Ocean impact on the intensification of cyclone Titli

Ocean impact on the intensiBcation of cyclone Titli

K MANEESHA1,* , V SIVA PRASAD

2 and K VENKATESWARARAO1

1National Institute of Oceanography, Regional Centre, Visakhapatnam, India.

2Andhra University, Visakhapatnam, India.*Corresponding author. e-mail: [email protected]

MS received 22 January 2021; revised 8 April 2021; accepted 15 April 2021

Tropical cyclones are the most devastating weather events which still needs to understand much due tothe uniqueness in their paths and intensiBcation locations. One such recent very severe cyclonic storm,Titli, is perfectly Bt for the individual case study. The cyclone Titli originated from a low-pressure areaformed over the southeast Bay of Bengal (BoB) and adjoining north of the Andaman Sea on 7th October2018 and intensiBed into a very severe cyclonic storm on 10th over a region of low saline water pool in thewestern Bay of Bengal. The addition of low salinity water in this region is from the lower Mahanadi basinand surroundings that received excess rainfall in September 2018. The low saline waters developed highstratiBcation and suppressed the upwelling at the cold-core eddy center and restricted to 50 m. Persistenthigh heat content and high internal energy are the primary sources of intensiBcation of cyclone Titli.Generally, intense cyclones enrich biomass’s enhancement in the ocean’s surface layer after their passage.But in the case of Titli, biomass enrichment did not happen due to persistent high stratiBcation, which isabout 3–4. Even the 80 knots (150 km/h) winds cannot break up the upper layer stratiBcation, sup-pressing the cyclone-induced upwelling. Observations show only a 0.2–0.4 mg/m3 rise in chlorophyll onthe surface and it may be due to the land drove nutrients and lightning. Data analyzed from the high-resolution global NEMO model also shows the intrusion of low saline waters oA India’s east coast. Thoughthe model is slightly overestimating the upwelling in the cold core eddy’s surface layers, it represents thebarrier caused by stratiBed waters. Both the observations and model datasets are well correlating, dailyanalyzed model data is well capturing the seasonal stratiBcation caused by the low saline waters in thenorthern Bay of Bengal. These high-resolution global ocean model datasets are useful and essential for theaccurate forecast of cyclones over the Bay of Bengal.

Keywords. Cold core eddy; upwelling; cyclones.

1. Introduction

Generally, about 6–7% of the global tropicalcyclones form over the Indian ocean (Nuemann1993). In the Indian Ocean, Bay of Bengal is muchprone to storms than the Arabian Sea because offreshwater input. This freshwater occupies the toplayers and holds high heat content, whereas in theArabian Sea, strong winds cause mixing and dis-tribution of heat, which may not support the

formation and intensiBcation of more cyclones(Shenoi et al. 2004; Vinay Chandran et al. 2008).We are now experiencing several climate changesall over the globe. In the recent climate-changingera, such changes are also evident in the IndianOcean region. Those increase the risk of extremeweather events like cyclones to certain areasunexpectedly due to ocean warming in many areas.Those include the recent 2014–2018 cyclones overthe northern Indian ocean. One of such cyclones,

J. Earth Syst. Sci. (2021) 130:164 � Indian Academy of Scienceshttps://doi.org/10.1007/s12040-021-01660-9 (0123456789().,-volV)(0123456789().,-volV)

we are discussing in this paper is Titli. Titli wasthe most destructive cyclonic storm to strike theIndian coast in 2018. The genesis of very severecyclonic storm (VSCS), Titli over the Bay ofBengal, took place 45 h after VSCS, Luban, overthe Arabian Sea. It was one of the rarest of rareevents that simultaneously two VSCSs developedover the Arabian Sea and Bay of Bengal. Titlioriginated as a low-pressure area over the south-east Bay of Bengal (BoB) and adjoining the northAndaman Sea on 7th October 2018. Moving nearlywest-northwestwards, it intensiBed into a deepdepression (DD) over east-central BoB on 8thOctober and further into a cyclonic storm (CS)‘Titli’ around noon of 9th October 2018. Later itmoved north-westwards and intensiBed into asevere cyclonic storm (SCS) in the early hours of10th, further moved north–northwestwards andintensiBed into a very severe cyclonic storm(VSCS) on 10th. It crossed near Palasa (18.80N/84.50E) on 11th as a VSCS with a wind speed of140–150 gusting to 165 kmph (www.imd.gov.in).An increase in the intensity and frequency ofstorms in recent years (2014–2018) indicates thatwe are already facing climate change. Even thoughthere are many improvements in observationmethods during cyclones, the past studies are notreliable with today’s weather. There is a need towork on recent individual storms since the cycloneformation, intensiBcation is a highly complex phe-nomenon and is an interaction between ocean andatmosphere. Similarly, cyclone Hudhud occurred inthe year 2014 was also intensiBed over the northernBay of Bengal during the post-monsoon season, butthe reasons behind the stratiBcation are differentfrom Titli. So, for each cyclone, the model has towork differently and thereby making the predictionmore complex. So it is necessary to understand therecent high-resolution coupled models and it has tobe veriBed how best they help to analyze theindividual tracks.The present study is on the cyclone Titli in

which we utilized the in-situ satellite and modeldata for analysis.

2. Data and methods

Cyclone track data has been taken from the IMDwebsite. The ocean data has been taken fromARMOR 3D data, which provides temperature-salinity vertical proBles to build a global 3D weeklytemperature, salinity Belds at a Spatial Resolution

of 0.250 9 0.250 grid resolution. Atmospheric datahas been taken from NCEP-NCAR reanalysisdatasets. Daily rainfall statistics have been takenfrom IMD. StratiBcation (S) has been estimatedfrom the stratiBcation factor which is derived fromthe maximum and reference Bruntvaisala fre-quencies (N). The maximum frequency is esti-mated from the vertical proBle and the referencefrequency is 3 cph.

S ¼ffiffiffiffiffiffiffiffiffiffiffi

Nmax

N o

r

where

N ¼

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

� g

qdqdZ

s

:

The model data has been taken from CMEMSglobal weakly coupled ocean–atmospheric dataassimilation, using 3D-var, which utilized GLO-CPL for 3d ocean forecast at 1/4th degree metoDce uniBed model coupled to an hourly NEMO v3.4 (Lea et al. 2015). NEMO run in ORACA025conBguration (Blockley et al. 2014) with a tripolargrid of 1/4th degree resolution and 75 verticallevels with 1 m resolution in the top 10 m of theocean, decreasing resolution in deeper parts. Oceanobservations include satellite SST data, daily in-situ Argo, moored buoys and gliders. Ocean heatcontent data is extracted from BHUVAN which iscomputed from ArtiBcial Neural Network modelsby utilizing remote-sensing sea-surface tempera-ture and sea surface height anomaly data. Atmo-spheric instability is calculated from equation x =Dp/Dt.

3. Results and discussions

Tropical cyclone formation and intensiBcation areunder the control of both atmospheric and oceanparameters jointly. The percentage of possession ofeach may vary from cyclone to cyclone. CycloneTitli is one of the best case studies to show theexact ocean role in the intensiBcation. In the pre-sent study, we had taken the spatial data up to500 m depth just 15 days before the formation ofcyclone Titli and overlaid the track. It is observedthat a freshwater plume oA the Orissa coast nearthe landfall of cyclone Titli and this plume exten-ded up to 150 km oAshore. This freshwater plume isdue to the high river discharge caused by the excessrainfall near Mahanadi during the monsoon season

164 of 9 J. Earth Syst. Sci. (2021) 130:164

of the year 2018. During the monsoon season(2008–2018), the annual rainfall graph under thelower Mahanadi basin, mainly in the Bhubaneswarand Paradeep regions, clearly shows the excessrainfall in 2018. This may be the reason behindwhy the high freshwater discharge oA the east coastnear Orissa (Bgure 1).Generally, this freshwater plume acts as a bar-

rier for mixing and traps heat in the ocean’s surfacelayers. The summer Coodwaters of the Ganges andBrahmaputra rivers discharging into the Bay ofBengal plus the Irrawaddy and Salween riversemptying into the Andaman Sea combine toinCuence the salinity of the surface waters overthousands of kilometers oAshore (Tomczak andGodfrey 1994). This fact was conBrmed during thesummer 2001 Bay of Bengal Ocean Process Studiescruise. The copious amounts of rainfall and riverrunoA freshened the Bay’s upper 30 m by 3–7 PSUcompared to the central Arabian Sea (PrasannaKumar et al. 2002). Similarly, in the western Bay ofBengal, the freshwater discharge due to excessrainfall from August to September 2018 over thelower Mahanadi basin (Bgure 1c), caused a highstratiBed layer that holds high heat contentfavourable for formation and intensiBcation ofpost-monsoon cyclones in this region.Figure 2 shows the pool of lower salinity water

along the east coast of India beyond the Godavariand after Mahanadi conCuence into the Bay ofBengal. Salinity decreased from 28 PSU at 150 kmoAshore to 22 PSU near the coast and this lowsalinity pool extended from 16.5� to 20oN along thecoast up to the longitudinal belt of 86oE.High stratiBcation has been observed over

north and northwestern parts of Bay of Bengal,i.e., almost greater than 3 where the central BayS varies from 1 to 2. Shallow mixed layer depthsbelow 10 m depict low entrainment due to thishigh stratiBed layer before cyclone Titli(Bgure 3).

During the last week (27th) of September of2018, the Cow of EICC is towards the south with aspeed of nearly 0.8 m/s and also the Cow is dividedinto several mesoscale eddies due to its instability.One such cold-core vortex is situated at the centerof the cyclone track before the cyclone (Bgure 4a),which is also evident through the vertical section’sobservation along the path. Figure 10(b) shows theupwelling of isotherms at the center of the cold-core eddy and the subsidence of isotherms at theeddy’s peripheral. But this upwelling could notreach the surface and it has been suppressed below50 m. This is due to the strong stratiBcation in thetop layers, which strongly inhibits the subsurfacewaters to mix with surface layers. The surfacewarm water layer with temperatures greater than28�C is due to the intrusion of fresh waters fromcoast to oAshore, the low saline water tongueextended up to 250 km oAshore (Bgure 4c). As thesurface layer stratiBcation reduces the strength ofupwelling (Prasanna Kumar et al. 2002; Carineet al. 2019), the density section clearly shows thesuppressed upwelling above 50 m (Bgure 10d). Thisphenomenon opposes cooling, which generallyhappens at the center of the cold-core eddy. So thetop layers are warming up, showing high internalenergy in the upper 50 m and the conditions beforethe cyclone are highly favourable for the formationand intensiBcation.The cyclone Titli formed on 8th October 2018

and started with a translation speed of 2.6 m/s andbecame a deep depression under a translation speedof 1.4 m/s. Later on, it continued with a hightranslation speed of 5.6 m/s up to 10th October2018 and after reaching the western part again, itslowed down to 3.2 m/s under high stratiBedwaters (Bgure 5).The interaction of the cyclone with the ocean

during intensiBcation on the 10th has been shownin Bgure 6. The cyclone circulation has been iden-tiBed from Bgure 6(a). It has undergone rapid

Figure 1. Cumulative rainfall during monsoon season under lower Mahanadi river basin. (a) Bhubaneswar, (b) Para deep, and(c) spatial rainfall pattern during September 2018 in mm over Orissa.

J. Earth Syst. Sci. (2021) 130:164 of 9 164

Figure 2. Spatial map of sea surface (a) salinity and (b) temperature over the Bay of Bengal before (26/09/2018) cyclone Titli.

Figure 3. Spatial map of (a) mixed layer depth (density threshold equivalent to 0.2�C) and (b) stratiBcation factor (S) over Bayof Bengal prior to (26/09/2018) cyclone Titli.

Figure 4. (a) Surface circulation over the Bay of Bengal and along-track vertical sections of (b) temperature (�C), (c) salinity(PSU), and (d) density (kg/m3) prior to Titli.


intensiBcation over the high temperatures greaterthan 30�C and stratiBcation greater than 3 at adistance of 200 km from the coast (Bgure 6b, c).The along-track section clearly shows the upwel-ling signature at the center of the cold-core. It

could not reach the upper layers due to persistenthigh stratiBcation due to low salinity water cap,evident from Bgure 6(b and f). Also, these resultsare compared with the ocean–atmospheric coupledmodel with NEMO 3.4 multiple layer ocean

Figure 5. Translation speed (m/s) and the corresponding pressure drop during Titli cyclone.

Figure 6. (a) Surface circulation (m/s), (b) surface salinity (PSU), (c) sea surface temperature (�C), (d) stratiBcation and along-track vertical sections of (e) temperature and (f) salinity prior to cyclone Titli.


conBguration simulated on daily intervals with0.25 resolution. The model could be able to catchthe low salinity patch very well though it is slightlyoverestimating. It clearly shows the low salinitywater along the coast at 18�N, higher SSTs around31�C, and cold-core eddy oA 16�N has beenobserved along the track (Bgure 7a–c).The stratiBcation factor has been underesti-

mated by the model with high stratiBcation on thetrack’s north right side. In contrast, the observa-tions show the maximum stratiBcation on the leftside. Both represent the stratiBed waters oA 18�Nand along-track vertical sections from model data,clearly illustrating the upwelling signatures in thecold core eddy and the intrusion of low saline waterfrom the coast (Bgure 7e, f), which prohibited theupwelling to surface layers. Higher temperaturesin the top 50 m have been observed from themodel data and are favourable for cyclone’sintensiBcation.Out of 100 m, the top 50 m layer contains high

heat content (estimated from remote sensing),crucial for cyclone formation and intensiBcation(Bgure 8).In the present cyclone, Titli started with low

translation speed up to deep depression and thetranslation speed increased at the time of rapidintensiBcation. There is a large drop in translation

speed at the location of high heat content. Thesurface waters with high heat content are respon-sible for the instability, ascending motion in theatmosphere, and accumulation of high relativehumidity (Bgure 9a) at different atmosphericlevels.These are helpful for the release of latent heat,

reduces the temperature contrast between theocean and atmosphere (Bgure 9b) and helps thedevelopment of cyclone further during and afterthe passage of cyclone along-track observational aswell as model data has been observed and notfound many changes in the upwelling along thetrack Bgure 10. The observed size of the cold coreeddy along the track is 2.5 9 4; it is extended up tothe coast due to strong EICC, whereas the modeldata underestimates the speed of the EICC and thesize of the eddy (3 9 2.5). Both the model andobservations show a stable upwelling signature,which is suppressed above 50 m. There is no neg-ative impact of the cold-core eddy on a cyclone;this may be due to the substantial intrusion of lowsaline water from the coast, which is evidentthrough observations and model data. Modelresults show the low saline (\30 PSU) waterintrusion up to 300 km from the coast, whereas theobservation shows 150 km (Bgure 11e–h). Themodel is underestimating the temperatures and

Figure 7. (a) Surface circulation (m/s), (b) surface salinity (PSU), (c) sea surface temperature (�C), (d) stratiBcation and along-track vertical sections of (e) temperature (�C), and (f) salinity (PSU) before cyclone Titli using model data.


overestimating the salinity in the top 50 m layer(Bgure 10a–d).

Biological response: Generally, in the Bay ofBengal coastal and open-ocean, biogeochemistrywas inCuenced by the cold-core eddies, rainfallevents and cyclones. Tropical cyclones (also knownas hurricanes or typhoons) can have devastatingeAects on human lives when passing over land.Still, they can enormously enhance another form oflife – ocean primary (phytoplankton) production(Lin et al. 2003). Once the cyclone is passed alongthe cyclone’s track, there will be cooling due to thewind-forced upwelling at the center. Such physicalchanges are expected to impact marine biogeo-chemistry. The cool, nutrient-rich water will cometo the surface, which reduces the surface temper-atures. It may vary from 1� to 10�C depending onthe strength of the cyclone. Injection of nutrientsfrom the subsurface will increase the production.After the cyclone departure, cold wakes, typicallymore evident to the right of the cyclone tracks(Chang and Anthes 1978; Price 1981; Cornillonet al. 1987; Monaldo et al. 1997; Bender and Ginis2000; Wentz et al. 2000; Lin et al. 2003; Maneeshaet al. 2011) are left behind. These cold wakes canexist in the ocean for days to weeks and generatecontinual feedback with the atmosphere (Emanuel2001; Lin et al. 2005). After the landfall, theland-drivenmaterial from the rainfall and the currentsalong the coast will aAect coastal biogeochemistry.

Figure 8. Spatial heat content (W/m2) at (a) 50 m and (b) 100 m over the western Bay of Bengal before the cyclone Titli.

Figure 9. Atmospheric (a) relative humidity (%) and (b) instability (pa/s) during cyclone Titli intensiBcation.

Figure 10. Along track vertical section of temperature (�C)(a–d) and salinity (PSU) (e–h) from observational (a, c, e, g),model (b, d, f, h) datasets during (a, b, e, f) and after (c, d, g,h) passage of cyclone.


Storms occurring in the world oceans typicallyhave a duration of only a few days. Still, theseperturbations in physical and biological eAects canlast up to several weeks (Madhu et al. 2002;Maneesha et al. 2011).The biological response of several recent

cyclones has been studied over the Bay of Bengal.But at present, for the cyclone Titli, the conditionis different though the winds were much morerobust, i.e., greater than 80 knots; they cannotbreak the stratiBcation in the western Bay ofBengal. This results in the suppression of upwellingin the surface layers significantly, less rise in thesurface chlorophyll levels over the Bay of Bengal.The minor variation, i.e., from 0.1 to 0.2 mg/m3 inchlorophyll, may be due to the land-drove nutri-ents and nitrogen Bxation due to lightning duringthe cyclone (Bgure 11).

4. Conclusions

A tropical cyclone is one of the devastating weatherevents which still needs to understand a lot due totheir uniqueness in tracks and intensiBcationlocations. The recent cyclone Titli is a perfect Btfor an individual case study because of its rapidintensiBcation and biological response. It dictatesthe ocean’s role in the sudden intensiBcation. Thecyclone Titli started in the central Bay of Bengaland moved northwestwards the highly stratiBed,low saline water pool and intensiBed rapidly. Thislow saline water results from the freshwater inCuxcaused by the heavy rainfall near the lowerMahanadi basin in September 2018. So, highlystratiBed upper layers in the western Bay of Bengalhaving high heat content in the top 50 m layer gavepositive feedback for the cyclone Titli’s rapid

intensiBcation. Even 80-knot cyclonic wind cannotbreak the surface layer stratiBcation, which can beevident through surface chlorophyll levels after thecyclone. High-resolution model results are wellcomparable with the observations and these mod-els are helpful for better prediction of cyclone trackand intensity.

Acknowledgements

The authors acknowledge the Director, CSIR-NIOand Scientist-in-Charge for their support and pro-viding the facilities. The authors are thankful toDr Rashmi Sharma and Dr Mini Raman, SpaceApplication Centre, Ahmedabad for their encour-agement and support. The authors thank ISRO forproviding funds through ISRO RESPOND ProjectOGP-204. This has the NIO contribution number6754.

Author statement

KM conceived the idea, made all the calculationsand interpretation, VSP and KV helped in thearrangement of data, Bgures and interpretationof physical, biological aspects, all the authorscontributed to the development of the Bnalmanuscript.

References

Bender M A and Ginis I 2000 Real-case simulations ofhurricane–ocean interaction using a high resolution coupledmodel: EAects on hurricane intensity; Mon. Wea. Rev. 128917–946.

Figure 11. Surface chlorophyll (mg/m3) variations over the Bay of Bengal (a) before (01/10/2018), (b) after (20/10/2018) thecyclone Titli, and (c) difference.


Blockley E W and Coauthors 2014 Recent development of theMet ODce operational ocean forecasting system: An over-view and assessment of the new Global FOAM forecasts;Geosci. Model Dev. 7 2613–2638, https://doi.org/10.5194/gmd-7-2613-2014.

Carine G van der Boog, Julie P, Henk D and Caroline A K2019 The impact of upwelling on the intensiBcation ofanticyclonic ocean eddies in the Caribbean Sea; Ocean Sci.Discuss., https://doi.org/10.5194/os-2019-51.

Chang S W and Anthes R A 1978 Numerical simulations of theocean’s nonlinear, baroclinic response to translating hurri-canes; J. Phys. Oceanogr. 8 468–480.

Cornillon P, Stramma L and Price J F 1987 Satellitemeasurements of sea surface cooling during hurricaneGloria; Nature 326 373–375.

Emanuel K A 2001 The contribution of tropical cyclones to theoceans’ meridional heat transport; J. Geophys. Res. 10614771–14781.

Lea D J, Mirouze I, Martin M J, King R R, Hines A, Walters Dand Thurlow M 2015 Assessing a new data assimilationsystem based on the Met ODce coupled atmosphere–land–ocean–sea ice model; Mon. Wea. Rev. 143 4678–4694.

Lin I I, Liu W T, Wu C C, Wong G T F, Chen C, Hu Z, LiangW D, Yang Y and Liu K K 2003 New evidence for enhancedocean primary production triggered by tropical cyclone;Geophys. Res. Lett. 30 1718, https://doi.org/10.1029/2003GL017141.

Lin I I, Wu C C, Emanuel K A, Lee I H, Wu C R and Pun I F2005 The interaction of supertyphoon maemi (2003) with awarm ocean eddy; Mon. Wea. Rev. 133 2635–2649.

Madhu N V, Maheswaran P A, Jyothibabu R, Sunil V,Revichandran C, Balasubramanian T, Gopalakrishnan T Cand Nair K C 2002 Enhanced biological production oAChennai triggered by October 1999 super cyclone (Orissa);Curr. Sci. 82 1472–1479.

Maneesha K, Sarma V V S S, Reddy N P C, Sadhuram Y,Murty T V R, Sarma V V and Kumar M D 2011 Meso-scale

atmospheric events promote phytoplankton blooms in thecoastal Bay of Bengal; J. Earth Syst. Sci. 120 773–782,https://doi.org/10.1007/s12040-011-0089-y.

Monaldo F M, Sikora T D, Babin S M and Sterner R E 1997Satellite imagery of sea surface temperature cooling in thewake of hurricane Edouard (1996); Mon. Wea. Rev. 1252716–2721.

Neumann C J 1993 Global Overview. Chapter 1; Global Guideto Tropical Cyclone Forecasting; WMO/TC-No. 560Report No. TCP-31 World Meteorological OrganizationGeneva Switzerland.

Prasanna Kumar S, Muraleedharan P M, Prasad N andMadhupratap M 2002 Why is the Bay of Bengal lessproductive during summer monsoon compared to theArabian Sea?; Geophys. Res. Lett. 29 2235, https://doi.org/10.1029/2002GL016013.

Prasanna Kumar S, Nuncio M, Jayu Narvekar, Ajoy Kumar,Sardesai S, Desouza S N, Mangesh Gauns, Ramaiah N andMadhuprat 2004 Are Eddies nature’s trigger to enhancebiological productivity in the Bay of Bengal?; Geophys. Res.Lett. 31 L07309, https://doi.org/10.1029/262003Gl019274.

Price J F 1981 Upper ocean response to a hurricane; J. Phys.Oceanogr. 11 153–175.

Shenoi S S C, Shankar D and Shetye S R 2004 Why is Bay ofBengal warmer than Arabian Sea during the summermonsoon?; Proceedings of the National SymposiumMETOC – 2004 on Emerging Trends in the Fields ofOceanography and Meteorology, 05–06 February 2004,pp. 87–93.

Tomczak M and Godfrey J S 1994 Regional oceanography: Anintroduction; Pergamon Press, Oxford.

Vinayachandran P N and Saji N H 2008 Mechanisms of southIndian Ocean intraseasonal cooling; Geophys. Res. Lett. 35L23607, https://doi.org/10.1029/2008GL035733.

Wentz F J, Gentemann C, Smith D and Chelton D 2000Satellite measurements of sea surface temperature throughclouds; Science 288(5467) 847–850.

Corresponding editor: N V CHALAPATHI RAO


Ocean impact on the intensification of cyclone Titli

Documents

Transcript of Ocean impact on the intensification of cyclone Titli