1 s2.0 S2225603221000059 Main

Available online at www.sciencedirect.
com
ScienceDirect
Tropical Cyclone Research and Review 10 (2021) 54e70
www.keaipublishing.com/tcrr
A HWRF-POM-TC coupled model forecast performance over North Indian

Ocean: VSCS TITLI & VSCS LUBAN
Akhil Srivastava a,*, V.S. Prasad b, Ananda Kumar Das a, Arun Sharma a
a
India Meteorological Department, Ministry of Earth Sciences, Government of India, New-Delhi, India
b
National Center for Medium Range Weather Forecasting, Ministry of Earth Sciences, Government of India, Uttar Pradesh, India
Available online 20 April 2021
Abstract
The HWRF-POM-TC coupled model is run operationally at India Meteorological Department (IMD). This study is first attempt to assess the
IMD's operational HWRF-POM-TC (Atmosphere-Ocean) coupled model forecast performance over North Indian Ocean (NIO). The two
cyclonic storms one each in Arabian Sea and Bay of Bengal were examined. Among them, VSCS LUBAN formed over Arabian Sea (AS) and
was followed by the formation of VSCS TITLI over Bay of Bengal (BoB). It constituted a rare case whereby two VSCS have formed in the north
Indian Ocean (NIO) simultaneously.
The HWRF-POM-TC modeling system, which was developed at National Centers for Environmental Prediction (NCEP) based on Non-
hydrostatic Mesoscale Model (NMM) dynamic core, was customized for NIO conditions. For the two storms, VSCS LUBAN & VSCS
TITLI, 28 and 15 consecutive 6-hourly HWRF model runs were performed. The HWRF-POM-TC coupled model showed great skill in fore-
casting of Track and Intensity for examined cyclones. The result shows that the model predicted the intensification and landfall of VSCS Luban
& Titli in agreement with the best track data as made available by Cyclone Warning Division (CWD), India Meteorological Department which is
also recognized as Regional Specialized Meteorological Center (RSMC) by WMO for NIO.
© 2021 The Shanghai Typhoon Institute of China Meteorological Administration. Publishing services by Elsevier B.V. on behalf of KeAi
Communication Co. Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Keywords: Ocean coupled hurricane WRF; POM-TC; HWRF-POM-TC
1. Introduction going on to mitigate the damages caused by tropical cyclones

by better prediction of track, intensity and landfall of the
Tropical cyclones (TC) are an example of a multi-hazard tropical cyclones. TC track prediction in NWP models has
event with impacts that can highly vary depending on land- been substantially improved over the past decades (Heming
fall location, wind speed, storm surge, and inland flooding et al., 2019), due to improved simulation of the large-scale
from precipitation (Pilkington and Mahmoud 2017). Based on environment through better and higher-resolution models and
the wind speed and pressure, tropical cyclones can cause un- better model initializations based on more advanced data
precedented damage to life and property. Efforts have been assimilation schemes (Tallapragada et al., 2008;
Gopalakrishnan et al., 2011 2012; Heming, 2016; Mohanty
et al., 2015; Rao et al., 2009; Srinivas et al., 2012). Owing
* Corresponding author. to origin of tropical cyclones over ocean and ocean acting as
E-mail address: [email protected] (A. Srivastava). the heat source for the tropical cyclones, there have also been
Peer review under responsibility of Shanghai Typhoon Institute of China major efforts directed towards the inclusion of ocean effects to
Meteorological Administration.
improve the tropical cyclone forecasts (Lin et al., 2008, 2012).
It is also well established fact that in most cases one of the
factors for evolution of tropical cyclone intensity is the heat
Production and hosting by Elsevier on behalf of KeAi exchange with the upper layer of the ocean under the core of
https://doi.org/10.1016/j.tcrr.2021.04.002
2225-6032/© 2021 The Shanghai Typhoon Institute of China Meteorological Administration. Publishing services by Elsevier B.V. on behalf of KeAi Communi-
cation Co. Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
A. Srivastava, V.S. Prasad, A.K. Das et al. Tropical Cyclone Research and Review 10 (2021) 54e70
the hurricane (Evans, 1993; Emanuel 1999). The SST cooling UTC of 9th October to 12 UTC of 10th October 2018 when
associated with tropical cyclone (Price 1981; Price et al., the maximum sustained wind speed increased from 40 knots to
1994; Neetu et al., 2012) reduces the heat flux available to 80 knots and the VSCS Luban went through RI phase during
the atmosphere (Schade, 2000) and plays important role in 0600 UTC of 9th October to 0600 UTC of 10th October 2018
limiting the intensification of tropical cyclone. The air-sea when the sustained maximum wind speed increased from 45
interaction is an important mechanism influencing the in- knots to 75 knots. The VSCS Titli recurved after landfall
tensity of tropical cyclones (Bender and Ginis, 2000). There- whereas the VSCS Luban changed its direction several times
fore, Sea surface temperature (SST) and its relative during its lifetime. Both the VSCS Luban and Titli moved at
importance in tropical cyclone maintenance and intensification slower speed (when compared to long period average of VSCS
has gained prominence in form of Ocean coupled atmospheric in AS and BoB respectively), however the speed of VSCS
models for better prediction of tropical cyclones. Luban was even slower between the two and its lifetime was
The HWRF modeling system evolved in NCEP and its approx. double of long period average of storms of same
evaluation displayed consistent improvement in tropical category formed over AS (210 h. Depression to Depression
cyclone predictions (Tallapragada et al.; 2014, 2016). There Stage of VSCS Luban against long period average of 107 h)
are various studies showing the skill of ocean coupled model (CWD, 2018a; CWD, 2018b). The VSCS Titli maintained the
for prediction of tropical cyclones (Kim et al., 2014; Yesubabu cyclonic storm intensity for considerable time (approx. 15
et al. 2020; Mehra et al., 2018). Few of the studies on tropical hrs.) after landfall even after not being in contact with sea
cyclones have been conducted over NIO. HWRF model per- surface (heat source) (CWD, 2018b). The details from best
formance in triple nested configuration significantly improved track are tabulated in Tables 1 and 2 for VSCS Luban & Titli.
tropical cyclone track & intensity prediction as compared to The relative position and intensity of two VSCS Luban &
double nested configuration was established over NIO (Das VSCS Titli is shown in Fig. 1(a) & (b) with the help of INSAT
et al., 2015). The comparative study between HWRF 3D satellite imagery based on 12 UTC of 9th October 2018 and
modeling system with WRF-ARW suggested better intensity 11 UTC of 10th October 2018.
prediction and better track forecast for longer lead times from
HWRF over NIO (Nadimpalli et al., 2020). Ability of HWRF 3. HWRF-POM-TC model description and methodology
modeling system to predict rapid intensification was discussed
in the study of tropical cyclone Phailin (Osuri et al., 2017), and In coupled models, airesea coupling can alter TC pre-
also importance of ocean coupling in tropical cyclone models dictions through different effects. A passing TC induces a cold
over Bay of Bengal has been established (Mohanty et al., wake in the upper ocean, due to the upwelling of cooler
2019). subsurface water and increased evaporation, forced by the
India Meteorological Department (IMD) has operational- strong surface wind. The vertical mixing mainly determines
ized Ocean coupled HWRF modelling system in triple nest the degree of ocean response to a tropical cyclone (Price 1981;
configuration with two ocean model components. The HWRF Ginis and Dikinov 1989). Therefore, for proper simulation of
model is coupled with the Princeton Ocean Model for Tropical the ocean interaction, the ocean model must have highly ac-
Cyclones (POM-TC) as well as Hybrid Coordinate Ocean curate representation of upper ocean mixed layer physics
Model (HYCOM). (Bender and Ginis, 2000). TCs can also cool the ocean surface
In this study, POM-TC coupled HWRF model analysis and through increased cloud cover with reduced downward
forecast is used to assess the track, intensity and other features shortwave radiation, similar to other atmospheric convective
of VSCS “Luban” and VSCS ‘Titli’. events such as equatorial waves and the Madden-Julian
Oscillation MJO (Feng et al., 2018). These negative feed-
2. Brief description of VSCS LUBAN and TITLI backs change the local thermodynamic structure of the TCs
through reduced upward surface heat fluxes, to discourage
VSCS Luban formed as Low-Pressure Area (LPA) on 5th rapid TC intensification (Mogensen et al., 2017.).
October over south-east Arabian Sea and finished its journey The atmosphere-ocean coupled HWRF modeling system is
on 15th October over Yemen and adjoining Saudi Arabia after developed and supported by the Environment Modeling Center
attaining maximum sustained surface wind speed (MSW) of (EMC) (Biswas et al., 2017). It is a primitive-equation, non-
135e145 kmph on 10th/11th October (CWD, 2018a). On the hydrostatic, coupled atmosphere-ocean model with an atmo-
other hand, VSCS Titli started its journey as low pressure area spheric component based on Non-hydrostatic Mesoscale Model
(LPA) on 7th October 2018 over southeast BoB and adjoining (NMM) dynamic core of the WRF model (WRF-NMM), with a
north Andaman Sea and completed its journey on 13th parent and two moving nest domains (Tallapragada et al.,
October over Gangetic West Bengal and adjoining Bangladesh 2016). The model's parent domain spans about 80 x 80 ,
& north BoB after attaining MSW of 140e150 kmph on 10th/ while the intermediate and inner nests, which move with the
11th October 2018 (CWD, 2018b). Both the cyclonic systems storm, cover areas of approximately 24 x 24 & 7 x7 ,
went through rapid intensification phase during their lifetime. respectively (Biswas et al., 2017). The stationary parent
Rapid Intensification (RI) is defined as the 24-h maximum domain has a grid spacing of 0.135 (about 18 km), the inter-
sustained surface wind speed rate equal to 30 knots. The mediate nest 0.045 (about 6 km), and the inner nest domain
VSCS Titli went through rapid intensification phase during 15 0.015 (about 2 km) (Biswas et al., 2017). The atmospheric
55
Table 1
Best Track parameters and intensity of VSCS Luban. Last column depicts the Rapid Intensification (RI) phase of the VSCS Luban. Yellow highlighted column
shows the period during which system maintained cyclonic storm intensity.
56
Table 2
Best Track parameters and intensity of VSCS Titli. Last column depicts the Rapid Intensification (RI) phase of the VSCS Titli. Yellow highlighted column shows
the time period during which system maintained cyclonic storm intensity.
component is run with 61 vertical levels with model top at Flux planetary boundary layer parameterization is used (Hong
10 mb (Biswas et al., 2017). The POM-TC Ocean model was and Pan, 1996). An upgraded GFDL model surface layer
developed at Princeton University and coupled to the WRF scheme (Kwon et al., 2010; Powell et al., 2003; Black et al.,
model at the University of Rhode Island (URI) (Yablonsky 2007), along with Noah land surface model (Chen and
et al., 2015a,b). The POM-TC ocean model is run in three Dudhia, 2001; Mitchell, 2005), is used to compute surface
dimensions with 1/12 (approximately 9-km) horizontal grid fluxes. Radiation physics are evaluated by the RRTMG scheme
spacing and 40 half-sigma vertical levels (Yablonsky et al., (Iacono et al., 2008; Thompson et al., 2014). In POM-TC,
2015a,b). The model physics includes a Scale Aware Simpli- Mellor-Yamada Level 2.5 (Mellor and Yamada 1982) turbu-
fied Arakawa-Schubert scheme for cumulus parameterization lence closure model is used for turbulence parameterization
(Han and Pan, 2011; Biswas et al., 2014; Han et al., 2017) and that uses a second moment turbulence closure sub model,
the Ferrier-Aligo cloud microphysics package (Rogers et al., which provides the vertical mixing coefficients and POM-TC
2001; Aligo et al., 2014) for explicit condensation. The uses Smagorinsky diffusivity (Smagorinsky 1963) for Hori-
Global Forecast System (GFS) Hybrid-Eddy Diffusivity Mass zontal diffusion (Biswas et al., 2017b).
57
Fig. 1. Relative position of VSCS Luban & VSCS Titli on (a) 12 UTC of 9th October 2018 and 11 UTC of 10th october 2018.
Table 3
List of Initial conditions for which coupled HWRF model was run for VSCS Luban & VSCS Titli. In the table AS¼> Arabian Sea; BoB¼> Bay of Bengal; DD¼>
Deep Depression; CS¼> Cyclonic Storm; SCS¼> Severe Cyclonic Storm; VSCS¼> very Severe Cyclonic Storm.
S. No. Initial Condition TC Name (basin) TC Name (basin)
Classification Classification
1. 7th October 2018 12 UTC LUBAN - AS - DD
2. 7th October 2018 12 UTC LUBAN - AS - DD
3. 8th October 2018 00 UTC LUBAN - AS - CS
6. 8th October 2018 18 UTC LUBAN - AS - CS TITLI e BOB - DD
7. 9th October 2018 00 UTC LUBAN - AS - CS TITLI e BOB - DD
8. 9th October 2018 06 UTC LUBAN - AS - CS TITLI e BOB - CS
9. 9th October 2018 12 UTC LUBAN - AS - CS TITLI e BOB - CS
10. 9th October 2018 18 UTC LUBAN - AS - SCS TITLI e BOB - CS
11. 10th October 2018 00 UTC LUBAN - AS - SCS TITLI e BOB - SCS
12. 10th October 2018 06 UTC LUBAN - AS - VSCS TITLI e BOB -VSCS
13. 10th October 2018 12 UTC LUBAN - AS - VSCS TITLI e BOB - VSCS
16. 11th October 2018 06 UTC LUBAN - AS - VSCS TITLI e BOB - SCS
17. 11th October 2018 12 UTC LUBAN - AS - VSCS TITLI e BOB - CS
18. 11th October 2018 18 UTC LUBAN - AS - VSCS TITLI e BOB - DD
19. 12th October 2018 00 UTC LUBAN - AS - VSCS TITLI e BOB - DD
20. 12th October 2018 06 UTC LUBAN - AS - SCS TITLI e BOB - DD
21. 12th October 2018 12 UTC LUBAN - AS - SCS
58
Fig. 2. Track forecast of real time POM-TC coupled HWRF run along with Best track from CWD-IMD for (a) VSCS Titli and (b) VSCS Luban.
Fig. 3. HWRF-POM-TC forecast track error with CWD long period average of forecast track.
Fig. 4. HWRF-POM-TC forecast and CWD forecast landfall point error.
59
Fig. 5. HWRF-POM-TC and CWD long term average of landfall time forecast error.
Fig. 6. HWRF forecasted Intensity and Best track Intensity (wind speed) for VSCS Titli (above panel) and VSCS Luban (Lower Panel).
60
The NCMRWF V14 GDAS analysis, the GDAS 6-h fore- VSCS Luban and VSCS Tilti. The maximum sustained wind
casts, IMD GFS forecasts and the storm message (TCVITAL) (MSW) at surface level in range of 17e27, 28e33, 34e47,
provided by cyclone warning Division (CWD-IMD) are used 48e63, 64e119 & 120 knots corresponds to Depression
to generate initial conditions for the POM-TC & HWRF (D), Deep Depression (DD), Cyclonic Storm (CS), Severe
coupled modeling system. The HWRF system contains a Cyclonic Storm (SCS), Very Severe Cyclonic Storm (VSCS)
forecast/analysis cycle in which a 6-h HWRF forecast from the and Super Cyclonic Storm (SuCS) respectively over North
previous cycle is used in a vortex relocation procedure, after Indian Ocean. In this study we attempt to assess the HWRF-
being adjusted for position, structure and intensity using the POM-TC coupled model performance for track, intensity
CWD-IMD storm message. After the vortex relocation, the and landfall forecast along with its utility to forecast different
initial conditions are further refined using the 3D hybrid synoptic and ocean SST characteristics.
variational data assimilation system (HDAS). The HDAS
utilizes a one-way hybrid procedure to assimilate conventional 4. Results and discussions
observation along with a few satellite radiances collected in
the local storm environment. The atmospheric component of The forecast of HWRF-POM-TC coupled model is evalu-
the coupled model uses NCMRWF GDAS analysis as back- ated against the track, intensity as well as location and time of
ground field for the first cycle and then the previous 6 h ocean landfall as provided by the best track data of CWD-IMD. The
coupled HWRF forecast are used as background fields for results are given in the following sub-sections.
successive GSI assimilation cycles. The IMD GFS forecasted
fields every 6 h are used to provide lateral boundary conditions 4.1. Track forecast
during each cycle. Prior to coupled model integration of the
HWRFePOM-TC, the GDEM3 (Generalized Digital Envi- The track of VSCS Titli and VSCS Luban as forecasted
ronmental Model - V3.0) climatology data together with GFS from operational HWRF-POM-TC coupled model along with
model SST output (Yablonsky and Ginis, 2008) are utilized to best track as provided by CWD-IMD is given in Fig. 2(a) and
create the ocean initial conditions which is referred as ocean (b).
spinup “phase 1”. Spinup “phase 1” uses 48 h of POM-TC The initial track as predicted by the model was showing
integration, keeping the SST constant for dynamic adjust- nearly westward movement in case of VSCS Titli. This was
ment of density field and for generation of dynamically largely due to spin-up time taken by the model. In the sub-
consistent currents (Yablonsky et al., 2015). The output of sequent real time cyclic 6 h run of coupled HWRF-POM-TC
“phase 1” is used as input for “phase 2 spinup” of POM-TC, model the track became aligned with the best track as pro-
during which cold wake at the ocean surface and currents vided by the best track data of CWD. In case of VSCS Luban,
before beginning of coupled model forecasts are generated by the track of the cyclone as predicted by the model was north of
72 h integration of POM-TC utilizing TCVITAL (Yablonsky the best track during initial hours and then was south of the
et al., 2015). The output of “phase 2” is used to initialize track during later forecast hours. Past five year (2014e18)
POM-TC in coupled HWRF model. average errors of CWD operational forecast were about 54, 86,
This HWRF-POM-TC ocean coupled model is run on six 102, 132, 156, 177, 219, 243, 238 and 284 km respectively. In
hourly bases based on 00, 06, 12 and 18 UTC initial conditions comparison to it the track forecast errors for 84 h lead period
to provide track and intensity forecasts along with surface at every 12 h were about 69, 90, 93, 120, 153, 220 and 405 km
wind and rain swaths valid up to 126 h. The model was inte- for VSCS Titli and for 120 h lead period at every 12 h were 42,
grated for different initial conditions as referred in Table 3 for 53, 76, 102, 125, 149, 169, 172, 181 and 215 km respectively
Fig. 7. HWRF-POM-TC forecast and long period average (2014e18) absolute intensity errors.
61
Fig. 8. HWRF-POM-TC forecast and long period average (2014e18) RMS intensity error for VSCS TITLI and VSCS LUBAN.
for VSCS Luban. These track errors are different and on the study by AAMC-WRF is limited to 3 runs based on initial
higher side as compared to the track errors as obtained by condition of 9th October 2018 for VSCS Titli and 13 runs
WRF model (AAMC-WRF) from Hong-kong observatory based on initial conditions during 8th October 2018 to 11th
(Kai-Kwong Hon, 2020). However, the data set used for the October 2018 for VSCS Luban. These initial conditions do not
Fig. 9. The SST Difference along the track with the passage of VSCS Titli over Bay of Bengal for different forecast hours (a) 06 h (b) 12 h (c)18 h (d) 24 h (e) 30 h
and (f) 42 h based on initial condition of 10th Oct 2018 00 UTC. Black circle corresponds to position of storm.
62
Fig. 10. The SST Difference along the track after the passage of VSCS Luban over Arabian Sea for different forecast hours (a) 06 h (b) 18 h (c) 30 h (d) 42 h (e)
54 h (f) 66 h (g) 78 h and (h) 90 h based on initial condition of 06 UTC of 10th Oct 2018. Black circle corresponds to position of storm.
63
correspond to entire lifecycle of VSCS Luban & VSCS Titli. 52.20 E) during 1100e1130 h IST (0530e0600 UTC) of 14th
The results from present study shows better mean track errors October 2018 (CWD, 2018a). The HWRF-POM-TC model
as compared to the mean track errors found from a study of forecast landfall point error during 48 h lead time in every 12 h
WRF model based on ARW core with resolution of 9 km, with respect to landfall point provided by CWD was 6, 15, 21
18 km and 27 km horizontal resolution (Osuri et al., 2013). and 77 km respectively for VSCS Titli. The HWRF-POM-TC
This also reconfirms the importance of finer resolution in model forecast landfall point error for 120 h lead time in every
cyclone track prediction (Osuri et al., 2013). 12 h with respect to landfall point provided by CWD was 25,
Fig. 3 shows the HWRF-POM-TC forecast track error with 15, 40, 102, 102, 70, 109, 44, 31 and 169 km respectively for
operational CWD forecast track long period average VSCS Luban. The Long term (2014e18) average landfall
(2014e18). For lead period up to 60hr the HWRF-POM-TC point error of CWD forecast during similar lead time is 26.5,
model provided slightly better track forecast as compared to 46.6, 44.1, 69.7, 88.9, 104.3 and 141.3 km respectively.
LPA forecast for VSCS Titli. However, the track forecast for The landfall point error is shown in Fig. 4. The results show
VSCS Luban consistently had lesser error across all lead that HWRF-POM-TC landfall point error was significantly
times. reduced as compared to CWD long period error average of
landfall point forecast in case of VSCS Titli for 36 h lead time.
4.2. Landfall point and landfall time forecast However, for VSCS Luban the landfall point error, except for
48 h and 60 h forecast, was less than CWD long period error
The VSCS Titli made landfall crossing north Andhra Pra- average.
desh and south Odisha coast near Palasa (18.80N/84.50 E) to The landfall time HWRF-POM-TC forecast error was 3.5,
the southwest of Gopalpur during 0430e0530 IST (around 00 3.5, 0.5 and 0.5 h for 48-h lead time in every 12 h respectively
UTC) of 11th October 2018 (CWD, 2018b). The VSCS Luban for VSCS Titli. The landfall time HWRF-POM-TC forecast
crossed Yemen and adjoining south Oman coasts (15.80N/ error for VSCS Luban was 2, 2, 2, 3, 3, 11, 9, 14, 4, and 0 h for
Fig. 11 (a) SST forecast (Top Panel left to right: - 06, 12 & 18 h forecast and bottom panel left to right: - 24, 30 & 36 h) for VSCS Titli based on initial condition of
00 UTC of 10th October 2018. (b) SST forecast for VSCS Luban based on initial condition of 10th October 2018 06 UTC for forecast hours (left to right:- 00, 06,
12, 18, 24, 30, 36, 42, 48, 54,60, 66,72,78 and 84 h).
64
Fig. 11. (continued)
120 h lead time in every 12 h. The long period average 4.3. Intensity forecast
(2014e18) of landfall time CWD forecast is 2, 2.9, 4.1, 5.1,
4.3, 5.8 and 5.8 h respectively for 84 h lead time in every 12 h. The temporal variation of intensity as forecasted from
The landfall time error is shown in Fig. 5. The HWRF-POM- HWRF-POM-TC model and intensity as given in best track by
TC model landfall time error in VSCS Titli forecast was bet- CWD for VSCS Titli & VSCS Luban are shown in Fig. 6(a)
ter during 36 h and 48 h forecast and for VSCS Luban time and (b) respectively. The time evolution of intensity (max
forecast error of HWRF-POM-TC were less than LPA for 60 h wind speed) as predicted from HWRF model closely follow
forecast. The lesser landfall point error and landfall time error the observed intensity as provided with best track data.
in case of LPA during first 24 h looks to be mainly due to better The HWRF-POM-TC intensity forecast for VSCS Titli
observational network (Radar Coverage) along the coast. closely followed the observed intensity (based on best
65
Fig. 11. (continued)
track). For VSCS Luban the HWRF-POM-TC model pre- Titli during all lead hours except for 12 h lead time
dicted the intensity close to the observed frequency during forecast. Whereas for the case of VSCS Luban, the ab-
the lifetime of VSCS Luban except for the duration between solute intensity errors were better during 36 h to 96 h
06 UTC of 10th October and 18 UTC of 11th October. forecast as compared to LPA.
During this period the HWRF-POM-TC model was not able The root means square error (RMSE) comparison of
to capture the high intensity of VSCS Luban. This was the HWRF forecast for VSCS Titli and VSCS Luban with
period after the Rapid Intensification of VSCS Luban. In this long period average (2014e18) as provided by CWD is
VSCS Luban case the HWRF-POM-TC model couldn't given in Fig. 8. The root means square error (RMSE) of
capture the rapid intensification phase and the model in- intensity forecast for VSCS Titli is also significantly less
tensity underestimated the tropical cyclone's intensity. This than long period average during all lead times except for
could also be due to data sparse region of central/west 12 h forecast. The RMSE of intensity in case of Luban
Arabian Sea and in future assimilation of satellite data might shows better results during 36 h to 96 h forecast. In ab-
improve the performance of model in this scenario over this solute intensity forecast error as well as intensity root
region. mean square error, the insignificant error in 84 h lead time
The error in absolute intensity (wind speed forecast) of could be due to very small members in the dataset during
VSCS Titli & VSCS Luban as predicted from HWRF- 84 h lead time for VSCS Titli.
POM-TC model along with long period average
(2014e18) of CWD forecasted absolute intensity error is 4.4. Sea surface temperature
shown in Fig. 7. The absolute intensity error as predicted
from HWRF-POM-TC shows improvement over long The HWRF-POM-TC model is run as an ocean coupled
period average forecast error of CWD in case of VSCS atmospheric model. The POM-TC ocean model provides the
66
Fig. 12. 700-500 hPa RH, 700 hPa geopotential height and 700 hPa winds based on initial condition of 9th October 2018 06 UTC for VSCS Titli for forecast hours
(a)00 h (b)12 h (c) 24 h (d) 36 h and (e) 48 h.
ocean forcing to the atmospheric model and vice versa. The For the case of VSCS Luban the effect of SST cooling
SST forecast of operational HWRF-POM-TC run is analyzed forecast is shown in Fig. 10 based on initial condition of 06
in this study. UTC of 10th October 2018. It is seen that the affect of SST
The images in Fig. 9 on SST shows that the time taken for cooling over the Arabian Sea along the track of VSCS Luban
maximum cooling with respect to VSCS Titli Cyclone passage was very prominent due to the slow motion of the VSCS
over Bay of Bengal was more i.e. the cooling lagged behind Luban.This in turn stopped the further intensification of VSCS
the passage of VSCS TITLI cyclone. Due to this, the effect of Luban near the coast and also it played role in weakening of
sea surface cooling was not prominent on VSCS Titli. This intensity of VSCS Luban near landfall. This was consistent
created favorable environment for VSCS Titli to maintain its with the observation of weakening of VSCS Luban before
intensity. landfall and it crossed coast as Cyclonic storm.
67
Fig. 13. 500 hPa relative vorticity, Geo potential height and winds based on initial condition of 11th October 2018 00 UTC for VSCS Titli for forecast hours (a) 00 h
(b) 12 h (c) 24 h and (d) 30 h.
The SST as forecasted from HWRF-POM-TC for VSCS The RI activity of the two systems during the same
Cyclones Titli and Luban are shown in Fig. 11(a) and (b) time window is analyzed based on the HWRF-POM-TC
respectively for above cases. The SST as predicted by the forecast. Fig. 12 shows the HWRF-POM-TC forecast for
model also shows that near the coast SST is cooler for VSCS VSCS Titli based on initial condition of 06 UTC of 09th
Luban as compared to that of VSCS Titli. October 2018. The plot shows forecast of 700-500 hPa
Relative Humidity (RH), 700 hPa Geopotential Height and
4.5. Some salient features of VSCS-Titli & VSCS-Luban 700 hPa winds.
The Fig. 12 shows the moisture convergence getting rein-
The VSCS Luban and VSCS Titli went through the period forced between the two VSCS with their interaction. The
of Rapid Intensification (RI) during their lifetime. The VSCS favorable interaction is mostly due to their position and east-
Titli went through the RI between 15 UTC of 9th October to west alignment. Both the systems were approximately in the
12 UTC of 10th October 2018 whereby its intensity increased East-West line between 100 N and 200 N due to their position. In
from 40 kts to 80 kts (CWD, 2018b). On the other hand, the next 24e48 h both the system moved and due to more north-
VSCS Luban went through the RI during 06 UTC of 9th ward movement of VSCS Titli it crossed the 200 N latitude and
October 2018 to 06 UTC of 10th October 2018, whereby its the favorable position for moisture convergence between two
intensity increased from 45 kts to 75 kts (CWD, 2018a). Both storms ended. This also stopped RI phase of the two storms.
the systems exhibited the RI feature over the similar time The VSCS Titli made landfall on 00UTC of 11th
window between 15 UTC of 9th October to 06th UTC of 10th October 2018 as very severe cyclonic storm and weakened
October. to deep depression only at 18 UTC of 11th October;
68
therefore maintaining cyclonic storm intensity for Atmospheric and Ocean Science (KBCAOS), University of
more than 15 h even after having no contact with the sea Allahabad for the support to complete this study.
surface which is a major source of energy for tropical
cyclones. References
The Fig. 13 shows the favorable interaction of trough at
500 hPa with that of VSCS Titli. This interaction of trough Aligo, E., Ferrier, B., Carley, J., Rodgers, E., Pyle, M., Weiss, S.J., Jirak, I.L.,
with VSCS Titli also helped VSCS Titli to maintain its 2014. “Modified Microphysics for Use in High Resolution NAM Fore-
cyclonic storm intensity for approximately 15 h after the casts”, 27 AMS Conference on Severe Local Storms.3-7 November,
Madison, WI.
landfall. However, no such interaction was seen for VSCS
Bender, M.A., Ginis, I., 2000. “Real-case simulations of hurricaneeocean
Luban which ultimately led to its weakening. interaction using a high-resolution coupled model: effects on hurricane
intensity”. Mon. Weather Rev. 128, 917e946.
Biswas, M.K., Bernardet, L., Dudhia, J., 2014. Sensitivity of hurricane fore-
5. Conclusion casts to cumulus parameterizations in the HWRF model. Geophys. Res.
Lett. 41, 9113e9119. https://doi.org/10.1002/2014GL062071.
In this study an attempt has been made to assess the per- Biswas, M.K., Carson, L., Newman, K., Bernardet, L., Kalina, E., Grell, E.,
formance of HWRF modelling system in coupled mode with Frimel, J., 2017a. Community HWRF Users' Guide V3.9a, p. 160.
Biswas, M.K., et al., 2017b. Hurricane Weather Research and Fore-Casting
POM-TC ocean model for the two tropical cyclones VSCS
(HWRF) Model: 2017 Scientific Documentation. NCAR/TN-544STR.
Titli and VSCS Luban over north Indian Ocean. Both of these Black, P.G., D'Asaro, E.A., Drennan, W.M., French, J.R., Sanford, T.B.,
VSCS cyclones formed simultaneously, one over Arabian Sea Terrill, E.J., Niiler, P.P., Walsh, E.J., Zhang, J., 2007. Air-sea exchange in
and other over Bay of Bengal. hurricanes: synthesis of observations from the coupled boundary layer air-
The results from the real time model output shows good sea transfer experiment. Bull. Amer. Meteor. Soc. 88, 357e374.
Chen, F., Dudhia, J., 2001. Coupling an advanced land surfaceehydrology
capability of POM-TC coupled HWRF model to predict the
model with the Penn State-NCAR MM5 modeling system. Part I: model
track and intensity of tropical cyclones over NIO. The mean description and implementation. Mon. Wea. Rev. 129, 569e585. https://
track errors, landfall point & time errors as well as intensity doi.org/10.1175/1520-0493(2001).129<0569:CAALSH>2.0.CO;2.
errors in both the cases were found to be in general better than CWD, 2018. IMDVery Severe Cyclonic Storm, ‘LUBAN’ over the Arabian
the long period average errors of operational forecasts. The Sea (06 e 15 October 2018): A Report, Cyclone Warning Division. India
Meteorological Department. http://rsmcnewdelhi.imd.gov.in/images/pdf/
results indicate that the HWRF-POM-TC coupled model can
publications/preliminary-report/luban.pdf. (Accessed 29 December 2020).
provide vital guidance on the prediction of storm track, in- CWD, 2018. IMD Very Severe Cyclonic Storm “Titli” over Eastcentral Bay of
tensity etc. for operational forecasters over NIO. Bengal (08-13 October 2018): A Report, Cyclone Warning Division. India
The HWRF-POM-TC model also provided the updated Meteorological Department. http://www.rsmcnewdelhi.imd.gov.in/images/
SST fields in the wake of passage of cyclone. The SST fields pdf/publications/preliminary-report/titli.pdf. (Accessed 29 December
2020).
also were able to explain the weakening of VSCS Luban
Das, A.K., Rao, Y.R., Tallapragada, V., Zhang, Z., Bhowmik, S.R., Sharma, A.,
before landfall. 2015. Evaluation of the hurricane weather research and forecasting
The HWRF-POM-TC model was able to capture the Rapid (HWRF) model for tropical cyclone forecasts over the north Indian ocean
Intensification (RI) phase of VSCS Titli on one part and failed (NIO). Nat. Hazards 75 (2), 1205e1221.
to predict the RI phase of VSCS Luban. More experiments are Emanuel, K., 1999. Thermodynamic control of hurricane intensity. Nature
401, 665e669. https://doi.org/10.1038/44326, 1999.
needed to verify the RI case over NIO. Moreover, with the
Evans Jenni, L., 1993. Sensitivity of tropical cyclone intensity to sea surface
assimilation of INSAT 3D/3DR radiance could probably pro- temperature. J. Clim. 6 https://doi.org/10.1175/1520-0442(1993)
vide better intensity prediction and would help to overcome 006<1133:SOTCIT>2.0.CO;2.
the data sparse region over Arabian Sea and Bay of Bengal. Feng, X., Haines, K., Liu, C., de Boisseson, E., Polo, I., 2018. Improved
This work is first assessment of one of the two operation- SSTeprecipitation intraseasonal relationships in the ECMWF coupled
climate reanalysis. Geophys. Res. Lett. 45 (8), 3664e3672.
alized ocean coupled HWRF model over NIO by India
Ginis, I., Dikinov, Kh Zh, 1989. Modelling of the Typhoon Virginia (1978)
Meteorological Department. forcing on the ocean. Meteor. Hydrol. 7, 53e60.
Gopalakrishnan, S.G., Marks, F.D., Zhang, X., Bao, J.W., Yeh, K.S., Atlas, R.,
2011. The experimental HWRF system: a study on the influence of hori-
Acknowledgement zontal resolution on the structure and intensity changes in tropical cyclones
using an idealized framework. Mon. Weather Rev. 139, 1762e1784.
We acknowledge the Implementation Agreement under Gopalakrishnan, S.G., Goldenberg, S., Quirino, T., Marks, F., Zhang, X.,
Indo-US MoU which made possible the implementation of Yeh, K.S., Atlas, R., Tallapragada, V., 2012. Towards improving high-
resolution numerical hurricane forecasting: influence of model horizontal
Ocean coupled HWRF modelling system at India Meteoro-
grid resolution, initialization, and physics. Weather Forecast. 27, 647e666.
logical Department. The authors also acknowledge Dr. Vijay Han, J., Pan, H.L., 2011. “Revision of convection and vertical diffusion
Tallapragada and Dr. Avichal Mehra from modelling division, schemes in the NCEP global forecast system”, wea. Forecasting 26,
NCEP, NOAA, USA for their constant support. The authors 520e533.
express sincere gratitude to Dr. K. J. Ramesh and Dr. M. Han, J., Wang, W., Kwon, Y.C., Hong, S., Tallapragada, V., Yang, F., 2017.
Updates in the NCEP GFS cumulus convection schemes with scale and
Mohapatra for their encouragement and help in this work. The
aerosol awareness. Weather Forecast. 32 (5), 2005e2017.
authors acknowledge CWD- IMD for providing observed Heming, J.T., 2016. Met Office Unified Model tropical cyclone performance
track, wind speed for validation of model simulation results. following major changes to the initialization scheme and a model upgrade.
The author also acknowledges K Banerjee Center for Weather Forecast. 31 (5), 1433e1449.
69
Heming, J.T., Prates, F., Bender, M.A., Bowyer, R., Cangialosi, J., Caroff, P., using the HWRF modelling system. Q. J. R. Meteorol. Soc. 143 (703),
Coleman, T., Doyle, J.D., Dube, A., Faure, G., Fraser, J., Howell, B.C., 678e690.
Igarashi, Y., McTaggart-Cowan, R., Mohapatra, M., Moskaitis, J.R., Pilkington, S.F., Mahmoud, H.N., 2017. Real-time application of the multi-
Murtha, J., Rivett, R., Sharma, M., Short, C.J., Singh, A.A., hazard hurricane impact level model for the atlantic basin. Front. Built
Tallapragada, V., Titley, H.A., Xiao, Y., 2019. Review of recent progress in Environ. 3, 67. https://doi.org/10.3389/fbuil.2017.00067.
tropical cyclone track forecasting and expression of uncertainties. Tropical Powell, M.D., Vickery, P.J., Reinhold, T.A., 2003. Reduced drag coefficient for
Cyclone Research and Review 8 (4), 181e218, 2019. high wind speeds in tropical cyclones. Nature 422, 279e283.
Hong, S.Y., Pan, H.L., 1996. Nonlocal boundary layer vertical diffusion in a Price, J.F., 1981. Upper ocean response to a hurricane. J. Phys. Oceanogr. 11
medium range-forecast model. Mon. Weather Rev. 124, 2322e2339. (2), 153e175. https://doi.org/10.1175/1520-0485(1981)011<0153:UOR-
Iacono, M.J., Delamere, J.S., Mlawer, E.J., Shephard, M.W., Clough, S.A., TAH>2.0.CO;2.
Collins, W.D., 2008. Radiative forcing by long-lived greenhouse gases : Price, J.F., Sanford, T.B., Forristall, G.Z., 1994. Observations and simulations
calculations with the AER radiative transfer models. J. Geophys. Res. 113, of the forced response to moving hurricanes. J. Phys. Oceanogr. 24,
D13103. https://doi.org/10.1029/2008JD009944. 233e260.
Kai-Kwong Hon, 2020. Tropical cyclone track prediction using a large-area Rao, D.V.B., Prasad, D.H., Srinivas, D., 2009. Impact of horizontal resolution
WRF model at the Hong Kong Observatory. Trop. Cycl. Res. Rev. 9 (1), and the advantages of the nested domains approach in the prediction of
67e74. https://doi.org/10.1016/j.tcrr.2020.03.002. ISSN 2225-6032. tropical cyclone intensification and movement. J. Geophys. Res. 114
Kim, H., Lozano, C., Tallapragada, V., Iredell, D., Sheinin, D., Tolman, H., (D11). Art. no. D11106.
Gerald, V., Sims, J., 2014. “Performance of ocean simulations in the Rogers, E., Black, T., Ferrier, B., Lin, Y., Parrish, D., DiMego, G., 2001.
coupled HWRFeHYCOM model”. J. Atmos. Ocean. Technol. 31 https:// Changes to the NCEP Meso Eta Analysis and Forecast System: Increase in
doi.org/10.1175/JTECH-D-13-00013.1. Resolution, New Cloud Microphysics, Modified Precipitation Assimila-
Kwon, Y.C., Lord, S., Lapenta, B., Tallapragada, V., Liu, Q., Zhang, Z., 2010. tion, Modified 3DVAR Analysis. Technical Procedures Bulletin. http://
Sensitivity of air-sea exchange coefficients (Cd and ch) on hurricane in- www.emc.ncep.noaa.gov/mmb/mmbpll/eta12tpb/.
tensity. In: 29th Conference on Hurricanes and Tropical Meteorology, Schade, L.R., 2000. Tropical cyclone intensity and sea surface temperature.
13C.1. J. Atmos. Sci. 57 (18), 3122e3130. https://doi.org/10.1175/1520-
Lin, I.I., Chen, C.H., Pun, I.F., Liu, W.T., Wu, C.C., 2008. Warm ocean 0469(2000)057<3122:TCIASS>2.0.CO;2.
anomaly, air sea fluxes, and the rapid intensification of tropical cyclone Smagorinsky, J., 1963. General circulation experiments with primitive equa-
Nargis. Geophys. Res. Lett. 36 (3). tions. Part I: the basic experiments. Mon. Wea. Rev. 91, 99e164. https://
Lin, I.I., Goni, G.J., Knaff, J., Forbes, C., Ali, M.M., 2012. Ocean heat content doi.org/10.1175/1520-0493(1963)091<0099:GCEWTP>2.3.CO;2.
for tropical cyclone intensity forecasting and its impact on storm surge. Srinivas, C.V., Rao, D.V.B., Yesubabu, V., Baskaran, R., Venkatraman, B.,
Nat. Hazards 66, 1481e1500. https://doi.org/10.1007/s11069-012-0214-5. 2012. Tropical cyclone predictions over the bay of Bengal using the high-
Mehra, A., Tallapragada, V., Zhang, Z., Liu, B., Zhu, L., Wang, W., Kim, H., resolution advanced research weather research and forecasting (ARW)
2018. A dvancing the state of the art in operational tropical cyclone model. Q. J. Roy. Meteorol. Soc. 139 (676), 1810e1825.
forecasting at ncep. Trop. Cycl. Res. Rev. 7 (1), 51e56. https://doi.org/ Tallapragada, V., Surgi, N., Liu, Q., Kwon, Y., Tuleya, R., O'Connor, W., 2008.
10.6057/2018TCRR01.06. ISSN 2225-6032. Performance of the advanced operational HWRF modeling system during
Mellor, G.L., Yamada, T., 1982. Development of a turbulence closure model pre-implementation testing and in real-time 2007 hurricane season. In:
for geophysical fluid problems. Rev. Geophys. 20, 851e875. Recorded Presentation, 28th Conference on Hurricanes and Tropical
Mitchell, K., 2005. The community Noah land surface model (LSM). http:// Meteorology, Orlando, FL: American Meteorological Society, 4A.5.
www.ral.ucar.edu/research/land/technology/lsm/noah/Noah_LSM_ Tallapragada, V., Kieu, C., Kwon, Y., Trahan, S., Liu, Q., Zhang, Z.,
USERGUIDE_2.7.1.pdf. Kwon, I.H., 2014. Evaluation of storm structure from the operational
Mogensen, K.S., Magnusson, L., Bidlot, J.R., 2017. Tropical cyclone sensi- HWRF during 2012 implementation. Mon. Weather Rev. 142 (11),
tivity to ocean coupling in the ECMWF coupled model. J. Geophys. Res. 4308e4325.
Oceans 122 (5), 4392e4412. Tallapragada, V., Kieu, C., Trahan, S., Liu, Q., Wang, W., Zhang, Z., Strahl, B.,
Mohanty, U.C., Osuri, K.K., Tallapragada, V., Marks, F.D., Pattanayak, S., 2016. Forecasting tropical cyclones in the western North Pacific basin
Mohapatra, M., Rathore, L.S., Gopalakrishnan, S.G., Niyogi, D., 2015. using the NCEP operational HWRF Model: model upgrades and evaluation
A great escape from the bay of bengal “super sapphireephailin” tropical of real-time performance in 2013. Weather Forecast. 31 (3), 877e894.
cyclone: a case of improved weather forecast and societal response for Thompson, G., Bernardet, L., Biswas, M., Holt, C., 2014. Explicitly-coupled
disaster mitigation. Earth Interact. 19 (17), 1e11. cloud physics and radiation parameterizations and subsequent testing in
Mohanty, S., Nadimpalli, R., Osuri, K.K., Pattanayak, S., Mohanty, U.C., HWRF. http://www.dtcenter.org/eval/hwrf_hdrf_hdgf/HFIP_
Sil, S., 2019. Role of sea surface temperature in modulating life cycle of GT2014Dec17.pdf.
tropical cyclones over bay of bengal. Trop. Cycl. Res. Rev. 8 (2), 68e83. Yablonsky, R.M., Ginis, I., 2008. Improving the ocean initialization of coupled
Nadimpalli, R., Osuri, K.K., Mohanty, U.C., Das, A.K., Kumar, A., Niyogi, D., hurricaneeocean models using feature-based data assimilation. Mon. Wea.
2020. Forecasting tropical cyclones in the Bay of Bengal using quasi- Rev. 136, 2592e2607. https://doi.org/10.1175/2007MWR2166.1.
operational WRF and HWRF modeling systems: an assessment study. Yablonsky, R.M., Ginis, I., Thomas, B., Tallapragada, V., Sheinin, D.,
Meteorol. Atmos. Phys. 132, 1e17, 2020. Bernardet, L., 2015a. “Description and analysis of the ocean component of
Neetu, S., Lengaigne, M., Vincent, E.M., Vialard, J., Madec, G., Samson, G., NOAA's operational hurricane weather research and forecasting model
Ramesh Kumar, M.R., Durand, F., 2012. “Influence of upper-ocean strat- (HWRF)”. J. Atmos. Ocean. Technol. 32 (1), 144e163.
ification on tropical cyclone-induced surface cooling in the Bay of Ben- Yablonsky, R.M., Ginis, I., Thomas, B., 2015b. Ocean modeling with flexible
gal”. J. Geophys. Res. 117, C12020. https://doi.org/10.1029/ initialization for improved coupled tropical cyclone-ocean model predic-
2012JC008433. tion. Environ. Model. Softw. 67, 26e30. https://doi.org/10.1016/j.env-
Osuri, K.K., Mohanty, U.C., Routray, A., Mohapatra, M., Niyogi, D., 2013. soft.2015.01.003. C (May 2015).
Real-time track prediction of tropical cyclones over the north Indian ocean Yesubabu, V., Kattamanchi, V.K., Vissa, N.K., Dasari, H.P., Sarangam, V.B.R.,
using the ARW model. J. Appl. Meteorol. Climatol. 52 (11), 2476e2492. 2020. Impact of ocean mixed-layer depth initialization on the simulation of
Osuri, K.K., Nadimpalli, R., Mohanty, U.C., Niyogi, D., 2017. Prediction of tropical cyclones over the Bay of Bengal using the WRF-ARW model.
rapid intensification of tropical cyclone Phailin over the Bay of Bengal Meteorol. Appl. 27, e1862 https://doi.org/10.1002/met.1862.
70

1 s2.0 S2225603221000059 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S2225603221000059 Main

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

1 s2.0 S2225603221000059 Main

Uploaded by

Copyright:

Available Formats

Available online at www.sciencedirect.

A HWRF-POM-TC coupled model forecast performance over North Indian

Keywords: Ocean coupled hurricane WRF; POM-TC; HWRF-POM-TC

1. Introduction going on to mitigate the damages caused by tropical cyclones

Fig. 4. HWRF-POM-TC forecast and CWD forecast landfall point error.

Fig. 11. (continued)

Fig. 11. (continued)

You might also like