Before 1 depicts the monthly number of tropical

Before further analysis, we studied the
climatology of tropical cyclones in South Indian Ocean. Figure 1 depicts the
monthly number of tropical cyclones averaged for the 1979-2012 period. In January, February
and March, more than 5 tropical cyclones are generated each month with the most
being in February, more than 3 tropical cyclones in April and December, about
1-1.5 tropical cyclones in May and November and about 0.5 tropical cyclones in
October. The remaining months very few tropical cyclones are generated (average
of less than 0.5 tropical cyclones per month). As the summary of these tropical
cyclones statistics, three-quarters of the total tropical cyclones formed over
the South Indian Ocean are observed in austral summer season (December through
March). Most of TCs originate from the
central tropical South Indian Ocean (east of Madagascar) and some from the
Mozambique Channel. Earlier studies suggest that TCs would increase
the frequency and intensity in response to global warming (Emanuel 1987, 2005); (Holland, 1997); (Knutson et al., 1998); (Webster et al., 2005); (Oouchi et al., 2006); (Tim Li et al., 2010). Even though warm
sea surface temperature (SST) is important factor in TC formation (Gray, 1968), numerous large
scale environment factors are also able to influence on TCs genesis and
development. Thus, an understanding of natural variability of TC activity on
interannual and even longer timescales is very essential to response the
possible effects of climate change on TC activity. As dominant interannual
modes in tropical oceans, ENSO and IOD may exert the great influence on TCs’
interannual variability. Particularly in northwest Pacific, northeast Pacific,
south Pacific, Atlantic, north Indian Ocean (NIO) and south Indian Ocean (SIO),
many studies reports, through induced large-scale atmosphere circulation, ENSO
modulates the frequency, intensity, and track of TCs in these ocean basins (Wang and Chan, 2002); (William and Young, 2007); (Wing et al., 2007); (Ng and Chan, 2012); ( Li, 2012); (Sumesh and Kumar, 2013); (Clifford et al., 2013);(Chung and Li, 2015); (Yu et al, 2016). Furthermore some studies described
TC activity in Indian Ocean related with ENSO e.g., (Ng & Chan, 2012), (Sumesh and Kumar,
2013), (Singh et al., 2001), (Camargo et al., 2007) and (Girishkumar and
Ravichandran, 2012);(Jury, 1993); (Jury et al., 1999); (Xie et al., 2002); (Kuleshov, 2003);(Kuleshov and de Hoedt, 2003),
IOD e.g. (Yuan & Cao, 2013a); ( Zhi Li et al., 2015)
convectively coupled equatorial waves including the Madden-Julian Oscillation (MJO)
e.g., (Bessafi and Wheeler, 2006), and the
stratospheric quasi-biennial oscillation (QBO); e.g.,(Jury,1993); (Jury et al., 1999). Ng and Chan, (2012) found that ENSO has significant
impact on large-scale parameters which provides less (more) favorable condition
for TC genesis and development over North Indian Ocean in an El Nino (La Nina)
year. During an El Nino (La Nina) year, weaker(stronger) low –level relative
vorticity, anomalous westerly (easterly) shear, less (more) Moist static
energy, weaker (stronger) 500-Zonal wind, lower (higher) 500geopotential height
and 800geopotential height, and weaker (stronger) divergence, over Bay of
Bengal during OND, could be observed (Ng and Chan, 2012). Ng an Chan, 2012 using accumulated
cyclone energy (ACE) and total numbers of TCs indicated El Niño (La Niña) had a
significant relationship to decreased (increased) TC activity over North Indian
Ocean. Sumesh and Kumar, 2013 found that the
air–sea interaction processes such as El Niño and El Niño Modoki events have
significant impacts on the tropical cyclones over North Indian Ocean, the
frequency of tropical cyclones is more (less) over Arabian Sea (Bay of Bengal)
during the El Niño Modoki years compared to the El Niño years. Singh et al., 2001b studied the impact
of ENSO and its relation to cyclonic activity in the Bay of Bengal during the
summer monsoon season (July–August). Their studies showed that the frequency of
formation of monsoon depressions in the Bay of Bengal increased during warm
ENSO phase. Singh et al., (2000)  reported a reduction in tropical cyclone
activity over the Bay of Bengal in severe cyclone months May and November during
warm phases of ENSO. Camargo et al., (2007) studied the effect
of ENSO on the genesis potential index in the world ocean and reported that
there is a shift in the genesis potential from the northern to southern part of
the Bay of Bengal between La Niña and El Niño year and it was mainly due to
wind shear. (Girishkumar and Ravichandran, (2012) reported that ENSO
significantly influences the frequency, genesis location and intensity of TC
during the primary TC peak season in the Bay of Bengal. During La Niña (El
Niño) regimes, TC activity is relatively more (less) pronounced and the genesis
location is shifted to the east (west) of 87°E in the Bay of Bengal. (Jury, 1993) found that the
frequency of TC genesis in the western SIO increases during the east phase of
QBO, but the impact of ENSO is not significant because of increased upper
westerly shear, in spite of convection being enhanced during El Niño summers. Xie et al., (2002) also examined TC day
in the western SIO with respect to in situ thermocline variability and found
that number of TC day is greatly increased in the offshore east of Madagascar. (Kuleshov and de Hoedt, (2003) demonstrated that TC
numbers were increased between 85?E and 105?E during La Niña years compared to
El Niño years. Bessafi and Wheeler, (2006) found that the MJO
extensively modulates low-level vorticity and vertical wind shear and further
modulates the number of TCs. Kuleshov et al.,(2008 , 2009), who observed a
shift in the geographic position of TC intensity depending on the ENSO phase,
that could be explained by changes in geographical distribution of relative
humidity and vorticity across the basin. Ho et al., (2006) observed a westward
shift of TC activity that they attributed to anomalous anticyclonic circulation
in the South east Indian Ocean during El Nino years. Not only ENSO influence, Ho et al., (2006) also examined the
importance of Madden-Julian Oscillation. (Yuan and Cao, 2013b) explained that tropical
cyclone activities over North Indian Ocean are closely related to IOD SST
anomalies. There are more North Indian Ocean TCs, especially more
westward-moving TCs west of 90°E in the Bay of Bengal, during IOD negative
phases and fewer North Indian Ocean TCs during positive IOD phases. SST heating
or cooling with an anomalous IOD pattern can modify atmospheric circulations,
which in turn influence North Indian Ocean TC activity. Also Yuan and Cao, (2013), described that, the
influence of IOD events on North Indian Ocean TC activity is through changing
conditions of TC genesis, and through modifications of steering flow that in
turn affect subsequent TC trajectories. Zhi Li et al., (2015), concluded that
negative IOD mainly affects South East Indian Ocean TC increase, also suggested
that relative humidity contributes mostly to the TCs increase, vertical wind
shear provides the secondary positive contribution and vorticity makes a weak
positive contribution. Mahala et al.,( 2015), studied the impacts
of ENSO and IOD on tropical activity in the Bay of Bengal and observed that
maximum frequency of TC is during La Nina years, negative