Abstract

Increasing tropical cyclone (TC) influence in the subtropical East Asia and decreasing TC activity in the South China Sea over the past few decades have been researched in previous studies. The singular value decomposition (SVD) of observational data and the Intergovernmental Panel on Climate Change (IPCC) climate change simulations in the Fourth Assessment Report (AR4) shows that the observed TC track changes are linked to the leading SVD mode of global sea surface temperature (SST) warming and the associated changes in large-scale steering flows. The selected five IPCC models can generally simulate the leading mode in their ensemble control run and prediction, suggesting the possible persistence of the reported track changes by 2040.

1. Introduction

Global mean land surface temperature and sea surface temperature (SST) have risen significantly over the past half century, and the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) concluded that most of the global surface temperature increase was very likely due to the observed increase in anthropogenic greenhouse gas concentrations (Solomon et al. 2007). Given the great societal and scientific concerns, the impact of global warming on tropical cyclone (TC) activity has been the subject of considerable investigation in recent years. While progress has been made in assessing the influence on TC intensity and rainfall (Knutson et al. 2010), relatively little is known about the possible change of TC tracks in a warming climate.

The TC tracks are controlled essentially by large-scale steering and propagation resulting from the interaction of a TC with its environment (Wang et al. 1998). Based on the output of the Geophysical Fluid Dynamics Laboratory (GFDL) climate model in the IPCC Third Assessment Report (AR3), Wu and Wang (2004) suggested that global warming may affect prevailing TC tracks over the western North Pacific (WNP). Using the best-track data from the Joint Typhoon Warning Center (JTWC), Wu et al. (2005) further showed that over the period 1965–2003 subtropical East Asia experienced increasing typhoon influence, but TC activity over the South China Sea has decreased considerably. Recently Tu et al. (2009) also found an abrupt increase in the influence of typhoons in the vicinity of Taiwan after 2000. Because of considerable interannual and interdecadal variations in TC activity over the WNP (Chan and Shi 1996; Wang and Chan 2002; Ho et al. 2004; Zhou and Chan 2007; Liu and Chan 2008), whether the observed TC track changes were linked to the ongoing global warming is still unknown. The objective of the present study is to address this issue using observational data and IPCC climate change simulations in the AR4.

2. Data and analysis method

Following Wu et al. (2005), TC tracks are measured with the frequency of TC occurrence, which is counted at a 6-h interval and measures how frequently TCs affect a grid of 2° latitude by 2° longitude. The best-track data from JTWC were used to calculate the frequency of TC occurrence in the WNP basin. Our analysis mainly covers the peak season (July–September) during the period 1965–2009 since the satellite monitoring of weather events first became routine in 1965. Wu et al. (2005) suggested that during the peak season the climatological mean flows can represent TC large-scale steering flows.

The singular value decomposition (SVD) analysis is used to obtain the leading mode associated with global warming. The SVD is conducted with the global SST field, large-scale steering flows in the WNP basin, and the frequency of TC occurrence. The monthly SST with 2° latitude by 2° longitude resolution and the monthly winds with 2.5° latitude by 2.5° longitude resolution are obtained from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis dataset. The large-scale steering flow is defined as the pressure-weighted mean flow from 850 to 300 hPa (Holland 1993). A five-point smoother is applied to the frequency of TC occurrence, wind fields, and SST to reduce interannual variability before the SVD analysis is performed. (To examine the relationship between the time series of the SVD mode and global warming, monthly anomalies of the land surface temperature, SST, and their combination, see http://www.ncdc.noaa.gov/cmb-faq/anomalies.html#anomalies.) The output of the control and global warming experiments from the IPCC climate models in the AR4 is used to examine the possible typhoon track change in the future.

3. A global warming mode associated with typhoon track changes

First the SVD analysis is conducted with global SST and large-scale steering flows over the period 1965–2009. The leading mode accounts for 47% of the total square covariance between the two fields, revealing an upward-trend pattern (Fig. 1). The SST shows a global-scale warming pattern with the most remarkable warming in the northern Indian Ocean and tropical Atlantic (Fig. 1a). In the large-scale steering flow, an anomalous cyclonic circulation is over eastern China (Fig. 1b). Wu et al. (2005) suggested that the anomalous pattern of steering flows over the western part of the WNP basin was responsible for the increased typhoon influence in the subtropical East Asia and decreased typhoon influence over the South China Sea since 1965.

Fig. 1.

Spatial patterns of the leading SVD mode of (a) observed global SST, (b) large-scale steering flows (vectors) and the frequency of TC occurrence (contours), (c) the standardized SVD time series in comparison with the combined global land and ocean temperature anomalies, (d) linear trends of the mean TC translation velocity (vectors) with 95% confidence (shading), and (e)–(f) time series of the annual number of TCs that entered the two boxes in (b) during 1965–2009. The contour intervals are 0.5 and 0.1 in (a) and (b), respectively, with zero contours suppressed. The blue and red lines in (e) and (f) are for the 5-yr running average and linear trends, respectively.

Fig. 1.

Spatial patterns of the leading SVD mode of (a) observed global SST, (b) large-scale steering flows (vectors) and the frequency of TC occurrence (contours), (c) the standardized SVD time series in comparison with the combined global land and ocean temperature anomalies, (d) linear trends of the mean TC translation velocity (vectors) with 95% confidence (shading), and (e)–(f) time series of the annual number of TCs that entered the two boxes in (b) during 1965–2009. The contour intervals are 0.5 and 0.1 in (a) and (b), respectively, with zero contours suppressed. The blue and red lines in (e) and (f) are for the 5-yr running average and linear trends, respectively.

The SVD analysis is further performed with the frequency of TC occurrence and global SST, which accounts for 28.5% of the total covariance and also shows an upward-trend pattern. The resulting spatial pattern of SST is nearly identical to that shown in Fig. 1a (figure not shown), suggesting a coupled mode in the large-scale steering flows, global SST, and frequency of TC occurrence. As shown in Fig. 1b, the spatial pattern of the frequency of TC occurrence is very similar to the track changes in Wu et al. (2005). In the WNP basin TCs generally take three prevailing tracks: the westward track to the South China Sea; the northwestward track to affect southeast China, Korea, and Japan; and the northeastward recurving track east of 130°E (Wu and Wang 2004; Wu et al. 2005). The negative anomalies over the central South China Sea indicate the decreased activity of the westward-moving TCs, while the positive anomalies extending from the Philippine Sea to the eastern coast of China indicate the enhanced TC activity in the vicinity of Taiwan and the coastal area of southeast China. Using a trajectory model Wu et al. (2005) confirmed that the track changes resulted mainly from the anomalous cyclonic circulation centered over eastern China, with eastward anomalous steering flows over the South China Sea and northward winds over the subtropical coastal region. The anomalous cyclonic circulation is consistent with the linear trends in the mean TC translation velocity over the western part of the basin during the period 1965–2009 (Fig. 1d).

The time series of the leading mode are highly correlated with the combined mean global land and ocean temperature anomaly (Fig. 1c). The correlation coefficients of the temperature anomaly with the SVD time series of the global SST, large-scale steering flows, and frequency of TC occurrence are 0.96, 0.87, and 0.89, respectively, suggesting a coupled mode in the global SST, large-scale steering flows, and frequency of TC occurrence, which is closely associated with the ongoing global warming. Note that the first two modes are statistically significant and the second mode is mainly associated with interannual variations of TC track in the WNP.

The changes in the TC numbers are consistent with the track changes in Fig. 1b. The numbers of TCs entering the vicinity of Taiwan (22°–27°N, 118°–123°E) and Hainan, China (14°–20°N, 109°–121°E), are counted over the period 1965–2009 (Figs. 1e and 1f), which roughly cover the most changes in the frequency of TC occurrence. The TC number near Hainan Island decreased significantly since 1965 with the significant increase near Taiwan Island since 1975.

4. Large-scale steering flow changes projected in climate models

Changes in large-scale steering flows associated with global warming are further examined with the projections of the 23 IPCC coupled general circulation models (CGCMs) in the AR4. The output of the control runs from these models is first evaluated. The simulated 700-hPa wind patterns in all of the control runs are compared with the observed by focusing on the summer mean monsoon trough and subtropical high since the two systems are important to TC tracks in the western North Pacific basin. Then the linear trends in the simulated 700-hPa wind field are calculated and compared with the observed. Based on their performance, five CGCMs are finally selected for assessing the influence of global warming on the large-scale steering flow in the future. These models are the Canadian Centre for Climate Modelling and Analysis (CCCma) Coupled General Circulation Model, version 3.1 (CGCM3.1); the GFDL Climate Model version 2.0 (GFDL CM2.0); the Model for Interdisciplinary Research on Climate 3.2, medium-resolution version [MIROC3.2(medres)]; the Hadley Centre Global Environmental Model version 1 (HadGEM1); and the third climate configuration of the Met Office Unified Model (HadCM3).

Figure 2 shows the leading SVD mode of the ensemble global SST and large-scale steering flows in the control and global warming experiments from the five selected models. The leading mode in the control experiments, which accounts for 40.4% of the covariance between the SST and large-scale steering flows, reveals the global-scale warming in SST and an anomalous cyclonic circulation in the large-scale steering flow during the period 1965–98. Although the center of the simulated anomalous cyclonic circulation in the control experiments shifts slightly southward to the coast of South China Sea, the five selected models are generally capable of simulating the observed features in global SST and the associated changes in large-scale steering flows over the western part of the WNP basin.

Fig. 2.

The spatial patterns of the leading SVD mode of (a) the 5-model ensemble global SST with contour intervals of 0.5, (b) large-scale steering flows over the western North Pacific basin, and (c) the standardized SVD time series in the control runs during 1965–98. (d)–(f) As in (a)–(c), respectively, but for the 5-model ensemble projection during the period 2001–40 under the A1B scenario.

Fig. 2.

The spatial patterns of the leading SVD mode of (a) the 5-model ensemble global SST with contour intervals of 0.5, (b) large-scale steering flows over the western North Pacific basin, and (c) the standardized SVD time series in the control runs during 1965–98. (d)–(f) As in (a)–(c), respectively, but for the 5-model ensemble projection during the period 2001–40 under the A1B scenario.

The SVD analysis is further conducted with their twenty-first-century ensemble projection under the A1B scenario. The SVD leading mode of the ensemble projection during the period 2001–40 accounts for 58.7% of the covariance between the global SST and large-scale steering flows with upward trends in their time series (Figs. 2d–f). The SST pattern indicates the most pronounced warming in the equatorial Pacific over the first 40 years of the twenty-first century. The wind pattern is fairly comparable to that shown in Fig. 2b, but the anomalous cyclonic circulation shifts northward by 15° latitudes. The resulting wind anomalies in the western part of the WNP basin generally agree with these shown in Figs. 1b and 2b, which are also comparable to the changes in TC translation vectors during the period 1965–2009 (Fig. 1d). Given the similar pattern of the projected changes in large-scale steering flows over the period 2001–40, it is suggested that increasing TC influence in the subtropical East Asia and decreasing TC influence over the South China Sea may persist by 2040. Note that this projection is different from Wu and Wang (2004). They derived the projected large-scale steering flow from the two global warming experiments of the GFDL climate model for the IPCC AR3 and argued that the prevailing TC tracks shift slightly southward during the period 2000–29.

5. Summary

While TC track changes in the WNP over the past several decades were detected (Wu et al. 2005; Tu et al. 2009), whether they were associated with the ongoing global warming is discussed in this study through the SVD analysis of observations and IPCC climate change simulations. With increasing TC influence over the subtropical East Asia and decreasing TC activity over the South China Sea, the observed TC track changes are linked to the SVD leading mode. The spatial patterns are characterized with global warming in SST, and the associated changes in the large-scale steering flows agree well with the linear trends in the TC translation velocity over the western part of the WNP basin.

The SVD analysis is also conducted with the ensemble control simulation of the five selected IPCC models in the AR4 and the associated projection under the A1B scenario. The models are generally capable of simulating the observed features in global SST warming and the changes of large-scale steering flows in the western part of the WNP basin. The projected changes in the large-scale steering flows are similar to those observed over the period 1965–2009, suggesting that the increasing TC influence in subtropical East Asia and decreasing TC influence over the South China Sea, which were observed over the period 1965–2009, will continue by 2040. It should be pointed out that changes in TC tracks are determined by changes in both large-scale steering flows and TC formation location. This study assumes little change in TC formation location in the future because of the lack of convincing projection. Thus, caution should be taken to interpret the projection in this study.

Acknowledgments

This research was jointly supported by the National Basic Research Program of China (2009CB421503), the National Natural Science Foundation of China (40875038), and the State Key Laboratory of Severe Weather of Chinese Academy of Meteorological Sciences (2010LASW-A03).

REFERENCES

REFERENCES
Chan
,
J. C. L.
, and
J.
Shi
,
1996
:
Long-term trends and interannual variability in tropical cyclone activity over the western North Pacific
.
Geophys. Res. Lett.
,
23
,
2765
2767
.
Ho
,
C. H.
,
J. J.
Baik
,
J. H.
Kim
,
D. Y.
Gong
, and
C. H.
Sui
,
2004
:
Interdecadal changes in summertime typhoon tracks
.
J. Climate
,
17
,
1767
1776
.
Holland
,
G. J.
,
1993
:
Tropical cyclone motion
.
Global guide to tropical cyclone forecasting, G. J. Holland, Ed., Tropical Cyclone Programme Rep. TCP-31 and WMO Tech. Doc. WMO/TD-560. [Available online at http://www.cawcr.gov.au/publications/BMRC_archive/tcguide/ch3/ch3_tableofcontents.htm.]
Knutson
,
T. R.
, and
Coauthors
,
2010
:
Tropical cyclones and climate change
.
Nat. Geosci.
,
3
,
157
163
,
doi:10.1038/ngeo779
.
Liu
,
K. S.
, and
J. C. L.
Chan
,
2008
:
Interdecadal variability of western North Pacific tropical cyclone tracks
.
J. Climate
,
21
,
4464
4476
.
Solomon
,
S.
,
D.
Qin
,
M.
Manning
,
M.
Marquis
,
K.
Averyt
,
M. M. B.
Tignor
,
H. L.
Miller
Jr.
, and
Z.
Chen
, Eds.,
2007
:
Climate Change 2007: The Physical Sciences Basis
.
Cambridge University Press, 996 pp
.
Tu
,
J. Y.
,
C.
Chou
, and
P. S.
Chu
,
2009
:
The abrupt shift of typhoon activity in the vicinity of Taiwan and its association with western North Pacific–East Asian climate change
.
J. Climate
,
22
,
3617
3628
.
Wang
,
B.
, and
J. C.
Chan
,
2002
:
How strong ENSO events affect tropical storm activity over the western North Pacific
.
J. Climate
,
15
,
1643
1658
.
Wang
,
B.
,
R.
Elsberry
,
Y.
Wang
, and
L.
Wu
,
1998
:
Dynamics in the tropical cyclone motion: A review
.
Chinese J. Atmos. Sci.
,
22
,
535
547
.
Wu
,
L.
, and
B.
Wang
,
2004
:
Assessing impacts of global warming on tropical cyclone tracks
.
J. Climate
,
17
,
1686
1698
.
Wu
,
L.
,
B.
Wang
, and
S.
Geng
,
2005
:
Growing typhoon influence on east Asia
.
Geophys. Res. Lett.
,
32
,
L18703
,
doi:10.1029/2005GL022937
.
Zhou
,
W.
, and
J. C.
Chan
,
2007
:
ENSO and South China Sea summer monsoon onset
.
Int. J. Climatol.
,
27
,
157
167
.