Indian Journal of Marine Sciences

Vol. 43(7), July 2014, pp.

Ocean - atmosphere interaction during Thane cyclone: a numerical study

using WRF
K. K. Vinod1, M. Soumya2, P.Tkalich1 & P. Vethamony2*
1
Tropical Marine Science Institute, National University of Singapore
2
CSIR-National Institute of Oceanography, Dona Paula, Goa, India

*[Email: mony@nio.org]

Received 19 September 2013; revised 7 November 2013

Cyclone 'Thane' developed over the southeast Bay of Bengal (BoB) at 88.5° E, 8.5° N during 25 – 31 December
2011.Simulations have been carried out using Weather Research and Forecasting (WRF) model to generate fine resolution winds that
prevailed over the BoB during this extreme weather event. Global Final analysis data from the National Centres for Environmental
Prediction (NCEP) having a spatial resolution of 1° and a temporal resolution of 6 h have been used to provide the initial and
boundary conditions. The model results (spatial resolution of 12km and temporal resolution of 1 h) have been validated with track
data and vertical observations. Maximum sustained wind speed during Thane cyclone was 15m/s, and made landfall over Tamil
Nadu coast. This study investigates the factors which reinforce cyclone genesis, its propagation and atmospheric parameters that
prevailed while the cyclone was passing through the coastal region. We have also studied the variations in sea level along the west
coast of India during the passage of Cyclone Thane in the Bay of Bengal and east coast of India.

[Keywords: Cyclone Thane, WRF model, atmospheric parameters]

Introduction 31 December 2011, it weakened into a low pressure

Cyclone is a natural disaster, which generates area over north Kerala. Storm surge of about 1m
serious threats to the coastal environment by strong height inundated the low lying coastal areas of
winds and wind induced storm surges. Tropical Tamil Nadu at the time of landfall of cyclone15. In
oceans are hearts for cyclones (10ºS to 15ºN)1. order to estimate the surge heights and its impact on
Globally, about 80% of all the tropical cyclones coastal areas, it is important to determine the
form near to or within the Intertropical cyclone parameters. The objective of present study
Convergence zone (ITCZ)2. In this region, unstable is to look into the role of various ocean atmosphere
flow is generated by latent heat release in deep parameters and their spatial distribution. In the
cumulus convection. ITCZ generates cyclonic present study, we examine various cyclone genesis
potential vorticity anomaly, which reverses the parameters and impact of cyclone on sea surface
meridonal potential vorticity gradient on its elevation.
poleward side and it satisfies the necessary
condition for the combined barotropic and Materials and Methods
baroclinic instabilities3 in lower troposphere. In Numerical simulations were carried out to
northern Indian Ocean, the cyclone generation generate the atmospheric parameters over the BoB
occurs mainly between April and December4. using the Weather Research and Forecasting
Several cyclonic response studies were carried out (WRF) model16. It is based on fully compressible,
over northern Indian Ocean by2, 5-14. non-hydrostatic Euler equations, third order Runge-
A very severe cyclonic storm ‘Thane’ Kutta (RK3) integration scheme and Arakawa C
developed over the Bay of Bengal, crossed the grid. Vertical mixing and diffusion was done by
Tamil Nadu coast, close to south of Cuddalore Asymmetric Convective Model with non-local
between 0100 and 0200 UTC of 30 December, upward mixing and local downward mixing
2011 with a wind speed of 35m/s. After the land (ACM2) scheme17 and surface physics by Pleim-
fall, the system rapidly weakened into a severe Xiu scheme.
cyclonic storm over north Tamil Nadu and into a The global Final analysis data from the National
deep depression over north interior Tamil Nadu. On Centre for Environmental Prediction (NCEP)
 INDIAN J. MAR. SCI., VOL. 43, NO. 7, JULY 2014

having a spatial resolution of 1° and a temporal Results and Discussions

resolution of 6 h have been used to provide the Model validation
initial and boundary conditions to the model. Fig. 2 shows the simulated cyclone track of
Terrain data has been taken from the US Geological Thane cyclone with the observed track data
Survey (USGS) which has a resolution of 0.9km. available at every 6h from the JTWC and IMD. The
The model domain extends from 75.23º E to simulated track was consistent with both JTWC and
89.75º E and 3.18ºN to 15.60º N (Fig. 2) with a IMD tracks. The simulated wind speed is compared
spatial resolution of 12km and the simulations were with the JTWC best track data and IMD
carried out from 23 Dec 2011 to 1 Jan 2012. A total observation during the cyclone period (Fig. 3). For
of 27 sigma levels were applied in the vertical the vertical validation, station selected is Bangalore
levels from surface to 100 hPa. observatory (77.58ºE, 12.96ºN) at 00h on 27 Dec
The observational data obtained from the Joint 2011. The vertical wind speed and direction and air
Typhoon Warning Centre (JTWC) best track and temperature match well with the observations (Fig.
Indian Meteorological Department (IMD) during 4).
the cyclone was used for model validation. Upper-
air sounding observations for the Bangalore region
were obtained from the website
http://weather.uwyo.edu/upperair/sounding.html
and used for the validation of simulated
atmospheric parameters at different vertical levels
(Fig. 4). Fig.3. Comparison of JTWC observations, IMD observations
and WRF simulated results of wind speed during the Thane
cyclone.

Fig. 4. Comparison of simulated wind speed and direction and

air temperature with sounding data at 77.58º E and 12.96 ºN.

Fig. 1. Locations showing the AWS and Sea level observations

along the coast of India Cyclone genesis parameters
The primary energy source driving tropical
cyclones is the latent heat release due to
condensation of water vapour18 and transfer of
sensible heat from the warm ocean heats the air in
contact with it, generates buoyancy and upward
motion at the surface which play an important role
in the initial stages of formation of the cyclone19.
The latent heat flux (LHF) and sensible heat flux
(SHF) exchanges between Ocean-atmosphere
system plays an important role in the development
and evolution of cyclone20,21. The fluxes are
strongly sensitive to surface and planetary
Fig. 2. Comparison of Cyclone Thane track between boundary layer parameterization22, and are
observed (JTWC, black line), (IMD, red line) and computed maximum at areas having maximum wind speed.;
(WRF, green line). i.e. during cyclones, the values of latent and
sensible heat fluxes are the maximum at the eye
walls (Fig. 5a&b) where strong convective
 VINOD et al.: OCEAN - ATMOSPHERE INTERACTION DURING THANE CYCLONE

activities take place, and is decreasing inwards and VWS = U850 – U200
outwards radially. The impact of storm intensity For the present study the VWS is calculated
depends mainly on the diameter of the cyclone. from the simulated winds at 850mb and 200mb.
For the present study we considered atmospheric The zonal shear is highly correlated with magnitude
parameters such as Latent heat, sensible heat, of the total shear at the low latitudes. A low vertical
surface friction velocity, Sea level pressure, zonal wind shear of 40m/s was observed on 29 Dec
Vertical wind shear, Mid Tropospheric Relative 2011 06h of the cyclone, as shown in Fig. 6(a).
Humidity (500mb) and Relative vorticity on 29 Since we are calculating only shear in speed, lower
Dec. 2011 at 6am. ACM2 scheme with the grid layer wind speed is more than that of upper layer,
resolution of 12km is used. From the Fig. 5 (a) it is but not too much difference was noticed. Mid
clear that the latent heat is minimum at the cyclonic tropospheric Relative humidity (500mb) plays an
centre and increases upto ~1500 W/m^2; the same important role in the development of deep
time the sensible heat varies upto 240 W/m^2 (Fig. convection, and is high over the region of
5b) and is confined to the southeast of eye; i.e., the cyclogenisis. Fig. 6(b) shows 85-100% of MTRH
intensity of cyclone depends on the latent heat from during the cyclone on 29 Dec 2011 06h, and a
the ocean. The surface friction velocity associated minimum at the eye. As a whole, the largest value
with the turbulence mixing increases during of mid-level moisture was at the eye wall and its
extreme events20. Fig. 5c shows the simulation surroundings. The moisture gradient is along the
output of the surface friction velocity (U*). Surface eye wall and follows the cyclonic path.
friction velocity is very weak within the eye of the According to Gray2, the region of positive
typhoon, indicating that this area is calm and the relative vorticity shows the chance of tropical
turbulence is weak and its high values are confined cyclogenisis and it increases with larger vorticity.
largely to the eye wall. Normally, the friction Relative vorticity is a measure of rotation of the
velocity values vary from zero to 0.4 m/s and environmental flow which characterizes the
having maximum at complex topography. kinematics of the fluid, and is calculated as the curl
of the velocity vector. A positive value means
anticlockwise and negative implies clockwise
rotation; its value is directly associated with the
quantity of velocity shear. The presence of a
northward (southward) relative vorticity gradient in
the environmental flow increases (reduces) kinetic
energy to the symmetric vortex and gyres 23 and the
vortex moves faster in the case of a positive relative
vorticity. Relative vorticity is calculated using the
dynamical equation:
RV = ( dV/dx ) - (dU/dy )

where, U = U850 – U200; and V = V850 – V200

Fig. 5. Simulated (a) latent heat (b) sensible heat (c) surface
friction velocity and (d) sea level pressure during cyclone The relative vorticity calculated from the
Thane on 29 Dec 2011 at 06 h. simulated wind data, a low relative vorticity value
of 0.0002m/s (not shown) before the cyclone
The VWS formed due to lateral forcing of
formation. Since the wind is strong, higher relative
surrounding atmosphere, creates a controlled deep
vorticity is needed to veer the environmental flow
convection, this advects heat from condensation to
in anticlockwise. From Fig. 5(c) shows the RV at
the upper tropospheric levels and that increases the
27 December 2001 before the land fall. It is clear
ventilation, responsible for tilt in upper and lower
that higher relative vorticity (0.002s-1) is produced
potential vorticity and mid-level flow warming in
at the vortex and it controls the cyclone. Higher
vortex. The tilt is the fast adjustment between
value of relative vorticity has a greater tendency to
asymmetric and asymmetric diabatic heating.
curl more the air flow. The RV during the cyclone
Cyclones are more likely to form in regions with
from 23/12/2011 to 1/1/2012 was shown in Fig.
low vertical wind shear (VWS)2, which inhibit
6(d).
intensification. The VWS is calculated as the
difference in zonal component of the wind between
the pressure levels 850mb and 200mb according to
 INDIAN J. MAR. SCI., VOL. 43, NO. 7, JULY 2014

their changes during extreme event. Sea level along

the northern part of west coast of India shows a
negative surge of ~20cm confirming the existence
of remote forcing from BoB to AS during Thane.

Acknowledgements
Authors thank Directors of CSIR-NIO, Goa and
TMSI, NUS for providing the facilities. The NIO
contribution no is 5494.

References
1. Sadhuram Y, Murthy T R, and Somayajulu Y,
Estimation of tropical cyclone heat potential in the
Fig. 6. simulated (a) Vertical wind shear (b) Mid Tropospheric Bay of Bengal and its role in the genesis and
Relative Humidity (500mb) (c) Relative vorticity and (d) intensification of storms. Indian Journal of Marine
Averaged Relative vorticity. Sciences, 2006. 35(2):132-138pp.
2. Gray. W, Hurricanes: Their formation, structure and
likely role in the tropical circulation. Meteorology
Sea level variation during Thane cyclone over the tropical oceans., 1979:155-218.
For the present study we analyzed the available 3. Ferreira R N and Schubert W H, Barotropic Aspects
data of sea level residual and wind speed over three of ITCZ Breakdown. Journal of the Atmospheric
Sciences, 1997. 54(2):261.
locations on the west coast of Indian Peninsula 4. Obasi G, O.,P. . WMO’s programme on tropical
(Fig. 1). From the observation it is noted a strong cyclone. 1977. Mausam.
negative surge along the Observation locations 5. Vinod Kumar K, Aboobacker V M, Saheed P P, and
were noted, negative surge of 22cm over Ratnagiri, Vethamony P, Coastal circulation along the central
20cm over Verem (Goa) and 18cm over Karwar, west coast of India during cyclone Phyan:
measurements and numerical simulations. Natural
but from the wind data we can confirm that the Hazards, 2012. 64(1):259-271.
blowing over the locations were not strong enough 6. Madsen H and Jakobsen F, Cyclone induced storm
to generate both the positive and negative surges surge and flood forecasting in the northern Bay of
(Fig. 7). From this we can confirm the existence of Bengal. Coastal Engineering, 2004. 51(4):277-296.
7. Sinha P C, Hydrodynamic numerical modelling of
remote force from BoB to Arabian Sea. storm surges — Application to the Bay of Bengal.
Sadhana, 1987. 10(3-4):433-444.
8. Kumar V S, Kumar K A, and Pednekar P,
Observation on oceanographic parameters of
nearshore waters off Yanam, during the cyclonic
storm.. CURRENT SCIENCE -BANGALORE., 2008.
94(11):1503-1508.
9. Rajesh. G, Joseph. K J, Harikrishnan. M, and
Premkumar. K, Observations on extreme
meteorological and oceanographic parameters in
Indian seas. CURRENT SCIENCE., 2005. 88(8).
10. Byju P and Prasanna Kumar S, Physical and
biological response of the Arabian Sea to tropical
cyclone Phyan and its implications. Marine
Environmental Research, 2011. 71(5):325-330.
11. Chinthalu G R, Seetaramayya P, Ravichandran M,
and Mahajan P N, Response of the Bay of Bengal to
Gopalpur and Paradip super cyclones during 15-31
October 1999. CURRENT SCIENCE -
Fig. 7. Observed wind speed and sea level residual along the BANGALORE., 2001. 81:283-291.
west coast of India 12. Sundar D, Shankar D, and Shetye S R, Sea level
during storm surges as seen in tide-gauge records
along the east coast of India. CURRENT SCIENCE.,
Conclusions 1999. 77(10):1325-1332.
Atmospheric circulation and surge during Thane 13. Dube S K, Rao A D, Sinha P C, Murty T S, and
cyclone have been analyzed using measurements Bahulayan N, Storm Surges in Bay of Bengal and
Arabian Sea: The problem and its prediction.
and numerical simulations. The atmospheric
Mausam., 1997. 48:283-204.
characteristics are well simulated by the WRF 14. Gopalakrishna V V, Murty V S N, Sarma M S S, and
model, and the results of horizontal and vertical Sastry J S, Thermal response of upper layers of Bay
atmospheric structures have been analyzed to study of Bengal to forcing of a severe cyclonic storm: A
 VINOD et al.: OCEAN - ATMOSPHERE INTERACTION DURING THANE CYCLONE

case study. Indian Journal of Marine Sciences., 19. Seetaramayya P and Master A H, Exchange of heat
1993. 22(1):8-11. and momentum between ocean and atmosphere in
15. IMD., Very Severe Cyclonic Storm “THANE” over relation to monsoon depression during MONEX-79.
the Bay of Bengal Meteorology and Atmospheric Physics, 1986. 35(1-
(25-31 December, 2011) : A Report India Meteorological 2):59-63.
Department. 2011. 20. Guo. D, Yu-Shu. Z, and Jian-Tong. L, The
16. Skamarock. W C, Klemp. J B, Dudhia. J, Gill. D O, Experiments of the Boundary Layer Schemes on
Barker. D M, Duda. M G, Huang. X-Y, Wang. W, Simulated Typhoon Part I. The Effect on the
and Powers. J G, A Description of the Advanced Structure of Typhoon. CHINESE JOURNAL OF
Research WRF Version 3 Note 475+STR. National ATMOSPHERIC SCIENCES., 2005(3):417-428.
Center for Atmospheric Research: Boulder, USA. 21. Byers. H R, General meteorology. New York
2008. (McGraw-Hill), 3rd Edition, 1959. Pp. x. 540; 223
17. Pleim J E, A Combined Local and Nonlocal Closure Figs.; 10 Tables. 74s. Quarterly Journal of the Royal
Model for the Atmospheric Boundary Layer. Part II: Meteorological Society, 1960. 86(367):123-123.
Application and Evaluation in a Mesoscale 22. Braun S A and Tao W-K, Sensitivity of High-
Meteorological Model. Journal of Applied Resolution Simulations of Hurricane Bob (1991) to
Meteorology and Climatology, 2007. 46(9):1396- Planetary Boundary Layer Parameterizations.
1409. Monthly Weather Review, 2000. 128(12):3941-3961.
18. Emanuel K A, An Air-Sea Interaction Theory for 23. Wang B and Li X, Propagation of a Tropical Cyclone
Tropical Cyclones. Part I: Steady-State Maintenance. in a Meridionally Varying Zonal Flow: An
Journal of the Atmospheric Sciences, 1986. Energetics Analysis. Journal of the Atmospheric
43(6):585-605. Sciences, 1995. 52(9):1421-1433.