Advanced Search
Article Contents

Anatomizing the Ocean's Role in Maintaining the Pacific Decadal Variability

Fund Project:

doi: 10.1007/s00376-013-3032-0

  • The role of ocean dynamics in maintaining the Pacific Decadal Variability (PDV) was investigated based on simulation results from the Parallel Ocean Program (POP) ocean general circulation model developed at the Los Alamos National Laboratory (LANL). A long-term control simulation of the LANL-POP model forced by a reconstructed coupled wind stress field over the period 1949-2001 showed that the ocean model not only simulates a reasonable climatology, but also produces a climate variability pattern very similar to observed PDV. In the Equatorial Pacific (EP) region, the decadal warming is confined in the thin surface layer. Beneath the surface, a strong compensating cooling, accompanied by a basin-wide-scale overturning circulation in opposition to the mean flow, occurs in the thermocline layer. In the North Pacific (NP) region, the decadal variability nonetheless exhibits a relatively monotonous pattern, characterized by the dominance of anomalous cooling and eastward flows. A term balance analysis of the perturbation heat budget equation was conducted to highlight the ocean's role in maintaining the PDV-like variability over the EP and NP regions. The analyses showed that strong oceanic adjustment must occur in the equatorial thermocline in association with the anomalous overturning circulation in order to maintain the PDV-like variability, including a flattening of the equatorial thermocline slpoe and an enhancement of the upper ocean's stratification (stability), as the climate shifts from a colder regime toward a warmer one. On the other hand, the oceanic response in the extratropical region seems to be confined to the surface layer, without much participation from the subsurface oceanic dynamics.
    摘要: The role of ocean dynamics in maintaining the Pacific Decadal Variability (PDV) was investigated based on simulation results from the Parallel Ocean Program (POP) ocean general circulation model developed at the Los Alamos National Laboratory (LANL). A long-term control simulation of the LANL-POP model forced by a reconstructed coupled wind stress field over the period 1949-2001 showed that the ocean model not only simulates a reasonable climatology, but also produces a climate variability pattern very similar to observed PDV. In the Equatorial Pacific (EP) region, the decadal warming is confined in the thin surface layer. Beneath the surface, a strong compensating cooling, accompanied by a basin-wide-scale overturning circulation in opposition to the mean flow, occurs in the thermocline layer. In the North Pacific (NP) region, the decadal variability nonetheless exhibits a relatively monotonous pattern, characterized by the dominance of anomalous cooling and eastward flows. A term balance analysis of the perturbation heat budget equation was conducted to highlight the ocean's role in maintaining the PDV-like variability over the EP and NP regions. The analyses showed that strong oceanic adjustment must occur in the equatorial thermocline in association with the anomalous overturning circulation in order to maintain the PDV-like variability, including a flattening of the equatorial thermocline slpoe and an enhancement of the upper ocean's stratification (stability), as the climate shifts from a colder regime toward a warmer one. On the other hand, the oceanic response in the extratropical region seems to be confined to the surface layer, without much participation from the subsurface oceanic dynamics.
  • 加载中
  • Alexander, M. A., C. Deser, and M. S. Timlin, 1999: The reemergence of SST anomalies in the North Pacific ocean. J. Climate, 12, 2419-2433.
    An, S. I., and B. Wang, 2000: Interdecadal change of the structure of the ENSO mode and its impact on the ENSO Frequency. J. Climate, 13, 2044-2055.
    Barnett, T. P., D. W. Pierce, M. Latif, D. Dommenget, and R. Saravanan, 1999: Interdecadal interactions between the tropics and midlatitudes in the Pacific basin. Geophys. Res. Lett., 26, 615-618.
    Battisti, D. S., and A. C. Hirst, 1989: Interannual variability in a tropical atmosphere-ocean model: Influence of the basic state, ocean geometry and nonlinearity. J. Atmos. Sci., 46, 1687-1712.
    Biondi, F., A. Gershunov, and D. R. Cayan, 2001: North Pacific decadal climate variability since 1661. J. Climate, 14, 5-10.
    Carton, J. A., G. Chepurin, and X. H. Cao, 2000a: A Simple Ocean Data Assimilation analysis of the global upper ocean 1950-1995. Part 2: results. J. Phys. Oceanogr., 30, 311-326.
    Carton, J. A., G. Chepurin, X. H. Cao, and B. Giese, 2000b: A Simple Ocean Data Assimilation analysis of the global upper ocean. 1950-95. Part I: Methodology. J. Phys. Oceanogr., 30, 294-309.
    Conkright, M. E., R. A. Locarnini, H. E. Garcia, T. D. O'Brien, T. P. Boyer, C. Stephens, and J. I. Antonov, 2002: World Ocean Atlas 2001: Objective Analyses, Data Statistics, and Figures, CD-ROM Documentation. National Oceanographic Data Center, Silver Spring, MD, 17 pp.
    Deser, C. and M. L. Blackmon, 1995: On the relationship between tropical and North Pacific sea surface temperature variations. J. Climate, 8, 1677-1680.
    Deser, C., M. A. Alexand er, and M. S. Timlin, 1996: Upper-ocean thermal variations in the North Pacific during 1970-91. J. Climate, 9, 1840-1855.
    DiNezio, P. N., A. C. Clement, G. A. Vecchi, B. J. Soden, B. P. Kirtman, and S.-K. Lee, 2009: Climate response of the equatorial Pacific to global warming. J. Climate, 22, 4873-4892.
    Francis, R. C., S, R. Hare, A. B. Hollowed, and W. S. Wooster 1998: Effects of interdecadal climate variability on the oceanic ecosystems of the NE Pacific. Fish. Oceanogr., 7, 1-21.
    Gu, D. F., and S. G. Philander, 1997: Interdecadal climate fluctuations that depend on exchanges between the tropics and extratropics. Science, 275, 805-807.
    Jin, F.-F., 1997: A Theory of interdecadal climate variability of the North Pacific ocean-atmosphere system. J. Climate, 10, 1821-1835.
    Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-year reanalysis project. Bull. Amer. Meteor. Soc., 77, 437-471.
    Kleeman, R., J. P. McCreary Jr., and B. A. Klinger, 1999: A mechanism for generating ENSO decadal variability. Geophys. Res. Lett., 26, 1743-1746.
    Latif, M., 1998: Dynamics of interdecadal variability in coupled ocean-atmosphere models. J. Climate, 11, 602-624.
    Latif, M., and T. P. Barnett, 1994: Causes of decadal climate variability over the North Pacific and North America. Science, 266, 634-637.
    Latif, M., and T. P. Barnett, 1996: Decadal climate variability over the North Pacific and North America: Dynamics and predictability. J. Climate, 9, 2407-2423.
    Lau, N.-C., and M. J. Nath, 1996: The role of the "atmospheric bridge" in linking tropical Pacific ENSO events to extratropical SST anomalies. J. Climate, 9, 2036-2057.
    Liu, Z. Y., 2012: Dynamics of interdecadal climate variability: A historical perspective. J. Climate, 25, 1963-1995.
    Liu, Z. Y., and H. J. Yang, 2003: Extratropical control of tropical climate, the atmospheric bridge and oceanic tunnel. Geophys. Res. Lett., 30, 1230, doi: 10.1029/2002GL016492.
    Liu, Z. Y., and M. Alexander, 2007: Atmospheric bridge, oceanic tunnel, and global climatic teleconnections. Rev. Geophys., 45, RG2005, doi: 10.1029/2005RG000172.
    Miller, A. J., D. R. Cayan, T. P. Barnett, N. E. Graham, and J. M. Oberhuber, 1994: Interdecadal variability of the Pacific Ocean: Model response to observed heat flux and wind stress anomalies. Climate Dyn., 9, 287-302
    Nakamura, H., G. Lin, and T. Yamagata, 1997: Decadal climate variability in the North Pacific during the recent decades. Bull. Amer. Meteor. Soc., 78, 2215-2225.
    Newman, M., G. P. Compo, and M. A. Alexander, 2003: ENSO-forced variability of the Pacific decadal oscillation. J. Climate, 16, 3853-3857.
    Nitta, T., and S. Yamada, 1989: Recent warming of tropical sea surface temperature and its relationship to the Northern Hemisphere circulation. J. Meteor. Soc. Japan, 67, 187-193.
    Pierce, D. W., T. P. Barnett, N. Schneider, R. Saravanan, D. Dommenget, and M. Latif, 2001: The role of ocean dynamics in producing decadal climate variability in the North Pacific. Climate Dyn., 18, 51-70.
    Schneider, N., A J. Miller, M. A. Alexand er, and C. Desser, 1999: Subduction of decadal North Pacific temperature anomalies: Observations and dynamics. J. Phys. Oceanogr. 29, 1056-1070.
    Smith, R., and P. Gent, 2002: Reference Manual for the Parallel Ocean Program (POP). A Technical Report of Los Alamos National Laboratory, 73 pp.
    Smith, T. M., and R. W. Reynolds, 2003: Extended reconstruction of global sea surface temperatures based on COADS data (1854-1997). J. Climate, 16, 1495-1510.
    Solomon, A., and D. X. Zhang, 2006: Pacific subtropical cell variability in coupled climate model simulations of the late 19th-20th century. Ocean Modelling, 15, 236-249.
    Suarez, M. J. and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45, 3283-3287.
    Thompson, C. J., and D. S. Battisti, 2001: A linear stochastic dynamical model of ENSO. Part II: Analysis. J. Climate, 14, 445-466.
    Timmerman, A., and F.-F. Jin, 2002: A nonlinear mechanism for decadal El Ni#cod#241; amplitude changes. Geophys. Res. Lett., 29, 1003, doi: 10.1029/2001GL013369.
    Trenberth, K. E., 1990: Recent observed interdecadal climate changes in the northern hemisphere. Bull. Amer. Meteor. Soc., 71, 988-993.
    Weng, W. J., and J. D. Neelin, 1999: Analytical prototypes for ocean-atmosphere interaction at midlatitude. Part II: Mechanisms for coupled gyre modes. J. Climate, 12, 2757-2774.
    Wu, L., Z. Liu, R. Gallimore, R. Jacob, Y. Zhong, and D. Lee, 2003: Pacific decadal variability: The Tropical Pacific mode and the North Pacific mode. J. Climate, 16, 1101-1120.
    Wu, L. X., and Z. Y. Liu, 2003: Decadal variability in the North Pacific: The eastern North Pacific mode. J. Climate, 16, 3111-3131.
    Wu, L. X., and C. Li, 2007: Warming of the North Pacific ocean: Local air-sea coupling and remote climatic impacts. J. Climate, 20, 2581-2601.
    Xu, W., T. P. Barnett, and M. Latif, 1998: Decadal variability in the North Pacific as simulated by a hybrid coupled model. J. Climate, 11, 297-312.
    Yang, H. J., and Z. Y. Liu, 2005: Tropical-extratropical climate interaction as revealed in idealized coupled climate model experiments. Climate Dyn., 24, 863-879.
    Yang, H. J., and Q. Zhang, 2008: Anatomizing the ocean's role in ENSO changes under global warming. J. Climate, 21, 6539-6555.
    Yeh, S.-W., B. Dewitte, B. Y. Kim, and Y. Noh, 2010: Role of the upper ocean structure in the response of ENSO-like SST variability to global warming. Climate Dyn., 35, 355-369, doi: 10.1007/s00382-010-0849-4
    Yu, J.-Y., C.-W. Chang, and J.-Y. Tu, 2011: Evaluation and improvement of a SVD-based empirical atmospheric model. Adv. Atmos. Sci., 28, 636-652, doi: 10.1007/s00376-010-0029-9.
    Zhang, X. B., J. Sheng, and A. Shabbar, 1998, Modes of interannual and interdecadal variability of Pacific SST. J. Climate, 11, 2556-2569.
    Zhang, Y., J. M. Wallace, and D. S. Battisti, 1997: ENSO-like interdecadal variability: 1900-93. J. Climate, 10, 1004-1020.
  • [1] YANG Haijun, ZHANG Qiong, 2003: On the Decadal and Interdecadal Variability in the Pacific Ocean, ADVANCES IN ATMOSPHERIC SCIENCES, 20, 173-184.  doi: 10.1007/s00376-003-0002-y
    [2] HAN Leqiong, LI Shuanglin, LIU Na, 2014: An Approach for Improving Short-Term Prediction of Summer Rainfall over North China by Decomposing Interannual and Decadal Variability, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 435-448.  doi: 10.1007/s00376-013-3016-0
    [3] Yun YANG, Jianping LI, Lixin WU, Yu KOSAKA, Yan DU, Cheng SUN, Fei XIE, Juan FENG, 2017: Decadal Indian Ocean Dipolar Variability and Its Relationship with the Tropical Pacific, ADVANCES IN ATMOSPHERIC SCIENCES, 34, 1282-1289.  doi: 10.1007/s00376-017-7009-2
    [4] GAN Bolan, WU Lixin, 2012: Possible Origins of the Western Pacific Warm Pool Decadal Variability, ADVANCES IN ATMOSPHERIC SCIENCES, 29, 169-176.  doi: 10.1007/s00376-011-0193-6
    [5] Liu Hailong, Zhang Xuehong, Li Wei, 2001: The Heat Balance in the Western Equatorial Pacific Warm Pool during the Westerly Wind Bursts: A Case Study, ADVANCES IN ATMOSPHERIC SCIENCES, 18, 882-896.
    [6] Tiejun XIE, Ji WANG, Taichen FENG, Ting DING, Liang ZHAO, 2023: Linkage of the Decadal Variability of Extreme Summer Heat in North China with the IPOD since 1981, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 1617-1631.  doi: 10.1007/s00376-023-2304-6
    [7] Peng LIU, Chung-Hsiung SUI, 2014: An Observational Analysis of the Oceanic and Atmospheric Structure of Global-Scale Multi-decadal Variability, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 316-330.  doi: 10.1007/s00376-013-2305-y
    [8] LIU Xiangcui, LIU Hailong, 2014: Heat Budget of the South-Central Equatorial Pacific in CMIP3 Models, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 669-680.  doi: 10.1007/s00376-013-2299-5
    [9] Xiuping YAO, Qin ZHANG, Xiao ZHANG, 2020: Potential Vorticity Diagnostic Analysis on the Impact of the Easterlies Vortex on the Short-term Movement of the Subtropical Anticyclone over the Western Pacific in the Mei-yu Period, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 1019-1031.  doi: 10.1007/s00376-020-9271-y
    [10] FENG Lin, WU Dexing, LIN Xiaopei, MENG Xiangfeng, 2010: The Effect of Regional Ocean-Atmosphere Coupling on the Long-term Variability in the Pacific Ocean, ADVANCES IN ATMOSPHERIC SCIENCES, 27, 393-402.  doi: 10.1007/s00376-009-8195-3
    [11] WANG Jia, ZHI Xiefei, and CHEN Yuwen, 2013: Probabilistic multimodel ensemble prediction of decadal variability of East Asian surface air temperature based on IPCC-AR5 near-term climate simulations, ADVANCES IN ATMOSPHERIC SCIENCES, 30, 1129-1142.  doi: 10.1007/s00376-012-2182-9
    [12] Huang Fei, Zhou Faxiu, Qian Xiaodan, 2002: Interannual and Decadal Variability of the North Pacific Blocking and Its Relationship to SST,Teleconnection and Storm Tracks, ADVANCES IN ATMOSPHERIC SCIENCES, 19, 807-820.  doi: 10.1007/s00376-002-0046-4
    [13] Guanghui ZHOU, Rong-Hua ZHANG, 2022: Structure and Evolution of Decadal Spiciness Variability in the North Pacific during 2004–20, Revealed from Argo Observations, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 953-966.  doi: 10.1007/s00376-021-1358-6
    [14] Qin XU, Jie CAO, 2021: Iterative Methods for Solving the Nonlinear Balance Equation with Optimal Truncation, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 755-770.  doi: 10.1007/s00376-020-0291-4
    [15] SUN Shufen, ZHANG Xia, 2004: Effect of the Lower Boundary Position of the Fourier Equation on the Soil Energy Balance, ADVANCES IN ATMOSPHERIC SCIENCES, 21, 868-878.  doi: 10.1007/BF02915589
    [16] Fu Congbin, Henry Diaz, Fan Huijun, 1992: Variability in Latent Heat Flux over the Tropical Pacific in Association with Recent Two ENSO Events, ADVANCES IN ATMOSPHERIC SCIENCES, 9, 351-358.  doi: 10.1007/BF02656945
    [17] ZHENG Fei, ZHANG Rong-Hua, ZHU Jiang, , 2014: Effects of Interannual Salinity Variability on the Barrier Layer in the Western-Central Equatorial Pacific: A Diagnostic Analysis from Argo, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 532-542.  doi: 10.1007/s00376-013-3061-8
    [18] YUAN Yuan, C. L. Johnny CHAN, ZHOU Wen, LI Chongyin, 2008: Decadal and Interannual Variability of the Indian Ocean Dipole, ADVANCES IN ATMOSPHERIC SCIENCES, 25, 856-866.  doi: 10.1007/s00376-008-0856-0
    [19] Shenming FU, Jie CAO, Xingwen JIANG, Jianhua SUN, 2017: On the Variation of Divergent Flow: An Eddy-flux Form Equation Based on the Quasi-geostrophic Balance and Its Application, ADVANCES IN ATMOSPHERIC SCIENCES, 34, 599-612.  doi: 10.1007/s00376-016-6212-x
    [20] Baofeng JIAO, Lingkun RAN, Na LI, Ren CAI, Tao QU, Yushu ZHOU, 2023: Comparative Analysis of the Generalized Omega Equation and Generalized Vertical Motion Equation, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 856-873.  doi: 10.1007/s00376-022-1435-5

Get Citation+

Export:  

Share Article

Manuscript History

Manuscript received: 17 February 2013
Manuscript revised: 22 August 2013
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Anatomizing the Ocean's Role in Maintaining the Pacific Decadal Variability

    Corresponding author: Jia-Yuh YU; 
  • 1. Department of Atmospheric Sciences, Chinese Culture University, Taipei 11114 ; 
  • 2. Graduate Institute of Earth Science, Chinese Culture University, Taipei 11114
Fund Project:  This work was supported by the National Science Council (NSC) in Taiwan (Grant Nos. NSC99-2111-M-034-001 and NSC100-2119-M-034-002). The second author (CWC) was also sponsored by an NSC postdoctoral fellowship (Grant No. NSC99-2811-M-034-002). The authors would like to thank the two anonymous reviewers for their critical comments and helpful suggestions. The NCEPNCAR reanalysis atmospheric data and the NOAA ERSST data were both downloaded from the NOAA/ESRL website at http://www.esrl.noaa.gov/psd/data/gridded/.

Abstract: The role of ocean dynamics in maintaining the Pacific Decadal Variability (PDV) was investigated based on simulation results from the Parallel Ocean Program (POP) ocean general circulation model developed at the Los Alamos National Laboratory (LANL). A long-term control simulation of the LANL-POP model forced by a reconstructed coupled wind stress field over the period 1949-2001 showed that the ocean model not only simulates a reasonable climatology, but also produces a climate variability pattern very similar to observed PDV. In the Equatorial Pacific (EP) region, the decadal warming is confined in the thin surface layer. Beneath the surface, a strong compensating cooling, accompanied by a basin-wide-scale overturning circulation in opposition to the mean flow, occurs in the thermocline layer. In the North Pacific (NP) region, the decadal variability nonetheless exhibits a relatively monotonous pattern, characterized by the dominance of anomalous cooling and eastward flows. A term balance analysis of the perturbation heat budget equation was conducted to highlight the ocean's role in maintaining the PDV-like variability over the EP and NP regions. The analyses showed that strong oceanic adjustment must occur in the equatorial thermocline in association with the anomalous overturning circulation in order to maintain the PDV-like variability, including a flattening of the equatorial thermocline slpoe and an enhancement of the upper ocean's stratification (stability), as the climate shifts from a colder regime toward a warmer one. On the other hand, the oceanic response in the extratropical region seems to be confined to the surface layer, without much participation from the subsurface oceanic dynamics.

摘要: The role of ocean dynamics in maintaining the Pacific Decadal Variability (PDV) was investigated based on simulation results from the Parallel Ocean Program (POP) ocean general circulation model developed at the Los Alamos National Laboratory (LANL). A long-term control simulation of the LANL-POP model forced by a reconstructed coupled wind stress field over the period 1949-2001 showed that the ocean model not only simulates a reasonable climatology, but also produces a climate variability pattern very similar to observed PDV. In the Equatorial Pacific (EP) region, the decadal warming is confined in the thin surface layer. Beneath the surface, a strong compensating cooling, accompanied by a basin-wide-scale overturning circulation in opposition to the mean flow, occurs in the thermocline layer. In the North Pacific (NP) region, the decadal variability nonetheless exhibits a relatively monotonous pattern, characterized by the dominance of anomalous cooling and eastward flows. A term balance analysis of the perturbation heat budget equation was conducted to highlight the ocean's role in maintaining the PDV-like variability over the EP and NP regions. The analyses showed that strong oceanic adjustment must occur in the equatorial thermocline in association with the anomalous overturning circulation in order to maintain the PDV-like variability, including a flattening of the equatorial thermocline slpoe and an enhancement of the upper ocean's stratification (stability), as the climate shifts from a colder regime toward a warmer one. On the other hand, the oceanic response in the extratropical region seems to be confined to the surface layer, without much participation from the subsurface oceanic dynamics.

1. Introduction
  • The Pacific Decadal Oscillation (PDO), first documented by (Francis et al., 1998) in an attempt to connect Alaskan salmon production cycles to Pacific climate change, is a basin-wide-scale climate variability pattern that involves phase changes of SST anomalies (SSTAs) in the North Pacific on interdecadal time scales, generally around 20-30 years (Zhang et al., 1997; Latif, 1998; Zhang et al., 1998; Biondi et al., 2001). A coupled global ocean-atmosphere study by (Wu et al., 2003) further showed that the decadal variability in the Pacific consists of two distinct modes: a decadal to bi-decadal tropical Pacific mode (TPM) and a multi-decadal North Pacific mode (NPM). Accordingly, the PDO is often considered to be the Pacific climate variability north of 20#cod#x000b0;N, while the term "Pacific decadal variability" (PDV) is regarded as the climate variability over the entire Pacific (including both the TPM and NPM). During the warm (cold) phase of PDV, warm (cold) SSTAs occur over the tropical eastern Pacific as well as the coastal areas of the extratropical eastern Pacific, while cold (warm) SST anomalies appear over the central North Pacific. Meanwhile, the associated changes in atmospheric and oceanic circulations include an intensified (weakened) Aleutian Low (Nitta and Yamada, 1989; Trenberth, 1990; Nakamura et al., 1997) a flattened (tilted) equatorial thermocline (Trenberth, 1990; Alexander et al., 1999; An and Wang, 2000) and a possible weakening (strengthening) of the subtropical cell (Kleeman et al., 1999; Liu and Alexander, 2007; Wu and Li, 2007).

    Various hypotheses have been proposed to account for the origin of PDV, which include extratropical, tropical, and coupled extratopical-tropical origins. Since PDV typically exhibits greater amplitudes of SSTA in the North Pacific than in the tropical Pacific (Trenberth, 1990; Deser and Blackmon, 1995; Zhang et al., 1997), one hypothesis therefore emphasizes the importance of extratropical dynamics in producing the decadal variability. For instance, by analyzing observational data and outputs from coupled ocean-atmosphere model experiments, Latif and Barnett (1994, 1996) proposed that the decadal variability can be induced by thermocline memory and positive ocean-atmosphere feedback within the extratropical North Pacific. Their conclusions are supported by some theoretical and model studies (Jin, 1997; Xu et al., 1998; Weng and Neelin, 1999; Pierce et al., 2001) although the latter have bypassed complex physical processes in the ocean-atmosphere system.

    Another hypothesis emphasizes the importance of tropical origin in that PDV is regarded as a long-lived ENSO mode naturally existing in the tropical ocean-atmosphere system (An and Wang, 2000; Timmerman and Jin, 2002). (Battisti and Hirst, 1989) first showed that the tropical decadal variability, in theory, can be maintained by a "delayed oscillator" mechanism similar to ENSO (Suarez and Schopf, 1988) but operating on longer time scales. (Thompson and Battisti, 2001) further showed that the inclusion of stochastic forcing (e.g., atmospheric noises on monthly or seasonal time scales) in a delayed oscillator model of ENSO may produce both interannual and decadal variability.

    In contrast to the hypotheses concerning origin of decadal variability from regional dynamics, (Gu and Philander, 1997) argued that the existence of PDV depends on coupled tropical-extratropical interactions. In their hypothesis, a warming of SST in the equatorial eastern Pacific weakens the easterly trade wind and upwelling which, in turn, further amplifies the warming in the equatorial region. The atmosphere responds to the warming with an intensification of the extratropical westerlies, leading to colder surface waters in the extratopical regions due to increased latent heat loss. The cold surface waters are then pumped downward and equatorward to arrive at the equatorial thermocline about 12 years later, providing a negative feedback to halt the warming and initiate the cold condition in the tropics. This process reverses the phase of PDV and allows another half cycle with opposing polarity to occur. In summary, the ventilation of SST between tropical and extratropical regions occurs through variations of the subtropical cell-a time-mean shallow overturning circulation consisting of a poleward warm surface Ekman flow, an equatorward cold compensating flow within the thermocline, and upwelling along the Equator [an effect known as the "oceanic tunnel" (e.g., Liu and Alexander, 2007)]; while the poleward and eastward propagation of the forced Rossby wave train serves to connect tropical and extratropical atmospheres [an effect known as the "atmospheric bridge" (e.g., Lau and Nath, 1996)]. Despite the Gu-Philander hypothesis being supported by a number of coupled model studies (Kleeman et al., 1999; Newman et al., 2003; Yang and Liu, 2005), some other studies nevertheless question the role of the oceanic tunnel since variations of the subtropical cell in the Pacific basin are found to be too weak to account for the decadal variability in the tropics (e.g., Schneider et al., 1999; Wu and Liu, 2003; Liu, 2012). This disagreement suggests that we need a detailed analysis of how the ocean reacts to the climate variability, especially on decadal timescales.

    To understand the ocean's role in maintaining PDV, numerical results from an ocean circulation model forced by a reconstructed coupled wind stress field are analyzed in this paper. In section 2, the data sources, ocean model, and experiment design used in the study are introduced. The simulated oceanic climatology and the associated decadal variability in the Pacific Ocean are presented in section 3. Section 4 provides a detailed term balance analysis of the perturbation heat budget equation to clarify the possible mechanisms responsible for the decadal variability in the equatorial Pacific and the North Pacific regions. Finally, the major findings of the study are summarized in section 5.

2. Data model and experiment design
  • Atmospheric data from the NCEP-NCAR (National Centers for Environmental Prediction-National Center for Atmospheric Research) Reanalysis dataset and the NOAA (National Oceanic and Atmospheric Administration) ERSST (Extended Reconstructed Sea Surface Temperature) dataset were used to derive the monthly wind stress field required for modeling the decadal variability in the Pacific Ocean. The NCEP-NCAR atmospheric data use a state-of-the-art analysis/forecast system to perform data assimilation using past data from 1948 to the present (Kalnay et al., 1996). These reanalysis data provide a global coverage of major meteorological variables at 17 standard pressure levels from 1000 hPa to 10 hPa with a horizontal resolution of 2.5#cod#x000b0;#cod#215;2.5#cod#x000b0;. The NOAA ERSST data combine the International Comprehensive Ocean-Atmosphere Data Set (ICOADS) SST data with statistical methods to reconstruct historical SST data dating back to 1854 (Smith and Reynolds, 2003). The SST data have a slightly finer horizontal resolution of 2#cod#x000b0;#cod#215;2#cod#x000b0;. In the present study, only the monthly NCEP-NCAR wind field at 1000 hPa and the NOAA ERSST data over the period 1949-2001 were utilized to reconstruct the wind stress field for the model experiment.

    In addition, the three-dimensional oceanic mean state (climatology) provided by the World Ocean Atlas (WOA) was adopted as the initial conditions for the model experiment. The WOA climatology is a digital product of the Ocean Climate Laboratory sponsored by the National Oceanographic Data Center (Conkright et al., 2002). This digital atlas contains global objectively analyzed fields of several commonly measured variables, including temperature, salinity, oxygen, dissolved inorganic nutrients, and chlorophyll at 33 standard depth levels from sea surface (0 m) to sea floor (5500 m). For consistency, all the aforementioned data were re-gridded into a uniform horizontal resolution of 2.5#cod#x000b0;#cod#215; 2.5#cod#x000b0; prior to the numerical calculation.

  • The Parallel Ocean Program (POP) model, developed at the Los Alamos National Laboratory (LANL) was utilized to reproduce the decadal variability in the Pacific Ocean. The LANL-POP model, designed for high-efficiency parallel computation, is a primitive equation ocean general circulation model coded in spherical coordinates. The model has complete parameterizations of major oceanic subgridscale physical processes such as horizontal tracer diffusion, horizontal viscosity and vertical mixing (Smith and Gent, 2002). In 2001, the LANL-POP model was officially adopted as the ocean component of the NCAR's CCSM (Community Climate System Model) whose architecture has been recognized as one of the most sophisticated climate models in the world.

    The current version of the LANL-POP model used in this study has 33 vertical levels from surface (0 m) to seafloor (5500 m) and a horizontal resolution of 2.5#cod#x000b0;#cod#215; 2.5#cod#x000b0;. A higher vertical resolution is applied in the first few hundreds meters (e.g., 15 levels in the upper 600 m) for a better representation of the upper ocean's structure. Table 1 summarizes the magnitudes of the major parameters (or constants) used in the model experiment.

  • To produce a climate variability pattern akin to observed PDV, a wind stress field containing decadal variability closely linked to the slowly-varying SST fluctuations was first required. As documented in (Yu et al., 2011), the coupled variability between wind stress and SST fields at decadal time scales can be nicely isolated from the remaining climate signals using the so-called "region-wise singular value decomposition" (RSVD) method [see section 2.2 of (Yu et al., 2011) for details]. Figure 1 shows the first RSVD mode (RSVD1) derived from a pair of observed SST and wind stress fields over the period 1949-2001. To isolate the climate variability of time scales longer than a year, the seasonal cycle within the SST and wind stress fields was removed prior to the RSVD calculation. Accordingly, RSVD1 accounts for nearly 60% of the total square covariance. As shown by the time series of the SST expansion coefficient (see lower panel of Fig. 1) RSVD1 demonstrates a combined feature of two distinct time scales, including a decadal climate shift from a colder regime to a warmer one (denoted by the two separated horizontal lines) and an interannual variability associated with the El Ni#cod#241;/La Ni#cod#241; events (denoted by a number of local maxima and minima of the undulating curve).

    The spatial pattern of RSVD1 (see upper panel of Fig. 1) further shows that the tropical and extratropical SSTA tends to vary in an out-of-phase fashion, i.e., when positive SSTA exists in the tropical Pacific, negative SSTA occurs in the central North Pacific. Moreover, SSTA off the coast of the western North America is generally out-of-phase with that in the central North Pacific. Associated with such an SSTA pattern, a strong cyclonic wind circulation emerges over the central North Pacific; meanwhile, strong northwesterly wind anomalies appear over the tropical Pacific. Overall, the climate vari-ability pattern shown in Fig. 1 agrees well with observations of PDO (Nakamura et al., 1997; Zhang et al., 1997, 1998; Barnett et al., 1999).

    Figure 1.  The first singular vectors of (a) wind (vectors) and SST (colored shading), and (b) the time series of wind and SST expansion coefficients, derived from the RSVD method. The two horizontal bars represent the mean expansion coefficients averaged over the cold regime (1949-76) and warm regime (1977-2001), respectively. The percentage of square covariance fraction (SCF) is displayed in the upper right corner.

    Since RSVD1 captures the major features of observed PDV, it was used as the basis mode to reconstruct the monthly wind stress field for the model experiment. In practice, we first used RSVD1 to reconstruct a time series of wind stress anomalies. These wind stress anomalies were then superimposed onto the monthly wind stress climatology to obtain a time series of the total wind stress field for the model experiment. Note that such a design allows us to focus on the ocean's response to the decadal wind stress variability, while still providing a reasonable representation of the seasonal and interannual variability (Miller et al., 1994). To obtain a stable (trendless) climate state, a 10-yr spin-up run forced by monthly wind stress climatology was conducted prior to the formal experiment run.

3. Model results
  • In this section, results from a 53-yr control simulation of the LANL-POP model are presented, with an emphasis on reproduction of the annual mean climatology (mean field) as well as the decadal variability (anomaly field) in the Pacific Ocean because both fields were required in order to conduct the heat budget analysis presented in section 4.

  • Figure 2 shows the annual mean climatology of SST and currents in the model's surface layer (0-10 m). As shown, the tropical Pacific exhibits a clear asymmetric SST pattern in the zonal direction, characterized by a vast area of warmer SST (#cod#62;26#cod#x000b0;C) over the tropical western Pacific and an elongated, east-west oriented area of cooler SST over the equatorial eastern Pacific. We note that the zonal asymmetry in SST (i.e., SST gradient) is a key factor for the observed easterly trade winds over the tropical Pacific. Associated with such a SST distribution, two anticyclonic circulations occur in the subtropical regions of both hemispheres, corresponding respectively to the North and South Pacific Gyres. Due to the coarse horizontal resolution (2.5#cod#x000b0;#cod#215; 2.5#cod#x000b0;), the narrow Kuroshio and equatorial counter currents are missing here. In practice, they can be nicely reproduced in a high-resolution (0.25#cod#x000b0;#cod#215; 0.25#cod#x000b0;) run using the same wind stress forcing (not shown).

    Figure 3 shows the vertical cross sections of the upper ocean's temperature (upper panel) and currents (lower panel) averaged over the equatorial zone between 10#cod#x000b0;S and 10#cod#x000b0;N. Of particular interest here is the existence of a sharp temperature gradient zone (also known as the "thermocline") below the warm surface layer. If one takes the 20#cod#x000b0;C isotherm (denoting approximately the depth of maximum vertical temperature gradient) as a typical depth of the equatorial thermocline, its depth ranges from 25-50 m in the eastern Pacific to 150-200 m in the western Pacific, consistent with the Simple Oceanographic Data Assimilation (SODA) analysis (not shown). Note that the downward thermocline tilt toward the west is a typical climatic feature that may exist all year long in the tropical oceans, especially in the Pacific. Forced by the easterly trade winds, the ocean's surface layer is dominated by the westward flows; while strong compensating eastward flows (corresponding roughly to the equatorial undercurrents) appear below the surface layer, with greatest amplitudes occurring at depths of around 200 m. As a result, strong upward motions (i.e., upwelling) take place in the equatorial eastern Pacific.

    Likewise, Fig. 4 shows the vertical cross sections of the upper ocean's temperature (upper panel) and currents (lower panel) but averaged over the extratropical zone between 30#cod#x000b0;-40#cod#x000b0;N. Due to less solar radiation input into the surface layer, the thermocline structure becomes less robust in the extratropical North Pacific compared with that in the equatorial Pacific. The upper part (0-600 m) of the extratropical ocean is dominated by the eastward flows, with amplitudes decreasing slowly downward, except in the eastern boundary where a deep downwelling can be observed. We also find that the average depth of the 11#cod#x000b0; isotherm in the extratropical North Pacific is about 400 m, which is close to the 11#cod#x000b0; isotherm in the equatorial Pacific. This feature implies that the differences in vertical temperature structure between the extratropical North Pacific and the equatorial Pacific regions lie mostly within the upper 400 m.

    Figure 2.  Annual mean climatology of SST (colored shading, units: #cod#x000b0;C) and surface currents (vectors, units: cm s-1) produced by a 53-yr control simulation of the LANL-POP model.

    Figure 3.  The same as Fig. 2, but showing the vertical cross sections of ocean temperature (upper panel) and currents (lower panel) in the upper 600 m averaged over the equatorial band of 10#cod#x000b0;S-10#cod#x000b0;N. For illustration purposes, the vertical velocity field has been multiplied by a factor of 105.

    Figure 4.  The same as Fig. 2, but showing the vertical cross sections of ocean temperature (upper panel) and currents (lower panel) in the upper 600 m averaged over the mid-latitude band of 30#cod#x000b0;-40#cod#x000b0;N. For illustration purposes, the vertical velocity field has been multiplied by a factor of 106.

  • Figure 5 shows the horizontal pattern of PDV-like variability in the model's surface layer (0-10 m). The data in this figure were obtained by subtracting the mean fields of SST and surface currents averaged over the period 1976-2000 from those over the period 1949-75. As shown in Fig. 5, the LANL-POP model generates a decadal variability pattern, includng SST and surface currents, very similar to the observed PDV (see Fig. B1 in Appendix B). For instance, the strong cooling in the central North Pacific, the warming in the tropical Pacific, as well as the warming off the coast of western North America are all well reproduced. Since surface currents are driven mainly by atmospheric winds, the surface currents shown in Fig. 5 exihibit a pattern very similar to the wind stress forcing (see Fig. 1).

    To understand the vertical structure of the PDV-like variability, Fig. 6 provides a cross section view in the upper 600 m averaged over the equatorial zone between 10#cod#x000b0;S and 10#cod#x000b0;N. As shown, a modest warming (with a maximum amplitude slightly over 0.6#cod#x000b0;C) occurs in the mixed layer (0-50 m) across almost the entire Pacific basin. Beneath the warm surface layer, a layer of strong cooling (with a maximum amplitude over 2.0#cod#x000b0;C) occurs in the thermocline layer (100-300 m), followed by a modest warming below 300 m, which is very similar to that reported by Solomon and Zhang (2006). When this temperature anomaly pattern is superimposed on the annual mean temperature climatology (see Fig. 3a), the slope of the thermocline tends to be "flattened". This means that the upward tilt of the thermocline to the east would become less (more) evident in a warm (cold) PDV regime, similar to the situation happening during the mature phase of an El Ni#cod#241; (La Ni#cod#241;) event. Within the thermocline layer (between depths of 100-250 m), upward and eastward motions occur between 170#cod#x000b0;E and 170#cod#x000b0;W while downward and westward motions appear to the east of 150#cod#x000b0;E, forming an anomalous overturning circulation in opposition to the mean zonal circulation (see Fig. 4b). We note that this anomalous overturning circulation appears to be responsible for the changes of themocline slope beneath the equatorial Pacific.

    The vertical structures of the ocean's temperature and currents in the extratropical North Pacific (see Fig. 7) become relative simple and straightforward as compared to those in the equatorial Pacific (see Fig. 6). As shown in Fig. 7, the PDV-like variability in the extratropical North Pacific features the dominance of cooling and eastward flows except near the eastern boundary where mild warming and upward motions are observed. In contrast to the equatorial Pacific, no large-scale anomalous overturning circulation exists in the extratropical North Pacific.

    Figure 5.  Decadal changes of the simulated SST (colored shading, units: #cod#x000b0;C) and surface currents (vectors, units: cm s-1) in the Pacific Ocean. The data in this figure were obtained by subtracting the annual mean climatology over the period 1976-2000 from that over the period 1950-75.

    Figure 6.  The same as Fig. 5, but showing the vertical cross sections of ocean temperature (upper panel) and currents (lower panel) in the upper 600 m averaged over the equatorial band of 10#cod#x000b0;S-10#cod#x000b0;N. For illustration purposes, the vertical velocity field has been multiplied by a factor of 105.

    Figure 7.  The same as Fig. 5, but showing the vertical cross sections of ocean temperature (upper panel) and currents (lower panel) in the upper 600 m depth averaged over the mid-latitude band of 30#cod#x000b0;-40#cod#x000b0;N. For illustration purposes, the vertical velocity field has been multiplied by a factor of 106 in the lower panel.

    In general, the oceanic climatology and the associated decadal variability (including SST and the upper ocean temperature profile) simulated by the LANL-POP model agree quite well with previous coupled ocean-atmosphere model studies (e.g., Kleeman et al., 1999; Wu and Liu, 2003; Yeh et al., 2010) or even with observations (e.g., Deser et al., 1996; Nakamura et al., 1997; Zhang et al., 1997, 1998). For comparison purposes, the decadal temperature variability derived from SODA analysis is desplayed in appendix A. Besides, the very different temperature and flow patterns between the North Pacific and equatorial Pacific regions suggets that the mechanisms responsible for the decadal variability in the extratropical North Pacific should be very different from those in the equatorial Pacific.

4. Analysis of the perturbation heat budget equation
  • To explore the role of ocean dynamics in maintaining the PDV-like variability shown in Figs. 5, 6 and 7, a term balance analysis of the perturbation heat budget equation [Eq. A2)] is carried out in this section (see appendix A for a detailed formulation of the perturbation heat budget equation). For brevity, two well-known areas of significant decadal variability are chosen for comparison: the equatorial Pacific (10#cod#x000b0;S-10#cod#x000b0;N 170#cod#x000b0;E-90#cod#x000b0;W) (EP) and the North Pacific (30#cod#x000b0;-50#cod#x000b0;N 150#cod#x000b0;E-140#cod#x000b0;W) (NP). Aside from the surface layer (0-10 m), results from the eighth model layer (150-175 m) are also evaluated to demonstrate the associated changes in the thermocline layer. Since the perturbation terms in Eq. (A2) tend to fluctuate around their respective means (Yang and Zhang, 2008), it is impractical to examine the term balances using their mean values over the studied period. Instead, we choose to compare the time series patterns of all perturbation terms to stratify the mechanisms responsible for the PDV-like variability.

  • Figure 8 shows the term balances of Eq. (A2) evaluated in the first model layer (0-10 m) averaged over the EP region. The time series of local temperature trend (i.e., the [#cod#952;'t] term in Fig. 8a) shows that the LANL-POP model produces a climate variability pattern very similar to the observed PDV (see Fig. 1b for comparison). As shown in Fig. 8, the PDV-like variability over the EP surface appears to come from a balance of four dominant terms. Among these terms, the horozontal temperature advection terms associated with changes of zonal currents (i.e., the term in Fig. 8c) and changes of meridional currents (i.e., the term in Fig. 8e) both play positive roles to maintain the PDV-like variability because their time series patterns, including interannual and interdecadal variability, behave qualitatively similar to that of [#cod#952;'t]. Since surface currents are driven mainly by atmospheric winds, the above results suggest the importance of air-sea interaction in shaping the decadal variability pattern over the EP surface. On the other hand, the surface heat flux exchange (i.e., the [Q'F] term in Fig. 8h) and the vertical temperature advection associated with changes of vertical temperature gradients (i.e., the term in Fig. 8f) both tend to oppose the above variability, especially from the [Q'F] term. The strong cancellation among the dominat terms leads to a modest decadal variability (warming) as shown over the EP surface (see Fig. 8a). Contributions from the remaining terms (e.g., , , and [R']) are minimal.

    Figure 8.  Term balances of the perturbation heat budget in Eq. (A2) evaluated in the first model layer (0-10 m) averaged over the EP region. The two horizontal lines in each panel denote the mean magnitudes over the periods 1950-75 and 1976-2000, respectively. All terms are in energy flux units of W m-2.

    Figure 9.  The same as Fig. 8, but showing the term balances in the eighth model layer (150-175 m) averaged over the EP region.122mm

    Likewise, Fig. 9 shows the term balances of Eq. (A2) but evaluated in the eighth model layer (150-175 m) over the EP region where a strong compensating cooling occurs (see Fig. 6a). In contrast to Fig. 8a, the time series of [#cod#952;'t] beneath the EP surface (see Fig. 9a) shows a strong cooling pattern. It apears that such a cooling pattern is maintained jointly by the two horizontal temperature advection terms, including temprature advection associated with changes of zonal temperature gradients (i.e., the term in Fig. 9b) and changes of zonal currents (i.e., the term in Fig. 9c). The vertical temperature advection associated with changes of vertical currents (i..e, the term in Fig. 9g) nonetheless tends to partially offset the above cooling through strong descending warming (w'#cod#60;0; ) in the thermocline layer. We also find that, while the term involves significant interannual variability (see Fig. 9f ), its contribution to the decadal variability seems very modest as compared with the aforementioned terms.

    To highlight the discrepancies in term balance between surface and thermocline layers, the decadal magnitude dififferences of all the terms shown in Figs. 8 and 9 are summarized in Table 2. These magnitude differences are obtained by subtracting the mean magnitudes of all terms averaged over the period 1976-2000 from those over the period 1950-75. In the surface (first) layer, a modest warming trend of [#cod#952;'t] exists over the EP surface, with a magnitude of 0.81 W m-2. It appears that the term makes the greatest contribution in maintaining the warming trend, with a magnitude of 2.86 W m-2, followed by the term with 1.20 W m-2. On the contrary, both the [Q'F] and terms tend to offset the above warming, with magnitudes of -2.37 W m-2 and -0.70 W m-2, respectively. In the thermocline (eighth) layer, the cooling trend of [#cod#952;'t] is striking, with a magnitude of -7.28 W m-2. Notably, the two zonal temperature advection terms, including (-5.93 W m-2) and (-2.57 W m-2), almost explain the cooling trend of [#cod#952;'t]. The dominance of the term suggests the importance of thermocline adjustment in maintaining the PDO-like variability. In this case (i.e., ), the slope of the equatorial thermocline would be flattened when the climate state shifts from a colder regime toward a warmer one, similar to what happens during a warm phase of an El Ni#cod#241; event. In spite of a weak cooling trend from (-0.98 W m-2), the combined effect from the two vertical temperature advection terms, however, plays a negative role due to descending warming from (2.71 W m-2).

    It is also noted that, despite the apparent discrepancies in the heat budget between surface and thermocline layers, the vertical temperature advection term associated with changes of vertical temperature gradients (i.e., ) is always negative in the upper 600 m ( O#cod#8764; 10-8#cod#x000b0;C s-1). Since is a small positive in the thermocline layer ( O#cod#8764;10-6 m s-1), a finite negative of implies a sizable positive of #cod#952; 'z ( O#cod#8764; 10-2#cod#x000b0;C m-1). The latter clearly indicates that the upper ocean's "stratification" (an equivelent to stability) would be notably strengthened when the climate state shifts from a colder regime toward a warmer one, consistent with the findings shown in (DiNezio et al., 2009).

    Figure 10.  The same as Fig. 8, but showing the term balances in the first model layer (0-10 m) averaged over the NP region.126mm

  • A similar term balance analysis technique was applied to the extratropical North Pacific region. Figure 10 shows the term balances of Eq. (A2) evaluated in the first model layer (0-10 m) averaged over the NP region. At first glance, the combined effect of and [Q'F] (see Figs. 10e and h, respectively) seems to explain the entire variability of [#cod#952;'t] (see Fig. 10a). However, when we examine their time series patterns more carefully, it appears that the decadal variability of [#cod#952;'t ] is controlled by the term while the interannual variability is dictated by the [Q'F] term. In practice, the former denotes the enhanced cold advection associated with the existence of an anomalous cyclonic circulation (i.e., an intensified Aleutian Low) over the NP region during the warm regime; while the latter is closely linked to fluctuations of surface sensible heat flux associated with El Ni#cod#241; and La Ni#cod#241; activities. We also note that the term balance patterns evaluated at the eighth model layer (150-175 m) are very similar to those in the surface layer except for much weaker amplitudes (not shown). Due to the lack of the [Q'F] term beneath the surface layer, the term now explains both the decadal and interannual variability of [#cod#952;'t].

    Similarly, Table 3 summarizes the decadal magnitude differences of all the terms in Eq. (A2) respectively in the first and eighth model layers. In the surface layer (0-10 m) of the NP region, the [#cod#952;'t] term shows a clear cooling trend, with a magnitude of -1.26 W m-2. This cooling trend appears to be maintained jointly by (-1.06 W m-2) and [Q'F] (-0.52 W m-2). In contrast, the term provides a modest warming (0.19 W m-2) to partially offset the cooling. We also find that, although the two vertical temperature advection terms ( and ) are individually sizable, their combined effect is very modest as their changes tend to oppose each other (e.g., -0.14 W m-2 and 0.18 W m-2).

    In the eighth model layer (150-175 m), the decadal magnitude difference of [#cod#952;'t] (-0.12 W m-2) is roughly an order of magnitude smaller than that in the first model layer. Due to the lack of the [Q'F] term, the term balances become rather simple and straightforward below the surface layer. The decadal change of [#cod#952;'t] is controlled only by the term. Contributions from the others seem minimal as their changes often oppose each other. For instance, a strong cancelation occurs between the two zonal temperature advection terms (e.g., 0.10 W m-2 and -0.07 W m-2 for the and terms, respectively) as well as between the two vertical temperature advection terms (e.g., -0.02 W m-2 and 0.04 W m-2 for the and terms, respectively). Moreover, the upper ocean's stability (denoted by the term) in the NP region remains nearly unchanged after the climate regime shift, which also indicates the lack of subsurface oceanic adjustment in describing the PDO-like variability in the extratropical region.

    As summarized in Tables 2 and 3, one should note that the net effect of horizontal temperature advection always plays positive roles to maintain the PDV warming (cooling) in the surface layer of the tropical (extratropical) Pacific; while the net effect of vertical temperature advection and the surface heat flux exchange term always tend to oppose the above decadal trends. Since the zonal temperature gradient of mean state (i.e., ) is significant in the tropical Paficic, the horizontal temperature advection term associated with changes of zonal currents (i.e., ) becomes important in determining the PDO warming there in addition to the temperature advection term associated with changes of meridional currents (i.e., ). On the contrary, the PDO cooling in the extratropical Pacific is determined solely by the term as the zonal temperature gradient of mean state is very small in these regions.

5. Concluding remarks
  • The role of ocean dynamics in maintaining PDV was investigated based on simulation results from the LANL-POP ocean general circulation model. To produce a climate variability pattern akin to the observed PDV, a wind stress field containing decadal coupled variability closely linked to the slowly-varying SST fluctuations was required. The RSVD method proposed by (Yu et al., 2011) was utilized to isolate the decadal mode from a pair of observed wind and SST data over the period 1949-2001. Since the major features of the observed PDV were largely reproduced by the first RSVD mode (see Fig. 1 in section 2.3), it was adopted as the basis mode to reconstruct a monthly wind stress field for the model experiment.

    Results from a 53-yr control simulation demonstrated that the LANL-POP model not only generates a reasonable climatology in the Pacific (see Figs. 2, 3 and 4), but also produces a climate variability pattern very similar to the observed PDV. For this PDV-like variability, distinct vertical structures are found between the EP and NP regions. Over the EP region, the PDV-like variability features a modest warming in the mixed layer (0-50 m), followed by a strong cooling in the thermocline layer (100-300 m) and a modest warming below 300 m (see Fig. 6). Associated with such a temperature pattern, a basin-wide-scale anomalous overturning circulation, with direction of travel in opposition to the mean zonal overturning circulation, exists in the thermocline layer (see Fig. 3). The PDV-like variability over the NP region, however, exhibits a relatively monotonous structure, characterized by the dominance of anomalous cooling and eastward flows except near the eastern boundary (see Fig. 7). As a result of this simple response, no basin-wide-scale anomalous overturning circulation exists in the NP region.

    To explore the mechanisms responsible for the PDV-like variability, a term balance analysis of the perturbation heat budget equation was conducted over the EP and NP regions. The results showed that, over the EP region, the surface warming comes from a balance of four dominant terms including the , , [Q'F] and terms. Among these terms, the first two play positive roles to sustain the warming trend while the latter two tend to oppose the warming. The strong cancellation among the four dominant terms leads to a modest warming over the EP surface (see Fig. 5). The dominance of indicated the importance of cross-equatorial current anomalies in shaping the PDV-like variability over the EP surface. This may also explain why the maximum amplitude of tropical decadal variability tends to occur at about 5#cod#x000b0;-10#cod#x000b0;S rather than at the Equator as the ENSO variability. On the other hand, the strong cooling in the thermocline layer results from an off-balance of three competing terms, including the , and terms, with the first two playing positive roles to sustain the cooling trend and the last one playing a negative role to partially offset the cooling. The predominance of the terms in the thermocline layer (see Table 2) indicates that the thermocline slpoe would be significantly flattened (because ) when the climate state shifts from a colder regime toward a warmer one. Besides, the upper ocean's stratification (stability) tends to be strengthened (because ). All these features suggests that strong oceanic adjustment must occur in the equatorial Pacific in order to maintain the PDV-like variability.

    Over the NP region, the term balance picture becomes quite different as the surface cooling is maintained jointly by the and [Q'F] terms, with the former controlling the decadal variability and the latter describing the interannual variability. The dominance of the term in controlling the decadal cooling trend over the NP region is associated with an enhanced Aleutian Low (and accordingly an enhanced cold advection) during the warm phase of PDV. Beneath the surface layer, the cooling trend is dictated by the term, with amplitude decreasing downward rapidly (see Table 3). In contrast to the EP region, the upper ocean's stratification (stability) in the NP region nonetheless remains nearly unchanged after the climate shift. In summary, the oceanic response to wind stress forcing in the extratropical region seems to be confined to the surface layer, without much participation from subsurface ocean dynamics.

    While the curent model design is able to reproduce a PDV-like variability very similar to observations, some caveats apply to the results presented here. First, the model resolution is too coarse to resolve some fine structures of the ocean, such as the Kuroshio current and its extensions whose effects might be of importance in initiating the reverse phase of PDV in the North Pacific sector (Latif and Barnett, 1994, 1996). Second, emerging evidence has shown that the tropical-extratropical interactions rely largely on atmospheric processes (Lau and Nath, 1996; Liu and Yang, 2003; Liu and Alexander, 2007) and such processes, also known as the "atmospheric bridge", cannot fully be represented without the explicit inclusion of atmospheric dynamics. This suggests that a high-resolution coupled model is required to address the full spectrum of air-sea interactions associated with PDV.

  • To explore the mechanisms responsible for the PDV-like variability as shown in Figs. 5, 6 and 7, a term balance analysis of the perturbation heat budget equation was conducted. Following (Yang and Zhang, 2008), the equation governing oceanic temperature perturbation can be written as:

    where an equilibrium mean state (i.e., ) has been subtracted in deriving Eq. (A1). In summary, the time rate change of local temperature perturbation (T't) is determined by three major effects, including the temperature advection terms associated with changes of zonal, meridional and vertical flows (i.e., and , respectively), the temperature advection terms associated with changes of zonal, meridional and vertical temperature gradients (i.e., and , respectively), as well as the net surface heat exchange term (Q'F). Note that the net surface heating exchange term includes sensible and latent heat fluxes across the ocean surface. As the nonlinear temperature advection terms (-u'#cod#952;'x-v'#cod#952;'y-w'#cod#952;'z) and the molecular diffusion terms (Ah#cod#952;'xx+Ah#cod#952;'yy+Av#cod#952;'zz) are much smaller than the others, they are treated as a residual term R' in Eq. (A1).

    Since the LANL-POP model employs a staggered-grid system in the vertical direction, with the velocity field stored at the integer levels and temperature field at the half-integer levels, it is more convenient to evaluate Eq. (A1) in a vertically-integrated format, which yields the following perturbation heat budget equation:

    The bracket symbol in Eq. (A2) represents a vertically-integrated quantity defined as

    where the subscript "k" represents the kth model layer, #cod#961; is the density of ocean water and cp is the ocean heat capacity. Note that all the terms in Eq. (A2) are in units of energy flux (W m-2). Since the four horizontal temperature advection terms in Eq. (A2), including , , and , change the size and sign of [#cod#952;'t] from remote sources, they are often referred to as the "non-local effects". In contrast, the two vertical temperature advection terms, including and , and the surface heat exchange term, [Q'F] alter the size and sign of [#cod#952; 't] from the nearby sources, they are regarded as the "local effects".

  • The SODA analysis is an oceanic gridded dataset consisting of several important field variables, including temperature, salinity, currents, and sea surface height. SODA was created by assimilating observed temperature and salinity data into an eddy-permitting forecast model based on the Parallel Ocean Program (POP1.3) forced by ERA40 winds (Carton et al., 2000a, 2000b). The SODA reanalysis data reconstruct historical ocean climate variability on space and time scales similar to those created by the atmospheric reanalysis projects at the NCEP. The current version used in the present study is SODA1.2, which covers the period 1958-2001 with a horizontal resolution of 0.5#cod#x000b0;#cod#215; 0.5#cod#x000b0;.

    Figure B1 shows the PDV derived from SODA by taking the difference between the two phases: 1976-2000 and 1958-75. Note that the latter phase (1958-75) is slightly shorter than the one taken in the model simulation (1950-75) due to the data length limit in SODA. Overall, the PDV derived from SODA, including SSTA and vertical temperature profiles in the EP and NP regions, behaves similarly to the modeled PDV-like variability shown in Figs. 5, 6, and 7. This confirms that our model design (i.e., an ocean model run forced by a coupled wind stress field) is able to reproduce a PDV similar to the real world.

    Figure B1.  The PDV derived from SODA reanalysis data (1976-2000 minus 1975-58) of (a) horizontal temperature variability, (b) vertical temperature variability in the EP region, and (c) vertical temperature variability in the NP (units: #cod#x000b0;C).

Reference

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint