ADVANCES IN ATMOSPHERIC SCIENCES, VOL. 28, NO.

5, 2011, 11091117

Oceanic Origin of A Recent La Nia-Like Trend n in the Tropical Pacic

ZHANG Liping 1 (
1 2

), WU Lixin1 (

), and YU Lisan2 (

Physical Oceanography Laboratory, Ocean University of China, Qingdao 266100

Department of Oceanography, Woods Hole Oceanography Institute, MA 02543, USA (Received 21 July 2010; revised 31 October 2010) ABSTRACT

Global ocean temperature has been rising since the late 1970s at a speed unprecedented during the past century of recordkeeping. This accelerated warming has profound impacts not only on the marine ecosystem and oceanic carbon uptake but also on the global water cycle and climate. During this rapid warming period, the tropical Pacic displays a pronounced La Nia-like trend, characterized by an intensication of west n east SST gradient and of atmospheric zonal overturning circulation, namely the Walker circulation. This La Nia-like trend diers from the El Nio-like trend in warm climate projected by most climate models, n n and cannot be explained by responses of the global water cycle to warm climate. The results of this study indicate that the intensication of the zonal SST gradient and the Walker circulation are associated with recent strengthening of the upper-ocean meridional overturning circulation. Key words: tropical Pacic, warming trend, hydrological cycle, subtropical tropical cell Citation: Zhang, L. P., L. X. Wu, and L. S. Yu, 2011: Oceanic origin of a recent La Ni a-Like trend in the n Tropical Pacic. Adv. Atmos. Sci., 28(5), 11091117, doi: 10.1007/s00376-010-0129-6.

1.

Introduction

Tropical Pacic Oceanatmosphere circulation can regulate global climate profoundly through atmospheric teleconnections (Alexander et al., 2002; Yang and Li, 2005; Wu et al., 2010; Xiao et al., 2010). A long-standing debate has been focusing on the trend of tropical Pacic SST westeast gradient in global warming, i.e., a La Ni a-like (enhanced zonal SST gran dient) trend or an El Ni o-like (reduced zonal SST gran dient) trend (Cubasch et al., 2001; Collins et al., 2005; Cane, 2005). The potential change of the atmospheric Walker circulation is also associated with this trend (Vecchi et al., 2006; Vecchi and Soden, 2007). Climatemodel projections and reconstructions have produced very diverse results, hindering the understanding of mechanisms and the ability to predict future tropical climate changes. Several competing hypotheses have been proposed to explain potential shift of the tropical coupled ocean atmosphere circulation in warm climate, including effects of latent heat cooling (Knutson and Manabe,
Corresponding

1995), cloud cover, and albedo feedbacks (Meehl and Washington, 1996) favoring El Nio-like SST response, n and oceanic dynamics favoring La Nia-like SST ren sponse (Clement et al., 1996; Seager and Murtugudde, 1997). Recent hypotheses have tended to link the tropical oceanatmospheric circulation with changes in the global hydrological cycle induced by global warming (Held and Soden, 2006; Vecchi and Soden, 2007). The 21st-century simulations of the Intergovernmental Panel on Climate Change (IPCC) models demonstrate a weakening trend of the Walker circulation associated with an El Ni o-like trend in the tropical Pacic. The n cause of the slowing Walker circulation has been suggested to be the dierence in the rate of increase of moistening and the rate of increase of precipitation in warm climates. A recent study attempted to unify these arguments (Karnauskas et al., 2009) by demonstrating that the potential mechanisms mentioned previously are at work but with relative strengths that vary seasonally. Therefore, these theories need not be inconsistent. While the cause of these trends remains elusive, the rapid warming of Earths climate

author: ZHANG Liping, ocean climate ping@yahoo.cn

China National Committee for International Association of Meteorology and Atmospheric Sciences (IAMAS), Institute of Atmospheric Physics (IAP) and Science Press and Springer-Verlag Berlin Heidelberg 2011

1110

A RECENT LA NINA-LIKE TREND IN THE TROPICAL PACIFIC

VOL. 28

and the accumulation of more reliable observational data during the past 30 years provide a unique opportunity to investigate possible mechanisms of the tropical oceanatmosphere circulation responses to warm climate (Chen et al., 2002; Soden et al., 2005; Mitas and Clement, 2005; Fu et al., 2006). In this study, analysis of the trend of tropical Pacic Ocean atmosphere response in warmer climate between 1980 and 2006 has been based on series of observation datasets, reanalysis data products, and climate-model data outputs. Meanwhile, we sought to determine the unied response between ocean and atmosphere and its associated potential mechanisms during this period. Available datasets are briey described in the following section. The trend of response of tropical Pacic Ocean and its associated atmosphere variation in the past three decades (19802006) are analyzed in section 3. A discussions of the unied trend in both ocean and atmosphere (i.e., whether or not they can be explained by hydrological cycle changes) is presented in section 4. In the following section 5 a possible mechanism responsible for the tropical Pacic trend in recent warm climate is proposed. Discussion and summary are presented in section 6.

2.

Datasets

Various ocean datasets were used to assess recent tropical Pacic variability: Hadley Center Sea Ice and Sea Surface Temperature (HadISST) dataset (Rayner et al., 2003), Kaplan Extended SST data (Kaplan et al., 1998), the National Oceanic and Atmospheric Administration (NOAA) Extended Reconstructed SST data (Smith et al., 2008), Simple Ocean Data Assimilation (SODA) subsurface temperature data (Carton and Giese, 2008), Ishiis data (Ishii et al., 2006), and German contribution to the Estimating the Climate and Circulation of the Ocean assimilation eort (GECCO) data (Stammer et al., 2004). To evaluate the associated atmosphere response from 1980 to 2006, atmospheric datasets were also adopted: Global Precipitation Climatology Project (GPCP) product (Adler et al., 2003), global ocean atmosphere ux data product developed recently by Woods Hole Oceanographic Institute (Yu and Weller, 2007), NCEP/NCAR Reanalysis product, and the ECWMF ERA40 Reanalysis product. Finally, 20thcentury simulations from the IPCC Fourth Assessment Report (AR4) were also used for comparison with ob-

Fig. 1. Time series of global mean SST (black line) and 30-year running trend (gray line) in the commonly used datasets: (a) HadISST, (b) ERSST, and (c) Kaplan SST. Right panel shows the corresponding seasonal trend of the zonal SST gradient dened as the dierence between west (5 S5 N, 148 E160 W) and east (5 S5 N, 130 W80 W) in the tropical Pacic in the past 30 years (19802006). Gray shading denotes 99% condence intervals based on the nonparametric Sen median slope method (Sen, 1968; Thiel, 1950; Salmi et al., 2002). Units for SST time series and trends are C and C (10 yr)1 , respectively.

NO. 5

ZHANG ET AL.

1111

servation data. In this study, nearly all available ocean atmosphere datasets were used to demonstrate a tropical Pacic trend from 1980 to 2006 and its potential mechanisms. 3. La Ni a-like warming trend in the Tropical n Pacic

The nonuniform amplitudes of global warming during the 137 years of data gathering agree well across the three widely used datasets, with maximum warming evident after 1980 (Figs. 1ac). Thirty years running trend further demonstrates that the warming after 1980 has a salient increasing velocity (Figs. 1ac). The tropical pacic trend during the rapid warming period (19802006) was considered with special regard to the anisotropy of warming time series. To estimate the strength of the equatorial Pacic zonal SST gradient, the dierence between SST averaged over the western equatorial Pacic Ocean (5 S 5 N, 148E160W) and the eastern equatorial Pacic Ocean (5 S5 N, 130W80 W), was computed

as x SST. The seasonal trend in the equatorial Pacic x SST agrees well with the three datasets (Figs. 1df), with a signicant strengthening every month. This indicates that tropical Pacic is characterized by a La Ni a-like trend during the rapid warming period. n However, a trend relating to the strong seasonal cycle, with positive peak in fall and negative peak in spring, is not signicant between 1980 and 2006, which diers from the seasonality of the long-term trend proposed by Karnauskas et al. (2009). In our study, the trend of peaks in winter is in phase with the seasonal development of a typical La Nia event, indicating that n dierent origins dominate. Next, we concentrated on the spatial distribution of SST trends. Global oceans underwent a general warming trend; these trends were stronger in the extratropical oceans, signifying the poleward amplication of global warming (Fig. 2a). However, the eastern tropical Pacic underwent a cooling trend, leading to an enhanced zonal SST gradient. A close inspection found that the zonal SST gradient increases at a rate of 0.22 C (10 yr)1 .

Fig. 2. LaNia-like trend in the tropical Pacic. (a) Linear trend of global SST. (b, c) Vertical n longitudinal prole of mean (white contours) and linear trend (shaded) of equatorial temperature in the SODA and Ishiis data. (d) Mean (shaded) and linear trend of equatorial upwelling in GECCO data. Units for temperature and upwelling trends are C and 104 cm s1 (10 yr)1 , respectively. The analyzed period is 1980 to 2006.

1112

A RECENT LA NINA-LIKE TREND IN THE TROPICAL PACIFIC

VOL. 28

The La Ni a-like trend was also revealed in the n changes of equatorial subsurface temperate from different reanalysis datasets, including the SODA (Carton and Giese, 2008) and the Ishii data (Ishii et al., 2006). In spite of dierences in magnitudes, both data clearly show deepening and shoaling of the equatorial thermocline in the west and the east, respectively (Figs. 2b, c). The maximum warming and cooling collocates with the mean thermocline, roughly around the isothermal of 20 C. The coherence of temperature trends in both surface and subsurface suggests a potential role of thermocline dynamics in the La Ni an like trend. This is further demonstrated by changes of equatorial upwelling from independent GECCO reanalysis data (Stammer et al., 2004). Overall, the upwelling intensied in the central equatorial Pacic and the eastern equatorial Pacic but decreased in the western equatorial Pacic during this period, which led to a strengthening of the equatorial thermocline (Fig. 2d). Consistent with the La Ni a-like trend,the atmon spheric circulation is characterized by a strengthening

of the Walker Circulation. The sea-level pressure (SLP) data shows a reduction in the west and intensication in the east at a rate of 0.2 hPa (10 yr)1 (Fig. 3a). The seasonal trend also exhibits a maximum trend of zonal SLP gradient peaks during winter (Fig. 4a), consistent with the SST seasonality. The enhanced surface convergence in the west corresponds to the SST increase, which led to an intensication of deep convection. This is demonstrated by the upward velocity at 500 hPa, which was enhanced at a rate of 0.004 Pa s1 (10 yr)1 over the warm pool region (Fig. 3a). In the east, the cooling trend corresponds to an enhancement of downward velocity and surface divergence. The intensication of the Walker Circulation coupled with the La Nia-like SST trend can be n also inferred from changes of the precipitation (Fig. 3b). The precipitation from the Global Precipitation Climatology Project (GPCP) data product (Adler et al., 2003) clearly displays an intensication trend with a magnitude of 0.2 mm d1 (10 yr)1 over the warm pool region, as a result of local warming. Compared

Fig. 3. Atmospheric circulation and precipitation trend. (a) Linear trend of sea level pressure [contours, units: hPa (10 yr)1 ] and vertical velocity (positive upward) at 500 hPa [shaded, units: 1/15 Pa s1 (10 yr)1 ]. (b) Linear trend of precipitation [units: mm d1 (10 yr)1 ].

Fig. 4. Trends of observed equatorial Pacic zonal (a) sea level pressure, (b) vertical velocity (positive upward), and (c) precipitation gradient from 1980 to 2006 as a function of calendar month. Gray shading denotes 99% condence intervals based on the nonparametric Sen median slope method. Units for sea level pressure, vertical velocity and precipitation trend are hPa (10 yr)1 , Pa s1 (10 yr)1 , and mm d1 (10 yr)1 , respectively.

NO. 5

ZHANG ET AL.

1113

Table 1. Fractional increase of Precipitation (P ), moisture (q), surface temperature (T ) from 1980 to 2006. The fractional increase of moisture is calculated based on WHOI OAux data for global ocean, and the ECWMF ERA40 Reanalysis product (19802002) for global ocean and land, respectively. Levels at 90% condence are given for each quantity. P/P (%) Global Ocean Global ocean and land 2.541.6 1.811.4 q/q (%) 3.351.7 4.422.6 T ( C) 0.30280.16 0.41470.2

with SST trend seasonality, the trend in peak of vertical velocity associated with precipitation displays some lags, implicating the passive response of atmosphere circulation (Figs. 4bc). 4. Can this trend be explained by global hydrological cycle changes during this period?

A La Ni a-like trend from 1980 to 2006 in the tropn ical Pacic has been conrmed. Can this trend be explained by changes in the global hydrological cycle during this period? To quantify hydrological cycle, we used the global oceanatmosphere ux data product developed recently by Woods Hole Oceanographic Institute (Yu and Weller, 2007). During this period over the global ocean, the evaporation increased by an average of 6.45% (E/E) and precipitation increased by an average of 2.54% (P/P ) (Fig. 5a). If the fractional increase in evaporation is further decomposed into the fractional increases in wind speed (U ) and humidity (q) as E/E = U/U + q/q , the evaporation increases due to wind speed change (based on NECP/NCAR Reanalysis data) is 3.1% (U/U ), and the humidity increases 3.35% (q/q = E/E U/U =6.45%3.1%). During this period, the global SST increases 0.3 C, therefore, the rate of moistening (q/q/T = E/E/T U/U/T ) increases 10%/ C, and precipitation (P/P/T ) increases 8%/ C. During this period the rate of moistening increase slightly exceeded the rate of the precipitation increase. The dierence is amplied if land data are included (Table 1). The dierence in the rate of increase of moistening and the rate of increase of precipitation from 1980 to 2000 is also captured by climate model simulations of 20th century (20C3M) organized by the program for Climate Model Diagnosis and Intercomparision (PCMDI) for the IPCC AR4. The models were forced by well-mixed greenhouse gases, aerosols, etc. All models showed an increase in surface air temperature (SAT) with considerable variability from 1980 to 2000 (Fig. 5b). In spite of a wide range of precipitation changes, these models display a strong coupling

Fig. 5. Hydrological cycle response in observation and climate models. (a) Time series of annual mean evaporation and precipitation averaged over the global ocean (units: mm d1 ). (b) Fractional increase of columnintegrated water vapor and precipitation versus increase of globally averaged surface air temperature ( C) during 19802000 simulations of the 24 IPCC AR4 climate models. (c) Distribution of linear trend of westeast equatorial SST contrast for 24 IPCC AR4 models. Red dot denotes the corresponding linear trend in observation.

1114

A RECENT LA NINA-LIKE TREND IN THE TROPICAL PACIFIC

VOL. 28

Fig. 6. Equatorial zonal SST gradient and Pacic STC. Scatter plot of the Pacic STC mass transport convergence [dened as the sum of mass transport within 2226.2 ( is isopycnal surface) at 9 N and 22.526.2 at 9 S] and equatorial zonal SST gradient from (a) GECCO and (b) SODA. Linear trend (shaded) and mean climatology (contours) of mass transport were integrated from coast to coast. Units for mean and linear trend are Sv and Sv (10 yr)1 , respectively.

between precipitation changes and temperature changes. The magnitude of the precipitation increase is roughly 1% K1 2% K1 . The fractional increase in total column water vapor is linearly proportional to the SAT increase, with a rate of 7% K1 , which resembles that expected from ClausiusClapeyron (C C) thermodynamic scaling. Although observations and climate models consistently indicate a greater rate of moistening increase than the rate of precipitation increase, the tropical Pacic responses do not show an El Nio-like response. n Observations indicate a La Ni a trend in the tropin cal Pacic (red dot in Fig. 5c), while climate models show either El Nio-like or La Ni a-like response withn n out preference (Fig. 5c). Therefore, the tropical Pacic Oceanatmosphere circulation changes may not be directly regulated by the global hydrological cycle responses to warm climate during this short period. 5. Ocean origin (STC) of La Ni a-like warmn ing trend

The recent La Ni a-like trend in the tropical Pacic n is associated with acceleration of upper-ocean shallow meridional overturning circulation. In the subtropical and tropical Pacic Ocean, water masses subduct into the thermocline in the eastern subtropics of both hemi-

spheres, enter into the equatorial thermocline through western boundary current and interior path, upwell to the surface in the eastern equatorial Pacic, and return to the subtropics through surface Ekman ow. These shallow, subtropicaltropical cells (STCs; McCreary and Lu, 1994; Liu et al., 1994) have been suggested to play an important role in the tropical climate variations (Gu and Philander, 1997; Kleeman et al., 1999). The equatorward mass convergence of STC was measured by total transport across 9 N and 9 S in both hemispheres based on the SODA and GECCO Ocean Reanalysis data (Schott et al., 2007; Schott et al., 2008). Analyses of both datasets indicate that the equatorial Pacic SST gradient displays a signicant correlation with the STC strength: a stronger STC corresponds to an intensication of the zonal SST gradient, and a weaker STC corresponds to a weakening of the zonal SST gradient (Fig. 6a). A stronger STC can pump more cold subtropical water into the equatorial thermocline, leading to cold surface anomalies in the eastern equatorial Pacic cold tongue and thus strengthening the zonal SST gradient. A weaker STC can pump less cold subtropical water into the equatorial thermocline, leading to warm surface anomalies in the eastern equatorial Pacic cold tongue and thus weakening the zonal SST gradient. Analysis of both SODA and GECCO data indicates that 1 Sv (106 m3

NO. 5

ZHANG ET AL.

1115

s1 ) increase of STC mass transport corresponds to 0.15C increase of the equatorial zonal SST gradient. In addition to year-to-year variability, an acceleration trend of STC was also revealed in the analysis of the two datasets, although there are some notable dierences between the datasets (Fig. 6b). The total equatorward transport across 9 N/S has an increasing trend of 1.41.1 Sv (10 yr)1 and 2.01.3 Sv (10 yr)1 in SODA and GECCO during 19802006, which may correspond to 0.2 C0.16C (GECCO) or 0.3 C0.2 C (SODA) increase of the equatorial zonal SST gradient (10 yr)1 . This is close to the trend observed [0.22 C0.17C (10 yr)1 ] in the equatorial Pacic. What drives a recent intensication of the STC? The direct forcing for the STC is the trade winds in both hemispheres that are associated with the atmospheric meridional overturning circulation, namely Hadley circulation. Both NCEP/NCAR and ERA40 atmospheric reanalysis data show an intensication of Hadley circulation during the past several decades (Quan et al., 2004; Tanaka et al., 2004; Mitas and Clement, 2005), although most climate models show either a negligible or a decreasing trend (Mitas and Clement, 2006). Although the discrepancies between atmospheric reanalysis and climate-model simulations and bias in both satellite observations and models raise uncertainties about the Hadley circulation changes in the past several decades, the La Nia-like SST trend n and the acceleration of oceanic meridional overturning circulation may be indicative of the acceleration of the Hadley cell. 6. Summary

ERSST dataset and the corresponding interior STC, has weakened in recent 50 years. Although both the study of Zhang and McPhaden (2006) and our study emphasize the inuence of the ocean bridge (STC) on tropical climate change, they dier with regard to studied time range. Indeed, the tropical Pacic SST trend during the past 50 years show great dierences among dierent datasets, with an El Ni o-like pattern n in the ERSST data and an La Nio-like pattern in the n HadISST and Kaplan SST datasets. The period and data dependence of the tropical Pacic trend indicate that the associated mechanism may include many aspects, with dierent dominating factors during dierent periods and therefore across datasets. As revealed by Seager and Murtugudde (1997), even with xed winds, the equatorial SST gradient can be strengthened by uniformly positive radiative forcing because of an increase in the oceans thermal stratication. Therefore, it is possible that an increased SST gradient can be a consequence of changes in the atmospheric energy and moisture budget, even if the SLP gradient and associated zonal winds are weakened (Vecchi et al., 2006). Further observations and multi-model comparisons are clearly necessary to improve our understanding of climate changes in the tropical Pacic.
Acknowledgements. This work was supported by the Major Project of National Science Foundation of China (Grant Nos. 40890150 and 40890155) and the National Science Foundation for Distinguished Young Scholars of China (Grant No. 40788002). The authors appreciate their discussions with Drs. Zhengyu LIU and Ping CHANG. REFERENCES Adler, R. F., and Coauthors, 2003: The version 2 global precipitation climatology project (GPCP) monthly precipitation analysis (1979Present). Journal of Hydrometeorology, 4, 11471167. Alexander, M. A., I. Blade, M. Newman, J. R. Lanzante, N. Lau, and J. Scott, 2002: The atmospheric bridge: The inuence of ENSO teleconnections on air-sea interaction over global oceans. J. Climate, 15, 2205 2231. Cane, M., 2005: The evolution of El Nio, past and fun ture. Earth and Planetary Science Letters, 230, 227 240. Carton, J. A., and B. Giese, 2008: A reanalysis of ocean climate using SODA. Mon. Wea. Rev., 136, 2999 3017. Chen, J. Y., B. E. Carlson, and A. D. Genio, 2002: Evidence for strengthening of the tropical general circulation in the 1990s. Science, 295, 838 841. Clement, A. C., R. Seager, M. A. Cane, and S. E. Zebiak, 1996: An ocean dynamical thermostat. J. Climate, 9, 21902196.

The tropical Pacic SST trend of recent rapid warming from 1980 to 2006 was analyzed based on various observations, reanalysis products, and IPCC models. The tropical Pacic was found to exhibit a La Nia-like trend characterized by warming in n the west and cooling in the east. The strengthened zonal SST gradient, coupled with intensied atmosphere Walker circulation, may be explained by acceleration of upper-ocean shallow meridional overturning circulation, without an apparent link to the changes in the global hydrological cycle during this period. Whether the current trend is a part of natural multidecadal variations or is a result of forcing by elevation of greenhouse gases, the mechanisms outlined here shed light on the processes and dynamics of tropical oceanatmosphere circulation changes in response to global warming. Zhang and McPhaden (2006) argued that tropical SST, characterized by an El Nio-like trend using the n

1116

A RECENT LA NINA-LIKE TREND IN THE TROPICAL PACIFIC

VOL. 28

Collins, M., and Coauthors, 2005: El Nino- or La Ninalike climate change? Climate Dyn., 24, 89104. Cubasch, U., and Coauthors, 2001: Projections of future climate change. Climate Change 2001: The Scientic Basis, Houghton et al., Eds., Cambridge University Press, 525582. Fu, Q., C. M. Johanson, J. M. Wallace, and T. Reichler, 2006: Enhanced mid-latitude tropospheric warming in satellite measurements. Science, 312, 1179. Gu, D., and S. G. H. Philander, 1997: Interdecadal climate uctuations that depend on exchanges between the tropics and extratropics. Science, 275, 805807. Held, I. M., and B. J. Soden, 2006: Robust responses of the hydrologic cycle to global warming. J. Climate, 19, 56865699. Ishii, M., M. Kimoto, K. Sakamoto, and S. I. Iwasaki, 2006: Steric sea level changes estimated from historical ocean subsurface temperature and salinity analyses. Journal of Oceanography, 62(2), 155170. Kaplan, A., M. Cane, Y. Kushnir, A. Clement, M. Blumenthal, and B. Rajagopalan, 1998: Analyses of global sea surface temperature 18561991. J. Geophys. Res., 103, 1856718589. Karnauskas, K., R. Seager, A. Kaplan, Y. Kushnir, and M. Cane, 2009: Observed strengthening of the zonal sea surface temperature gradient across the equatorial Pacic Ocean. J. Climate, 22, 43164321. Kleeman, R., J. P. McCreary, and B. Klinger, 1999: A mechanism for generating ENSO decadal variability. Geophys. Res. Lett., 26, 17431746. Knutson, T. R., and S. Manabe, 1995: Time-mean response over the tropical Pacic to increased CO2 in a coupled ocean-atmosphere model. J. Climate, 8, 21812199. Liu, Z., S. G. H. Philander, and R. C. Pacanowski, 1994: A GCM study of tropical-subtropical upper-ocean water exhange. J. Phys. Oceanogr., 24, 26062623. McCreary, J. P., and P. Lu, 1994: On the interaction between the subtropical and the equatorial oceans: The subtropical cell. J. Phys. Oceanogr., 24, 466497. Meehl, G. A., and W. M. Washington, 1996: El Nion like climate change in a model with increased atmospheric CO2 concentration. Nature, 382, 5660. Mitas, C. M., and A. Clement, 2005: Has the Hadley cell been strengthening in recent decades? Geophys. Res. Lett., 32, L03809, doi: 10.1029/2004GL021765. Mitas, C. M., and A. Clement, 2006: Recent behavior of the Hadley cell and tropical thermodynamics in climate models and reanalyses. Geophys. Res. Lett., 33, L01810, doi: 10.1029/2005GL024406. Quan, X., H. F. Diaz, and M. P. Hoerling, 2004: Change in the tropical Hadley cell since 1950. The Hadley Circulation: Past, Present, and Future, Diaz and Bradley, Eds., Cambridge Univ. Press, 85120. Rayner, N., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res.,

108(D14), 4407, doi: 10.1029/2002JD002670. Salmi, T., A. Mtt, P. Anttila, T. Ruoho-Airola, and a a T. Amnell, 2002: Detecting trends of annual values of atmospheric pollutants by the Mann-Kendall test and Sens slope estimatesThe Excel template application MAKESENS. Publication on Air Quality. Finnish Meteorological Institute, Helsinki, No. 31, 35pp. Schott, F. A., L. Stramma, W. Wang, B. S.Giese, and R. Zantopp, 2008: Pacic subtropical cell variability in the SODA 2.0.2/3 assimilation. Geophys. Res. Lett., 35, L10607, doi: 10.1029/2008GL033757. Schott, F. A., W. Wang, and D. Stammer, 2007: Variability of the Pacic subtropical cells in the 50-year ECCO assimilation. Geophys. Res. Lett., 34, L05604, doi: 10.1029/2006GL028478. Seager, R., and R. Murtugudde, 1997: Ocean dynamics, thermocline adjustment, and regulation of tropical SST. J. Climate, 10, 521533. Sen, P. K., 1968: Estimates of the regression coecient based on Kendalls tau. J. Amer. Stat. Assoc., 63, 13791389. Smith, T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore, 2008: Improvements to NOAAs historical merged landocean surface temperature analysis (18802006). J. Climate, 21, 22832296. Soden, B. J., D. L. Jackson, V. Ramaswamy, D. Schwarzkopf, and X. Huang, 2005: The radiative signature upper tropospheric moistening. Science, 310, 841844. Stammer, D., K. Ueyoshi, A. Kohl, W. B. Large, S. Josey, and C. Wunsch, 2004: Estimating air-sea uxes of heat, freshwater and momentum through global ocean data assimilation. J. Geophys. Res., 109, C05023, doi: 10.1029/2003JC002082. Tanaka, H. L., N. Ishizaki, and A. Kitoh, 2004: Trend and interannual variability of Walker, monsoon and Hadley circulations dened by velocity potential in the upper troposphere. Tellus, 56, 250269. Theil, H., 1950: A rank-invariant method of linear and polynomial regression analysis. Proceedings of the Royal Netherlands Academy of Science, 53A, 1397 1412. Vecchi, G. A., B. J. Soden, A. T. Wittenberg, I. M. Held, A. Leetmaa, and M. J. Harrison, 2006: Weakening of tropical Pacic atmospheric circulation due to anthropogenic forcing. Nature, 441, 7376. Vecchi, G. A., and B. J. Soden, 2007: Global Warming and the weakening of the tropical circulation. J. Climate, 20, 43164340. Wu, S., L. X. Wu, Q. Y. Liu, and S.-P. Xie, 2010: Development processes of the Tropical Pacic Meridional Mode. Adv. Atmos. Sci., 27(1), 9599, doi: 10.1007/s00376-009-8067-x. Xiao, Y., Z. Zhang, J. He, L. Qi and Y. Ren, 2010: Multi-scale interactions between the Indian Ocean dipole and tropical Pacic sea surface temperature anomalies. Chinese Journal of Atmospheric Sciences, 34(3), 483494, doi: 10.1006/9895(2010) 03-0483-12.

NO. 5

ZHANG ET AL.

1117

(in chinese) Yang, H., and C. Li, 2005: Eect of the tropical PacicIndian ocean temperature anomaly mode on the South Asia high. Adv. Atmos. Sci., 29, 99110. Yu, L., and R. A. Weller, 2007: Objectively analyzed air-sea heat ux for the global ice-free oceans (1981

2005). Bull. Amer. Meteor. Soc., 88, 527539. Zhang, D., and M. J. McPhaden, 2006: Decadal variability of the shallow Pacic meridional overturning circulation: Relation to tropical sea surface temperatures in observations and climate change models. Ocean Modelling, 15, 250273.