The influences of ENSO and the subtropical Indian Ocean ...ufdcimages.uflib.ufl.edu/IR/00/00/12/00/00001/AshMatyasIJOC2012.pdfet al., 2003; Reason and Keibel, 2004; Ho et al., 2006;

INTERNATIONAL JOURNAL OF CLIMATOLOGYInt. J. Climatol. 32: 41–56 (2012)Published online 5 November 2010 in Wiley Online Library(wileyonlinelibrary.com) DOI: 10.1002/joc.2249

The influences of ENSO and the subtropical Indian OceanDipole on tropical cyclone trajectories in the southwestern

Indian Ocean

Kevin D. Asha* and Corene J. Matyasb

a Hazards and Vulnerability Research Institute, Department of Geography, University of South Carolina, 709 Bull Street, Columbia, SC 29208,USA

b Department of Geography, University of Florida, 3141 Turlington Hall, Gainesville, FL 32611, USA

ABSTRACT: El Nino-Southern Oscillation (ENSO) is known to associate with variability of tropical cyclone (TC)trajectories in the southwestern Indian Ocean. However, consideration of ENSO phase alone does not account for allvariability of TC tracks within this region. This study demonstrates that the subtropical Indian Ocean Dipole (SIOD)sea-surface temperature anomaly pattern is also significantly associated with variability in southwestern Indian OceanTC tracks. Hierarchical cluster analysis is employed to group TC trajectories by their initial and final positions. Medianmonthly values of the Nino-3.4 index and Subtropical Dipole Index corresponding to the life cycles of TCs in each groupare compared using non-parametric analysis of variance. The results suggest that both ENSO and SIOD are significantlyassociated with different types of southwestern Indian Ocean TC trajectories. Furthermore, significant interactions of ENSOand SIOD phases are found to influence certain types of TC tracks using contingency table tests. During simultaneous warmENSO and negative SIOD phases, TCs moving across the southwestern Indian Ocean tend to follow more southward orsoutheastward tracks. During neutral or cool ENSO and positive SIOD phases, TCs moving through the southwestern IndianOcean tend toward more westward trajectories. These findings suggest that use of an SIOD index in addition to an ENSOindex could improve intraseasonal to seasonal statistical prediction of southwestern Indian basin TC activity. Copyright 2010 Royal Meteorological Society

KEY WORDS tropical cyclones; ENSO; subtropical Indian Ocean dipole; trajectories; cluster analysis; tropical temperatetroughs

Received 17 June 2010; Revised 4 October 2010; Accepted 7 October 2010

1. Introduction

Tropical cyclones (TCs) are a recurring phenomenon inthe southwestern Indian Ocean, most frequently duringTC season from November to April, and the inhabitedregions of the western rim of the basin are particularlyprone to repeated TC impacts. TCs frequently pass nearthe small islands of Mauritius and Reunion and cancause severe damage and societal disruption, thoughloss of human life is typically mitigated by the well-organised and executed TC preparedness and warningsystems of these islands (Parker, 1999; Roux et al.,2004). Madagascar has endured many devastating TCimpacts (Jury et al., 1993; Naeraa and Jury, 1998;Chang-Seng and Jury, 2010b), and the population’ssocial vulnerability to negative impacts from TC strikesis increased due to widespread livelihood dependenceon agriculture and the lesser economic status of thecountry (Brown, 2009). Mozambique is, likewise, in aprecarious economic and climatological position, though

* Correspondence to: Kevin D. Ash, Hazards and VulnerabilityResearch Institute, Department of Geography, University of South Car-olina, 709 Bull Street, Columbia, SC 29208, USA.E-mail: [email protected]

TC landfalls are not as frequent as in Madagascar(Vitart et al., 2003; Reason and Keibel, 2004; Reason,2007; Klinman and Reason, 2008; Silva et al., 2010).TC impacts also extend over the open ocean throughdisruption of busy shipping lanes across the entiresouthern Indian Ocean, threatening the lives of vesselcrews and their valuable cargoes as they steam betweenmaritime economic hubs in Europe, the Americas, SouthAsia, and the Far East (Chang-Seng and Jury, 2010b).

Large scale, low-frequency modes of ocean-atmospherevariability related to TC genesis in the southern IndianOcean have been studied previously and are rela-tively well understood. The El Nino-Southern Oscillation(ENSO), the Madden-Julian Oscillation, and the IndianOcean Dipole, or Zonal Mode, are all known to influencethe spatial pattern and frequency of TC genesis in thesouthern Indian Ocean (Kuleshov and de Hoedt, 2003;Bessafi and Wheeler, 2006; Ho et al., 2006; Camargoet al., 2007a; Kuleshov et al., 2008; Leroy and Wheeler,2008; Chan and Liu, 2009; Kuleshov et al., 2009; Vitartet al., 2010). However, given that TCs forming in thecentral or southeastern Indian Ocean may traverse thou-sands of kilometers to the west to impact the aforemen-tioned countries on the western fringe of the basin, it is

Copyright 2010 Royal Meteorological Society

42 K. D. ASH AND C. J. MATYAS

also important to improve understanding of TC trajectoryvariability in relation to ocean-atmosphere variability atintraseasonal to seasonal time scales. Less research hasspecifically focused on the topic of intraseasonal or sea-sonal southwestern Indian Ocean TC track variability,save for the work of Vitart et al. (2003) which suggestedan ENSO link to TC track variability in the region. Out-side the body of southern Indian Ocean TC research,ENSO, the subtropical Indian Ocean Dipole (SIOD), andtropical temperate troughs (TTTs) have all been shownto play important roles in air–sea interactions within thesouthwestern Indian Ocean which could influence TCtrajectory variability on intraseasonal to seasonal timescales. In this study, we specifically consider the influ-ence of ENSO and SIOD on southwestern Indian OceanTC tracks, and use the previously established TTT frame-work for a broader understanding of the mechanism bywhich ENSO and SIOD may alter TC track patterns inthis region.

It has been demonstrated in each of the world’s princi-pal TC regions that ENSO can alter the global or regionalatmospheric circulation to either change the spatial pat-terns of TC genesis or affect their directions of movement(Gray and Sheaffer, 1991; Landsea, 2000; Chu, 2004;Camargo et al., 2010). The importance of ENSO influ-ence on the coupled ocean-atmosphere system in thesouthern Indian Ocean alone is very well documented(van Loon and Rogers, 1981; Pan and Oort, 1983; Meehl,1987; Karoly, 1989; Nicholson, 1997; Chambers et al.,1999; Klein et al., 1999; Reason et al., 2000; Venzkeet al., 2000; Larkin and Harrison, 2002; Hermes and Rea-son, 2005; Yoo et al., 2006; Huang and Shukla, 2007a).Just as in the northern and southern Pacific Oceans, thereis not a significant basin-wide correlation in the southernIndian Ocean between ENSO and TC frequency (Jury,1993); instead it is the spatial patterns of genesis anddirections of movement that exhibit links to ENSO (Rev-ell and Goulter, 1986; Lander, 1994). ENSO has beendemonstrated as a significant predictor of southern IndianTC activity at weekly to monthly time scales and is nowutilised in operational statistical prediction of TC activ-ity in the region (Leroy and Wheeler, 2008; Vitart et al.,2010).

During El Nino, TC genesis is more frequent over thesouthwestern Indian Ocean (west of about 75 °E–85 °E)than in the eastern ocean (Evans and Allan, 1992;Kuleshov and de Hoedt, 2003; Ho et al., 2006; Kuleshovet al., 2008). Favourable conditions for deep convec-tion in the western ocean are generated following theoccurrence of easterly lower tropospheric wind anoma-lies south and west of Sumatra during early austral spring(late September–November), which excite a westward-propagating oceanic Rossby wave that, coupled withincreased poleward Ekman transport associated with theeasterly anomalies in the tropical eastern part of thebasin, induce an unusually deep pool of warm sea sur-face temperature anomalies (SSTA) centered near 15 °S,60 °E (Chambers et al., 1999; Jury et al., 1999; Rea-son et al., 2000; Xie et al., 2002; Jury and Huang,

2004). Cooler SSTA in the southeastern Indian Oceanand anticyclonic vorticity from the lower troposphericeasterly anomalies account for a reduction of TC gen-esis, most notably in late austral spring and early sum-mer (Ho et al., 2006; Camargo et al., 2007a; Kuleshovet al., 2008; Kuleshov et al., 2009). Conversely, duringan ENSO cool event, the western portion of the basinis characterised by reduced TC genesis while the eastern(east of about 75 °E–85 °E) experiences more, with posi-tive SSTA, increased mid-tropospheric relative humidity,and increased lower tropospheric cyclonic vorticity con-tributing to an environment conducive to more frequentTC genesis (Wolter, 1987; Ho et al., 2006; Camargoet al., 2007a; Kuleshov et al., 2008; Chan and Liu, 2009;Kuleshov et al., 2009).

As for the tracks of southern Indian TCs in relationto ENSO, Vitart et al. (2003) note that zonal steeringflow averaged over 850–200 hPa across the tropical andsubtropical portions of the basin is more westerly (east-erly) during El Nino (La Nina). Consequently, Mozam-bique should be at greater risk for landfall during LaNina, whereas westerly steering flows are suggestive ofincreased incidences of re-curving TCs just east of Mada-gascar during El Nino (Jury and Pathack, 1991; Vitartet al., 2003; Reason and Keibel, 2004; Ho et al., 2006;Camargo et al., 2007a; Kuleshov et al., 2009). Chang-Seng and Jury (2010a) agree that La Nina is associ-ated with increased frequency of longer-lived and moreintense TCs in the southwestern Indian basin. However,though ENSO is the most important source of climatevariability in the global tropics and subtropics, it does notaccount for all variability in the global tropical and sub-tropical ocean–atmosphere system, nor does it account intotality for the variability of trajectories amongst south-western Indian TCs (Fauchereau et al., 2003; Vitart et al.,2003; Huang and Shukla, 2007b; Klinman and Reason,2008).

During El Nino in austral summer, tropical temperatetroughs (TTTs) influence atmospheric conditions in theTC regions of the southwestern Indian Ocean (Harangozoand Harrison, 1983; Lyons, 1991; van den Heever et al.,1997; Washington and Todd, 1999). The TTTs frequentlyshift 30° –35° eastward and collocate with the deepwarm pool over the tropical western ocean (Lindesayet al., 1986; Mason and Jury, 1997; Cook, 2000; Tysonand Preston-Whyte, 2000; Nicholson, 2003; Fauchereauet al., 2009; Pohl et al., 2009; Manhique et al., 2011).This eastward shift of the regional tropical convectivemaximum and accompanying poleward outflow is con-sistent with the non-linear El Nino teleconnection frame-work of Hoerling et al. (1997), and places within abroader synoptic context the positive zonal wind anoma-lies observed in previous studies of the tropical and sub-tropical southwestern Indian basin (Reason et al., 2000;Yoo et al., 2006). Heretofore, the ENSO-TTT frameworkhas not been applied to the southwestern Indian TC gene-sis or track research, as only Mavume et al. (2009) brieflymention TTTs as potentially important in relation to TCtracks within the Mozambique Channel.

Copyright 2010 Royal Meteorological Society Int. J. Climatol. 32: 41–56 (2012)

ENSO AND SIOD INFLUENCES ON SOUTHWESTERN INDIAN TC TRAJECTORIES 43

The SIOD has also been identified as an importantsource of ocean–atmosphere variability in the southernIndian Ocean. It is represented by the second empiri-cal orthogonal function of southern Indian Ocean tropi-cal and subtropical SSTA (Behera et al., 2000; Beheraand Yamagata, 2001; Qian et al., 2002; Suzuki et al.,2004; Huang and Shukla, 2007b), and SIOD phaseshave been suggested to exhibit variations independentof the simultaneous ENSO phase (Behera and Yam-agata, 2001; Fauchereau et al., 2003; Washington andPreston, 2006; Huang and Shukla, 2007b). The pat-tern is most prominent during the austral warm seasonand is characterised in the positive (negative) mode bycool (warm) SSTA in the southeastern Indian Ocean,whereas the southwestern Indian Ocean is simultaneouslywarm (cool). The mechanisms for cooling in the easternpole during a positive mode are strengthened southeast-erly trade winds and resultant enhanced ocean surfaceevaporation and mixing, while the western warm poledevelops concurrently with increased poleward Ekmantransport of warm SSTs from tropical latitudes com-bined with decreased equatorward cold air advectionand ocean surface evaporation (Reason, 1999; Venzkeet al., 2000; Behera and Yamagata, 2001; Qian et al.,2002; Hermes and Reason, 2005; Chiodi and Harrison,2007; Huang and Shukla, 2007b). In negative mode, theSSTA poles are generally reversed with cold air advec-tion and equatorward Ekman pumping occurring over thesouthwestern Indian Ocean in conjunction with more fre-quent cold frontal passages, while warm air advection islocated more frequently over the southeastern part of thebasin.

There is evidence in the literature to suggest that east-ward shifts of the African TTTs are not only associatedwith ENSO, but also with the negative phase of the SIOD.Reason (2002) and Fauchereau et al. (2009) noted thestrong similarities between the positive SIOD SST andwind anomaly patterns, and the atmospheric anomaliesassociated with TTTs over southern Africa. This is impor-tant because it raises the possibility of local SSTA influ-ence in the variability of TTTs, which have often beenattributed to the influence from ENSO. During a negativeSIOD phase, a persistent eastward shift of TTTs over thesouthwestern Indian Ocean should influence TCs to fol-low more re-curving poleward or even eastward tracks(Parker and Jury, 1999; Chang-Seng and Jury, 2010b).Therefore, if negative SIOD is associated with eastwardphase shifting of TTTs, and if SIOD phases can exhibitvariations independent of the simultaneous ENSO phase(Fauchereau et al., 2003; Washington and Preston, 2006;Huang and Shukla, 2007b), then an interactive consid-eration of both ENSO and SIOD is warranted to betteraccount for variability in southwestern Indian Ocean TCtrajectories.

In Section 2, we describe the clustering procedureemployed to reduce a sample of 191 TCs to 6 typesbased on their genesis locations within 3 subregionsof the southern Indian basin main development region(54 °E–110 °E) and then further subdivided by their

directions of movement from each genesis subregion.We also outline the analysis of variance (ANOVA) testsused to infer differences of association with SIOD andENSO between the final six TC trajectory groups. Section3 conveys the clustering procedure results, presents theANOVA test results, and compares different TC trajectorygroups that share common genesis regions. Expandeddiscussion of the results and potential applications aregiven in Section 4, and Section 5 summarises theimportant findings from this study and offers possibledirections for future research.

2. Data and methods

2.1. Tropical cyclone data

The TC trajectories in this study are from the JointTyphoon Warning Center (JTWC) best-track dataset forthe Southern Hemisphere (Chu et al., 2002). Because thisresearch focuses on the southwestern Indian Ocean, weconsider only those storms which passed west of 90 °Elongitude during their respective life cycles. The 90thmeridian is chosen as the west–east boundary in accor-dance with the official forecasting areas of responsibil-ity assigned by the World Meteorological Organization(WMO) to Regional Specialized Meteorological Centers(RSMC) La Reunion (west of 90 °E) and Perth (east of90 °E) (Caroff, 2009). The JTWC best-track data are pre-ferred over the RSMC La Reunion best-track archivefor this study because the La Reunion dataset does notextend as far to the east of 90 °E as the JTWC datadue to the RSMC administrative boundary. By using theJTWC archive, we include more complete trajectoriesof TCs that formed east of 90 °E, crossed that merid-ian, and continued westward into the southwestern Indianbasin.

Additionally, we considered only TCs that reached amaximum lifetime intensity of at least 30 m s−1 maxi-mum 1-minute sustained wind. This proviso addresses TCdata quality concerns through inclusion of stronger andbetter organised TCs, which are less prone to position-fix errors than weaker, poorly organised tropical systems(Yip et al., 2006). These cautious steps are deemed nec-essary due to the lack of complete and permanent geosta-tionary satellite coverage over the southern Indian regionprior to May 1998 (Kossin et al., 2007; Chang-Seng andJury, 2010a; Kuleshov et al., 2010). Despite the limitedsatellite coverage, Knaff and Sampson (2009) note thatJTWC TC data are suitable for analysis from about 1980.In keeping with their suggestion, we assembled a sampleof 191 TCs from the recent thirty-year period spanning1979–2008. An equally important consideration of study-ing TCs from this period is to research principally theocean–atmosphere interactions operating after the knownIndo-Pacific region climate shift of 1976–1977, as spa-tial patterns of ocean–atmosphere variability associatedwith ENSO and SIOD interactions are known to differ onmulti-decadal time scales (Trenberth, 1990; Zinke et al.,2004; Terray and Dominiak, 2005).



2.2. Clustering procedure

To test for influences from ENSO and SIOD, we firstbreak the 191 trajectories into groups using clusteranalysis. Cluster analysis is widely used in geophysicalresearch for classification or data exploration, and awell-structured clustering solution, whether or not itreveals the natural modality of the data, can aid in thediscernment of mechanisms that shape differences andsimilarities of complex events or phenomena (Gong andRichman, 1995; Wilks, 2006). Cluster analysis has beenemployed recently as a tool to research variability inTC trajectories in the northwestern Pacific (Elsner andLiu, 2003; Camargo et al., 2007b; Choi et al., 2009),the northeastern Pacific (Camargo et al., 2008), and theNorth Atlantic (Elsner, 2003; Kossin et al., 2010). Thepresent study represents the first known application ofcluster analysis to analyse TC track variability in thesouthwestern Indian Ocean.

We implement a two-stage agglomerative hierarchicalclustering procedure using a Euclidean distance measureand the group average linkage method. Agglomerativehierarchical clustering begins with each observation asits own group and then iteratively joins the two clos-est groups until there is one group that includes everyobservation (Lattin et al., 2003). The cophenetic corre-lation may then be used as a diagnostic of the strengthof a clustering structure in tandem with a dendrogram(or a tree diagram) that illustrates the structure of link-ages carried out by the clustering algorithm. Copheneticcorrelation, therefore, indicates how well the visual struc-ture illustrated by the dendrogram reflects real cluster-ing structures in the data. If the cophenetic correlationapproaches or exceeds 0.8, visual inspection of the den-drogram may allow for discernment of an appropriatenumber of clusters by cutting the dendrogram where theaverage dissimilarity between the groups jumps consider-ably between grouping iterations (Romesburg, 1984). Atboth stages of the clustering procedure for this study, thecophenetic correlations exceed 0.7. By this method, thefinal number of clusters extracted may not always reflectthe true modality of the data, but does facilitate analysisand interpretation of the data (Wilks, 2006).

The two-stage clustering solution employed in ourstudy allows for incorporation of known physical mech-anisms related to ENSO that influence variability in spa-tial patterns of southern Indian TC genesis, while alsofurther subdividing the TCs by the directions of theirmovement. It has been shown previously in observationalstudies that El Nino (La Nina) is associated with moreTCs forming west (east) of about 75 °E–85 °E (Ho et al.,2006; Camargo et al., 2007a; Kuleshov et al., 2008).This is also theoretically justified in that a favourablepool of warm SSTA and an unusually deep thermoclineare present in the southwestern Indian Ocean duringstrong ENSO warm events (Xie et al., 2002; Jury andHuang, 2004; Camargo et al., 2007a; Kuleshov et al.,2009), while unfavourable negative SSTAs, lower tro-pospheric easterly anomalies, and elevated mean sealevel pressure (MSLP) are present in the southeasternIndian basin (Reason et al., 2000; Larkin and Harri-son, 2002). It follows that a stratification of the longi-tudes of TC genesis points should echo the establishedregional physical characteristics of ENSO. Thus, the firststage is a univariate clustering considering only the ini-tial longitude of each TC, from which five clusters areextracted.

Qualitative comparison of the TC genesis clusters(Figure 1) with previous research suggests that stage1 of the analysis approximates the stratification of TCgenesis associated with ENSO influences as describedabove (refer to Figures 3 and 4 from Kuleshov et al.,2008, and Figure 2(c) from Ho et al., 2006). The threelarge groups in the main development region of thesouthern Indian Ocean contain 89% of the 191 TCs inthe initial cluster analysis. The western (eastern) groupis in the favourable genesis area during a warm (cool)phase of ENSO. Whereas Ho et al. (2006) placed theboundary separating the west and east genesis regions at75 °E, Kuleshov et al. (2008) suggested 85 °E. Therefore,in stage 1 of the analysis, the central group representsa region of transition where the differing influences ofENSO phases on TC genesis are not well defined. As thethree large groups across the main body of the southernIndian basin comprise the majority of TCs in the sample,

Figure 1. Map of five cluster solution of southwestern Indian Ocean tropical cyclones clustered by initial longitude. Only the three maindevelopment subregions (western, central, eastern) are considered in subsequent analysis. Note that this study only considers TCs which formed

or passed west of 90 °E during their life cycles.



Figure 2. Six TC trajectory clusters, arranged according to group size, within the main development regions for the southern IndianOcean between 54 °E and 110 °E. a) C1, eastern genesis/southwest–south movement; b) C2, central/west-southwest; c) C3, western/

west-southwest; d) C4, western/south-southeast; e) C5, eastern/west; f) C6, central/south-southeast.

only these are carried forth for the second stage ofclustering.

The second stage subsequently applies a bivariatecluster analysis within each of the three genesis clustersin the main TC development regions bounded by 54 °Eand 110 °E. Each genesis group is further subdividedinto eastward and westward trajectories using the finallatitude and longitude of each respective TC. One ofthe principal goals of this study is to investigate therelationship of ENSO and SIOD to the directions of TCmovements. The second stage of clustering therefore isintended to further subdivide TCs within each region thatmove mostly westward and threaten land from those thatmove southward or eastward and remain at sea. The two-stage solution, incorporating both the initial and finalgeographic locations of the TCs, results in six groupsof trajectories within the main body of the tropical andsubtropical southern Indian Ocean. These are presentedin greater detail in Section 3.1.

2.3. Analysis of variance

The six trajectory groups are compared in terms of theirmedian monthly values of SSTA indices representingboth ENSO and SIOD through ANOVA. Each TC isassigned the monthly values of the index correspond-ing to the month(s) spanning each TC’s life cycle. If astorm’s life cycle bridged two months, the two index val-ues are averaged. For ENSO, five-month running meansof standardised Nino-3.4 region anomalies (N3.4, 5°N,5 °S, 170 °W, 120 °W) are used (data obtained onlineat http://www.cpc.noaa.gov/data/indices/), and for SIODthe Subtropical Dipole Index (SDI) is used (obtained athttp://www.jamstec.go.jp/res/ress/behera/iosdindex.html).SDI is calculated by subtracting the east pole SSTA(18 °S–28 °S, 90 °E–100 °E) from the west pole SSTA(27 °S–37 °S, 55 °E–65 °E) (Behera and Yamagata, 2001).

A non-parametric rank test, the Kruskal-Wallis (KW)test with ties adjustment (Kruskal and Wallis, 1952;Higgins, 2004), is used to compare the medians of N3.4



Table I. Six main groups of TC trajectories, ranked by number of TCs in each group. Group median initial and final longitudesare in decimal degrees east, and median final latitudes are in decimal degrees south.

Clusterdesignation

Numberof TCs

Median initiallongitude (°E)

Genesisregion

Median finallatitude (°S)

Median finallongitude (°E)

Direction ofmovement

C1 40 94.3 Eastern 23.4 82.4 SW/SC2 39 78.0 Central 25.2 56.1 W/SWC3 39 64.3 Western 26.9 49.8 W/SWC4 22 66.1 Western 26.5 68.0 S/SEC5 16 99.1 Eastern 18.7 58.0 WC6 14 81.2 Central 20.5 83.6 S/SE

and SDI for the six principal trajectory groups. The nullhypotheses for the tests are that the median values of N3.4and SDI are not significantly different across all six TCclusters. The alternative hypotheses are that there existsignificant differences in at least one pair of TC groups intheir median values of N3.4 or SDI. To ensure the resultsof the KW tests are appropriately interpreted, Modified-Levene Equal-Variance Tests are also applied (Brown andForsythe, 1974). Dunn’s rank sums procedure is usedto test statistical significance in multiple comparisons(Dunn, 1964).

To complement the ANOVA tests, SSTA compositesare also constructed for each cluster using the NationalOceanic and Atmospheric Administration (NOAA) Opti-mum Interpolation (OI) SST Version 2 (Reynolds et al.,2002). An SST average is extracted based on the ini-tial and final dates for each respective TC’s life cycle.Monthly SST averages are also created based on theperiod 1981–2008. To obtain SST anomaly compositesfor each TC, the 26-year monthly composite is subtractedfrom each respective TC event composite. For exam-ple, TC Jaya’s life cycle spanned March–April of 2007.Therefore, the SSTA composite map for Jaya is the SSTcomposite for that TC’s life cycle minus the combinedMarch–April 26-year SST composite.

Finally, to test the interaction of ENSO and SIODphases with respect to the TC trajectory types, threecontingency tables are constructed stratifying TCs thatformed in the same TC genesis subregions by six cat-egories of ENSO and SIOD interactions: ENSO warmevent with positive SIOD (E+S+), ENSO warm eventwith negative SIOD (E+S−), ENSO neutral with posi-tive SIOD (E∼S+), ENSO neutral with negative SIOD(E∼S−), ENSO cool event with positive SIOD (E−S+),and ENSO cool event with negative SIOD (E−S−).Months in which each TC occurred are assigned as ENSOwarm, neutral, or cool phases based on a criteria of atleast six consecutive months with +/− 0.4 °C anomaliesfor five-month running means of the standardised N3.4index. This is a very similar methodology for determin-ing ENSO phase as outlined by Trenberth (1997) andimplemented recently in Lau et al. (2008) and Antico(2009). The principal difference is that our base period forstandardisation of the N3.4 index is limited to the studyperiod of 1979–2008. Fisher’s Exact Test provides exactp-values for contingency table tests, thus it is applied

such that robust results may be obtained despite the smallsample sizes (Higgins, 2004).

3. Results

3.1. Cluster analysis

The two-stage cluster analysis assigns the 170 maindevelopment region TCs into 6 groups of trajectories,according to their initial longitudes and final latitudesand longitudes. They are ranked by the number of TCswithin each group, designated simply as C1 through C6(Table I). Almost 70% of the 170 TCs are containedwithin the 3 largest groups (C1-C3), and these display anet southwesterly component of movement. This is to beexpected given that typical background conditions wouldallow for such an average motion around the northwestfringes of the southern Indian subtropical anticyclone.

The largest group is C1 with 40 TCs. These formedbetween 87 °E and 110 °E and followed southwestwardand southward trajectories (Figure 2(a)). C2 is comprisedof 39 TCs which formed between 73 °E and 87 °E andmoved westward or southwestward (Figure 2(b)). GroupC2 threatened land with relative frequency (16 of 39within 200 km of Madagascar, Mauritius, La Reunion, orMozambique) while C1 storms remained over the openocean. The third largest group, C3 (39 TCs), originatedbetween 54 °E and 73 °E with a large proportion (33of 39) that moved westward or southwestward within200 km of inhabited areas (Figure 2(c)).

The remaining 30% of TCs comprise the 3 smallergroups. These represent marked departures from the typi-cal net southwestward movement, and are therefore likelycandidates for association with notable departures fromthe more typical southwestward steering flow regime,such as might occur with strong phases of ENSO orSIOD. C4 (Figure 2(d)), in contrast to C3, developedbetween 54 °E and 73 °E with a dominant displacementsouthward and southeastward. Though the 22 TCs ingroup C4 formed in close proximity to the inhabitedislands in the southwestern Indian Ocean, they seldompassed within 200 km of Madagascar or the Mascareneislands. Having formed between 87 °E and 110 °E, groupC5 (Figure 2(e)) is distinguished from C1 by longer dis-tance tracks with more westward trajectories, 10 of whichpassed west of 60 °E and threatened populated regions.



Table II. Median standardised anomalies for Nino-3.4 (N3.4)sea surface temperature and Subtropical Dipole Index (SDI).

Cluster ID Nino-3.4 SDI

C1 −0.05 −0.19C2 −0.04 −0.06C3 −0.22 0.34C4 0.73 −0.70C5 −0.37 0.77C6 0.16 0.14

The smallest of the 6 trajectory clusters, C6, formedbetween 73 °E and 87 °E (Figure 2(f)). Direction of move-ment is south and southeastward, keeping these systemswell out to sea.

3.2. KW ANOVA

3.2.1. ANOVA results

The clustering procedure using initial longitude and finallatitude and longitude allows TC trajectory clusters withcommon geographic origins and diverging paths to becompared. This is accomplished through ANOVA, wherewe compare the median values of the N3.4 index and theSDI (Table II) first across all groups and then in multiplepair-wise comparisons. The results of the KW tests areto reject H0 of no difference between the median valuesof both N3.4 and SDI (α = 0.001, p-values 0.00019 and0.00012, respectively). Modified Levene Equal-Variancetests are also employed, and H0 of homoscedasticity isnot rejected for either N3.4 or SDI (α = 0.05, p-values0.88 and 0.22, respectively). Thus, there is statisticalevidence to support the assertion that SSTs in the N3.4and SDI regions are significantly and contemporaneouslyassociated with different types of southwestern Indian TCtrajectories.

Several other variables were tested using the sameKW ANOVA method. Previous studies suggested a link-age between the tropical Indian Ocean Dipole or ZonalMode (Saji et al., 1999; Webster et al., 1999) and south-ern Indian TC activity (Leroy and Wheeler, 2008; Chanand Liu, 2009; Vitart et al., 2010). Klinman and Rea-son (2008) and Chang-Seng and Jury (2010a) noted thatvariability of the Southern Hemisphere planetary wavescould also affect southwestern Indian TC trajectories.Vitart et al. (2010) include the Trans-Nino Index (Tren-berth and Stepaniak, 2001) as a significant predictor ofsouthern Indian TC activity in association with ENSOvariability, consistent with recent findings of associationbetween ENSO variability and TC activity in the NorthAtlantic and North Pacific basins (Kim et al., 2009; Chenand Tam, 2010; Kim et al., 2010). Accordingly, we car-ried out the same ANOVA procedure using TNI, DipoleMode Index (DMI) (Saji et al., 1999), and an Antarc-tic Oscillation Index (AAO) (obtained from NOAA athttp://www.cpc.noaa.gov/products/precip/CWlink/dailyao index/aao/aao index.html). We did not find significant

Table III. Multiple comparison Z-value tests for Nino-3.4region stratified by the six TC trajectory clusters.

Nino-3.4 C1 C2 C3 C4 C5 C6

C1 0.7929 1.0209 3.2288b 1.8009 0.4907C2 0.2266 3.8834b 1.1933 1.0618C3 4.0759b 1.0205 1.2265C4 4.2297b 2.0611a

C5 1.8720C6

a Indicates significant differences of median Nino-3.4 values at α =0.05.b Indicates significance at α = 0.01.

Table IV. Multiple comparison Z-value tests for SDI regionstratified by the six TC trajectory clusters.

SDI C1 C2 C3 C4 C5 C6

C1 0.6004 1.9028 2.1531a 3.1125b 1.0592C2 1.3024 2.6538b 2.6414b 0.6178C3 3.7465b 1.6118 0.3485C4 4.5597b 2.6445b

C5 1.6232C6

a Indicates significant differences of median SDI values at α = 0.05.b Indicates significance at α = 0.01.

differences between the six TC trajectory types basedupon these three additional indices (results not shown).

3.2.2. Western region TC types: C3 versus C4

To ascertain exactly which TC types are more signif-icantly associated with phases of ENSO and/or SIOD,Dunn’s rank-sum procedure is employed for multiplecomparisons of the median values of N3.4 and SDI acrossthe 6 trajectory clusters. Results of the multiple compar-isons for N3.4 (Table III) and SDI (Table IV) indicatethat TC type C4 is significantly more associated with ElNino and the negative SIOD mode than the other 5 tra-jectory types. The SSTA composites for C4 corroboratethese results. The composite SSTA pattern in the IndianOcean (Figure 3(d)) is suggestive of the negative SIOD,which is characterised by warm anomalies from Mada-gascar arcing eastward and southeastward toward south-western Australia (Reason, 1999; Behera et al., 2000;Behera and Yamagata, 2001; Qian et al., 2002; Suzukiet al., 2004). Positive SSTA exceeding 1 °C are also evi-dent between 120 °W and 170 °W in the N3.4 region(Figure 4(d)).

El Nino is well known to associate with westerly windanomalies over the southwestern Indian basin during anaustral warm season (Vitart et al., 2003; Todd et al.,2004; Yoo et al., 2006). TTTs are, likewise, associatedwith westerly anomalies over the same geographic region,with vertical extent from the boundary layer up to500 hPa and a west–east signature that can extend across



Figure 3. Sea surface temperature anomaly composites for the southern Indian Ocean corresponding to the six TC trajectory clusters presentedin Figure 2. SDI West (27 °S–37 °S, 55 °E–65 °E) and SDI East (18 °S–28 °S, 90 °E–100 °E) regions are shown as boxes in each image. Positive

(negative) values are symbolised by dark solid (light broken) lines with units in°C.

the entire southern Indian Ocean (Todd and Washington,1999). Reason (2002) found the TTTs also to vary inassociation with SSTA dipoles similar in pattern to theSIOD of Behera and Yamagata (2001), and the deepestconvection coincident with TTTs has likewise beenfound to shift from southern Africa to the southwesternIndian Ocean in association with warm ENSO events(Fauchereau et al., 2009; Pohl et al., 2009; Manhiqueet al., 2011). Given the high significance of El Ninoand negative SIOD for C4 relative to all other TCtrajectory types, coupled with the propensity of TTTsto extend eastward over Madagascar and the westernocean during both of these phases, this suggests thatC4 type TCs are highly likely to occur when ENSO isin warm phase and SIOD is simultaneously in negativemode. This 30° –35° eastward shift in the regional tropical

convective maximum and the related eastward shift ofpoleward convective outflow, and the subtropical jetstream, provides a logical explanation by which C4-type TCs are often swept south and southeastward awayfrom Madagascar. This is consistent with the synoptic-scale teleconnections of Hoerling et al. (1997) and theS-moving TC kinematic and thermodynamic profile ofChang-Seng and Jury (2010b).

Having established the ocean–atmospheric connec-tions for C4, it is now appropriate to compare with C3,which forms in the same western region, yet approachesinhabited regions with greater frequency. This group issignificantly different from C4 in median values of N3.4and SDI (Tables III and IV), suggesting that C3 occursmore frequently during La Nina or a positive SIOD phase.The SSTA composite maps for C3 (Figures 3(c) and 4(c))



Figure 4. Sea surface temperature anomaly composites for the equatorial Pacific Ocean corresponding to the six TC trajectory clusters presentedin Figure 2. Nino-3.4 region (5 °N–5 °S, 120 °W–170 °W) is shown with a box in each image. Positive (negative) values are symbolised by dark

solid (light broken) lines with units in°C.

depict warm anomalies south and southeast of Madagas-car, and cool anomalies in the equatorial Pacific. SSTAcomposite difference maps for C4 and C3 (Figures 5(a)and 6(a)) further support the importance of the SDI andN3.4 regions in distinguishing between these trajectorytypes. Enhanced easterly and southeasterly trade windsacross the subtropical and tropical southern Indian basinduring La Nina tend to coincide with more westwardmoving TCs that would threaten inhabited land masseswith higher frequency such as those in group C3 (Reasonet al., 2000; Vitart et al., 2003; Chang-Seng and Jury,2010a and 2010b). Furthermore, in accordance with apositive SIOD phase, warm SSTA in the southwesternIndian Ocean are associated with increased convectiveprecipitation over southern Africa as the TTTs are more

frequently anchored to the Angola thermal low (Todd andWashington, 1999; Reason, 2001; Reason, 2002). For C3-type TCs that turn poleward of 25 °S late in their lifecycle, interaction with TTTs over the African subconti-nent would help explain late re-curvature.

Further supporting the difference between TC tracksbelonging to groups C3 and C4, the result of the Fisher’sExact Test indicates a highly significant associationbetween the TC types (C3 and C4) and the type ofinteraction between ENSO and SIOD (p-value = 0.0007,Table V). When ENSO was in warm phase and SIODwas simultaneously negative, type C4 TCs occurredfrequently, and type C3 TCs seldom occurred. Thesefindings support the assertion that type C4 storms arehighly associated with antiphasing of ENSO (warm)



Figure 5. Sea surface temperature anomaly composite differences in the southern Indian Ocean for (a) C4 minus C3; (b) C1 minus C5; and(c) C6 minus C2. SDI West (27 °S–37 °S, 55 °E–65 °E) and SDI East (18 °S–28 °S, 90 °E–100 °E) regions are shown as boxes in each image.

Positive (negative) values are symbolised by dark solid (light broken) lines with units in°C.

Table V. Southern Indian Ocean western subregion tropicalcyclones by type and ENSO and SIOD interactions. E+represents a warm ENSO event, E∼ is a neutral event, andE− a cool event. S+ signifies a positive SIOD event and S−

a negative SIOD event.

C3 C4 Total

E+S+ 5 2 7E+S− 2 12 14E∼S+ 13 2 15E∼S− 5 4 9E−S+ 3 0 3E−S− 6 2 8Total 34 22 56

and SIOD (negative) wherein conditions are favourablefor TC–trough interaction over the southwestern IndianOcean. In contrast, type C3 storms occur significantlymore often than C4 storms when ENSO is neutraland SIOD is positive. In this situation, the Mascareneanticyclone and associated trade winds are present overthe central and southwestern Indian basin, and TTTstypically remain anchored over the African subcontinent.These results underscore the insufficiency of an opposingsymmetrical El Nino–La Nina paradigm in discerning

influences on southwestern Indian TC trajectories withoutconsideration of ENSO-independent variability (Vitartet al., 2003; Klinman and Reason, 2008).

3.2.3. Eastern region TC types: C1 versus C5

The TC trajectory type that develops in the eastern regionwith a typically westward motion, C5, is most stronglyassociated with the positive SIOD mode relative to theother clusters. Readily identifiable positive SIOD signalsare apparent both in the high significance of C5 in themultiple comparisons (Table IV), and in the SSTA com-posite maps (Figure 3(e)) which show spatially coherentwarm anomalies off the southeast coast of Africa intothe southwestern Indian Ocean, as well as cool anoma-lies extending eastward from Madagascar along 15 °S,25 °S. A La Nina SSTA pattern is also visible across theequatorial central Pacific (Figure 4(e)). The cool SSTAsin the southern Indian region are largely induced bydecreased air-to-sea latent heat flux and Ekman trans-port associated with strong trade winds and concomitantequatorward advection of relatively dry mid-latitude airmasses across a tight north–south SST gradient (Beheraand Yamagata, 2001; Hermes and Reason, 2005; Chiodiand Harrison, 2007; Huang and Shukla, 2007b). Thepresence of these SSTA infers anomalously strong tradewinds across the southern Indian Ocean north of 25 °S



Figure 6. Sea surface temperature anomaly composite differences in the equatorial Pacific Ocean for (a) C4 minus C3; (b) C1 minus C5; and(c) C6 minus C2. Nino-3.4 region (5 °N–5 °S, 120 °W–170 °W) is shown with a box in each image. Positive (negative) values are symbolised

by dark solid (light broken) lines with units in°C.

which could aid in steering TCs on longer duration andlower latitude westward tracks, as in Vitart et al. (2003).Therefore, when a positive SIOD phase occurs in tan-dem with an ENSO neutral or cool phase, the TCsof type C5 tend to remain at lower latitudes becausethe seasonally strong subtropical southern Indian anticy-clone precludes repeated northward intrusions by plan-etary waves (L’Heureux and Thompson, 2006) whichthen fail to foster eastward extension of convective clus-ters and their poleward outflows associated with theAngola thermal trough and South Indian ConvergenceZone (SICZ).

While TCs in group C5 develop in the southeasternIndian basin, they occasionally threaten land as theyprogress far westward. TCs in the counterpart group C1also develop in the east but re-curve into the middlelatitudes in the central or eastern ocean. The SSTA com-posite maps for C1 (Figures 3(a) and 4(a)) do not exhibita strong ENSO warm or cool pattern in the Pacific. How-ever, the data depict a northwest to southeast regionof warm SSTA in the central and southeast subtropicalsouthern Indian Ocean which resembles the spatial SSTApattern of the SIOD negative mode. Maps of the differ-ences between the C1 and C5 composites (Figures 5(b)and 6(b)) show marked differences in SSTA patternsin both the SIOD and ENSO regions. Specifically, the

presence of warm SSTA in the central and eastern tropi-cal and subtropical southern Indian basin and cool SSTAin the southwest subtropical Indian Ocean indicates weak-ened trade winds and increased frequency of frontal intru-sions. This would result in stronger cold air advectionpromoting decreased air-to-sea latent heat flux in thesubtropical southern ocean (Behera and Yamagata, 2001;Chiodi and Harrison, 2007). Also, reduced Ekman trans-port from the relatively weak trade winds to the north,in conjunction with increased air-to-sea heat flux fromwarm advection ahead of cold fronts, would allow forwarm SSTA patterns associated with the negative SIODphase or El Nino. Similar to TC type C4, troughs pen-etrating farther north into the central and southeasternIndian basin would influence C1 TCs to re-curve moreabruptly into the mid-latitudes than C5 TCs.

Following the same methodology as for types C3 andC4 above, a contingency table of ENSO and SIODinteractions is constructed for types C1 and C5 andFisher’s Exact Test is again applied (Table VI). Resultsindicate a significant association between the two easternregion trajectory types and ENSO and SIOD phaseinteractions, albeit with a more liberal alpha level (α =0.1, p-value = 0.097). Type C5 TC trajectories, thewestward moving group, did not occur at all coincidentwith any ENSO warm phase during our study period,



Table VI. As in Table V, but for eastern subregion tropicalcyclones.

C1 C5 Total


and only twice when ENSO was neutral but SIOD wasin negative mode. Type C5 storms occurred most oftenduring positive SIOD phases with only two during LaNina when SIOD was negative. The data suggest for theeastern region that it is not sufficient to ascribe TCs oflong duration that move consistently westward only toLa Nina. Contemporaneous consideration of SIOD phaseis especially relevant when ENSO is neutral, and it isphysically consistent that a positive SIOD phase wouldinfluence more westward moving TCs, as seen in bothtrajectory types C3 and C5. Finally, it is also evident inanalysis of group C1 that a warm ENSO event combinedwith a negative SIOD phase results in a higher frequencyof re-curving TC trajectories than during an ENSO warmevent combined with a positive SIOD mode.

3.2.4. Central region TC types: C2 versus C6

The TC group with genesis in the central region andwestward and southwestward movement is C2. C2 is onlysignificantly different from C4 in the N3.4 comparisons(Table III), and is only significantly different than C5in the SDI comparisons (Table IV). However, the latterresult should not be interpreted to mean that C2 exhibitsa strong association with SIOD that is opposite in sign tothe strong association between C5 and the positive SIODphase. The statistical significance is in a strong positiveassociation (C5) compared to a very weak association(C2), not a strong positive association compared to astrong negative association. It is clear from both theSDI values and the SSTA composite map (Figure 3(b))that in this study C2 TCs are not associated obviouslywith either phase of the SIOD. The equatorial PacificSSTA composite map for C2 (Figure 4(b)) displays arecognisable La Nina pattern east of 180°, though theN3.4 values are not statistically significant in multiplecomparisons.

As in the previous sections, it is useful to compareC2 to the counterpart central region TC type C6, whichis characterised by more eastward trajectories than C2.These two groups are not statistically different in com-parisons of their median values of N3.4 and SDI. Norare there contiguous warm SSTA in the equatorial Pacific(Figure 4(f)) to suggest an El Nino association. The com-posite difference maps of C6 minus C2 (Figures 5(c)and 6(c)) identify the equatorial Pacific as a region of

Table VII. As in Table V, but for central subregion tropicalcyclones.

C2 C6 Total


note in distinguishing the two TC types, but there is nota significant statistical nor an obvious geospatial differ-ence that would allow for more concise interpretation.Finally, a contingency table is constructed stratifying theTCs by type and by configurations of ENSO and SIOD(Table VII). The results of Fisher’s test also suggestno significant association to distinguish TC trajectoriesin types C2 and C6 by their frequencies during differ-ent interactions of ENSO and SIOD phases (p-value =0.559). When separated from the western and easternregions, the central region TC trajectories do not displayobvious modes relating re-curving and eastward tracks toEl Nino and negative SIOD, nor westward tracks to neu-tral ENSO and positive SIOD or La Nina. Given the smallsize of group C6 and that half of the TCs in the groupwere early or late season storms (September–October,or April–May), perhaps it is not surprising that no clearinteractive signal between ENSO and SIOD is found inthe central region. Both ENSO and SIOD are known toexhibit stronger signals at the height of austral summeras they are phase locked to the seasonal cycle.

4. Discussion

In this study, we build on the ENSO-TC track frameworkof Vitart et al. (2003) and Klinman and Reason (2008) byconsidering interactions of the ENSO and SIOD phases.We find that southward and southeastward moving TCs(types C1 and C4) are more likely to occur, particularly inthe southwestern Indian Ocean, when ENSO is in warmphase and SIOD is simultaneously in negative mode.Furthermore, westward and southwestward moving TCs(C3 and C5) are likely during a cool ENSO event (relativeto warm ENSO), but are most likely to occur whenENSO is neutral and SIOD is in positive mode. A keyimplication of these results is that periods of either strongwesterly or easterly steering flow in the southwesternIndian Ocean should not be attributed solely to anyparticular ENSO phase, but the simultaneous SIOD phaseshould be considered as well.

The SIOD-TC trajectory relationship proposed in thispaper may be applied to TC Favio of 2007 to explainits unusual direction of motion south of Madagascarand back to the northwest over Mozambique. Klinmanand Reason (2008) noted that such a westward track



during a weak El Nino year did not follow the ENSO-TCsteering model of Vitart et al. (2003), and suggested thatmonthly data might be more appropriate than seasonaldata in discerning southern Indian TC track variability.Using monthly data, the SDI was strongly positive (1.35)in February 2007, which is consistent with both thewarm SSTA south of Madagascar and persistent tradewinds which allowed Favio to remain at a lower latitude(rather than re-curving into higher latitudes) and provideda favourable thermodynamic environment for it to re-intensify southwest of Madagascar. While the example ofFavio is well suited to demonstrate how SIOD mode canassist in understanding TC track behaviour when ENSOis neutral or weakly in warm or cool phase, it must also beacknowledged that SIOD mode, like ENSO phase, is nota perfect predictor of southern Indian TC track direction.

This study suggests that inclusion of an index forSIOD mode could add predictive power in statisticalmodelling of southwestern Indian TC occurrence atintraseasonal or even seasonal temporal scales. Indirectevidence in support of this assertion exists in the peer-reviewed literature. Leroy and Wheeler (2008) includethe second rotated principal component of Indo-PacificSSTA as a significant predictor (variable SST2) in theirlogistic regression model of Southern Hemisphere weeklyTC genesis and activity. Their SST2 variable is verysimilar in construction and spatial pattern (shown intheir Figure 5(b), which was adapted from Drosdowskyand Chambers, 2001) to the subtropical SSTA dipolepatterns identified in numerous previous studies (Reason,1999; Behera and Yamagata, 2001; Qian et al., 2002;Suzuki et al., 2004; Huang and Shukla, 2007b). In theupdated operational statistical model (Vitart et al., 2010),the SST2 variable is replaced by the tropical Dipole ModeIndex of Saji et al. (1999). Our results suggest that anindex capturing the subtropical dipole phenomenon (suchas SDI) could be at least as important as the tropical modein modelling of southwestern Indian TC occurrence ortrack direction. As a proxy measure of the variability instrength and position of the TTTs and southern Indiansubtropical anticyclone at subseasonal intervals, it isphysically consistent that SDI should be significantlyassociated with variability of southwestern Indian TCtrajectories and, therefore, occurrence at subseasonalintervals. As Hermes and Reason (2005) observe, theSIOD anomalies may not always be adequately capturedby the SDI as constructed by Behera and Yamagata(2001). More research is needed to index SIOD-likepatterns with varying spatial signatures before these canmost effectively be applied in prediction of TC activityor track direction.

The strong tendencies for certain TC trajectories tooccur with certain configurations of ENSO and SIOD(Tables V and VI) also raise the possibility that theseconfigurations are not only specific to types of TC trajec-tories, but actually reflect preferred ocean–atmospherepatterns during southern Indian TC season. Using thesame methodology as described in Section 2.1 for assign-ment of El Nino, neutral, and La Nina months, SDI

Figure 7. Percentage of months during November–March over theperiod 1979–2007 when ENSO was in cool phase (ENSO−), neutralphase (ENSO∼), or warm phase (ENSO+) and SDI was simultaneously

in negative or positive mode.

was negative 64% of the time coincident with El Nino(Figure 7) during November–March over the period1979–2007, whereas SDI was positive 61% of the timecoincident with neutral ENSO conditions. La Nina andboth positive and negative SIOD modes occurred withequal frequency during the study period. These monthlypercentages are similar to daily percentages for coin-cidences of Antarctic Oscillation (AAO) and ENSOphases presented in Carvalho et al. (2005) (seen in theirFigure 6). There is a strong possibility that the AAO orSouthern Annular Mode (SAM) is associated with bothENSO and SIOD and subsequently with TTTs and TCtrajectory variability through the modification of plane-tary waves (Hermes and Reason, 2005; L’Heureux andThompson, 2006; Klinman and Reason, 2008; Manhiqueet al., 2011; Chang-Seng and Jury, 2010a).

5. Conclusions

The goal of this work was to investigate the influencesof both the SIOD and ENSO on southwestern IndianOcean TC trajectories. We grouped TCs by their gene-sis regions and subsequent trajectories and then employednon-parametric ANOVA to test for differences of monthlyvalues of N3.4 and SDI corresponding to each TC’s lifecycle. The results indicate that both N3.4 and SDI areassociated with significant differences in the directions ofTC trajectories in the study region. Specifically, our worksuggests that southern Indian basin TCs originating inthe regions 54 °E–73 °E or 87 °E–110 °E, and followingwestward and southwestward trajectories (southward andsoutheastward trajectories) are significantly more associ-ated with a cool or neutral ENSO phase and a positiveSIOD mode (warm ENSO and negative SIOD mode).We do not find the same statistically significant rela-tionships in the central ocean approximately bounded by73 °E–87 °E.

The results presented in this study go beyond anENSO-only model of southwestern Indian TC trajectoryvariability. When considered together with ENSO, the



SIOD (using the SDI) allows a clearer understandingthat when ENSO is in warm phase and SIOD is neg-ative the regional ocean–atmosphere patterns stronglyfavour a shift of tropical temperate troughs (TTTs) overthe southwestern Indian Ocean coincident with the warmSSTA pool. In this situation, the TTTs are positionedto frequently steer TCs away from inhabited areas onthe western rim of the basin. When ENSO is neutralor in cool phase and SIOD is positive, TTTs do notshift persistently eastward over the southwestern IndianOcean and well developed tropical systems are frequentlysteered along more westward and southwestward tracks.It is important to discern the differences in these con-figurations and could be of great use in the future toprovide forewarning perhaps a month or more in advanceof enhanced potential for TC strikes to the heavily popu-lated and socially vulnerable nations of Madagascar andMozambique.

In broad terms, much additional work is needed tounderstand the asymmetrical and non-linear tropical andextratropical interactions that produce the SIOD SSTApatterns, and how these bear influence on southwesternIndian TC trajectories at intraseasonal and seasonal timescales. In particular, understanding of the relationshipbetween TTTs and TC activity could benefit from deeperinquiry as this topic has not been comprehensively exam-ined in the literature. Future research should also explorealternative indexing procedures to most effectively cap-ture the geographic location and extent as well as mag-nitude of SIOD events, and test the application of thesealternative SIOD indexes for potential improvement ofoperational intraseasonal to seasonal southwestern Indianbasin TC prediction schemes.

Acknowledgments

The authors extend thanks to the two anonymous review-ers for their constructive comments. We also thank PeterWaylen and Tim Fik for their valuable suggestions dur-ing the earlier stages of this work. This research wasundertaken at the University of Florida as part of thelead author’s M.S. thesis. He wishes to thank the UFGeography Department for funding and support.

ReferencesAntico PL. 2009. Relationships between autumn precipitation

anomalies in southeastern South America and El Nino eventclassification. International Journal of Climatology 29: 719–727,DOI:10.1002/joc.1734.

Behera SK, Salvekar PS, Yamagata T. 2000. Simulation of interannualSST variability in the tropical Indian Ocean. Journal of Climate13: 3487–3499, DOI:10.1175/1520-0442(2000)013<3487:SOISVI>2.0.CO;2.

Behera SK, Yamagata T. 2001. Subtropical SST dipole events in thesouthern Indian Ocean. Geophysical Research Letters 28: 327–330,DOI:10.1029/2000GL011451.

Bessafi M, Wheeler MC. 2006. Modulation of South Indian Oceantropical cyclones by the Madden Julian Oscillation and convectivelycoupled equatorial waves. Monthly Weather Review 134: 638–656,DOI:10.1175/MWR3087.1.

Brown MB, Forsythe AB. 1974. Robust tests for the equality ofvariances. Journal of the American Statistical Association 69:364–367.

Brown ML. 2009. Madagascar’s cyclone vulnerability and the globalvanilla economy. The Political Economy of Hazards and Disasters,Jones EC, Murphy AD. (eds). Altamira Press: New York.

Camargo SJ, Emanuel KA, Sobel AH. 2007a. Use of a genesispotential index to diagnose ENSO effects on tropical cyclone genesis.Journal of Climate 20: 4819–4834, DOI:10.1175/JCLI4282.1.

Camargo SJ, Robertson AW, Barnston AG, Ghil M. 2008. Clusteringof eastern North Pacific tropical cyclone tracks: ENSO andMJO effects. Geochemistry Geophysics Geosystems 9: Q06V05,DOI:10.1029/2007GC001861.

Camargo SJ, Robertson AW, Gaffney SJ, Smyth P, Ghil M. 2007b.Cluster analysis of typhoon tracks. Part I: General properties. Journalof Climate 20: 3635–3653, DOI:10.1175/JCLI4188.1.

Camargo SJ, Sobel AH, Barnston AG, Klotzbach PJ. 2010. Theinfluence of natural climate variability on tropical cyclones, andseasonal forecasts of tropical cyclone activity. Global perspectiveson tropical cyclones: From science to mitigation, 2nd edn. Chan JCL,Kepert JD (eds). World Scientific: New York.

Caroff P. 2009. Operational practices for best-track elaboration atRSMC La Reunion. Presented 5 May 2009 at the InternationalBest-Track Archive for Climate Stewardship (IBTrACS) Workshop,NOAA National Climatic Data Center: Asheville, NC.

Carvalho LM, Jones C, Ambrizzi T. 2005. Opposite phases ofthe Antarctic Oscillation and relationships with intraseasonal tointerannual activity in the tropics during the austral summer. Journalof Climate 18: 702–718, DOI:10.1175/JCLI-3284.1.

Chambers DP, Tapley BD, Stewart RH. 1999. Anomalous warming inthe Indian Ocean coincident with El Nino. Journal of GeophysicalResearch 104: 3035–3047, DOI:10.1029/1998JC900085.

Chan JCL, Liu KS. 2009. Interannual variations of tropical cycloneactivity in the Southern Hemisphere. Paper presented at 9th Inter-national Conference on Southern Hemisphere Meteorology andOceanography, Australian Meteorological Society: Melbourne, Vic-toria, Australia. http://www.bom.gov.au/events/9icshmo/manuscripts/M1500 Chan.pdf

Chang-Seng DS, MR Jury. 2010a. Tropical cyclones in the SWIndian Ocean. Part 1: inter-annual variability and statisticalprediction. Meteorology and Atmospheric Physics 106: 149–162,DOI:10.1007/s00703-009-0055-2.

Chang-Seng DS, MR Jury. 2010b. Tropical cyclones in the SW IndianOcean. Part 2: structure and impacts at the event scale. Meteorologyand Atmospheric Physics 106: 163–178, DOI:10.1007/s00703-010-0059-y.

Chen G, Tam C-Y. 2010. Different impacts of two kinds ofPacific Ocean warming on tropical cyclone frequency over thewestern North Pacific. Geophysical Research Letters 37: L01803,DOI:10.1029/2009GL041708.

Chiodi AM, Harrison DE. 2007. Mechanisms of summertime southernIndian Ocean sea surface temperature variability: On the importanceof humidity anomalies and the meridional advection of water vapor.Journal of Climate 20: 4835–4852, DOI:10.1175/JCLI4271.1.

Choi K-S, Kim B-J, Choi C-Y, Nam J-C. 2009. Cluster analysis oftropical cyclones making landfall on the Korean peninsula. Advancesin Atmospheric Sciences 26: 202–210, DOI:10.1007/s00376-009-0202-1.

Chu JH, Sampson CR, Levine AS, Fukada E. 2002. The Joint TyphoonWarning Center tropical cyclone best-tracks, 1945–2000. NavalResearch Laboratory Reference Number NRL/MR/7540-02-16.

Chu P-S. 2004. ENSO and tropical cyclone activity. Hurricanes andTyphoons: Past, Present, and Future. Murnane RJ, Liu K-B. (eds).Columbia University Press: New York.

Cook KH. 2000. The South Indian Convergence Zone and interannualrainfall variability over southern Africa. Journal of Climate13: 3789–3804, DOI:10.1175/1520-0442(2000)013<3789:TSICZA>2.0.CO;2.

Drosdowsky W, Chambers LE. 2001. Near-Global sea surfacetemperature anomalies as predictors of Australian seasonalrainfall. Journal of Climate 14: 1677–1687, DOI:10.1175/1520-0442(2001)014<1677:NACNGS>2.0.CO;2.

Dunn OJ. 1964. Multiple comparisons using rank sums. Technometrics6: 241–252.

Elsner JB. 2003. Tracking hurricanes. Bulletin of the AmericanMeteorological Society 84: 353–356, DOI:10.1175/BAMS-84-3-353.

Elsner JB, Liu K-B. 2003. Examining the ENSO-typhoon hypothesis.Climate Research 25: 43–54, DOI:10.3354/cr025043.

Evans JL, Allan RJ. 1992. El Nino/Southern Oscillation modificationto the structure of the monsoon and tropical cyclone activity in



the Australasian region. International Journal of Climatology 12:611–623, DOI:10.1002/joc.3370120607.

Fauchereau N, Pohl B, Reason CJC, Rouault M, Richard Y. 2009.Recurrent daily OLR patterns in the Southern Africa/SouthwestIndian Ocean region, implications for South African rainfall and tele-connections. Climate Dynamics 32: 575–591, DOI:10.1007/s00382-008-0426-2.

Fauchereau N, Trzaska S, Richard Y, Roucou P, Camberlin P. 2003.Sea-surface temperature co-variability in the southern Atlantic andIndian Oceans and its connections with the atmospheric circulationin the Southern Hemisphere. International Journal of Climatology23: 663–677, DOI:10.1002/joc.905.

Gong X, Richman MB. 1995. On the application of cluster analysisto growing season precipitation data in North America east ofthe Rockies. Journal of Climate 8: 897–931, DOI:10.1175/1520-0442(1995)008<0897:OTAOCA>2.0.CO;2.

Gray WM, Sheaffer JD. 1991. El Nino and QBO influences ontropical cyclone activity. Teleconnections linking worldwide climateanomalies. Glantz MH, Katz RW, Nicholls N. (eds). CambridgeUniversity Press: Cambridge, UK; 535.

Harangozo SA, Harrison MSJ. 1983. On the use of synoptic data inindicating the presence of cloud bands over southern Africa. SouthAfrican Journal of Science 79: 413–414.

Hermes JC, Reason CJC. 2005. Ocean model diagnosis of inter-annual coevolving SST variability in the South Indian andSouth Atlantic Oceans. Journal of Climate 18: 2864–2882,DOI:10.1175/JCLI3422.1.

Higgins JJ. 2004. An Introduction to Modern Nonparametric Statistics.Brooks/Cole: Pacific Grove, CA; 366.

Ho C-H, Kim J-H, Jeong J-H, Kim H-S, Chen D. 2006. Variationof tropical cyclone activity in the South Indian Ocean: El Nino-Southern Oscillation and Madden-Julian Oscillation effects. Journalof Geophysical Research 111: D22101, DOI:10.1029/2006JD007289.

Hoerling MP, Kumar A, Zhong M. 1997. El Nino, La Nina, andthe nonlinearity of their teleconnections. Journal of Climate 10:1769–1786, DOI:10.1175/1520-0442(1997)010<1769:ENOLNA>2.0.CO;2.

Huang B, Shukla J. 2007a. Mechanisms for the interannual variabilityin the tropical Indian Ocean. Part I: The role of remote forcingfrom the tropical Pacific. Journal of Climate 20: 2917–2936,DOI:10.1175/JCLI4151.1.

Huang B, Shukla J. 2007b. Mechanisms for the interannual variabilityin the tropical Indian Ocean. Part II: Regional processes. Journal ofClimate 20: 2937–2960, DOI:10.1175/JCLI4169.1.

Jury MR. 1993. A preliminary study of climatological associationsand characteristics of tropical cyclones in the SW IndianOcean. Meteorology and Atmospheric Physics 51: 101–115,DOI:10.1007/BF01080882.

Jury MR, Huang B. 2004. The Rossby wave as a key mechanism ofIndian Ocean climate variability. Deep-Sea Research Part I: Oceano-graphic Research Papers 51: 2123–2136, DOI:10.1016/j.dsr.2004.06.005.

Jury MR, Pathack B. 1991. A study of climate and weather variabilityover the tropical Southwest Indian Ocean. Meteorology andAtmospheric Physics 47: 37–48, DOI:10.1007/BF01025825.

Jury MR, Pathack B, Parker B. 1999. Climatic determinants andstatistical prediction of tropical cyclone days in the Southwest IndianOcean. Journal of Climate 12: 1738–1746.

Jury MR, Pathack B, Wang B, Powell M, Raholijao N. 1993. Adestructive tropical cyclone season in the SW Indian Ocean: January-February 1984. South African Geographical Journal 75: 53–59.

Karoly DJ. 1989. Southern Hemisphere circulation features associatedwith El Nino-Southern Oscillation events. Journal of Climate2: 1239–1252, DOI:10.1175/1520-0442(1989)002<1239:SHCFAW>2.0.CO;2.

Kim H-M, Webster PJ, Curry JA. 2009. Impact of shifting patterns ofPacific Ocean warming on North Atlantic tropical cyclones. Science325: 77–80, DOI:10.1126/science.1174062.

Kim J-H, Ho C-H, Chu P-S. 2010. Dipolar redistribution ofsummertime tropical cyclone genesis between the PhilippineSea and the northern South China Sea and its possiblemechanisms. Journal of Geophysical Research 115: D06104,DOI:10.1029/2009JD012196.

Klein SA, Soden BJ, Lau N-C. 1999. Remote sea surface temperaturevariations during ENSO: Evidence for a tropical atmosphericbridge. Journal of Climate 12: 917–932, DOI:10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.

Klinman MG, Reason CJC. 2008. On the peculiar storm trackof TC Favio during the 2006–2007 Southwest Indian Ocean

tropical cyclone season and relationships to ENSO. Meteorologyand Atmospheric Physics 100: 233–242, DOI:10.1007/s00703-008-0306-7.

Knaff JA, Sampson CR. 2009. Southern hemisphere tropical cycloneintensity forecast methods at the Joint Typhoon Warning Center, PartI: Control forecasts based on climatology and persistence. AustralianMeteorological and Oceanographic Journal 58: 1–7.

Kossin JP, Camargo SJ, Sitkowski M. 2010. Climate modulation ofNorth Atlantic hurricane tracks. Journal of Climate 23: 3057–3076,DOI:10.1175/2010JCLI3497.1.

Kossin JP, Knapp KR, Vimont DJ, Murnane RJ, Harper BA. 2007.A globally consistent reanalysis of hurricane variability andtrends. Geophysical Research Letters 34: L04815, DOI:10.1029/2006GL028836.

Kruskal WH, Wallis WA. 1952. Use of ranks in one-criterion varianceanalysis. Journal of the American Statistical Association 47:583–621.

Kuleshov Y, de Hoedt G. 2003. Tropical cyclone activity in theSouthern Hemisphere. Bulletin of the Australian Meteorological andOceanographic Society 16: 135–137.

Kuleshov Y, Fawcett R, Qi L, Trewin B, Jones D, McBride J,Ramsay H. 2010. Trends in tropical cyclones in the South IndianOcean and the South Pacific Ocean. Journal of Geophysical Research115: D01101, DOI:10.1029/2009JD012372.

Kuleshov Y, Ming FC, Qi L, Chouaibou I, Hoareau C, Roux F. 2009.Tropical cyclone genesis in the Southern Hemisphere and itsrelationship with the ENSO. Annales Geophysicae 27: 2523–2538.

Kuleshov Y, Qi L, Fawcett R, Jones D. 2008. On tropical activity inthe Southern Hemisphere: Trends and the ENSO connection. Geo-physical Research Letters 35: L14S08, DOI:10.1029/2007GL032983.

Lander MA. 1994. An exploratory analysis of the relationship betweentropical storm formation in the Western North Pacific and ENSO.Monthly Weather Review 122: 636–651.

Landsea CW. 2000. El Nino-Southern Oscillation and the seasonal pre-dictability of tropical cyclones. El Nino: Impacts of Multiscale Vari-ability on Natural Ecosystems and Society. Diaz HF, Markgraf V.(eds). Cambridge University Press: Cambridge, UK.

Larkin NK, Harrison DE. 2002. ENSO warm (El Nino) and cold(La Nina) event life cycles: Ocean surface anomaly patterns, theirsymmetries, asymmetries, and implications. Journal of Climate 15:1118–1140, DOI:10.1175/1520-0442(2002)015<1118:EWENOA>2.0.CO;2.

Lattin JM, Carroll JD, Green PE. 2003. Analyzing Multivariate Data.Brooks/Cole: Pacific Grove, CA; 556.

Lau N-C, Leetmaa A, Nath MJ. 2008. Interactions between theresponses of North American climate to El Nino-La Nina and to thesecular warming trend in the Indian-Western Pacific Oceans. Journalof Climate 21: 476–494, DOI:10.1175/2007JCLI1899.1.

Leroy A, Wheeler MC. 2008. Statistical prediction of weekly tropicalcyclone activity in the Southern Hemisphere. Monthly WeatherReview 136: 3637–3654, DOI:10.1175/2008MWR2426.1.

L’Heureux ML, Thompson DWJ. 2006. Observed relationshipsbetween the El Nino-Southern Oscillation and the extratropi-cal zonal-mean circulation. Journal of Climate 19: 276–287,DOI:10.1175/JCLI3617.1.

Lindesay JA, Harrison MSJ, Haffner MP. 1986. The SouthernOscillation and South African rainfall. South African Journal ofScience 82: 196–198.

Lyons SW. 1991. Origins of convective variability over equatorialsouthern Africa during austral summer. Journal of Climate 4: 23–39,DOI:10.1175/1520-0442(1991)004<0023:OOCVOE>2.0.CO;2.

Manhique AJ, Reason CJC, Rydberg L, Fauchereau N. 2011. ENSOand Indian Ocean sea surface temperatures and their relationshipswith tropical temperate troughs over Mozambique and the SouthwestIndian Ocean. International Journal of Climatology 31: 1–13,DOI:10.1002/joc.2050.

Mason SJ, Jury MR. 1997. Climatic variability and change over south-ern Africa: a reflection on underlying processes. Progress in PhysicalGeography 21: 23–50, DOI:10.1177/030913339702100103.

Mavume AF, Rydberg L, Rouault M, Lutjeharms JRE. 2009. Clima-tology and landfall of tropical cyclones in the South-West IndianOcean. Western Indian Ocean Journal of Marine Science 8: 15–36.

Meehl GA. 1987. The annual cycle and interannual variability in thetropical Pacific and Indian Ocean regions. Monthly Weather Review115: 27–50, DOI:10.1175/1520-0493(1987)115<0027:TACAIV>2.0.CO;2.



Naeraa M, Jury MR. 1998. Tropical cyclone composite struc-ture and impacts over eastern Madagascar during January-March 1994. Meteorology and Atmospheric Physics 65: 43–53,DOI:10.1007/BF01030268.

Nicholson SE. 1997. An analysis of the ENSO signal in thetropical Atlantic and western Indian Oceans. InternationalJournal of Climatology 17: 345–375, DOI:10.1002/(SICI)1097-0088(19970330)17 : 4<345::AID-JOC127>3.0.CO;2–3.

Nicholson SE. 2003. Comments on “The South Indian Conver-gence Zone and interannual rainfall variability over south-ern Africa” and the question of ENSO’s influence on south-ern Africa. Journal of Climate 16: 555–562, DOI:10.1175/1520-0442(2003)016<0555:COTSIC>2.0.CO;2.

Pan YH, Oort AH. 1983. Global climate variations connected with seasurface temperature anomalies in the eastern equatorial Pacific Oceanfor the 1958-73 period. Monthly Weather Review 111: 1244–1258,DOI:10.1175/1520-0493(1983)111<1244:GCVCWS>2.0.CO;2.

Parker BA, Jury MR. 1999. Synoptic environment of compositetropical cyclones in the South-West Indian Ocean. South AfricanJournal of Marine Science 21: 99–115.

Parker D, 1999. Criteria for evaluating the condition of a tropicalcyclone warning system. Disasters 23: 193–216, DOI:10.1111/1467-7717.00113.

Pohl B, Fauchereau N, Richard Y, Rouault M, Reason CJC. 2009.Interactions between synoptic, intraseasonal and interannualconvective variability over southern Africa. Climate Dynamics 33:1033–1050, DOI:10.1007/s00382-008-0485-4.

Qian W, Hu H, Deng Y, Tian J. 2002. Signals of interannual andinterdecadal variability of air-sea interaction in the basin-wide IndianOcean. Atmosphere-Ocean 40: 1480–9214, DOI:10.3137/ao.400302.

Reason CJC. 1999. Interannual warm and cool events in thesubtropical/mid-latitude South Indian Ocean region. GeophysicalResearch Letters 26: 215–218, DOI:10.1029/1998GL900285.

Reason CJC. 2001. Subtropical Indian Ocean SST dipole eventsand southern African rainfall. Geophysical Research Letters 28:2225–2227, DOI:10.1029/2000GL012735.

Reason CJC. 2002. Sensitivity of the southern African Circulationto dipole sea-surface temperature patterns in the South IndianOcean. International Journal of Climatology 22: 377–393,DOI:10.1002/joc.744.

Reason CJC. 2007. Tropical cyclone Dera, the unusual 2000/01 tropicalcyclone season in the Southwest Indian Ocean and associated rainfallanomalies over southern Africa. Meteorology and AtmosphericPhysics 97: 181–188, DOI:10.1007/s00703-006-0251-2.

Reason CJC, Allan RJ, Lindesay JA, Ansell TJ. 2000. ENSO andclimatic signals across the Indian Ocean basin in the globalcontext: Part I, Interannual composite patterns. InternationalJournal of Climatology 20: 1285–1327, DOI:10.1002/1097-0088(200009)20 : 11<1285::AID-JOC536>3.0.CO;2-R.

Reason CJC, Keibel A. 2004. Tropical cyclone Eline and its unusualpenetration and impacts over the southern African mainland.Weather and Forecasting 19: 789–805, DOI:10.1175/1520-0434(2004)019<0789:TCEAIU>2.0.CO;2.

Revell CG, Goulter SW. 1986. South Pacific tropical cyclones and theSouthern Oscillation. Monthly Weather Review 114: 1138–1145.

Reynolds RW, Rayner NA, Smith TM, Stokes DC, Wang W. 2002.An improved in situ and satellite SST analysis for cli-mate. Journal of Climate 15: 1609–1625, DOI:10.1175/1520-0442(2002)015<1609:AIISAS>2.0.CO;2.

Romesburg HC. 1984. Cluster Analysis for Researchers. Wadsworth:North Carolina; 334.

Roux F, Chane-Ming F, Lasserre-Bigorry A, Nuissier O. 2004.Structure and evolution of intense tropical cyclone Dinanear La Reunion on 22 January 2002: GB-EVTD analysisof single Doppler radar observations. Journal of Atmosphericand Oceanic Technology 21: 1501–1518, DOI:10.1175/1520-0426(2004)021<1501:SAEOIT>2.0.CO;2.

Saji NH, Goswami BN, Vinayachandran PN, Yamagata T. 1999. Adipole mode in the tropical Indian Ocean. Nature 401: 360–363.

Silva JA, Eriksen S, Ombe ZA. 2010. Double exposure in Mozam-bique’s Limpopo River basin. The Geographical Journal 176: 6–24,DOI:10.1111/j.1475–4959.2009.00343.x.

Suzuki R, Behera SK, Iizuka S, Yamagata T. 2004. Indian Oceansubtropical dipole simulated using a coupled general circu-lation model. Journal of Geophysical Research 109: C09001,DOI:10.1029/2003JC001974.

Terray P, Dominiak S. 2005. Indian Ocean sea surface temperature andEl Nino-Southern Oscillation: A new perspective. Journal of Climate18: 1351–1368, DOI:10.1175/JCLI3338.1.

Todd M, Washington R. 1999. Circulation anomalies associatedwith tropical-temperate troughs in southern Africa and theSouth West Indian Ocean. Climate Dynamics 15: 937–951,DOI:10.1007/s003820050323.

Todd MC, Washington R, Palmer PI. 2004. Water vapour transportassociated with tropical-temperate trough systems over southernAfrica and the Southwest Indian Ocean. International Journal ofClimatology 24: 555–568, DOI:10.1002/joc.1023.

Trenberth KE. 1990. Recent observed interdecadal climate changes inthe Northern Hemisphere. Bulletin of the American MeteorologicalSociety 71: 988–993, DOI:10.1175/1520-0477(1990)071<0988:ROICCI>2.0.CO;2.

Trenberth KE. 1997. The definition of El Nino. Bulletin of the AmericanMeteorological Society 78: 2771–2777.

Trenberth KE, Stepaniak DP. 2001. Indices of El Nino evolution.Journal of Climate 14: 1697–1701.

Tyson PD, Preston-Whyte RA. 2000. The Weather and Climate ofSouthern Africa. Oxford University Press: Oxford, UK; 396.

van den Heever SC, D’Abreton PC, Tyson PD. 1997. Numericalsimulation of tropical-temperate troughs over southern Africa usingthe CSU RAMS model. South African Journal of Science 93:359–365.

van Loon H, Rogers JC. 1981. The Southern Oscillation. PartII: Associations with changes in the middle troposphere inthe Northern winter. Monthly Weather Review 109: 1163–1168,DOI:10.1175/1520-0493(1981)109<1163:TSOPIA>2.0.CO;2.

Venzke S, Latif M, Villwock A. 2000. The Coupled GCM ECHO-2.Part II: Indian Ocean response to ENSO. Journal of Climate 13:1371–1383.

Vitart F, Anderson D, Stockdale T. 2003. Seasonal forecasting of trop-ical cyclone landfall over Mozambique. Journal of Climate 16:3932–3945, DOI:10.1175/1520-0442(2003)016<3932:SFOTCL>2.0.CO;2.

Vitart F, Leroy A, Wheeler MC. 2010. A comparison of dynamicaland statistical predictions of weekly tropical cyclone activity in theSouthern Hemisphere. Monthly Weather Review 138: 3671–3682,DOI:10.1175/2010MWR3343.1.

Washington R, Preston A. 2006. Extreme wet years over southernAfrica: Role of Indian Ocean sea surface temperatures. Journal ofGeophysical Research 111: D15104, DOI:10.1029/2005JD006724.

Washington R, Todd M. 1999. Tropical-Temperate links in south-ern African and Southwest Indian Ocean satellite-derived dailyrainfall. International Journal of Climatology 19: 1601–1616,DOI:10.1002/(SICI)1097-0088(19991130)19 : 14<1601::AID-JOC407>3.3.CO;2-S.

Webster PJ, Moore AM, Loschnigg JP, Leben RR. 1999. Coupledocean-atmosphere dynamics in the Indian Ocean during 1997-98.Nature 401: 356–360.

Wilks DS. 2006. Statistical Methods in the Atmospheric Sciences.Elsevier: Boston; 627.

Wolter K. 1987. The Southern Oscillation in surface circulation andclimate over the tropical Atlantic, eastern Pacific, and Indian Oceansas captured by cluster analysis. Journal of Climate and AppliedMeteorology 26: 540–558.

Xie S-P, Annamalai H, Schott FA, McCreary, JP Jr. 2002. Struc-ture and mechanisms of South Indian Ocean climate vari-ability. Journal of Climate 15: 864–878, DOI:10.1175/1520-0442(2002)015<0864:SAMOSI>2.0.CO;2.

Yip CL, Wong KY, Li PW. 2006. Data complexity in tropicalcyclone positioning and classification. Data Complexity in PatternRecognition. Basu M, Ho TK. (eds). Springer-Verlag: Heidelberg,DOI:10.1007/978-1-84628-172-3 13.

Yoo S-H, Yang S, Ho C-H. 2006. Variability of the Indian Oceansea surface temperature and its impacts on Asian-Australianmonsoon climate. Journal of Geophysical Research 111: D03108,DOI:10.1029/2005JD006001.

Zinke J, Dullo W-C, Heiss GA, Eisenhauer A. 2004. ENSO andIndian Ocean subtropical dipole variability is recorded in acoral record off southwest Madagascar for the period 1659to 1995. Earth and Planetary Science Letters 228: 177–194,DOI:10.1016/j.epsl.2004.09.028.


Documents

The influences of ENSO and the subtropical Indian Ocean ...ufdcimages.uflib.ufl.edu/IR/00/00/12/00/00001/AshMatyasIJOC2012.pdfet al., 2003; Reason and Keibel, 2004; Ho et al., 2006;