Impact of atmospheric and oceanic conditions on the frequency and genesis location of tropical cyclones over the western North Pacific in 2004 and 2010
Impact of Atmospheric and Oceanic Conditions on the Frequency and Genesis Location of Tropical Cyclones over the Western North Pacific in 2004 and 2010
Pan SONG 0 2
Jiang ZHU 1
Zhong ZHONG 2
Linlin QI 0
Xiaodan WANG 0
0 Beijing Institute of Aeronautical Meteorology , Beijing 100085
1 International Center for Climate and Environment Science, Institute of Atmospheric Physics
2 College of Meteorology and Oceanography, PLA University of Science and Technology , Nanjing 211101
This study examines the impact of atmospheric and oceanic conditions during May-August of 2004 and 2010 on the frequency and genesis location of tropical cyclones over the western North Pacific. Using the WRF model, four numerical experiments were carried out based on different atmospheric conditions and SST forcing. The numerical experiments indicated that changes in atmospheric and oceanic conditions greatly affect tropical cyclone activity, and the roles of atmospheric conditions are slightly greater than oceanic conditions. Specifically, the total number of tropical cyclones was found to be mostly affected by atmospheric conditions, while the distribution of tropical cyclone genesis locations was mainly related to oceanic conditions, especially the distribution of SST. In 2010, a warmer SST occurred west of 140◦E, with a colder SST east of 140◦E. On the one hand, the easterly flow was enhanced through the effect of the increase in the zonal SST gradient. The strengthened easterly flow led to an anomalous boundary layer divergence over the region to the east of 140◦E, which suppressed the formation of tropical cyclones over this region. On the other hand, the colder SST over the region to the east of 140◦E led to a colder low-level air temperature, which resulted in decreased CAPE and static instability energy. The decrease in thermodynamic energy restricted the generation of tropical cyclones over the same region.
tropical cyclone; SST; numerical simulation; western North Pacific
The western North Pacific (WNP) is the world’s most
active tropical cyclone basin (Emanuel, 2005; Peduzzi et al.,
2012; Lin et al., 2013). The interannual variability of
tropical cyclone activities, including the frequency, intensity,
location and landfalling, is complicated and has been found to
be closely related to large-scale circulation, including ENSO
(Chan, 1985, 2000; Lander, 1994; Wang and Chan, 2002;
Camargo and Sobel, 2005; Wada and Chan, 2008), MJO
(Liebmann et al., 1994; Nakazawa, 2006; Nakano et al., 2015),
quasi-biennial oscillation (Chan, 1985; Lau and Chan, 1993),
Asian–Pacific Oscillation (Zhou et al., 2008), North Pacific
oscillation (Wang et al., 2007), Antarctic oscillation (Ho et
al., 2005; Wang and Fan, 2007), North Atlantic oscillation
(Zhou and Cui, 2014), and Hadley circulation (Zhou and Cui,
∗ Corresponding author: Pan SONG
2008). In addition, many researchers consider global
warming, and related SST change, to be a key driver of changes
in tropical cyclone activity (Webster et al., 2005; Emanuel et
al., 2008; Sugi et al., 2002, 2012; Gleixner et al., 2014;
Scoccimarro et al., 2014; Holland and Bruye`re, 2014). However,
the exact impacts that changes in SST may have on tropical
cyclone activity remain ambiguous, especially when changes
are not uniform.
The effects of climate change on tropical cyclones have
been a prominent issue for a number of years. In recent
decades, SST in major tropical cyclone generation regions
has increased several tenths of a degree Celsius (Santer et
al., 2006). As a result of the recent increase in the
capabilities of climate models, such models have captured some of
the essential physical relationships that govern the links
between the climate and tropical cyclones. Early climate model
simulations, however, suggested some ambiguity in changes
of tropical cyclone characteristics induced by warmer SST.
While many models projected fewer tropical cyclones
globally (Sugi et al., 2002; Bengtsson et al., 2007; Gualdi et and discussed in section 4, followed by a summary in section
al., 2008; Knutson et al., 2010), other climate models sug- 5.
gested some increase in future numbers (Broccoli and
Manabe, 1990; Haarsma et al., 1993; Emanuel, 2013). Yoshimura
and Sugi (2005) investigated impacts of SST warming on 2. Choice of abnormal years for tropical
cythe tropical cyclone climatology using a high-horizontal- clone activity, and the climate background
resolution AGCM. The results of numerical experiments in
which SST was uniformly higher by 2 K demonstrated that 2.1. Analysis of tropical cyclone frequency
the changes in SST had a relatively small influence on the Tropical cyclone numbers were analyzed using the best
tropical cyclone frequency. Similar to Yoshimura and Sugi track dataset of the China Meteorological Administration
(2005), Held and Zhao (2011) carried out an experiment in (CMA) (http://tcdata.typhoon.gov.cn/en/index.html) for the
which SST was uniformly higher by 2 K, and found that the period 2000–2012 to identify abnormal years for tropical
cytropical cyclone frequency decreased by 10%. Zhao et al. clone activity. Figure 1 shows that the annual total number
(2013) compared Hurricane Working Group model responses of tropical cyclones was highest in 2004 and lowest in 2010.
for various simulations. They found that most of the models In addition, the average number of tropical cyclones during
showed decreases in global tropical cyclone frequency when May–August was also highest in 2004 and lowest in 2010.
the model was run with 2 K higher SST. In addition to the Therefore, 2004 and 2010 were chosen as abnormal years for
experiments with a uniformly higher SST, Sugi et al. (2002) this study. The red and blue lines in Fig. 1 show the same
patfound that the regional change in tropical cyclone frequency tern, meaning that the variation in the total number of tropical
was closely related to the distribution of the SST anomaly and cyclones during May–August (MJJA) can also indicate the
the change in convective activity associated with it. Chen and variation in the total number of tropical cyclones in a year.
Huang (2006) pointed out that when the West Pacific warm The numbers of tropical cyclones for each month between
pool was warm, the tropical cyclone numbers over the WNP May and August of 2000 to 2012 are given in Table 1. The
were lower than those when it was cold. Additionally, when total number of tropical cyclones in 2004 and 2010 was 34
the West Pacific warm pool was warm, the locations of tropi- and 18, respectively. The number of tropical cyclones during
cal cyclones were mostly in the northwest, whereas the tropi- MJJA in 2004 and 2010 was 20 and 8, respectively. For this
cal cyclones occurred in the southeast more frequently when reason, we chose the months of MJJA in 2004 and 2010 as
the warm pool was relatively cold. These studies show that the case study period.
the exact influence of SST changes on tropical cyclone activ- In addition to the difference in total tropical cyclone
ity is still a matter for debate. Moreover, the uniform increase numbers, we considered whether there was any difference
or decrease in SST and non-uniform SST changes may affect in the activity region or distribution of the genesis location
the tropical cyclone activity differently. of tropical cyclones between the two years. Figure 2 shows
The horizontal scale of tropical cyclone ranges from the distribution of tropical cyclone genesis location over the
hundreds of kilometers to thousands of kilometers, both of WNP from the best track dataset. There are clear differences
which are small relative to the scale of global climate mod- in tropical cyclone numbers and locations between 2004 and
els. However, tropical cyclones always involve substantial 2010. Tropical cyclone numbers in 2004 are considerably
energy exchange and complex thermodynamic mechanisms, greater than those in 2010, and the locations are scattered
which cannot be depicted in detail even by the highest res- across a range from 100◦E to 180◦E in 2004, whereas in
olution global climate models. The WRF model, as one of 2010 they are mainly located to the west of 140◦E. In this
the most popular numerical models in studying mesoscale study, we sought to determine why such great differences in
weather systems, has been widely used in tropical cyclone total numbers and locations existed between the two years.
case studies (Fierro et al., 2009; Cha and Wang, 2013; Sun We also sought to determine what the differences were in the
et al., 2014a). Shen et al. (2010) and Wang et al. (2012)
attempted the seasonal prediction of tropical cyclones over the
WNP in 2006 and indicated that WRF was capable of
tropical cyclone seasonal forecasting. However, they did not
discuss the impacts of atmospheric environment change and SST
change on tropical cyclone activity.
In the present study, the impacts of the different
atmospheric environments and SST between 2004 and 2010
on tropical cyclone activity were examined using the WRF
model. The rationale for the choice of 2004 and 2010 as
abnormal years for tropical cyclone activity is introduced in
section 2. After a brief description of the model and experiments
in section 3, the results of the experiments and the possible Fig. 1. Tropical cyclone numbers recorded in the CMA best
reasons for the changes in tropical cyclone activities due to track dataset over the WNP during the period 2000–2012:
anchanges in atmospheric environments and SST are presented nual total number (red line); total during MJJA (blue line).
Number of Tropical Cyclones
Fig. 2. The tracks and locations from the best track dataset during MJJA (a) 2004 and (b) 2010 over the WNP. Different
colors represent different intensities: weaker than tropical depression (<TD); tropical depression (TD); tropical storm (TS);
severe tropical storm (STS); typhoon (TY); severe typhoon (STY); and super typhoon (SuperTY).
atmospheric and oceanic situation that influenced the tropical the subtropical high. In 2010, the subtropical high extended
cyclone activity between these two abnormal years. southward to 10◦N. In general, tropical cyclones do not
originate in such low latitude regions, so the locations of tropical
2.2. Atmospheric and oceanic background of 2004 and cyclones in 2010 were located mostly in the west of the sub
2010 tropical high, and fewer tropical cyclones were activated in
We examined the differences in the subtropical high, total.
WNP monsoon trough and SST during MJJA between 2004 The western North Pacific monsoon trough is a low
presand 2010. The datasets used were all from the NCEP FNL (fi- sure belt formed by convergence of the southwest monsoon
nal) Analysis (http://rda.ucar.edu/datasets/ds083.2/) at 1◦ res- or cross-equatorial flow and southeast trade wind in the south
olution and 6-h intervals. of the subtropical high. It is a part of the ITCZ. The WNP
The forecasting of circulation is a very important part in monsoon trough is a region of considerable convective
acthe process of tropical cyclone seasonal forecasting (Wang et tivity, and such activities are conducive to the generation of
al., 2012). In the typhoon season, the southeast trade winds low pressure perturbations and therefore the generation and
in the south of the western Pacific subtropical high directly development of tropical cyclones (Gao et al., 2008).
Figinfluence the main genesis locations of tropical cyclones, ures 3c and d show the monsoon trough in 2004 and 2010.
as well as their generation, development and track (Lei and The monsoon trough in 2004 was clearly stronger than it was
Chen, 2001; Huang et al., 2013). Figures 3a and b demon- in 2010. The monsoon trough extended eastward to 150◦E
strate the circulation at 500 hPa in 2004 and 2010. In 2004, with more tropical cyclones in 2004; while in 2010, it did not
the subtropical high (area of 588 dagpm line) was weaker reach 130◦E and therefore the tropical cyclone locations were
than that in 2010. The western ridge was near 140◦E in 2004, further west, and fewer tropical cyclones were activated in
while it expanded to 122◦E in 2010. Downward air-flows in total.
regions controlled by a subtropical high are unfavorable for From the analysis of atmospheric circulations, we can
the formation of tropical cyclones. Therefore, tropical cy- see that the atmospheric conditions were clearly distinct
beclones always activate in the south and west of the west- tween the two years in question. The location and strength
ern Pacific subtropical high (Ren et al., 2007; Sun, 2011). of the subtropical high and monsoon troughs strongly
influThis explains why the locations of tropical cyclones in 2004 enced the tropical cyclone activity. Following the description
were scattered over the broad region in the south and west of of these different atmospheric situations, we next examined
Fig. 3. The average MJJA (a, b) geopotential height (contours; interval of solid contours: 4 dagpm; units: dagpm)
at 500 hPa and (c, d) stream fields at 850 hPa with tropical cyclone genesis locations (red dots) in (a, c) 2004 and
(b, d) 2010. Fields were derived from NCEP FNL; tropical cyclone genesis locations were from best track dataset
whether the oceanic situation in 2004 and 2010 was also
noticeably different. Previous research has indicated that
tropical cyclone activity is closely related to SST (Holland, 1997;
Hoyos et al., 2006; Dare and McBride, 2011; Sun et al.,
2014b). Therefore, this paper also analyzes the variation in
the SST in 2004 and 2010.
Figure 4 shows the distribution of oceanic heat (SST)
difference between 2004 and 2010 (expressed as 2010 minus
2004). Figure 4 shows that there was a significant difference
in SST distribution between 2004 and 2010. The SST
difference of the warm pool [located at approximately (0◦–20◦N,
110◦–150◦E)] and the ocean to the east of the warm pool were
totally opposite. The warm pool in 2010 was warmer than
that in 2004, while the ocean to the east of the warm pool
was colder than that in 2004.
The SST in 2004 and 2010 was also compared with the
climatology (not shown). The results were consistent with
those in Fig. 4, e.g., the warm pool in 2004 had a negative
anomaly, but there was a positive anomaly in 2010. The
observations of tropical cyclone genesis locations in 2004 and
2010 show that the tropical cyclone numbers over the east of
the warm pool in 2010 were lower than those of 2004. This
indicates that the number of tropical cyclones formed in the
east of the warm pool decreases when the warm pool has a
warm anomaly. This relationship between different thermal
3. Model and experiments ment than disagreement on the sign of the model response be
3.1. Model configuration tween different tracking schemes. Nevertheless, it was pos
3.2. Experiments (2) A maximum surface wind speed in a centered 7◦ × 7◦
From the analyses above, we can see that both oceanic
and atmospheric conditions were significantly different
between 2004 and 2010. There seems to be a close link
between the difference in SST and tropical cyclone locations in
2004 and 2010. Based on these observations, we next
considered what influence the SST changes may have on tropical
cyclone activity when the atmospheric conditions remain the
same. To answer this question, four experiments were
The atmospheric initial and lateral boundary conditions
and oceanic SST forcing of EXP1 were from the 2004 FNL
dataset. EXP2 was the same as EXP1 except that the SST
forcing was taken from 2010. The atmospheric initial and
lateral boundary conditions and oceanic SST forcing of EXP3
were from the 2010 FNL dataset. EXP4 was the same as
EXP3 except that the SST forcing was taken from 2004.
3.3. Criteria for selecting tropical cyclones
An essential first step in the analysis was to select
a method for detecting and tracking the cyclones in the
model outputs. A number of such schemes have been
developed over the years. For example, the model and
basinthreshold dependent scheme (Camargo and Zebiak, 2002),
the structure-based threshold scheme (Walsh et al., 2007),
and the circulation-based scheme (Tory et al., 2013). These
schemes have some key differences but also share many
common characteristics: (1) near-surface wind speed; (2)
lowlevel relative vorticity; (3) a warm core; (4) a difference in
wind speed between the upper and low levels; and (5)
dura(5) A local temperature anomaly at 300 hPa that is greater
than that at 850 hPa;
(6) An average speed that is larger at 850 hPa than at 300
(7) A duration of at least 1.5 days (for six-hourly output).
Camargo and Zebiak (2002) pointed out that the tracks
obtained by these criteria are usually very short. Visual
examination of the corresponding relative vorticity fields showed
that the cyclone structure is visible well before and after the
detection criteria are met. This suggests that relaxing the
detection criteria would produce longer tracks. To verify the
given method, it was used to detect tropical cyclones from the
FNL dataset ranging from 0000 UTC 1 May 2004 to 1800
UTC 31 August 2004. Figure 5 shows the result from the
model and the best track. The results show that the method is
able to successfully detect the tropical cyclone locations and
Results and discussion
The numbers of tropical cyclones and tropical cyclone
“dots” simulated by the numerical experiments are shown
Fig. 5. Tracks detected from the FNL dataset using the Camargo
(2002) method (blue) and the tracks of the best track dataset
from the CMA (red).
in Table 2. A tropical cyclone “dot” is a grid location that
meets all seven of the criteria mentioned above. A longer
track was obtained by tracking backward and forward based
on the tropical cyclone dot. Both the number of tropical
cyclones and number of tropical cyclone dots were the most in
EXP1, while in EXP3 they were both the least. These
results agreed well with the CMA best track dataset. EXP1
simulated three fewer tropical cyclones than the best track of
2004 and EXP3 simulated two fewer tropical cyclones than
the best track of 2010, i.e., the simulated tropical cyclone
numbers were lower than observed, which is in agreement
with the result of Wang et al. (2012). Comparison of EXP1
and EXP2 indicates that the SST of 2010 (EXP2) reduced the
number of tropical cyclone dots by nearly 15% and tropical
cyclone numbers by nearly 29%. Meanwhile, comparison of
EXP3 and EXP4 indicates that the SST of 2010 (EXP3)
decreased the number of tropical cyclone dots by nearly 35%
and tropical cyclone numbers by nearly 25%. Comparison of
EXP1 and EXP4 indicates that the atmospheric environment
of 2010 (EXP4) reduced the number of tropical cyclone dots
by nearly 44% and tropical cyclone numbers by nearly 53%.
Meanwhile, comparison of EXP2 and EXP3 indicates that
the atmospheric environment of 2010 (EXP3) decreased the
number of tropical cyclone dots by nearly 57% and tropical
cyclone numbers by nearly 50%. Thus, we can conclude that
such changes in atmospheric conditions and oceanic
conditions (SST) greatly affect tropical cyclone activity. Moreover,
the roles of changes in atmospheric conditions were slightly
greater than those of oceanic conditions.
The tracks and locations of tropical cyclones simulated in
EXP1 and EXP3 are shown alongside the best track results
in Fig. 6. EXP1 was a simulation of 2004, while EXP3 was
a simulation of 2010. Although the simulated tracks were
not able to remain consistent with the best track one by one,
the results nevertheless captured general features of tracks
Fig. 6. The (a, b) tracks and (c, d) locations of model outputs (blue) and best track dataset (red) in (a, c) 2004 (EXP1)
and (b, d) 2010 (EXP3).
Notes: TC, tropical cyclone; Obs, observation. The second and fourth rows
in the last column denote observed number of tropical cyclones in 2004 and
and the distribution of locations. The tropical cyclone genesis
locations simulated in EXP1 were essentially in accordance
with the best track results for 2004, with a slight westward
deviation. The results of EXP3 show that the tropical cyclone
genesis locations in 2010 were mostly located to the west of
Figure 7 shows the tracks and locations of the four
experiments. The number of tropical cyclones that originated
from east of 140◦E in EXP2 was lower than in EXP1 (Fig.
7e). In other words, SST in 2010 reduced the number of
tropical cyclones originating from east of 140◦E. This is in
agreement with the best track results, which show that the
locations in 2010 were in the west. The number of
tropical cyclones originating from the region east of 140◦E was
higher in EXP4 than in EXP3 (Fig. 7f). In other words, SST
in 2004 increased the number of tropical cyclones
originating from east of 140◦E. Again, this is in agreement with the
best track results, which show that the tropical cyclone
genesis locations in 2004 were scattered widely both east and west
of 140◦E. Figure 7g shows many tropical cyclones
originating from east of 140◦E in both EXP1 and EXP4 with SST in
2004. The atmospheric conditions in 2010 reduced the total
number of tropical cyclones. Figure 7h shows that tropical
cyclones in EXP2 and EXP3 with SST in 2010 were mainly
scattered west of 140◦E. The atmospheric conditions in 2004
increased the total number of tropical cyclones. These model
results were all in accordance with the observations.
The comparisons between the experiments reported
above indicate that any changes of atmospheric conditions
and oceanic conditions can greatly affect tropical cyclone
activity. We now attempt to explain this process by analyzing
the impacts of their changes on the convective activity and its
Figure 8 shows the differences in average atmospheric
conditions (geopotential height at 500 hPa: H500), oceanic
conditions (SST) and precipitation rate. We can see good
agreement between the pattern of the SST anomaly and the
pattern of the precipitation difference in both Figs. 8a and
b. The pattern correlation (8◦–30◦N) reaches 0.44 and 0.42,
respectively. All the pattern correlation coefficients in this
paper passed the u-test (Mann and Whitney, 1947), with a
significance level of 0.05. The regions with a positive SST
anomaly have a positive precipitation difference, and vice
versa. Thus, over the region east of 140◦E, the patterns of
changes in tropical cyclone numbers, precipitation and the
SST anomaly resemble each other. There is also good
agreement between the H500 and the pattern of the precipitation
difference in both Figs. 8c and d. The pattern correlation (8◦–
30◦N) reaches −0.38 and −0.43, respectively. The regions
with a positive H500 anomaly have a negative precipitation
difference. This suggests a chain of links from the
geopotential height anomaly at 500 hPa and the SST anomaly to
convective activity (or precipitation), from convective activity to
circulation, and from circulation to cyclogenesis. Besides, we
can see that the spatial patterns of SST and precipitation rate
in Figs. 8a and b possess apparent differences in zonal
direction, while the spatial patterns of H500 and precipitation rate
in Figs. 8c and d do not show any obvious zonal difference.
To understand the changes in convective activity, we
examined the changes in the distribution of the mean fields of
quantities associated with convection. Figure 9 shows the
difference in the wind field at 850 hPa and the vertical
velocity at 500 hPa (W500). In Fig. 9a, there is weak anticyclonic
circulation in the east of the warm pool between 10◦N and
20◦N, and weak anticyclonic circulation in the same regions
shown in Fig. 9b. Zhao et al. (2013) showed that decreases
in global tropical cyclone frequency for the 2 K higher SST
run were most closely related to 500 hPa vertical velocity.
Kim et al. (2014), using GFDL CM2.5, obtained the same
result; that the reduction of tropical cyclone frequency was
strongly related to weakening of vertical velocity in the
midtroposphere. In the 500 hPa vertical velocity field in Fig. 9a,
there is a downward W500 in the east of the warm pool where
an anticyclone is located in the lower level, which is not
suitable for tropical cyclone generation. In Fig. 9b, there is also
downward W500 in the east of the warm pool with anticyclonic
circulation in the lower level, which is not suitable for
tropical cyclone formation. The anticyclone and downward W500
can be seen in Figs. 9c and d over the region to the west of
the warm pool, and also to the east of the warm pool. We can
conclude that the atmospheric circulation triggered by the
atmospheric condition and SST in 2010 was not conductive to
In addition to the dynamic variables, Figs. 10 and 11
show the patterns of differences in average CAPE, static
stability, relative humidity at 700 hPa, and OLR. In Fig. 10a,
negative anomalous CAPE with a negative SST difference
is apparent in the east of the warm pool. Over these
regions, anomalous convection is negative (Fig. 8a). On the
other hand, the same relationship can be seen in Figs. 10b
and 8b. The pattern correlation in Figs. 10a and b reach 0.54
and 0.66, respectively. Here, the static stability is defined as
the difference in potential temperature at 500 hPa and 1000
hPa (denoted as THETA for convenience). Figure 10c shows
negative anomalous static stability east of the warm pool,
and Fig. 10d also shows negative anomalous static stability
in the same region. The pattern correlation in Figs. 10c and
d reaches 0.82 and 0.81, respectively. In Fig. 10e, there are
positive and negative relative humidity centers in the east of
Fig. 7. The simulated (a–d) tracks and (e–h) locations in the four experiments.
Fig. 8. Difference in average SST (a, b) (contours; above zero: solid line; below zero: dashed line; interval: 0.4◦C),
geopotential height at 500 hPa (c, d) (contours; above zero: solid line; below zero: dashed line; interval: 0.5 dagpm)
and precipitation rate (shaded; units: mm d−1). Pattern correlations between SST, geopotential height at 500 hPa and
precipitation are also shown in parentheses.
the warm pool with a negative anomalous SST. In Fig. 10f,
there are also positive and negative relative humidity centers
in the region where the anomalous SST is negative.
Therefore, the pattern correlation in Figs. 10e and f only reaches
−0.04 and 0.12, respectively. In Fig. 10g, there is positive
anomalous OLR in the warm pool and east of the warm pool,
whereas Fig. 10h shows positive anomalous OLR in the east
of the warm pool with a negative anomalous SST and
negative anomalous OLR in the warm pool with a positive
anomalous SST. The pattern correlation in Figs. 10g and h reaches
−0.20 and −0.16, respectively. Moreover, there are zonal
differences in all of these quantities’ distributions. The
consistency between the distribution of SST and these quantities
associated with convection means that the distribution of SST
influenced the distribution of tropical cyclone genesis
Figure 11 shows the difference in CAPE, static stability,
relative humidity at 700 hPa, OLR, and H500, in EXP4
versus EXP1, and EXP3 versus EXP2. Compared with Fig. 10,
we can see that the distributions of these quantities in Fig.
11 possess no obvious zonal differences. The differences in
CAPE, static stability, and relative humidity at 700 hPa are
negatively correlated with the differences in H500, while the
differences in OLR are positively correlated with the
differences in H500. The pattern correlations in Figs. 11a–h reach
−0.22, −0.25, −0.17, −0.24, −0.39, −0.53, 0.34 and 0.36,
respectively. This means that the increase in H500 in 2010
restrained the convective activity.
Besides, we also calculated the pattern correlation with
SST and H500 of the relative vorticity at 850 hPa and
divergence at 200 hPa, separately. Figure 10 shows the roles of
SST fields (EXP1 versus EXP2, EXP3 versus EXP4) in
tropical cyclone activity in the WNP, and Fig. 11 shows the roles
of the atmospheric environment (EXP1 versus EXP4, EXP2
versus EXP3) in tropical cyclone activity in the WNP. We
averaged the pattern correlation in Figs. 10a and b to represent
the impacts of SST on CAPE. Similarly, we averaged the
pattern correlation in Figs. 11a and b to represent the impacts of
the atmospheric environment (H500) on CAPE. The results
are shown in Table 3. It is clear that changes in atmospheric
Fig. 9. Difference in the average stream field (vectors; units: m s−1) at 850 hPa and vertical velocity (colored shading;
units: m s−1) at 500 hPa.
CAPE THETA RH
environment have greater impacts on changes in RH, OLR,
relative vorticity and divergence, compared with changes in
ocean conditions. However, changes in ocean conditions have
greater impacts on changes in CAPE and THETA, compared
with changes in atmospheric environment.
The above analysis may help us to better understand why
colder SST over the region to the east of 140◦E in 2010
decreased tropical cyclone numbers over the same region. In
2010, a greater SST warming occurred west of 140◦E;
meanwhile, there was colder SST east of 140◦E. As a result, the
zonal SST gradient increased across the whole region (90◦–
180◦E). The increased zonal SST gradient strengthened the
easterly flow, which led to an increase in boundary layer
divergence over the region to the east of 140◦E (wind fields in
Figs. 9a and b). As we know, tropical cyclones originate from
tropical disturbances (Lau and Lau, 1990; Fu et al., 2007);
therefore, the anomalous boundary layer divergence over that
region suppressed the formation of tropical cyclones. In
addition to the unfavorable dynamic conditions, the colder SST
east of 140◦E led to a colder low-level air temperature, which
resulted in a decrease in CAPE and static instability energy
(Figs. 10a–d). The decrease in thermodynamic energy
restricted the generation of tropical cyclones.
In this study, the influence of different atmospheric and
oceanic conditions on tropical cyclone frequency and
location in the WNP was examined using the WRF model. Four
Fig. 10. Difference in (a, b) CAPE (units: J kg−1), (c, d) static stability (units: ◦), (e, f) relative
humidity (units: percent) at 700 hPa, and (g, h) OLR (units: W m−2) and SST (contours; above
zero: solid line; below zero: dashed line; interval: 0.4◦C). Pattern correlations between CAPE,
static stability, relative humidity at 700 hPa, OLR and SST are also shown in parentheses.
Fig. 11. Difference in (a, b) CAPE (units: J kg−1), (c, d) static stability (units: ◦), (e, f) relative
humidity (units: percent) at 700 hPa, and (g, h) OLR (units: W m−2) and geopotential height at
500 hPa (contours; above zero: solid line; below zero: dashed line; interval: 0.5 dagpm). Pattern
correlations between CAPE, static stability, relative humidity at 700 hPa, OLR and geopotential
height at 500 hPa are also shown in parentheses.
experiments that employed different atmospheric initial and
lateral boundary conditions, as well as different SST forcing,
were performed for the MJJA period of 2004 and 2010.
The results of the experiments showed that changes in
atmospheric environment have greater impacts on changes in
RH at 700 hPa, OLR, relative vorticity at 850 hPa, and
divergence at 200 hPa, compared with changes in ocean
conditions. However, changes in ocean conditions have greater
impacts on changes in CAPE and static stability. The
differences in the locations of tropical cyclones in 2004 and 2010
were closely related to the distribution of the SST anomaly,
and the change in its associated convective activity. In the
warmer (colder) region, low-level static instability enhanced
(weakened), mid–low-level moisture increased (decreased),
and CAPE increased (decreased). The increase (decrease) of
low-level instability energy and moisture directly affected the
convective activity. Accompanied by low-level cyclonic
(anticyclonic) circulation and upward (downward) vertical
velocity in the mid-troposphere, tropical cyclone activity over
this region was promoted (restrained).
It has been found that the total number of tropical
cyclones was mostly affected by atmospheric conditions, while
the distribution of tropical cyclone genesis locations was
mainly influenced by oceanic conditions, especially the
distribution of SST. The physical processes responsible for the
decreased numbers of tropical cyclones over the region to the
east of 140◦E in 2010, caused by colder SST over the same
region, are discussed below. In 2010, on the one hand, the
easterly flow was enhanced through the effect of the increase
of the zonal SST gradient. The strengthened easterly led
to anomalous boundary layer divergence over the region to
the east of 140◦E. The anomalous boundary layer divergence
suppressed the formation of tropical cyclones. On the other
hand, the colder SST over the region to the east of 140◦E
led to colder low-level air temperature, which resulted in
decreased CAPE and static instability energy. The decrease in
thermodynamic energy restricted the generation of tropical
Although the simulated numbers and locations of
tropical cyclones in 2004 and 2010 were close to the best track
on the seasonal timescale, there were still clear differences
on the monthly timescale. Model errors accumulated in
longterm simulations and the choice of parameterization schemes
may contribute to this issue. Improving our understanding of
the mechanisms through which changes of atmospheric
conditions and oceanic conditions influence tropical cyclone
activity requires more detailed analyses and experiments in the
Acknowledgements. This work was supported by the
Chinese Academy of Sciences’ Project “Western Pacific Ocean System:
Structure, Dynamics and Consequences” (Grant No. XDA10010405),
the National High Technology Research and Development Program
of China (863 program) (Grant No. 2012AA091801), the National
Natural Science Foundation of China (Grant Nos. 41205044 and
41205075), and the Natural Science Foundation of Jiangsu Province
(Grant No. BK2012062).
Bengtsson , L. , K. I. Hodges , M. Esch , N. Keenlyside , L. Kornblueh , J. J. Luo , and T. Yamagata , 2007 : How may tropical cyclones change in a warmer climate ? Tellus , 59 ( 4 ), 539 - 561 .
Broccoli , A. J. , and S. Manabe , 1990 : Can existing climate models be used to study anthropogenic changes in tropical cyclone climate? Geophys . Res . Lett., 17 ( 11 ), 1917 - 1920 .
Camargo , S. J. , and S. E. Zebiak , 2002 : Improving the detection and tracking of tropical cyclones in atmospheric general circulation models . Wea. Forecasting , 17 ( 6 ), 1152 - 1162 .
Camargo , S. J. , and A. H. Sobel , 2005 : Western North Pacific tropical cyclone intensity and ENSO . J. Climate, 18 ( 15 ), 2996 - 3006 .
Cha , D. H. , and Y. Q. Wang , 2013 : A dynamical initialization scheme for real-time forecasts of tropical cyclones using the WRF model . Mon. Wea. Rev. , 141 ( 3 ), 964 - 986 .
Chan , J. C. L., 1985 : Tropical cyclone activity in the northwest Pacific in relation to the El Ni n˜o/Southern Oscillation phenomenon . Mon. Wea. Rev. , 113 ( 4 ), 599 - 606 .
Chan , J. C. L., 2000 : Tropical cyclone activity over the western North Pacific associated with El Nin˜o and La Nin˜a events . J. Climate , 13 ( 16 ), 2960 - 2972 .
Chen , G. H. , and R. H. Huang , 2006 : The effect of warm pool thermal states on tropical cyclone in West Northwest Pacific . J. Trop. Meteor., 22 ( 6 ), 527 - 532 . (in Chinese)
Chou , M. D. , and M. J. Suarez , 1994 : An efficient thermal infrared radiation parameterization for use in general circulation models . NASA Tech. Memo , 84 pp.
Dare , R. A. , and J. L. McBride , 2011 : Sea surface temperature response to tropical cyclones . Mon. Wea. Rev. , 139 ( 12 ), 3798 - 3808 .
Emanuel , K. A. , 2005 : Divine Wind, the History and Science of Hurricanes. Oxford University Press, New York, 296 pp.
Emanuel , K. A. , 2013 : Downscaling CMIP5 climate models shows increased tropical cyclone activity over the 21st century . Proceedings of the National Academy of Sciences of United States of America , 110 ( 30 ), 12219 - 12224 .
Emanuel , K. A. , R. Sundararajan , and J. Williams , 2008 : Hurricanes and global warming: Results from downscaling IPCC AR4 simulations . Bull. Amer. Meteor. Soc. , 89 ( 3 ), 347 - 367 .
Fierro , A. O. , R. F. Rogers , F. D. Marks , and D. S. Nolan , 2009 : The impact of horizontal grid spacing on the microphysical and kinematic structures of strong tropical cyclones simulated with the WRF-ARW model . Mon. Wea. Rev. , 137 ( 11 ), 3717 - 3743 .
Fu , B. , T. Li , M. Peng , and F. Weng , 2007 : Analysis of tropical cyclone genesis in the western North Pacific for 2000 and 2001 . Wea. Forecasting, 22 ( 4 ), 763 - 780 .
Gao , J. Y. , X. Z. Zhang , Z. H. Jiang , and L. J. You , 2008 : Anomalous western North Pacific monsoon trough and tropical cyclone activities . Acta Oceanologica Sinica , 30 ( 3 ), 35 - 47 . (in Chinese)
Gleixner , S. , N. Keenlyside , K. I. Hodges , W. L. Tseng , and L. Bengtsson , 2014 : An inter-hemispheric comparison of the tropical storm response to global warming . Climate Dyn. , 42 ( 7-8 ), 2147 - 2157 .
Gualdi , S. , E. Scoccimarro , and A. Navarra , 2008 : Changes in tropical cyclone activity due to global warming: Results from a high-resolution coupled general circulation model . J. Climate , 21 ( 20 ), 5204 - 5228 .
Haarsma , R. J. , J. F. B. Mitchell, and C. A. Senior , 1993 : Tropical disturbances in a GCM . Climate Dyn., 8 ( 5 ), 247 - 257 .
Held , I. M. , and M. Zhao , 2011 : The response of tropical cyclone statistics to an increase in CO2 with fixed sea surface temperatures . J. Climate , 24 ( 20 ), 5353 - 5364 .
Ho , C. H. , J. H. Kim , H. S. Kim , C. H. Sui , and D. Y. Gong , 2005 : Possible influence of the Antarctic Oscillation on tropical cyclone activity in the western North Pacific . J. Geophys. Res ., 110 (D19), D19104 .
Holland , G. J. , 1997 : The maximum potential intensity of tropical cyclones . J. Atmos. Sci. , 54 ( 21 ), 2519 - 2541 .
Holland , G. , and C. L. Bruye `re, 2014 : Recent intense hurricane response to global climate change . Climate Dyn. , 42 ( 3-4 ), 617 - 627 .
Hong , S. Y. , Y. Noh , and J. Dudhia , 2006 : A new vertical diffusion package with an explicit treatment of entrainment processes . Mon. Wea. Rev. , 134 ( 9 ), 2318 - 2341 .
Horn , M. , and Coauthors , 2014 : Tracking scheme dependence of simulated tropical cyclone response to idealized climate simulations . J. Climate , 27 ( 24 ), 9197 - 9213 .
Hoyos , C. D. , P. A. Agudelo , P. J. Webster , and J. A. Curry , 2006 : Deconvolution of the factors contributing to the increase in global hurricane intensity . Science , 312 ( 5770 ), 94 - 97 .
Huang , L. N. , J. Y. Gao , J. Sun, and J. L. Wu , 2013 : Abnormal climatic features of accumulated cyclone energy over the Northwest Pacific . Meteorological Monthly , 39 ( 8 ), 995 - 1003 . (in Chinese)
Janjic ´ , Z. I. , 1994 : The step-mountain eta coordinate model: Further developments of the convection, viscous sublayer, and turbulence closure schemes . Mon. Wea. Rev. , 122 ( 5 ), 927 - 945 .
Kim , H. S. , G. A. Vecchi , T. R. Knutson , W. G. Anderson , T. L. Delworth , A. Rosati , F. R. Zeng , and M. Zhao , 2014 : Tropical cyclone simulation and response to CO2 doubling in the GFDL CM2.5 high-resolution coupled climate model . J. Climate , 27 ( 21 ), 8034 - 8054 .
Knutson , T. R. , and Coauthors, 2010 : Tropical cyclones and climate change . Nature Geoscience , 3 ( 3 ), 157 - 163 .
Lander , M. A. , 1994 : An exploratory analysis of the relationship between tropical storm formation in the western North Pacific and ENSO. Mon . Wea. Rev., 122 ( 4 ), 636 - 651 .
Lau , K. H. , N. C. Lau , 1990 : Observed structure and propagation characteristics of tropical summertime synoptic scale disturbances . Mon. Wea. Rev. , 118 ( 9 ), 1888 - 1913 .
Lau , R. , and M. Y. Chan , 1993 : Equatorial stratospheric flow patterns and quasi-biennial/pentaennial oscillations . East Asia and Western Pacific Meteorology and Climate , 2 , 31 - 38 .
Lei , X. T. , and L. S. Chen , 2001 : An overview on the interaction between tropical cyclone and mid-latitude weather systems . J. Trop. Meteor. , 17 ( 4 ), 452 - 461 . (in Chinese)
Liebmann , B. , H. H. Hendon , and J. D. Glick , 1994 : The relationship between tropical cyclones of the western Pacific and Indian Oceans and the Madden-Julian oscillation . J. Meteor. Soc. Japan , 72 ( 3 ), 401 - 412 .
Lin , I. I. , and Coauthors, 2013 : An ocean coupling potential intensity index for tropical cyclones . Geophys. Res. Lett. , 40 ( 9 ), 1878 - 1882 .
Mann , H. B. , D. R. Whitney , 1947 : On a test of whether one of two random variables is stochastically larger than the other . Ann. Math. Statis. , 18 ( 1 ), 50 - 60 .
Mlawer , E. J. , S. J. Taubman , P. D. Brown , M. J. Iacono , and S. A. Clough , 1997 : Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave . J. Geophys. Res ., 102 , 16 663 - 16 682.
Nakano , M. , Sawada , M. , Nasuno , T. , and Satoh , M. , 2015 : Intraseasonal variability and tropical cyclogenesis in the western North Pacific simulated by a global nonhydrostatic atmospheric model . Geophys. Res. Lett. , 42 , 565 - 571 .
Nakazawa , T. , 2006 : Madden-Julian oscillation activity and typhoon landfall on Japan in 2004 . Sola, 2 , 136 - 139 .
Peduzzi , P. , B. Chatenoux , H. Dao , A. D. Bono , C. Herold , J. Kossin , F. Mouton , and O. Nordbeck , 2012 : Global trends in tropical cyclone risk . Nature Climate Change , 2 ( 4 ), 289 - 294 .
Ren , S. L. , Y. M. Liu , and G. X. Wu , 2007 : Interactions between typhoon and subtropical anticyclone over western pacific revealed by numerical experiments . Acta Meteorologica Sinica , 65 ( 3 ), 329 - 340 . (in Chinese)
Rogers , E. , T. Black , B. Ferrier , Y. Lin , D. Parrish , and J. DiMego , 2001 : Changes to the NCEP Meso Eta Analysis and Forecast System : Increase in resolution, new cloud microphysics, modified precipitation assimilation, modified 3DVAR analysis . NWS Technical Procedures Bulletin, No . 488 . [Available online at http://www.emc.ncep.noaa.gov/mmb/mmbpll/eta22 tpb/]
Santer , B. D. , and Coauthors, 2006 : Forced and unforced ocean temperature changes in Atlantic and Pacific tropical cyclogenesis regions . Proceedings of the National Academy of Sciences of United States of American , 103 ( 38 ), 13 905 - 13 910.
Scoccimarro , E. , S. Gualdi , G. Villarini , M. Zhao , K. Walsh , and A. Navarra , 2014 : Intense precipitation events associated with landfalling tropical cyclones in response to a warmer climate and increased CO2 . J. Climate, 27 ( 12 ), 4642 - 4654 .
Shen , X. Y. , W. D. Zhu , J. Du, and W. Y. Pan , 2010 : The seasonal forecasting experiment of typhoon from July to September of 2006 . Scientia Meteorologica Sinica , 30 ( 5 ), 676 - 683 . (in Chinese)
Skamarock , W. C. , and Coauthors, 2008 : A description of the Advanced Research WRF version 3. NCAR Tech. Note NCAR/ TN-475+STR , 113 pp.
Sugi , M. , A. Noda , and N. Sato , 2002 : Influence of the global warming on tropical cyclone climatology: An experiment with the JMA global model . J. Meteor. Soc. Japan , 80 ( 2 ), 249 - 272 .
Sugi , M. , H. Murakami , and J. Yoshimura , 2012 : On the mechanism of tropical cyclone frequency changes due to global warming . J. Meteor. Soc. Japan , 90 , 397 - 408 .
Sun , L. , 2011 : Analysis of features and causation for tropical cyclone activities over the Western North Pacific in 2010 . Meteorological Monthly, 37 ( 8 ), 929 - 935 . (in Chinese)
Sun , Y. , Z. Zhong , L. Wei, and Y. J. Hu , 2014a : Why are tropical cyclone tracks over the western North Pacific sensitive to the cumulus parameterization scheme in regional climate modeling? A case study for Megi ( 2010 ). Mon . Wea. Rev., 142 ( 3 ), 1240 - 1249 .
Sun , Y. , Z. Zhong , L. Yi , Y. Ha , and Y. M. Sun , 2014b : The opposite effects of inner and outer sea surface temperature on tropical cyclone intensity . J. Geophys. Res.: Atmos. , 119 ( 5 ), 2193 - 2208 .
Tory , K. J. , S. S. Chand , R. A. Dare , and J. L. McBride , 2013 : The development and assessment of a model-, grid-, and basinindependent tropical cyclone detection scheme . J. Climate , 26 ( 15 ), 5493 - 5507 .
Wada , A. , and J. C. L. Chan, 2008 : Relationship between typhoon activity and upper ocean heat content . Geophys. Res. Lett. , 35 ( 17 ), L17603.
Walsh , K. J. E. , M. Fiorino , C. W. Landsea , and K. L. Mcinnes , 2007 : Objectively determined resolution-dependent threshold criteria for the detection of tropical cyclones in climate models and reanalyses . J. Climate , 20 ( 10 ), 2307 - 2314 .
Wang , B. , and J. C. L. Chan, 2002 : How strong ENSO events affect tropical storm activity over the western North Pacific . J. Climate, 15 ( 13 ), 1643 - 1658 .
Wang , H. J. , and K. Fan , 2007 : Relationship between the Antarctic oscillation in the western North Pacific typhoon frequency . Chinese Science Bulletin , 52 ( 4 ), 561 - 565 .
Wang , H. J. , J. Q. Sun , and K. Fan , 2007 : Relationships between the North Pacific Oscillation and the typhoon /hurricane frequencies. Science in China Series D: Earth Sciences , 50 ( 9 ), 1409 - 1416 .
Wang , Y. Q. , T. Y. Song , J. Liang, and W. Y. Pan , 2012 : Simulation of seasonal tropical cyclone activity over the western North Pacific by using the WRF model . Transactions of Atmospheric Sciences , 35 ( 1 ), 24 - 31 . (in Chinese)
Webster , P. J. , G. J. Holland , J. A. Curry, and H. R. Chang , 2005 : Changes in tropical cyclone number, duration, and intensity in a warming environment . Science , 309 ( 5742 ), 1844 - 1846 .
Yoshimura , J. , and M. Sugi , 2005 : Tropical cyclone climatology in a high-resolution AGCM-Impacts of SST warming and CO2 increase . SOLA , 1 , 133 - 136 .
Zhao , M. , and Coauthors , 2013 : Response of global tropical cyclone frequency to a doubling of CO2 and a uniform SST warming-A multi-model intercomparison . U. S. CLIVAR Hurricane Workshop , Geophysical Fluid Dynamics Laboratory, Princeton, US, 5 - 7 June 2013 .
Zhou , B. T ., and X. Cui , 2008 : Hadley circulation signal in the tropical cyclone frequency over the western North Pacific . J. Geophys. Res.: Atmos. , 113 (D16), D16107 .
Zhou , B. T ., and X. Cui , 2014 : Interdecadal change of the linkage between the North Atlantic Oscillation and the tropical cyclone frequency over the western North Pacific . Science China Earth Sciences , 57 ( 9 ), 2148 - 2155 .
Zhou , B. T ., X. Cui, and P. Zhao , 2008 : Relationship between the Asian-Pacific oscillation and the tropical cyclone frequency in the western North Pacific . Science in China Series D: Earth Sciences , 51 ( 3 ), 380 - 385 .