspective,” Cli. Dyn., Vol. 54, 2020, pp. 2433–2446.
15S -3 [3] V. Yesubabu, C.V. Srinivas, G. Basha, D. Hari Prasad, S. Langodan, M.
80E 90E 100E 110E 120E V. Ratnam, and I. Hoteit, “A diagnostic study of extreme precipitation
over Kerala during August 2018,” Atm. Sci. Lett., 2019, pp. 1–26.
0.01m/s [4] Saji Mohandas, Timmy Francis, Vivek Singh, A. Jayakumar, John P.
George, A. Sandeep, Prince Xavier, E.N. Rajagopal, “NWP perspective
of the extreme precipitation and flood event in Kerala (India) during
August 2018,” Dyn. Atm Oce., Vol. 91, 2020, pp. 1–21.
Fig. 12. 5-8days bandpass-filtered Specific humidity vertically integrated from [5] Rosa Vellosa Lyngwa and Munir Ahmad Nayak, “Atmospheric river
surface (1000 hPa) to 600 hPa overlaid with 5-8days bandpass-filtered ERA5 linked to extreme rainfall events over Kerala in August 2018,” Atm.
winds at 700 hPa from 14th to 17th August 2018. Res., Vol. 253, 2021, pp. 1–13.
[6] P. Vijaykumar, S. Abhilash, A.V. Sreenath, U.N. Athira, K. Mohanaku-
mar, B.E. Mapes, B. Chakrapani, A.K. Sahai, T.N. Niyas, O.P. Sreejith,
“Kerala floods in consecutive years - Its association with mesoscale
treme events in the state. The 12-20 days mode (Quasi- cloudburst and structural changes in monsoon clouds over the west coast
of India,” Weather and Climate Extremes, Vol. 33, 100339, 2021.
biweekly Oscillation) and 20-60 days mode (Madden- [7] K. Wyrtki, “An equatorial jet in the indian ocean. Science, vol. 181,
Julian Oscillation) have characterized the zonal winds in 1973, pp. 262–264.
2019, thereby justifying their contribution to the extreme [8] C. F. M. Raupp and P. L. Silva Dias, “Excitation mechanism of mixed
Rossby-Gravity waves in the equatorial atmosphere: Role of the non
event of the year, but not exclusively. Hence, the 5-8 linear interactions among equatorial waves,” J. Atm. Sci., Vol. 62, 2005.
days high frequency mode is solely responsible for the [9] Michael Dickinson and John Molinari, “Mixed Rossby–Gravity waves
extreme precipitation in Kerala. and western pacific tropical cyclogenesis. part i: Synoptic evolution,” J.
of Atm. Sci., Vol. 59, 2002.
4) The high-frequency mode manifested as westward- [10] S. R. Kiran, “High-Frequency Mixed Rossby-Gravity Waves in the Mid-
propagating high frequency tropical atmospheric waves Troposphere Triggered Kerala Floods of 2018,” J. Extreme Events, Vol.
of characteristic phase speed of nearly 12m/s, which 7, 2021, pp. 1–17.
[11] Xin Yan, Song Yang, Teng Wang, Eric D. Maloney, Shaorou Dong, Wei
originated near east equatorial Indian Ocean or tropical Wei and Shan He, “Quasi-biweekly oscillation of the Asian monsoon
West Pacific and travelled to the east coast of Africa, rainfall in late summer and autumn: different types of structure and
and coincide with the same period of extreme rainfall propagation,” Cli. Dyn., vol. 53, 2019, pp. 6611–6628.
[12] Raghu Murtugudde, Richard Seager, and Prasad Thoppil, “Arabian sea
events over Kerala. Moreover, the waves appeared as response to monsoon variation,” Paleoceanography, Vol. 22, 2007.
cyclonic and anti-cyclonic circulations trapped to the [13] Sulochana Gadgil, “The monsoon system: Land–sea breeze or the
equator, which dilated the wind field and transported ITCZ?,” J. Earth Syst. Sci., Vol. 127, 2018, pp. 1–29.
6You can also read
