Sea Surface Temperature Variability Over the Tropical Indian Ocean During the ENSO and IOD Events in 2016 and 2017

atmosphere

Article Sea Surface Temperature Variability over the Tropical Indian Ocean during the ENSO and IOD Events in 2016 and 2017

Sartaj Khan 1,2 , Shengchun Piao 1,2, Guangxue Zheng 1 , Imran Ullah Khan 1 , David Bradley 3, Shazia Khan 4 and Yang Song 1,2,*

1 College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China; [email protected] (S.K.); [email protected] (S.P.); [email protected] (G.Z.); [email protected] (I.U.K.) 2 Acoustic Science Technology Laboratory, Harbin Engineering University, Harbin 150001, China 3 School of Marine Science and Ocean Engineering, University of New Hampshire, Durham, NH 03824, USA; [email protected] 4 Department of Biological Sciences, International Islamic University, Islamabad 44000, Pakistan; [email protected] * Correspondence: [email protected]; Tel.: +86-158-4650-8908

Abstract: 2016 and 2017 were marked by strong El Niño and weak La Niña events, respectively, in the tropical East Pacific Ocean. The strong El Niño and weak La Niña events in the Pacific significantly impacted the sea surface temperature (SST) in the tropical Indian Ocean (TIO) and were followed by extreme negative and weak positive Indian Ocean Dipole (IOD) phases in 2016 and 2017, which triggered floods in the Indian subcontinent and drought conditions in East Africa. The IOD is an irregular and periodic oscillation in the Indian Ocean, which has attracted much attention in the last Citation: Khan, S.; Piao, S.; Zheng, two decades due to its impact on the climate in surrounding landmasses. Much work has been done G.; Khan, I.U.; Bradley, D.; Khan, S.; in the past to investigate global climate change and its impact on the evolution of IOD. The dynamic Song, Y. Sea Surface Temperature behind it, however, is still not well understood. The present study, using various satellite datasets, Variability over the Tropical Indian examined and analyzed the dynamics behind these events and their impacts on SST variability in the Ocean during the ENSO and IOD TIO. For this study, the monthly mean SST data was provided by NOAA Optimum Interpolation Events in 2016 and 2017. Atmosphere Sea Surface Temperature (OISST). SST anomalies were measured on the basis of 30-year mean daily 2021, 12, 587. https://doi.org/ climatology (1981–2010). It was determined that the eastern and western poles of the TIO play quite 10.3390/atmos12050587 different roles during the sequence of negative and positive IOD phases. The analysis of air-sea

Academic Editors: Ke Fan and interactions and the relationship between wind and SST suggested that SST is primarily controlled by Rocky Talchabhadel wind force in the West pole. On the other hand, the high SST that occurred during the negative IOD phase induced local convection and westerly wind anomalies via the Bjerknes feedback mechanism. Received: 23 March 2021 The strong convection, which was confined to the (warm) eastern equatorial Indian Ocean was Accepted: 28 April 2021 accompanied by east–west SST anomalies that drove a series of downwelling Kelvin waves that Published: 1 May 2021 deepened the thermocline in the east. Another notable feature of this study was its observation of weak upwelling along the Omani–Arabian coast, which warmed the SST by 1 ◦C in the summer of Publisher’s Note: MDPI stays neutral 2017 (as compared to 2016). This warming led to increased precipitation in the Bay of Bengal (BoB) with regard to jurisdictional claims in region during the summer of 2017. The results of the present work will be important for the study of published maps and institutional affil- monsoons and may be useful in predicting both droughts and floods in landmasses in the vicinity of iations. the Indian Ocean, especially in the Indian subcontinent and East African regions.

Keywords: sea surface temperature; Indian Ocean dipole; El Niño; La Niña

Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article 1. Introduction distributed under the terms and conditions of the Creative Commons Sea surface temperature (SST) is one of the most critical oceanic parameters. It plays a Attribution (CC BY) license (https:// major role in the development of atmospheric events in the Indian Ocean both on regional creativecommons.org/licenses/by/ and global scales [1–3]. The evolution of SST anomalies in the Indian Ocean mainly involves 4.0/).

Atmosphere 2021, 12, 587. https://doi.org/10.3390/atmos12050587 https://www.mdpi.com/journal/atmosphere Atmosphere 2021, 12, 587 2 of 20

coupled ocean-atmosphere processes [4,5] that are either (i) generated by the large-scale atmospheric forces linked with El Niño/La Niña-Southern Oscillation (ENSO) in the tropical eastern Pacific [6–8] or (ii) brought about by an internal independent ocean mechanism such as the Indian Ocean Dipole (IOD) [4,9–13], either of which can affect the interannual variability of SST. The Indian Ocean is considered the warmest ocean in the world in April–May (the Indian Ocean warm pool [14]) and is a major cause of monsoon rainfall [15]. However, the size of the warm pool (which maximizes in April–May) has been reduced by the Somalia–Oman upwelling and also in part by increased latent heat in the Arabian Sea (AS) [16,17]. These warm/cold SST anomalies occur in the western AS due to weak/strong upwellings, respectively, and are the main cause of increased/decreased precipitation anomalies for Indian summer monsoon rainfall [18]. In fact, these SST anomalies, driven by various ocean dynamics (e.g., horizontal and vertical advection, surface-based energy flows, horizontal and vertical wind turbulence), are mainly responsible for causing extreme weather conditions in the Indian Ocean during the monsoon season. They also influence weather and climate over adjoining land areas [19,20]. These extreme weather events in the Indian Ocean may lead to drought situations if break conditions continue for a few weeks. The longevity of active conditions can lead to heavy rain and severe floods [21,22]. These increased drought conditions have a robust relationship with SST variations—particularly in tropical regions—and related variations in atmospheric circulation and rainfall [23]. The dynamics of such variations in monsoon rainfall and other extreme weather conditions are not well understood. However, it is well recognized that the atmosphere interacts with the upper ocean (rather than the surface alone), suggesting that forecasts of the monsoon in the Indian Ocean can be improved by considering upper ocean parameters [24]. Recent studies have focused on the Pacific Ocean as a prospective player in modulating global warming trends due to its huge volume [25–27]. The ENSO is an irregularly periodic variation in SST and wind over the tropical eastern Pacific Ocean, with major global socioeconomic and environmental impacts [28]. As can be seen from Figure1, 2016 was marked by strong El Niño conditions in which maximum positive SST anomalies were confined to the Niño 3.4 region. 2017, in turn, was marked by weak La Niña conditions in which maximum negative SST anomalies were confined to the same region. The 2016 El Niño was one of the first powerful El Niño events of the 21st century and one of the three strongest documented since 1950—together with those of 1997/98 and 1982/83 [29]. Warm conditions persisted, especially from October 2015 to April 2016, when the El Niño impact on global climate was at its peak. Positive equatorial SST anomalies continued across most of the Pacific and Indian oceans, while negative SST anomalies prevailed most of the year in 2017 (Figure1a,b). The large positive SST anomalies in the eastern Pacific reached a historical high during late 2015 and early 2016. The powerful El Niño in 2016 and weak La Niña in 2017 in the Pacific significantly influenced SST in the Indian Ocean; they were followed by strong negative and weak positive IOD events in 2016 and 2017, respectively. Previous research demonstrated the significant climatic effects of IOD, including the severe East Africa floods that occurred during two extreme positive IOD events in 1994 [30] and 1997 [12,31]. In this study, two contrasting years (2016 and 2017) were chosen to investigate the impact of ENSO and IOD on SST variability in the tropical Indian Ocean (TIO) using various satellite datasets. The dynamics of these two years were unusual and were not thoroughly explored. In that two-year period, a strong El Niño was followed by an extreme negative IOD (in 2016), and a weak La Niña was followed by a positive IOD (in 2017). El Niño in the Pacific Ocean usually favors positive IOD in the Indian Ocean, while La Niña favors negative IOD. These extreme events had a significant impact on Indian Ocean SST variability; they caused significant flooding and above-average rainfall in many parts of Australia, Indonesia, and Bangladesh, and drought conditions in East Africa [29,32]. On the other hand, the positive IOD, in combination with a weak La Niña in 2017, was associated with major climate events, including dry summer conditions in much of Australia, above average rains in the Horn of Africa late in the year after Atmosphere 2021, 12, 587 3 of 20

an extended period of drought, and monsoon ﬂoods in the Indian subcontinent [33,34]. Another interesting feature of this study was its examination of the cooling/warming in the SST cycle during the summers of 2016/2017. The impact of this was analyzed using data on monsoon rainfall distribution in the Bay of Bengal (BoB) region. The results of these analyses will improve our understanding of the climate system, as we examined it from a better perspective. SST climatology is an essential prerequisite for the ocean modeling community. It may also be useful to study the climate dynamics affecting Atmosphere 2021, 12, x FOR PEER REVIEWdroughts, ﬂoods, severe rainfall events, etc. experienced by landmasses, particularly in the3 of 22 Indian subcontinent and East Africa. This, in turn, would ultimately serve those parts of society whose livelihoods rely on agriculture.

Figure 1. Yearly mean SST anomalies of world oceans in 2016 (a) and 2017 (b). The black boxes correspond to the Niño Figure3.4 1. areaYearly (5◦ meanS–5◦ N, SST 240–290 anomalies◦ E) in of the world East Pacific oceans Ocean. in 2016 Monthly (a) and average 2017 (b SST). The anomalies black boxes in Niño correspond 3.4 region to are the shown Niño 3.4 area (5°within S–5° theN, 240–290° map. E) in the East Pacific Ocean. Monthly average SST anomalies in Niño 3.4 region are shown within the map. The rest of the paper is structured as follows: Section2 describes details and analyses of various satellite datasets and measurement of ENSO and IOD indices. The evolution In this study, two contrasting years (2016 and 2017) were chosen to investigate the of the IOD in the Indian Ocean in 2016 and 2017 is presented in Section 3.1. Section 3.2 impactdiscusses of ENSO the analysis and IOD of on SST SST variability variability and relatedin the tropical mechanisms Indian in the Ocean equatorial (TIO) Indianusing var- iousOcean satellite and datasets. the Arabian The Sea dynamics region. Section of thes 3.3e two provides years the were precipitation unusual and variability were not and thor- oughlySSS circulationexplored. inIn 2016 that and two-year 2017. Section period,4 summarizes a strong El the Niño main was findings. followed by an extreme negative IOD (in 2016), and a weak La Niña was followed by a positive IOD (in 2017). El Niño2. in Materials the Pacific and Ocean Methods usually favors positive IOD in the Indian Ocean, while La Niña favors2.1. negative Study Area IOD. These extreme events had a significant impact on Indian Ocean SST variability;The they area undercaused investigation significant wasflooding situated and in theabove-average TIO. The physical rainfall extent in many of the TIOparts of Australia,is shown Indonesia, in Figure2 and. As Bangladesh, shown in the and figure, drou theght black conditions boxes indicate in East the Africa areas [29,32]. under On investigation: eastern tropical Indian Ocean (ETIO), western tropical Indian Ocean (WTIO), the other hand, the positive IOD, in combination with a weak La Niña in 2017, was asso- and the Arabian Sea (AS). The AS is the northwest part of the TIO, with land boundaries ciatedin thewith west, major north climate and east. events, It is including surrounded dry by summer India to conditions the east, Pakistan in much and of Iran Australia, to abovethe average north, and rains the Arabianin the Horn Peninsula of Africa to the la west.te in Thethe Gulfyear ofafter Oman an isextended situated inperiod the of drought,northwest and cornermonsoon of the floods AS. The in the sea connectsIndian su withbcontinent the Persian [33,34]. Gulf viaAnother the Gulf interesting of Oman fea- tureand of this the Straitstudy of was Hormuz. its examination The Gulf of Adenof the links cooling/warming it with the RedSea in the in the SST southwest. cycle during The the summersBoB is theof 2016/2017. Indian Ocean’s The northeastern impact of this extension, was analyzed surrounded using on thedata west on andmonsoon northwest rainfall distributionby India, onin the northBay of by Bengal Bangladesh, (BoB) and region on the. The east results by Myanmar of these and analyses the Andaman will improve and our Nicobarunderstanding Islands of of India. the climate system, as we examined it from a better perspective. SST climatology is an essential prerequisite for the ocean modeling community. It may also be useful to study the climate dynamics affecting droughts, floods, severe rainfall events, etc. experienced by landmasses, particularly in the Indian subcontinent and East Africa. This, in turn, would ultimately serve those parts of society whose livelihoods rely on agriculture. The rest of the paper is structured as follows: Section 2 describes details and analyses of various satellite datasets and measurement of ENSO and IOD indices. The evolution of the IOD in the Indian Ocean in 2016 and 2017 is presented in Section 3.1. Section 3.2 discusses the analysis of SST variability and related mechanisms in the equatorial Indian Ocean and the Arabian Sea region. Section 3.3 provides the precipitation variability and SSS circulation in 2016 and 2017. Section 4 summarizes the main findings.

2. Materials and Methods 2.1. Study Area

Atmosphere 2021, 12, x FOR PEER REVIEW 4 of 22

The area under investigation was situated in the TIO. The physical extent of the TIO is shown in Figure 2. As shown in the figure, the black boxes indicate the areas under investigation: eastern tropical Indian Ocean (ETIO), western tropical Indian Ocean (WTIO), and the Arabian Sea (AS). The AS is the northwest part of the TIO, with land boundaries in the west, north and east. It is surrounded by India to the east, Pakistan and Iran to the north, and the Arabian Peninsula to the west. The Gulf of Oman is situated in the northwest corner of the AS. The sea connects with the Persian Gulf via the Gulf of Oman and the Strait of Hormuz. The Gulf of Aden links it with the Red Sea in the southwest. The BoB is the Indian Ocean’s northeastern extension, surrounded on the west and Atmosphere 2021, 12, 587 northwest by India, on the north by Bangladesh, and on the east by Myanmar and4 ofthe 20 Andaman and Nicobar Islands of India.

Figure 2. Geographical extent of the TIO (23.5◦ S–23.5◦ N, 30–120◦ E). The black boxes represent Figure 2. Geographical extent of the TIO (23.5° S–23.5° N, 30–120° E). The black boxes represent the areas of study: ETIO (10◦ S–Eq, 90–110◦ E), WTIO (10◦ S–10◦ N, 50–70◦ E), AS (12–24◦ N, the areas◦ of study: ETIO (10°◦ S–Eq, 90–110°◦ E), WTIO (10° S–10° N, 50–70° E), AS (12–24° N, 55–75° E),55–75 and EBoB), and (10–20° BoB (10–20N, 80–100°N, 80–100E). The BoBE). Theregion BoB is regionincluded is includedin this study in this for studyanalysis for of analysis precipi- of tationprecipitation variability. variability. 2.2. Data Source and Analysis 2.2. Data Source and Analysis The dynamics of SST variability in the Indian Ocean are different for different seasons andThe regions. dynamics Therefore, of SST monthlyvariability mean in the of Indian SST dataOcean for are 2016 different and 2017for different were analyzed seasons andseparately regions. for Therefore, each region monthly in this mean paper. of SST The data monthly for 2016 average and 2017 SST andwere SST analyzed anomalies sep- aratelydata in for the each selected region regions in this werepaper. derived The monthly from NOAA average OISST SST and (Optimum SST anomalies Interpolation data in theSea selected Surface Temperature)regions were derived blended from product, NOAA Version OISST 2.1. (Optimum This product Interpolation included Sea satellite Sur- faceobservations Temperature) as well blended as advanced product, very Version high-resolution 2.1. This radiometerproduct incl (AVHRR)uded satellite and advanced observa- tionsmicrowave as well scanning as advanced data, very and high-resol was availableution from radiometer 1981 on, (AVHRR) with a 0.25 and◦ × advanced0.25◦ spatial mi- crowaveresolution scanning and daily data, temporal and was interval available [35–37 from]. The 1981 anomalies on, with of a SST 0.25° were × 0.25° measured spatial on reso- the lutionbasis of and the daily daily temporal mean climatology interval [35–37]. of 30 years The (1981–2010). anomalies of The SST analyses were measured of the data on were the basisperformed of the indaily MATLAB mean climatology R2017a. Before of 30 analysis, years (1981–2010). the data were The interpolatedanalyses of the to data eliminate were performedmissing data in fromMATLAB the NOAA R2017a. daily Before SST analysis, dataset. the data were interpolated to eliminate missingThe data monthly from meanthe NOAA sea air daily temperature SST dataset. (SAT) data were provided by NCEP Global DataThe Assimilation monthly mean System sea (GDAS), air temperature available (SAT) at 2.5 data◦ × were2.5◦ horizontalprovided by resolution. NCEP Global The Datatemporal Assimilation coverage System of SAT (GDAS), included available daily mean at 2.5° values × 2.5° horizontal from 1979 resolution. to the present The withtem- poralvarying coverage pressure of upSAT to included 12 levels daily from 1000mean to va 50lues mb from (millibars). 1979 toIn the this present study, with the monthlyvarying pressuremean SAT up data to 12 were levels computed from 1000 in the to 50 selected mb (millibars). regions at In 1000 this mb study, pressure the monthly level, or aboutmean SAT1 atmospheric data were pressure. computed In in the the tropical selected oceans, region thes at propagation 1000 mb pressure of the ﬂuctuations level, or about in the 1 atmosphericthermocline ispressure. a crucial In factor the tropical in maintaining oceans, thethe cyclespropagation of ENSO of andthe fluctuations IOD. The changes in the thermoclinein the thermocline is a crucial depths factor can in be maintaining represented the by thecycles changes of ENSO in the and depths IOD. The of an changes isotherm in layer. In this study, the depth data of the isotherm layer were provided by Global Ocean Data Assimilation System (GODAS; [38]). GODAS depends on continuous real-time data from the Global Ocean Observing System. For ILD, the criteria used was the depth range where the temperature of given depth (z) is within 0.8 ◦C of the surface temperature-i.e., ILD=depth where T(z) ≥ SST-∆T, and ∆T = 0.8 ◦C[39]. In addition, in this study, the precipitation data and wind speed at 10 m above sea surface (U10) were used. The monthly mean precipitation data (mm d−1) were derived from the Global Precipitation Climatology Project Version 2.3 (GPCPv2.3) for reanalysis. The GPCPv2.3 data were available with 2.5◦ × 2.5◦ horizontal resolution [40,41]. The long-term dataset of precipitation anomalies (1948–2019) were collected from NOAA precipitation reconstruction (PREC). The climatology of precipitation anomalies was based on time period of 1979–1998 over oceans. The Woods Hole Oceanographic Institution (WHOI) Objectively Analyzed Air Sea Fluxes (OAFlux) project provided the mean wind speed at Atmosphere 2021, 12, 587 5 of 20

10 m above sea surface datasets. The OAFlux is an ongoing research and development project for global air-sea ﬂuxes (http://oaﬂux.whoi.edu, accessed on 13 July 2020). The daily and monthly means of wind speed data were available at 0.25◦ × 0.25◦ horizontal resolution from July 1987 onward. The monthly mean wind speed data were computed in the selected regions for 2016–2017 and missing data were removed for better quality. The monthly mean Sea Surface Salinity (SSS) data were obtained from the SMAPv3 (Soil Moisture Active Passive version 3; [42]), and are accessible online at (www.remss.com/ missions/smap, accessed on 18 Septemer 2020). Although SMAP was designed to measure space soil moisture, its L-band radiometer can also be used to measure SSS.

2.3. Measurement of Oceanic Niño Index The oceanic Niño index (ONI) is a key oceanic variable that describes the El Niño/La Niña phases in the Pacific Ocean. The two phases last several months each, typically occurring with varying intensity per period every few years. El Niño and La Niña are phenomena in the tropical Pacific Ocean described as five consecutive three-month running means of SST anomalies in the Niño 3.4 area (5◦ S–5◦ N, 170–120◦ W) that is above or below the threshold of +0.5 ◦C or −0.5 ◦C, respectively. This standard of measure is known as the ONI. The selection criteria of the El Niño/La Niña events were consistent with NOAA Oceanic Niño Index, in which the most recent three-month average SST anomalies in the Niño 3.4 area are considered. If the area is more than 0.5 ◦C above or below average for that period, El Niño or La Niña conditions are considered to be in progress. Here, we used data from NOAA Extended Reconstructed Sea Surface Temperature version 5 (ERSSTv5; [43]) to measure the strength of ONI from 1990 to 2020, as presented in Figure3. The data are automatically updated each month and are freely available online at NOAA’s website. The ONI values indicated that strong El Niño conditions developed during November 2014 ◦ Atmosphere 2021, 12, x FOR PEER REVIEWand persisted for 19 months before decaying in May 2016. The high value of ONI (>2.56 of 22C),

which was observed from November 2015 to February 2016, represents one of the most powerful El Niños since 1990 (Figure3).

FigureFigure 3. 3.TimeTime series series of the of the ENSO ENSO index index and and SST SSTanom anomaliesalies in Niño in Niño 3.4 Region. 3.4 Region. Warm Warm and cold and cold periodsperiods are are based based on on a athreshold threshold of of +/ +/−0.5−0.5 °C ◦forC forthe the ONI ONI (3 month (3 month running running mean mean of ofSST SST anomalies anomalies in inthe the Niño Niño 3.4 3.4 Region) Region) and and represented represented by by red red and and blue blue colors, colors, respecti respectively.vely. Dotted Dotted lines lines repre- represent sentstandard standard deviations deviations of theof the series. series. The The 2015–2016 2015–2016 El El Niño Niño was was the the ﬁrst first powerful powerful El El Niño Niño of of the the 21st 21st century and one of the three strongest El Niño events since 1950. century and one of the three strongest El Niño events since 1950. 2.4. Measurement of Dipole Mode Index The dipole mode index (DMI) is a key oceanic parameter that describes the strength of positive/negative IOD phases in the Indian Ocean. As shown in Figure 4a,b, average SST anomalies were estimated for boxes in the ETIO—bounded by (10° S–Eq, 90–110° E)— and WTIO—bounded by (10° S–10° N, 50–70° E)—respectively [11]. In an IOD year, the DMI is expected to be higher than one standard deviation and should remain so for 3 to 4 months. The west-east SST anomalies are positive during a positive IOD year and vice versa. The DMI has been widely used in IOD studies examining its mechanism [11,44,45], predictability [46–50], and effect on climate [51,52]. Here, the DMI was obtained with the averaged SST anomaly of ERSST.V5 datasets [43], from 1990 to 2020, with solid red and blue bars for positive and negative IOD, respectively (Figure 4c). DMI indicated that a strong negative IOD event appeared between June and October 2016, with abnormally warm SST in the ETIO and relatively cool conditions in the WTIO. According to the NOAA OISSTv2 datasets [53], it was the strongest negative IOD event since 1980.

Atmosphere 2021, 12, 587 6 of 20

2.4. Measurement of Dipole Mode Index The dipole mode index (DMI) is a key oceanic parameter that describes the strength of positive/negative IOD phases in the Indian Ocean. As shown in Figure4a,b, average SST anomalies were estimated for boxes in the ETIO—bounded by (10◦ S–Eq, 90–110◦ E)— and WTIO—bounded by (10◦ S–10◦ N, 50–70◦ E)—respectively [11]. In an IOD year, the DMI is expected to be higher than one standard deviation and should remain so for 3 to 4 months. The west-east SST anomalies are positive during a positive IOD year and vice versa. The DMI has been widely used in IOD studies examining its mechanism [11,44,45], predictability [46–50], and effect on climate [51,52]. Here, the DMI was obtained with the averaged SST anomaly of ERSST.V5 datasets [43], from 1990 to 2020, with solid red and blue bars for positive and negative IOD, respectively (Figure4c). DMI indicated that a Atmosphere 2021, 12, x FOR PEER REVIEW 7 of 22 strong negative IOD event appeared between June and October 2016, with abnormally warm SST in the ETIO and relatively cool conditions in the WTIO. According to the NOAA OISSTv2 datasets [53], it was the strongest negative IOD event since 1980.

Figure 4. Time series of average SST anomalies in the (a) ETIO (10◦ S–Eq, 90–110◦ E), (b) WTIO (10◦ S–10◦ N, 50–70◦ E), Figure 4. Timeand ( cseries) IOD index of average (DMI=WTIO SST− anomaliesETIO). Data werein the derived (a) ETIO fromExtended (10° S–Eq, Reconstructed 90–110° E), Sea Surface(b) WTIO Temperature (10° S–10° version N, 50–70° E), and (c) IOD5 index (ERSST.V5) (DMI=WTIO from 1990− toETIO). 2020. The Data red were arrows derived indicate from that negative Extended IOD Reconstructed events occurred duringSea Surface El Niño. Temperature The weak version 5 (ERSST.V5)positive from IOD 1990 event to in 2020. 2017 isThe indicated red arrows by black indicate arrow. that negative IOD events occurred during El Niño. The weak positive IOD event in 2017 is indicated by black arrow.

3. Results and Discussion 3.1. Evolution of Dipole Structure in 2016 and 2017 Figure 5a shows the development of the strong IOD in 2016, with the western equatorial Indian Ocean being unusually cold and the eastern Indian Ocean being unusually warm [11]. The two black boxes are the WTIO and ETIO in the equatorial Indian Ocean. Following the El Niño event in the Pacific, the Indian Ocean experienced frequent warming as a result of suppressed atmospheric convection and enhanced surface energy flux [54]. This basin-wide warming was noted from January–April 2016, in which the positive SST anomalies persisted in both poles of the equatorial Indian Ocean. The SST anomalies over the western Indian Ocean fell to 0.5 °C below average after the quick demise of the powerful 2015/16 El Niño in April 2016 (Figure 5a).

Atmosphere 2021, 12, 587 7 of 20

3. Results and Discussion 3.1. Evolution of Dipole Structure in 2016 and 2017 Figure5a shows the development of the strong IOD in 2016, with the western equatorial Indian Ocean being unusually cold and the eastern Indian Ocean being unusually warm [11]. The two black boxes are the WTIO and ETIO in the equatorial Indian Ocean. Following the El Niño event in the Paciﬁc, the Indian Ocean experienced frequent warming as a result of suppressed atmospheric convection and enhanced surface energy ﬂux [54]. This basin-wide warming was noted from January–April 2016, in which the positive SST Atmosphere 2021, 12, x FOR PEER REVIEWanomalies persisted in both poles of the equatorial Indian Ocean. The SST anomalies8 of 22 over the western Indian Ocean fell to 0.5 ◦C below average after the quick demise of the powerful 2015/16 El Niño in April 2016 (Figure5a).

Figure 5. EvolutionEvolution of of SST SST anomalies anomalies (°C) (◦C) in in the the TIO TIO in in 2016 2016 and and 2017 2017 with with respect respect to to 1981–2010 1981–2010 reference reference period, period, ( (aa)) yearly yearly mean of of SST SST anomalies anomalies in in 2016, 2016, (b (b) )yearly yearly mean mean of of SST SST anomalies anomalies in in 2017. 2017. Th Thee monthly monthly distribution distribution of ofSST SST anomalies anomalies in 2016in 2016 and and 2017 2017 is shown is shown within within the themap. map. The The two two boxes boxes represen representt WTIO WTIO (10° (10 S–10°◦ S–10 N,◦ 50–70°N, 50–70 E) ◦andE) andETIO ETIO (10° (10S–Eq,◦ S–Eq, 90– 110° E). Data obtained from NOAA OISSTv2.1. 90–110◦ E). Data obtained from NOAA OISSTv2.1.

However,However, the the SST anomalies at the ETIO remained warmer than average during 2016,2016, which which is is consistent with with the persistent negative IOD warming in 1992, 1998, and 2010,2010, following the ElEl NiñoNiño eventsevents ofof 1991/1992, 1991/1992, 1997/1998,1997/1998, and 2009/2010,2009/2010, respectively (Figures(Figures 33 andand5 5a).a). The The ETIO ETIO warmed warmed up up again again during during the the summer summer of of 2016, 2016, hitting hitting 1 1 ◦°CC aboveabove average average in in September September 2016, 2016, establishing establishing the most most extreme extreme negative IOD event since 1980 [32]—and [32]—and in in the the last last 63 63 years years during during June June to to September September [55]. [55 The]. The DMI DMI exceeded exceeded −1 °C,−1 ◦aC historical, a historical low, low, in inJuly July 2016 2016 (Figure (Figure 4c).4c). The The temporal temporal variation variation of SSTSST anomaliesanomalies showedshowed that the 20162016 dipoledipole beganbegan during during early early summer summer in in June June and and decayed decayed in in November. Novem- ber.In contrast, In contrast, a weak a weak positive positive IOD IOD episode episode existed existed for for most most of of 2017, 2017, withwith SSTSST anomalies marginallymarginally below below the the 1981–2010 1981–2010 average average over over the the ETIO ETIO and and above above the the WTIO WTIO (Figure (Figure 55b).b ). ColdCold SST SST anomalies suppressed suppressed atmospheric atmospheric convection convection in in the the east pole whereas warm SSTSST anomalies anomalies enhanced enhanced convection convection in in the the we westst pole. pole. This This anomalous anomalous state state of of the the ocean- ocean- atmosphereatmosphere system is referred to as a positi positiveve IOD [11]. [11]. The The unusual unusual warming warming and and cooling cooling inin SST, SST, which occurred in 2016 and 2017, had had a strong influence inﬂuence on the climate in the surroundingsurrounding sea sea and and land land areas areas [29,33]. [29,33].

3.2.3.2. SST SST Variability Variability and and Associ Associatedated Mechanisms Mechanisms in in the the TIO TIO 3.2.1. The Equatorial Indian Ocean Region 3.2.1. The Equatorial Indian Ocean Region SST variability in the TIO during 2016–2017 is shown in Figure6. As shown in the SST variability in the TIO during 2016–2017 is shown in Figure 6. As shown in the ﬁgure, the seasonal cycle of SST in the western and eastern poles was not consistent, which figure, the seasonal cycle of SST in the western and eastern poles was not consistent, which usually indicates variability. In the WTIO, the SST cooled in the summer and predominantly warmed in the pre-monsoon months. The pre-monsoon warming was most likely due to clear skies, reduced winds, and an increase in solar insolation (absorbed in the upper layer), which induced thermal stratification and suppressed turbulent mixing. This was analogous to the suppression of turbulent kinetic energy [56], which caused the thermocline depth to rise with a shallow isothermal layer depth (ILD), which was reduced (~30 to 40 m) in April–May (see Figure 7). On examining seasonal averages, the pre-monsoon warming periods were dissimilar and varied in both 2016 and 2017. Significant variations in the SST cycle were observed during the pre-monsoon; SST was 1 °C warmer in 2016 than in 2017. The SST reached about 31 °C in April 2016 (versus about 30 °C in 2017). SST warming in 2016 was consistent with the extreme El Niño that began in October 2015 and decayed in May 2016. The powerful El Niño significantly influenced the equatorial

Atmosphere 2021, 12, x FOR PEER REVIEW 9 of 22

Pacific and Indian oceans in which positive SST anomalies persisted until May 2016. The surface wind at 10 m above sea surface (U10) and SAT at 1 atmospheric pressure were around 4 m s−1 and 28.5 °C, respectively (see Figure 8a for surface winds and Figure 9a for SAT). However, the wind speed showed an increasing trend as the monsoon set in. SST grew cool and hit 27 °C in summer 2016 (versus 28 °C in 2017) from July to September, Atmosphere 2021, 12, 587 8 of 20 due to strong surface winds (wind > 8 m s−1) and cooling in the air temperature (<26 °C). During this period, the effect of wind force was high (as compared to the pre-monsoon period), and the turbulent mixing caused the cool surface waters to sink into a deeper layer,usually resulting indicates in variability.a deeper Inthermocline the WTIO, the (up SST to cooled 70 m). in theIn summercomparison, and predominantly the temperature of warmed in the pre-monsoon months. The pre-monsoon warming was most likely due to the sea surface in this region was nearly 1 °C warmer in 2017 than in 2016, which was clear skies, reduced winds, and an increase in solar insolation (absorbed in the upper layer), consistentwhich induced with our thermal analysis stratification of positive and suppressed and negative turbulent IOD mixing. events This in was 2017 analogous and 2016 with correspondingto the suppression high ofand turbulent low SSTs kinetic in the energy west [56 pole.], which The caused associated the thermocline anomalously depth westerly surfaceto rise wind with stress a shallow caused isothermal upwelling layer in depth the (ILD),WTIO. which This was upwelling reduced pulled (~30 to 40cool m) subsurface in watersApril–May upward, (see increasing Figure7). On the examining zonal SST seasonal gradient averages, between the west pre-monsoon and east, warming and reinforcing the periodsnegative were IOD dissimilar pattern. and Significant varied in both correl 2016ation and 2017. was Significantobserved variationsbetween in the the SST SST and the cycle were observed during the pre-monsoon; SST was 1 ◦C warmer in 2016 than in 2017. wind,The representing SST reached about a strong 31 ◦C coupling in April 2016 between (versus the about ocean 30 ◦ Cand in 2017).the atmosphere SST warming in in the west pole.2016 These was consistentvariability with patterns the extreme highlight El Niño the that fundamental began in October role 2015of wind and decayedon the SST in in the westernMay 2016.basin. The In powerful conclusion, El Niño the significantly mechanisms influenced describing the equatorial the air-sea Pacific feedback and Indian (as well as the oceansrelationship in which between positive SSTwind anomalies and thermo persisteddynamic until May parameters) 2016. The surface showed wind that at SST is 10 m above sea surface (U10) and SAT at 1 atmospheric pressure were around 4 m s−1 and mainly ◦driven by wind force in this region. 28.5LikeC, the respectively western (seepole, Figure the eastern8a for surface pole also winds experiences and Figure9 SSTa for cooling SAT). However, in the monsoon the wind speed showed an increasing trend as the monsoon set in. SST grew cool and seasonhit 27 and◦C warming in summer in 2016 the (versus pre-monsoon 28 ◦C in 2017)season. from As July can to be September, seen in dueFigure to strong 6a,b, the SST peakedsurface in windsApril (windin both > 8years m s−1 and) and crossed cooling in30.5 the °C air in temperature 2016 (versus (<26 29◦C). °C During in 2017). this But the easternperiod, pole the played effect of a wind rather force different was high role (as comparedduring the to thesequence pre-monsoon of negative period), and and positive IODthe events. turbulent A mixingslight causeddecrease the coolin the surface SST waterswas noted to sink in into this a deeper region layer, during resulting the inmonsoon perioda deeper in which thermocline the SST (up reached to 70 m). 29 In °C comparison, in 2016 (and the 28 temperature °C in 2017) of thein July–September. sea surface in The this region was nearly 1 ◦C warmer in 2017 than in 2016, which was consistent with our SAT in this region remained high in 2016 over the entire period, as opposed to 2017, in analysis of positive and negative IOD events in 2017 and 2016 with corresponding high and whichlow SAT SSTs inwas the more west pole.pronounced The associated during anomalously the first half westerly of the surface year. wind It is stressinteresting caused to high- lightupwelling that SAT in was the WTIO. similar This in upwellingboth poles pulled (except cool during subsurface the waterspre-monsoon upward, increasingperiod) throughout the2017. zonal However, SST gradient the betweenSAT over west the and eastern east, andbasin reinforcing was warmer the negative than in IOD the pattern. west by more thanSignificant 1.5 °C in correlation the summer was of observed 2016 (Figure between 9a,b). the SST This and warming the wind, representingin the air temperature a strong was consistentcoupling with between the thestrong ocean El and Niño the which atmosphere occurred in the westin the pole. eastern These Pacific variability in early patterns 2016. That highlight the fundamental role of wind on the SST in the western basin. In conclusion, the El Niño impacted the eastern equatorial Indian Ocean region and triggered an unusual mechanisms describing the air-sea feedback (as well as the relationship between wind and strongthermodynamic negative IOD parameters) in the Indian showed Ocean that SST in is the mainly summer driven of by 2016. wind force in this region.

Figure 6.Figure Yearly 6. Yearly mean mean SST SSTin the in theTIO TIO based based on on NOAA NOAA OISSTv2.0 for for 2016 2016 (a) ( anda) and 2017 2017 (b). Monthly(b). Monthly mean SSTmean cycle SST of thecycle of the WTIO andWTIO ETIO and ETIOregions regions is depicted is depicted on on the the map. map.

Atmosphere 2021Atmosphere, 12, x2021 FOR, 12 PEER, 587 REVIEW 9 of 20 10 of 22

Atmosphere 2021, 12, x FOR PEER REVIEW 11 of 22

Figure 7. As in Figure6, but for isothermal layer depth (ILD) in 2016 ( a) and 2017 (b). The changes in the thermocline depth Figure 7. As in Figure 6, but for isothermal layer depth (ILD) in 2016 (a) and 2017 (b). The changes in the thermocline depth areare reflected reﬂected by by the the changes in in the the ILD. ILD. Data Data obtained obtained from GODAS.from GODAS.

Figure 8. FigureAs in Figure 8. As in 6, Figure but6 for, but neutral for neutral wind wind speed speed at at 10 10 m m above above sea sea surface surface (U10) (U10) in 2016 in 2016 ( a) and (a) 2017 and ( b2017). Data (b). provided Data provided by the WHOIby the OAFlux WHOI OAFlux project. project.

The strong negative IOD event led to unusual warming in the eastern pole during the summer of 2016. The SST was higher (by 2 °C) than in the western pole, which marked the east-west asymmetry in the Indian Ocean. A close relation between SST and SAT was noted for both 2016 and 2017 and in both poles. However, SAT cooled by around 2 °C more than SST and was more pronounced during the pre-monsoon season (April–May). During the negative IOD event that occurred in the summer of 2016, anomalously high SSTs appeared off the Sumatran coast, inducing local convection and western wind anomalies through the Bjerknes feedback [57]. The surface wind patterns were similar in the east and west poles in 2017; however, weak surface wind (wind < 6 m s−1) in the eastern basin and strong surface wind in the western basin (wind > 8 m s−1) appeared more pronounced from May to September 2016 (Figure 8a,b). This anomalous increase in wind speed reduced SST (up to 2 °C) in the western basin in 2016. This basin-wide variability in SST that appeared in summer 2016 highlighted the fundamental role of wind anomalies that caused the extreme negative IOD event in the Indian Ocean. The strong convection that was confined to the (warm) eastern equatorial Indian Ocean was associated with the east-west SST anomalies that drove a series of downwelling Kelvin waves that deepened the thermocline in the east. As a result, the deepening thermocline reduced upwelling efficiency and warm SSTs in the east. These effects are noted in Figure 7a, in which a deepened isotherm layer was observed in 2016 off the Sumatran coast due to a weak upwelling event. The deep ILD crossed 100 m along the Sumatran coast during the monsoon and post-monsoon periods (not shown). A weak positive IOD episode existed in most of 2017, with SST anomalies marginally below the 1981–2010 average over the eastern and above the western equatorial Indian Ocean (Figure 5b). Cold SST anomalies suppressed atmospheric convection in the eastern pole, while warm SST anomalies enhanced convection in the western pole. Winds blew westward over the equatorial Indian Ocean and from the southwest off the Sumatran coast, favoring coastal upwelling. This upwelling led to a reduction in SST of −1 °C. The thermocline depth tilted upward (up to 60 m) in the summer of 2017 in the east. In summary, negative IOD phases were characterized by warmer SST anomalies, enhanced convection, and deeper thermoclines in the

Atmosphere 2021, 12, x FOR PEER REVIEW 12 of 22

Atmosphere 2021, 12, 587 10 of 20 eastern pole, while positive IOD phases were characterized by cooler SST anomalies, reduced convection, and shallower thermoclines in the western pole.

Figure 9. Monthly mean SAT variability at 1 atmospheric pressure in the equatorial tropical Indian Figure 9. Monthly mean SAT variability at 1 atmospheric pressure in the equatorial tropical Indian Ocean in 2016 (a) and 2017 (b). Data provided by NCEP Global Data Assimilation System (GDAS) Ocean in 2016 (a) and◦ 2017 ◦(b). Data provided by NCEP Global Data Assimilation System (GDAS) andand available available at at 2.5° 2.5 ×× 2.5°2.5 horizontalhorizontal resolution. resolution. Like the western pole, the eastern pole also experiences SST cooling in the monsoon 3.2.2. The Arabian Sea Region season and warming in the pre-monsoon season. As can be seen in Figure6a,b, the SST peakedThe inAS April is a innorthwestern both years and part crossed of the 30.5TIO ◦andC in has 2016 a (versusmonsoon 29 climate.◦C in 2017). The Butstrong the seasonalityeastern pole of played the SST a rather is the different result of role the during combined the sequence effects of of oceanic negative and and atmospheric positive IOD processesevents. A at slight the air-sea decrease interface in the SST(mainly was notedcontrolled in this by region seasonal during changes the monsoon in incoming period solar in radiation)which the and SST oceanic reached and 29 ◦ atmosphericC in 2016 (and circulation. 28 ◦C in 2017) The inannual July–September. cycle of monsoons The SAT mainly in this exhibitsregion remained bimodal distri high inbution 2016 and over significantly the entire period, affects as th opposede upper tothermal 2017, instructure, which SAT which was ismore mainly pronounced responsible during for regional the ﬁrst half circulation of the year. and It heat/salt is interesting transport to highlight in the thatArabian SAT Sea was [17].similar The in annual both poles cycle (except of SST duringin the theAS pre-monsoonconsists primarily period) of four throughout stages: 2017.(1) a warming However, stagethe SAT from over about the February eastern basin to May; was (2) warmer cooling than from in theMay west to August; by more (3) than warming 1.5 ◦C infrom the Augustsummer to of October; 2016 (Figure and 9(4)a,b). cooling This warmingfrom October in the to air January. temperature This pattern was consistent is in contrast with the to thestrong annual El Niño cycle which of the occurredSST in most in the other eastern regions Paciﬁc of the in earlyworld 2016. ocean, That which El Niño display impacted only twothe easternphases: equatorialwarming during Indian pre-summer Ocean region and and summer; triggered and an unusualcooling during strong negativeautumn and IOD winter.in the IndianAll available Ocean inevidence the summer suggests of 2016. that this unusual behavior of the AS is due to the influenceThe strongof the southwest negative IOD monsoon event led (summer to unusual season) warming that dominates in the eastern the poleAS during during the the northernsummer ofhemispheric 2016. The SSTsummer. was higher The energetic (by 2 ◦C) thancirculation in the westernof wind pole, during which this marked period theis knowneast-west to asymmetryhave an effect in the on Indian the SST. Ocean. In the A closecoastal relation regions, between upwelling SST and typically SAT was occurs, noted whichfor both brings 2016 up and colder 2017 andwater in and both then poles. spreads However, offshore SAT [58], cooled whereas by around in the 2 open◦C more sea, than the lossSST of and energy wasmore and heat pronounced on the surface during lowe the pre-monsoonrs the SST. These season changes (April–May). in SSTs Duringand wind the negative IOD event that occurred in the summer of 2016, anomalously high SSTs appeared off the Sumatran coast, inducing local convection and western wind anomalies through the Bjerknes feedback [57]. The surface wind patterns were similar in the east and west poles in 2017; however, weak surface wind (wind < 6 m s−1) in the eastern basin and strong

Atmosphere 2021, 12, 587 11 of 20

surface wind in the western basin (wind > 8 m s−1) appeared more pronounced from May to September 2016 (Figure8a,b). This anomalous increase in wind speed reduced SST (up to 2 ◦C) in the western basin in 2016. This basin-wide variability in SST that appeared in summer 2016 highlighted the fundamental role of wind anomalies that caused the extreme negative IOD event in the Indian Ocean. The strong convection that was conﬁned to the (warm) eastern equatorial Indian Ocean was associated with the east-west SST anomalies that drove a series of downwelling Kelvin waves that deepened the thermocline in the east. As a result, the deepening thermocline reduced upwelling efﬁciency and warm SSTs in the east. These effects are noted in Figure7a, in which a deepened isotherm layer was observed in 2016 off the Sumatran coast due to a weak upwelling event. The deep ILD crossed 100 m along the Sumatran coast during the monsoon and post-monsoon periods (not shown). A weak positive IOD episode existed in most of 2017, with SST anomalies marginally below the 1981–2010 average over the eastern and above the western equatorial Indian Ocean (Figure5b). Cold SST anomalies suppressed atmospheric convection in the eastern pole, while warm SST anomalies enhanced convection in the western pole. Winds blew westward over the equatorial Indian Ocean and from the southwest off the Sumatran coast, favoring coastal upwelling. This upwelling led to a reduction in SST of −1 ◦C. The thermocline depth tilted upward (up to 60 m) in the summer of 2017 in the east. In summary, negative IOD phases were characterized by warmer SST anomalies, enhanced convection, and deeper thermoclines in the eastern pole, while positive IOD phases were characterized by cooler SST anomalies, reduced convection, and shallower thermoclines in the western pole.

3.2.2. The Arabian Sea Region The AS is a northwestern part of the TIO and has a monsoon climate. The strong seasonality of the SST is the result of the combined effects of oceanic and atmospheric processes at the air-sea interface (mainly controlled by seasonal changes in incoming solar radiation) and oceanic and atmospheric circulation. The annual cycle of monsoons mainly exhibits bimodal distribution and significantly affects the upper thermal structure, which is mainly responsible for regional circulation and heat/salt transport in the Arabian Sea [17]. The annual cycle of SST in the AS consists primarily of four stages: (1) a warming stage from about February to May; (2) cooling from May to August; (3) warming from August to October; and (4) cooling from October to January. This pattern is in contrast to the annual cycle of the SST in most other regions of the world ocean, which display only two phases: warming during pre-summer and summer; and cooling during autumn and winter. All available evidence suggests that this unusual behavior of the AS is due to the influence of the southwest monsoon (summer season) that dominates the AS during the northern hemispheric summer. The energetic circulation of wind during this period is known to have an effect on the SST. In the coastal regions, upwelling typically occurs, which brings up colder water and then spreads offshore [58], whereas in the open sea, the loss of energy and heat on the surface lowers the SST. These changes in SSTs and wind over the ocean may have an impact on the weather and climate of the adjacent landmasses [8,59]. The seasonal cycles of SST, SAT, wind speed and ILD over the AS in 2016 and 2017 are shown in Figure 10a–d, respectively. As shown in Figure 10a, the SST in this region experienced a semiannual cycle in SST circulation, where low SST occurred in both the summer and winter seasons. The winter minimum temperature reached 26 ◦C in February. However, the SST warmed during the pre-monsoon season and crossed 30 ◦C in May. It can be seen that SST was warmer by almost 4 ◦C. The SAT also peaked, reaching 31 ◦C in May (Figure 10b). The available climatology data suggest that the skies over the AS were clear prior to the onset of the summer monsoon, which means that it received a large amount of heat from solar radiation [60]. The winds were weak (Figure 10c) and therefore the latent heat loss was small, resulting in a large heat gain by the AS [61]. The pre-monsoon warming induced thermal stratification and suppressed turbulent mixing, causing the thermocline to rise with a shallow depth of isotherm layer (ILD < 30 m) in May Atmosphere 2021, 12, 587 12 of 20

(Figure 10d). In winter, the wind effect was weak compared to summer; convective mixing caused the cool surface waters to sink into a deeper layer, with larger ILD amplitude that reached 90 m in February. Interestingly, the seasonal cycle of SST distribution showed a slight increase in SST in 2016 as compared to 2017. This warming in the SST in 2016 may have been the result of large-scale atmospheric forcing linked to the presence of well-known Atmosphere 2021, 12, x FOR PEER REVIEW 14 of 22 strong El Niño conditions that appeared in the Paciﬁc in early 2016, impacting the tropical Indian Ocean.

FigureFigure 10. 10. MonthlyMonthly mean mean (a) sea (a) surface sea surface temperature temperature (SST), (b (SST),) sea air (b temperature) sea air temperature (SAT) at 1 at- (SAT) at 1 atmo- mosphericspheric pressure, pressure, ( c)) neutral neutral wind wind at 10 at m 10 above m above sea surface sea surface (U10), (U10),and (d) andisothermal (d) isothermal layer layer depth depth (ILD) in the Arabian Sea region (12–24° N, 50–75° E) during 2016 and 2017. (ILD) in the Arabian Sea region (12–24◦ N, 50–75◦ E) during 2016 and 2017. A remarkable feature of the AS circulation is the existence of the strong upwelling along the Omani-Arabian coast that occurs during the summer season. A typical aspect of

Atmosphere 2021, 12, 587 13 of 20

The summer monsoon in the AS expresses some of the strongest and most balanced wind forces. The period is characterized by strong winds, moist air, and a decrease in solar insulation due to cloud cover. The surface wind showed an increasing trend as the monsoon began (average wind speed reached 10 m s−1 in July 2017) and the summer minimum temperature dropped to around 27 ◦C in 2016 (versus 28 ◦C in 2017). This fall in SST during summer 2016 may be linked to the interannual variability of extreme negative IOD events that appeared in summer 2016 and impacted the western Indian Ocean with negative SST anomalies. Interestingly, mean SAT patterns were similar to SST but lower in amplitude—except for the monsoon season, in which most of the AS regions had SATs above 28 ◦C. During this period, the effect of wind force was high as compared to the pre-monsoon, and the turbulent mixing caused cool surface waters to sink into a deeper layer, resulting in a deeper thermocline. The average ILD reached 60 m in summer 2016 (Figure 10d). A remarkable feature of the AS circulation is the existence of the strong upwelling along the Omani-Arabian coast that occurs during the summer season. A typical aspect of the upwelling mechanism is the presence of an undercurrent flowing opposite to the surface current [62]. The interesting element is that the cooling (or warming) of the SST occurs due to strong (or weak) upwelling variations that occur in the same region. The area of the undercurrent is characterized by isotherm downsloping at subsurface levels and upsloping of isotherms at surface levels toward the coast. It is important to mention here that the weak upwelling that occurs along the Omani-Arabian coast in summer usually affects and warms the SST, leading to an increased rainfall in the west coast of India and western BoB [18]. During pre-summer and summer seasons, SST in the western AS region is very sensitive to upwelling fluctuations due to the shallow mixed layer (~15–30 m) [16]. SST in the AS was cooler by almost 1 ◦C from July to October in 2016 (as compared to 2017). Cooling and warming of the SST is primarily due to strong or weak (respectively) upwellings that occur along the Omani-Arabian coast near the Ras al Hadd region between 22◦ N, 60◦ E band in late summer and early autumn in 2016 and 2017, respectively. In the upwelling, water of about 25 ◦C may have originated below the pycnocline. It was, therefore, colder than the surface water (which was 30 ◦C or more) and formed a shallow thermocline with reduced isothermal depth. This summer cooling in the northern AS caused dramatic changes in thermocline characteristics, where a small change in their strength might have a significant effect on the SST and hence on the monsoon precipitation [8,59]. The strong and weak upwellings can also be seen in Figure5a,b, in which negative and positive SST anomalies prevailed along the Omani-Arabian coast in 2016 and 2017, respectively. During the 2016 monsoon season, high surface winds were noted in the AS, except in July (Figure 10c). The high wind speed increased the upwelling efficiency along the Omani-Arabian coast, forming a deep thermocline. In 2017, conversely, low surface winds were noted for the same period, and the lower wind speeds decrease the upwelling efficiency, forming a shallow thermocline. The depth of the thermocline varied from the sea surface, depending on multiple factors, including tide conditions and direction, the speed of coastal winds, and currents. These remarkable changes in SST in 2016 and 2017 (and their effect on summer precipitation in the BoB) were analyzed. The following section discusses that analysis.

3.3. Precipitation Variability and SSS Circulation As discussed earlier, SSTs in the Indian Ocean were significantly impacted by El Niño/La Niña phases in the Pacific Ocean, and by the evolution of the positive/negative IOD episodes in the tropical equatorial Indian Ocean in 2016–2017. During an El Niño, the Niño 3.4 region gets relatively warmer (as was observed until April in 2016). This may have an adverse impact on the Indian monsoon. In the late 1800s, Gilbert Walker investigated drought in India and determined that drought conditions were connected to shifts in ENSO. Several studies in the past have shown the relationship between SST in the TIO and the atmosphere in the surrounding regions [10,63–66]. Significant climatic impacts Atmosphere 2021, 12, 587 14 of 20

of extreme positive IOD events on severe East African floods have also been extensively studied [12,30,31]. According to a World Meteorological Organization report [29], the negative IOD event that occurred in June–September 2016 was associated with above average rainfall in many parts of Australia, Indonesia, and Bangladesh, as well as dry conditions in East Africa. On the other hand, there were significant weather and climate events in 2017, including dry conditions in summer in most of Australia, above average rainfall in the Horn of Africa late in the year after an extended period of drought, and monsoon floods in the Indian subcontinent [33]. This study further elucidated the impact of ENSO and IOD on precipitation variability in the Indian Ocean and surrounding areas in 2016 and 2017. As discussed in the previous section, the weak and strong upwellings Atmosphere 2021, 12, x FOR PEER REVIEW that occurred along the Omani-Arabian coast led to respective warming and cooling in 16 of 22 the SST, which affected precipitation and rainfall activity in the western coast of India and the BoB. Time series of precipitation anomalies (1948–2019) in the WTIO, ETIO, and BoB regions and 2017.are shown The inmonthly Figures 11 mean–13, respectively. precipitation The BoBanomalies region is in added 2016 to and confirm 2017 the are impact presented withinof the unusualfigures. warming The climatology and cooling is in based the SST on on a precipitation20-year (1979–1998) variability reference in 2016 and period. The associated2017. The monthly precipitation mean precipitation effects on anomalies SSS variability in 2016 and are 2017 shown are presented in Figure within 14. It the became figures. The climatology is based on a 20-year (1979–1998) reference period. The associated evidentprecipitation that the extreme effects on El SSS Niño variability in the are Pacific shown during in Figure early 14. It2016 became directly evident affected that the rainfall distribution,extreme Elwith Niño above in the normal Pacific during rainfall early over 2016 the directly ETIO affected and below rainfall normal distribution, rainfall with over the WTIO.above In the normal presence rainfall of overEl Niño, the ETIO there and was below an excess normal in rainfall rainfall over distribution the WTIO. that In the reached 10 mmpresence d−1 (5 ofmm El Niño,d−1) in there February was an in excess the east in rainfall pole distribution(versus 5 mm that d reached−1 in the 10 west mm dpole)−1 (not −1 −1 shown).(5 mm The d associated) in February drop in the in east SSS pole in these (versus re 5gions mm d is inshown the west in pole)Figu (notre 14a,b. shown). The The low SSS duringassociated early 2016 drop in in WTIO SSS in these and regionsETIO may is shown have in been Figure due 14a,b. to Theexcessive low SSS precipitation during early over 2016 in WTIO and ETIO may have been due to excessive precipitation over the evaporation. the evaporation.

FigureFigure 11. Time 11. Time series series (1948–2019) (1948–2019) of of precipitation precipitation anomalies anomalies in the in ETIOthe ETIO region. region. Datasets Datasets provided pro- videdby by NOAA NOAA precipitation precipitation reconstruction reconstruction (PREC). (PREC). The climatology The climatology is based on is a timebased period on a coveringtime period covering1979–1998. 1979–1998. The monthly The monthly distribution distribution of precipitation of precipitation anomalies inanomalies 2016 and 2017in 2016 is shown and 2017 within is shownthe within ﬁgure. the figure.

Atmosphere 2021, 12, x FOR PEER REVIEW 17 of 22

Atmosphere 2021, 12, 587 15 of 20

Atmosphere 2021, 12, x FOR PEER REVIEW 18 of 22

Figure 12. As in Figure 11, but for WTIO region. Figure 12. As in Figure 11, but for WTIO region.

Figure 13. As in Figure 11, but for BoB region. Figure 13. As in Figure 11, but for BoB region.

Summer 2016 revealed a different variability with increased precipitation in the eastern pole of the tropical equatorial Indian Ocean, and reduced precipitation in the western pole. Excessive precipitation in the ETIO during the summer of 2016 may have been associated with the extreme negative IOD phase in June–August that triggered flooding in various parts of India, Nepal, and Bangladesh [29]. As shown in Figure 12, there was a decrease in precipitation in the WTIO during the summer and autumn of 2016. Below normal precipitation (<2 mm d−1) in June–October was mainly due to the strong negative IOD in 2016, which brought drought conditions in East Africa and reduced EASR (East African Short Rains) in October–December (not shown). Recently, Lu et al. 2018 [32] investigated the reduction of EASR by 1 mm d−1 in 2016, which included a 50% reduction in normal rainfall in some regions. Interestingly, the corresponding effect—low precipitation on surface salinity stratification—was not observed in the WTIO. Instead, there was a small decline in SSS in 2016 compared to 2017. The rise in surface salinity in 2017 is most likely due to high evaporation, which could have occurred due to a warmer SST in 2017 in the WTIO (Figures 5 and 6). High evaporation adds more moisture to the atmosphere, contributing to the increased precipitation and flooding observed in East Africa by the end of 2017 [33]. In the AS region, the precipitation patterns were similar and no significant variations were noted over the entire period of 2016–2017 (not shown). However, the monthly average SSS remained high for the whole span of 2016 (compared to 2017) and was more pronounced in September (not shown). This was because SSTs in the AS covaried with the ONI observed in the Niño 3.4 region (5° S–5° N, 240–290° E). A positive SST anomaly was induced in the AS during the strong El Niño in the Pacific in 2016, which increased the evaporation rate and surface salinity due to fresh water loss. In the BoB, warm SST anomalies persisted over the central region from January to May 2016 followed by the strong El Niño of 2016 (Figure 5a, the monthly variation is not shown). However, the effect of high SST on the surface salinity was not detected in this region (Figure 14c). Instead, there was a small decline in SSS (up to 0.5) from January to

Atmosphere 2021, 12, x FOR PEER REVIEW 19 of 22

May in 2016, as compared to 2017. The dynamic behind this, however, could not be investigated in this study. As shown in Figure 13, precipitation patterns were almost identical in both years, with no significant differences except in June and September (in which positive precipitation anomalies were higher in 2017 than in 2016). The highest precipitation levels were observed during the monsoon season in both years, albeit with more in 2017. The corresponding effect—enhanced precipitation in the monsoon season—can be seen in Figure 14c, in which a slight decrease in SSS was observed in 2017, as compared to 2016. Atmosphere 2021, 12, 587 16 of 20 The low precipitation rate observed during the negative IOD and the high precipitation rate observed during the positive IOD in the BoB region are both consistent with the findings of Chanda et al. 2018 [67].

Figure 14. SSS variability in the TIO in the areas under investigation during 2016 and 2017: (a) ETIO (10◦ S–Eq, 90–110◦ E), ◦ ◦ ◦ ◦ ◦ Figure(b) WTIO 14. (10SSS variabilityS–10 N, 50–70 in theE), TIO and in (thec) BoB areas (10–20 underN, investig 80–100ationE). Dataduring obtained 2016 and from 2017: the ( SMAPv3a) ETIO (10° (Soil S–Eq, Moisture 90–110° Active E), (Passiveb) WTIO version (10° S–10° 3). N, 50–70° E), and (c) BoB (10–20° N, 80–100° E). Data obtained from the SMAPv3 (Soil Moisture Active Passive version 3). Summer 2016 revealed a different variability with increased precipitation in the eastern 4.pole Conclusions of the tropical equatorial Indian Ocean, and reduced precipitation in the western pole. Excessive precipitation in the ETIO during the summer of 2016 may have been associated In 2016, the strong El Niño in the Pacific Ocean induced an extreme negative IOD in with the extreme negative IOD phase in June–August that triggered flooding in various the Indian Ocean. The following year, a weak La Niña in the same region induced a weak parts of India, Nepal, and Bangladesh [29]. As shown in Figure 12, there was a decrease positive IOD. These unusual events had a major impact on the SST and were followed by in precipitation in the WTIO during the summer and autumn of 2016. Below normal above-average flooding in the Indian subcontinent and drought conditions in East Africa precipitation (<2 mm d−1) in June–October was mainly due to the strong negative IOD in [29,33]. The current study investigated the dynamics behind these events and their effect 2016, which brought drought conditions in East Africa and reduced EASR (East African on SST variability in the TIO. Short Rains) in October–December (not shown). Recently, Lu et al. 2018 [32] investigated the reductionIt was noted of EASR that during by 1 mm the d strong−1 in 2016, El Niño which in includedthe Pacific a 50%in early reduction 2016, the in normalIndian Oceanrainfall warmed in some due regions. to suppressed Interestingly, atmospheri the correspondingc convection effect—lowand positive precipitation SST anomalies on whichsurface persisted salinity stratification—wasin both the east and not west observed poles of inthe the Equatorial WTIO. Instead, Indian thereOcean. was Our a smallanal- ysesdecline indicated in SSS inthat 2016 high compared SSTs along to 2017. the Su Thematran rise in coast surface in the salinity summer in 2017 of 2016 is most induced likely localdue toconvection high evaporation, and western which wind could anomalies have occurredvia the Bjerknes due to feedback a warmer mechanism SST in 2017 [57]. in Thesethe WTIO a led (Figuresto an increase5 and6 in). wind High speed evaporation and reduced adds moreSST in moisture the western to the basin, atmosphere, leading to thecontributing east-west toSST the anomalies. increased precipitationThe wind anomal and floodingies drove observed a series inof East Kelvin Africa waves by the which end reducedof 2017 [ 33the]. upwelling efficiency and warmed the SST in the east. On the other hand, a weakIn positive the AS IOD region, phase the precipitationpersisted throughout patterns were2017; similarSST anomalies and no significant were below variations the 30- yearwere average noted over(reference the entire period: period 1981–2010) of 2016–2017 over the (not eastern shown). pole However,and above thethe monthlywestern poleaverage of the SSS equatorial remained Indian high for Ocean. the whole The extr spaneme of 2016El Niño (compared also influenced to 2017) andrainfall was distri- more butionpronounced in the inTIO. September It was observed (not shown). that, Thisduring was El because Niño, there SSTs inwas the above AS covaried normalwith rainfall the overONI the observed ETIO, inand the below Niño normal 3.4 region rainfall (5◦ S–5 ov◦erN, the 240–290 WTIO.◦ TheE). A corresponding positive SST anomaly low sea sur- was faceinduced salinity, in the found AS duringin the ETIO the strongat the beginning El Niño in of the 2016, Pacific was inmainly 2016, due which to excess increased precip- the itation.evaporation The negative rate and IOD surface in 2016 salinity was due characterized to fresh water by above-average loss. rainfall in Bangla- desh Inand the dry BoB, conditions warm SST in anomaliesEast Africa persisted [29,32]. over the central region from January to May 2016 followed by the strong El Niño of 2016 (Figure5a, the monthly variation is not shown). However, the effect of high SST on the surface salinity was not detected in this region (Figure 14c). Instead, there was a small decline in SSS (up to 0.5) from January to May in

2016, as compared to 2017. The dynamic behind this, however, could not be investigated in this study. As shown in Figure 13, precipitation patterns were almost identical in both years, with no signiﬁcant differences except in June and September (in which positive precipitation anomalies were higher in 2017 than in 2016). The highest precipitation levels were observed during the monsoon season in both years, albeit with more in 2017. The corresponding effect—enhanced precipitation in the monsoon season—can be seen in Figure 14c, in which a slight decrease in SSS was observed in 2017, as compared to 2016. Atmosphere 2021, 12, 587 17 of 20

The low precipitation rate observed during the negative IOD and the high precipitation rate observed during the positive IOD in the BoB region are both consistent with the ﬁndings of Chanda et al. 2018 [67].

4. Conclusions In 2016, the strong El Niño in the Pacific Ocean induced an extreme negative IOD in the Indian Ocean. The following year, a weak La Niña in the same region induced a weak positive IOD. These unusual events had a major impact on the SST and were followed by above-average flooding in the Indian subcontinent and drought conditions in East Africa [29,33]. The current study investigated the dynamics behind these events and their effect on SST variability in the TIO. It was noted that during the strong El Niño in the Pacific in early 2016, the Indian Ocean warmed due to suppressed atmospheric convection and positive SST anomalies which persisted in both the east and west poles of the Equatorial Indian Ocean. Our analyses indicated that high SSTs along the Sumatran coast in the summer of 2016 induced local convection and western wind anomalies via the Bjerknes feedback mechanism [57]. These a led to an increase in wind speed and reduced SST in the western basin, leading to the east-west SST anomalies. The wind anomalies drove a series of Kelvin waves which reduced the upwelling efficiency and warmed the SST in the east. On the other hand, a weak positive IOD phase persisted throughout 2017; SST anomalies were below the 30-year average (reference period: 1981–2010) over the eastern pole and above the western pole of the equatorial Indian Ocean. The extreme El Niño also influenced rainfall distribution in the TIO. It was observed that, during El Niño, there was above normal rainfall over the ETIO, and below normal rainfall over the WTIO. The corresponding low sea surface salinity, found in the ETIO at the beginning of 2016, was mainly due to excess precipitation. The negative IOD in 2016 was characterized by above-average rainfall in Bangladesh and dry conditions in East Africa [29,32]. The strong El Niño and the extreme negative IOD also influenced the SST in the Ara- bian Sea region in 2016. It was observed that SST in the Arabian Sea remained high during early 2016 and decreased in the summer period, as compared to 2017. The increase in SST during early 2016 was mainly due to the strong El Niño, which warmed the Indian Ocean from January to May 2016. However, the summer SST in the western equatorial region of the Indian Ocean and the Arabian Sea was most likely reduced by the development of a negative IOD phase. Another important feature of this study was the observation of increased upwelling along the Omani-Arabian coast during the summer and post-summer periods of 2016—and the corresponding decreased upwelling in the same region in 2017. The low upwelling in 2017 warmed the SST by almost 1 ◦C, which increased precipitation rates in the BoB region. We found that, in the BoB region, precipitation was low during negative IOD and high during positive IOD, which is consistent with the findings of Chanda et al. in 2018 [67]. The SST variability is robust in the TIO in spite of the co-occurrence with the strong El Niño/La Niña cycle. The central role of the TIO in modulating the regional climate suggests that the variability of the SST in this basin has serious implications for both the highly populated regions around this basin and the ocean modeling community.

Author Contributions: Formal analysis, S.K. (Sartaj Khan) and Y.S; Funding acquisition, S.P. and Y.S; Methodology, S.K. (Sartaj Khan) and G.Z; Resources, S.P. and Y.S; Supervision, S.P. and Y.S; Validation, Y.S., G.Z. and D.B; Writing—original draft, S.K. (Sartaj Khan) and I.U.K.; Writing—review and editing, S.K. (Shazia Khan) I.U.K. and D.B. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Atmosphere 2021, 12, 587 18 of 20

Acknowledgments: We would like to thank David Bradley and two anonymous reviewers for their insightful and constructive comments and suggestions. The authors also wish to acknowledge the Acoustic Science and Technology Laboratory of Harbin Engineering University for analysis and graphics. This work was supported by an international postgraduate scholarship awarded by the Chinese Scholarship Council (CSC) and National Key Research and Development Program of China, No. 2016YFC1400100. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Dong, L.; McPhaden, M.J. Unusually warm Indian Ocean sea surface temperatures help to arrest development of El Niño in 2014. Sci. Rep. 2018, 8, 2249. [CrossRef] 2. Kug, J.S.; Kang, I.S. Interactive feedback between ENSO and the Indian Ocean. J. Clim. 2006, 19, 1784–1801. [CrossRef] 3. Okumura, Y.M.; Deser, C. Asymmetry in the duration of El Niño and La Niña. J. Clim. 2010, 23, 5826–5843. [CrossRef] 4. Murtugudde, R.; Busalacchi, A.J. Interannual variability of the dynamics and thermodynamics of the tropical Indian Ocean. J. Clim. 1999, 12, 2300–2326. [CrossRef] 5. Vinayachandran, P.N.; Iizuka, S.; Yamagata, T. Indian Ocean dipole mode events in an ocean general circulation model. Deep Sea Res. Part II 2002, 49, 1573–1596. [CrossRef] 6. Walker, G.T. Correlation in seasonal variations of weather. IX. A further study of world weather. Mem. Indian Meteorol. Dep. 1924, 24, 275–332. 7. Krishnamurthy, V.; Shukla, J. Intraseasonal and interannual variability of rainfall over India. J. Clim. 2000, 13, 4366–4377. [CrossRef] 8. Arpe, K.; Dümenil, L.; Giorgetta, M.A. Variability of the Indian monsoon in the ECHAM3 model: Sensitivity to sea surface temperature, soil moisture, and the stratospheric quasi-biennial oscillation. J. Clim. 1998, 11, 1837–1858. [CrossRef] 9. Keerthi, M.; Lengaigne, M.; Vialard, J.; de Boyer Montégut, C.; Muraleedharan, P. Interannual variability of the Tropical Indian Ocean mixed layer depth. Clim. Dyn. 2013, 40, 743–759. [CrossRef] 10. Reverdin, G.; Cadet, D.L.; Gutzler, D. Interannual displacements of convection and surface circulation over the equatorial Indian Ocean. Q. J. R. Meteorol. Soc. 1986, 112, 43–67. [CrossRef] 11. Saji, N.H.; Goswami, B.N.; Vinayachandran, P.N.; Yamagata, T. A dipole mode in the tropical Indian Ocean. Nature 1999, 401, 360–363. [CrossRef] 12. Webster, P.J.; Moore, A.M.; Loschnigg, J.P.; Leben, R.R. Coupled ocean–atmosphere dynamics in the Indian Ocean during 1997–98. Nature 1999, 401, 356–360. [CrossRef][PubMed] 13. Murtugudde, R.; McCreary, J.P., Jr.; Busalacchi, A.J. Oceanic processes associated with anomalous events in the Indian Ocean with relevance to 1997–1998. J. Geophys. Res. Ocean. 2000, 105, 3295–3306. [CrossRef] 14. Joseph, P.V. Monsoon variability in relation to equatorial trough activity over India and west Paciﬁc Oceans. Mausam 1990, 41, 291–296. 15. Niñomiya, K.; Kobayashi, C. Precipitation and moisture balance of the Asian summer monsoon in 1991. J. Meteorol. Soc. Jpn. 1999, 77, 77–99. [CrossRef] 16. De Boyer Montégut, C.; Mignot, J.; Lazar, A.; Cravatte, S. Control of salinity on the mixed layer depth in the world ocean: 1. General description. J. Geophys. Res. Ocean. 2007, 112, C06011. [CrossRef] 17. Murtugudde, R.; Seager, R.; Thoppil, P. Arabian Sea response to monsoon variations. Paleoceanography 2007, 22, PA4217. [CrossRef] 18. Izumo, T.; Montégut, C.B.; Luo, J.J.; Behera, S.K.; Masson, S.; Yamagata, T. The role of the western Arabian Sea upwelling in Indian monsoon rainfall variability. J. Clim. 2008, 21, 5603–5623. [CrossRef] 19. Suppiah, R. Relationships between Indian Ocean sea surface temperature and the rainfall of Sri Lanka. J. Meteorol. Soc. Jpn. 1988, 66, 103–112. [CrossRef] 20. Kripalani, R.H.; Kumar, P. Northeast monsoon rainfall variability over south peninsular India vis-à-vis the Indian Ocean dipole mode. Int. J. Climatol. J. R. Meteorol. Soc. 2004, 24, 1267–1282. [CrossRef] 21. Joseph, S.; Sahai, A.K.; Goswami, B.N. Eastward propagating MJO during boreal summer and Indian monsoon droughts. Clim. Dyn. 2009, 32, 1139–1153. [CrossRef] 22. Joseph, S.; Sahai, A.K.; Chattopadhyay, R.; Sharmila, S.; Abhilash, S.; Rajeevan, M.; Mandal, R.; Dey, A.; Borah, N.; Phani, R. Extremes in June rainfall during the Indian summer monsoons of 2013 and 2014: Observational analysis and extended-range prediction. Q. J. R. Meteorol. Soc. 2016, 142, 1276–1289. [CrossRef] 23. Trenberth, K.E.; Jones, P.D.; Ambenje, P.; Bojariu, R.; Easterling, D.; Klein Tank, A.; Parker, D.; Rahimzadeh, F.; Renwick, J.A.; Rusticucci, M.; et al. Chapter 3: Observations: Surface and atmospheric climate change. In Climate Change; Cambridge University Press: Cambridge, UK; New York, NY, USA, 2007; pp. 235–336. 24. Venugopal, T.; Ali, M.M.; Bourassa, M.A.; Zheng, Y.; Goni, G.J.; Foltz, G.R.; Rajeevan, M. Statistical evidence for the role of southwestern Indian Ocean heat content in the Indian summer monsoon rainfall. Sci. Rep. 2018, 8, 12092. [CrossRef] 25. Meehl, G.A.; Arblaster, J.M.; Fasullo, J.T.; Hu, A.; Trenberth, K.E. Model-based evidence of deep-ocean heat uptake during surface-temperature hiatus periods. Nat. Clim. Chang. 2011, 1, 360–364. [CrossRef] Atmosphere 2021, 12, 587 19 of 20

26. Brown, P.T.; Li, W.; Xie, S.-P. Regions of significant influence on unforced global mean surface air temperature variability in climate models. J. Geophys. Res. Atmos. 2015, 120, 480–494. [CrossRef] 27. Dai, A.; Fyfe, J.C.; Xie, S.-P.; Dai, X. Decadal modulation of global surface temperature by internal climate variability. Nat. Clim. Chang. 2015, 5, 555–559. [CrossRef] 28. McPhaden, M.J.; Zebiak, S.E.; Glantz, M.H. ENSO as an integrating concept in Earth science. Science 2006, 314, 1740–1745. [CrossRef] 29. World Meteorological Organization (WMO). Statement on the State of the Global Climate in 2016; World Meteorological Organization: Geneva, Switzerland, 2017. 30. Behera, S.K.; Krishnan, R.; Yamagata, T. Unusual ocean-atmosphere conditions in the tropical Indian Ocean during 1994. Geophys. Res. Lett. 1999, 26, 3001–3004. [CrossRef] 31. Birkett, C.; Murtugudde, R.; Allan, T. Indian Ocean climate event brings floods to East Africa’s lakes and the Sudd Marsh. Geophys. Res. Lett. 1999, 26, 1031–1034. [CrossRef] 32. Lu, B.; Ren, H.L.; Scaife, A.A.; Wu, J.; Dunstone, N.; Smith, D.; Wan, J.; Eade, R.; MacLachlan, C.; Gordon, M. An extreme negative Indian Ocean Dipole event in 2016: Dynamics and predictability. Clim. Dyn. 2018, 51, 89–100. [CrossRef] 33. World Meteorological Organization (WMO). Statement on the State of the Global Climate in 2017; World Meteorological Organization: Geneva, Switzerland, 2018. 34. Suthinkumar, P.S.; Babu, C.A.; Varikoden, H. Spatial Distribution of Extreme Rainfall Events during 2017 Southwest Monsoon over Indian Subcontinent. Pure Appl. Geophys. 2019, 176, 5431–5443. [CrossRef] 35. Huang, B.; Liu, C.; Banzon, V.; Freeman, E.; Graham, G.; Hankins, B.; Smith, T.; Zhang, H.M. Improvements of the Daily Optimum Interpolation Sea Surface Temperature (DOISST) Version 2.1. J. Clim. 2020, 34, 2923–2939. [CrossRef] 36. Banzon, V.; Smith, T.M.; Chin, T.M.; Liu, C.Y.; Hankins, W. A long-term record of blended satellite and in situ sea-surface temperature for climate monitoring, modeling and environmental studies. Earth Syst. Sci. Data 2016, 8, 165–176. [CrossRef] 37. Reynolds, R.W.; Smith, T.M.; Liu, C.; Chelton, D.B.; Casey, K.S.; Schlax, M.G. Daily high-resolution-blended analyses for sea surface temperature. J. Clim. 2007, 20, 5473–5496. [CrossRef] 38. Behringer, D.W.; Xue, Y. Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean. In Proceedings of the Eighth Symposium on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface, AMS 84th Annual Meeting, Washington State Convention and Trade Center, Seattle, WA, USA, 11–15 January 2004. 39. Vissa, N.K.; Satyanarayana, A.N.V.; Prasad Kumar, B. Comparison of mixed layer depth and barrier layer thickness for the Indian Ocean using two different climatologies. Int. J. Climatol. 2013, 33, 2855–2870. [CrossRef] 40. Adler, R.F.; Huffman, G.J.; Chang, A.; Ferraro, R.; Xie, P.P.; Janowiak, J.; Rudolf, B.; Schneider, U.; Curtis, S.; Bolvin, D.; et al. The version-2 global precipitation climatology project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeorol. 2003, 4, 1147–1167. [CrossRef] 41. Huffman, G.J.; Adler, R.F.; Bolvin, D.T.; Gu, G. Improving the global precipitation record: GPCP version 2.1. Geophys. Res. Lett. 2009, 36, L17808. [CrossRef] 42. Meissner, T.; Wentz, F.J.; Le Vine, D.M. The Salinity Retrieval Algorithms for the NASA Aquarius Version 5 and SMAP Version 3 Releases. Remote Sens. 2018, 10, 1121. [CrossRef] 43. Huang, B.; Thorne, P.W.; Banzon, V.F.; Boyer, T.; Chepurin, G.; Lawrimore, J.H.; Menne, M.J.; Smith, T.M.; Vose, R.S.; Zhang, H.M. Extended reconstructed sea surface temperature, version 5 (ERSSTv5): Upgrades, validations, and intercomparisons. J. Clim. 2017, 30, 8179–8205. [CrossRef] 44. Saji, N.H.; Yamagata, T. Structure of SST and surface wind variability during Indian Ocean Dipole mode year: COADS observations. J. Clim. 2003, 16, 2735–2751. [CrossRef] 45. Rao, A.S.; Behera, S.K.; Masumoto, Y.; Yamagata, T. Interannual subsurface variability in the tropical Indian Ocean with a special emphasis on the Indian Ocean dipole. Deep Sea Res. Part II Top. Stud. Oceanogr. 2002, 49, 1549–1572. [CrossRef] 46. Luo, J.J.; Masson, S.; Behera, S.; Yamagata, T. Experimental forecasts of the Indian Ocean Dipole using a coupled OAGCM. J. Clim. 2007, 20, 2178–2190. [CrossRef] 47. Song, Q.; Vecchi, G.A.; Rosati, A.J. Predictability of the Indian Ocean sea surface temperature anomalies in the GFDL coupled model. Geophys. Res. Lett. 2008, 35, L02701. [CrossRef] 48. Shi, L.; Hendon, H.H.; Alves, O.; Luo, J.J.; Balmaseda, M.; Anderson, D. How predictable is the Indian Ocean Dipole? Mon. Weather Rev. 2012, 140, 3867–3884. [CrossRef] 49. Liu, H.; Tang, Y.; Chen, D.; Lian, T. Predictability of the Indian Ocean Dipole in the coupled models. Clim. Dyn. 2017, 48, 2005–2024. [CrossRef] 50. Doi, T.; Storto, A.; Behera, S.K.; Navarra, A.; Yamagata, T. Improved prediction of the Indian Ocean dipole mode by use of subsurface ocean observations. J. Clim. 2017, 30, 7953–7970. [CrossRef] 51. Ashok, K.; Guan, Z.; Yamagata, T. Impact of the Indian Ocean Dipole on the relationship between the Indian monsoon rainfall and ENSO. Geophys. Res. Lett. 2001, 28, 4499–4502. [CrossRef] 52. Behera, S.K.; Luo, J.J.; Masson, S.; Rao, S.A.; Sakuma, H.; Yamagata, T. A CGCM study on the interaction between IOD and ENSO. J. Clim. 2006, 19, 1688–1705. [CrossRef] 53. Reynolds, R.W.; Rayner, N.A.; Smith, T.M.; Stokes, D.C.; Wang, W. An improved in situ and satellite SST analysis for climate. J. Clim. 2002, 15, 1609–1625. [CrossRef] Atmosphere 2021, 12, 587 20 of 20

54. Weare, B.C. A statistical study of the relationships between ocean surface temperatures and the Indian monsoon. J. Atmos. Sci. 1979, 36, 2279–2291. [CrossRef] 55. Lim, E.P.; Hendon, H.H. Causes and predictability of the negative Indian Ocean Dipole and its impact on La Niña during 2016. Sci. Rep. 2017, 7, 12619. [CrossRef][PubMed] 56. Kantha, L.H.; Clayson, C.A. An improved mixed layer model for geophysical applications. J. Geophys. Res. Ocean. 1979, 99, 25235–25266. [CrossRef] 57. Bjerknes, J. Atmospheric teleconnections from the equatorial Paciﬁc. Mon. Weather Rev. 1969, 97, 163–172. [CrossRef] 58. Düing, W.; Leetmaa, A. Arabian Sea cooling: A preliminary heat budget. J. Phys. Oceanogr. 1980, 10, 307–312. [CrossRef] 59. Shukla, J. Effect of Arabian sea-surface temperature anomaly on Indian summer monsoon: A numerical experiment with the GFDL model. J. Atmos. Sci. 1975, 32, 503–511. [CrossRef] 60. Hastenrath, S.; Lamb, P. Climatic Atlas of the Indian Ocean. Part I. Surface Circulation and Climate; University of Wisconsin Press: Madison, WI, USA, 1979; p. 109. 61. Rao, R.R.; Sivakumar, R. Seasonal variability of near-surface thermal structure and heat budget of the mixed layer of the tropical Indian Ocean from a new global ocean temperature climatology. J. Geophys. Res. Ocean. 2000, 105, 995–1015. [CrossRef] 62. Yoshida, K.; Mao, H.L. A theory of upwelling of horizontal extent. J. Mar. Res. 1957, 16, 40–54. 63. Saha, K. Zonal anomaly of sea surface temperature in equatorial Indian Ocean and its possible effect upon monsoon circulation. Tellus 1970, 22, 403–409. [CrossRef] 64. Goddard, L.; Graham, N.E. Simulation skills of the SST-forced global climate variability of the NCEP-MRF9 and the Scripps-MPI ECHAM3 model. J. Clim. 1999, 13, 3657–3679. 65. Nicholls, N. All-India summer monsoon rainfall and sea surface temperatures around northern Australia and Indonesia. J. Clim. 1995, 8, 1463–1467. [CrossRef] 66. Nicholson, S.E.; Kim, J. The relationship of the El Niño–Southern oscillation to African rainfall. Int. J. Climatol. J. R. Meteorol. Soc. 1997, 17, 117–135. [CrossRef] 67. Chanda, A.; Das, S.; Mukhopadhyay, A.; Ghosh, A.; Akhand, A.; Ghosh, P.; Ghosh, T.; Mitra, D.; Hazra, S. Sea surface temperature and rainfall anomaly over the Bay of Bengal during the El Niño-Southern Oscillation and the extreme Indian Ocean Dipole events between 2002 and 2016. Remote Sens. Appl. Soc. Environ. 2018, 12, 10–22. [CrossRef]