• ADVANCES IN ATMOSPHERIC SCIENCES, 2017, 34(12): 1426-1436
    doi: 10.1007/s00376-017-6176-5
    The Tropical Pacific-Indian Ocean Associated Mode Simulated by LICOM2.0
    Xin LI1,2,, Chongyin LI1,2

    Abstract:

    Oceanic general circulation models have become an important tool for the study of marine status and change. This paper reports a numerical simulation carried out using LICOM2.0 and the forcing field from CORE. When compared with SODA reanalysis data and ERSST.v3b data, the patterns and variability of the tropical Pacific-Indian Ocean associated mode (PIOAM) are reproduced very well in this experiment. This indicates that, when the tropical central-western Indian Ocean and central-eastern Pacific are abnormally warmer/colder, the tropical eastern Indian Ocean and western Pacific are correspondingly colder/warmer. This further confirms that the tropical PIOAM is an important mode that is not only significant in the SST anomaly field, but also more obviously in the subsurface ocean temperature anomaly field. The surface associated mode index (SAMI) and the thermocline (i.e., subsurface) associated mode index (TAMI) calculated using the model output data are both consistent with the values of these indices derived from observation and reanalysis data. However, the model SAMI and TAMI are more closely and synchronously related to each other.

    Key words: ocean general circulation model; numerical simulation; tropical Pacific-Indian Ocean associated mode; subsurface ocean temperature anomaly;
    摘要: 大洋环流模式已成为研究海洋状态及其变化的重要工具. 本文利用中科院大气所的LICOM2.0模式和来自CORE的强迫场进行了一个数值模拟, 并重点分析了热带太平洋-印度洋联合模(以下简称联合模)特征在模式中的表现. 与SODA再分析资料和ERSST. V3b等观测资料进行对比发现, 数值试验很好地再现了联合模的形态和变率. 即, 当热带中西印度洋和中东太平洋异常偏暖/冷时, 热带东印度和西太平洋相应地偏冷/暖. 这进一步证实了联合模是热带太平洋—印度洋非常重要的一个模态, 不仅在表层海温异常场上显著, 而且在次表层海温异常场上表现更为明显. 研究还发现, 模式海温资料计算得到的表层联合模指数和温跃层(次表层)联合模指数与从观测资料中计算得到的都较为一致. 然而, 模式中这两个指数相关性更好而且更加同步.
    关键词: 大洋环流模式 ; 数值模拟 ; 热带太平洋-印度洋联合模 ; 次表层海温异常

    1. Introduction

    El Niño-Southern Oscillation (ENSO) is an important interannual climate system. Since (Bjerknes, 1969) noted that ENSO is a result of air-sea interaction, it has been the subject of much detailed research that has revealed its physical mechanisms and global-scale climate impacts. However, (Saji et al., 1999) found that a similar interannual zonal oscillation of SST anomalies (SSTAs) also exists in the equatorial Indian Ocean and named it the Indian Ocean dipole (IOD). Since then, many papers have discussed its mechanisms and impacts. Originally, the IOD was thought to be the result of ocean-atmosphere interaction in the Indian Ocean itself (Saji et al., 1999; Webster et al., 1999). Gradually, however, research has focused on the relationship between the IOD and ENSO. For example, the asymmetric SSTA between the eastern and western Indian Ocean in 1997/98 was probably triggered by the strong El Niño of that year (Yu and Rienecker, 1999; Ueda and Matsumoto, 2000), as ENSO can affect the sea surface wind field through the anti-Walker circulation over the equator (Yu and Rienecker, 1999). Based on statistical analysis, (Li and Mu, 2001) reported a close relationship between the SSTA dipole mode in the equatorial Indian Ocean and ENSO (which can also be treated as a dipole) in the Pacific. By analyzing the SSTA features of the Indian Ocean in the warm and cold phases of the ENSO cycle, (Yan et al., 2001) showed that a significant dipole oscillation phenomenon develops in the Indian Ocean during ENSO events. In addition, SST variations in the Indian Ocean commonly play an important role in the development of El Niño events (Annamalai et al., 2005; Izumo et al., 2010; Yuan et al., 2011, 2013). Thus, the IOD can also influence ENSO (Annamalai et al., 2005; Yuan, 2005; Yuan et al., 2011). Taken together, the above studies indicate that there are strong interactions between the Indian and Pacific oceans, and hence we should consider the SSTA field of the tropical Pacific and Indian oceans as a whole.

    Based on this idea, the tropical Pacific-Indian Ocean SSTA was analyzed by (Ju et al., 2004) using EOF decomposition. The first mode shows that the SSTA in the equatorial central-western Indian Ocean and central-eastern Pacific is opposite to that in the equatorial western Pacific and eastern Indian Ocean. As this mode contributes over 50% of the total variance, it is referred to as the tropical Pacific-Indian Ocean temperature anomaly mode or the Pacific-Indian Ocean associated mode (PIOAM), and this has been further researched in other studies (e.g., Wu et al., 2005; Yang and Li, 2005). This mode is meaningful because it better reflects the SSTA differences between the east and west in the tropical basin, and has an important influence on the South Asian high and even on the Asian climate (Yang and Li, 2005; Yang et al., 2006). It is also closely associated with the evolution and propagation of the subsurface ocean temperature anomaly (SOTA) (Wu and Li, 2009).

    The above studies were based mainly on SST data. However, the ocean temperature anomaly in the subsurface zone, especially in the thermocline, is much stronger than that at the surface (Qian et al., 2004; Chao et al., 2005). The temperature anomalies usually appear first in the thermocline and then propagate along this layer. For example, the SOTA in the equatorial western Pacific and its eastward propagation are an important driver of El Niño (Li and Mu, 1999). In fact, the variation in the SOTA in the western Pacific warm pool is closely associated with the entire ENSO cycle, and they interact with each other (Li and Mu, 2000). Moreover, the IOD is also more prominent in the subsurface zone than at the surface, and it appears as a dipole in a real physical sense (Chao et al., 2005). Thus, (Li et al., 2013) further studied the tropical Pacific-Indian Ocean temperature anomaly mode in the thermocline, analyzing the monthly thermocline temperature anomaly (TOTA) over the period 1958-2007 and the weekly sea surface height (SSH) anomaly between 1992 and 2011 in the tropical Pacific-Indian Ocean using the EOF method. Both of the first two modes show coupled variations between the tropical Indian Ocean and the Pacific. That is, when the subsurface temperature in the tropical central-western Indian Ocean and central-eastern Pacific is abnormally warmer/colder, the subsurface temperature in the tropical eastern Indian Ocean and western Pacific is abnormally colder/warmer. This is seen as a major tripole pattern and is referred to as the tropical Pacific-Indian Ocean thermocline temperature anomaly associated mode. This mode shows a good correlation with both ENSO and the IOD, and its evolution is closely related to the propagation of the TOTA. This mode also has a high positive correlation with that defined by the SSTA, but can better represent the spatial distribution and temporal variation of the associated features between temperature anomalies in the tropical Indian Ocean and Pacific. Considering the spatial patterns of SST and SSH in the first modes of multi-variable EOF of SST, SSH and surface wind stress both resemble a tripole, and (Lian et al., 2014) named the dominant mode in the tropical Pacific-Indian Ocean the Indo-Pacific tripole. For simplicity, these modes are referred to collectively as the PIOAM in this paper.

    As observational data are sparse in the ocean, especially in the subsurface zone, extended simulations produced by ocean general circulation models (OGCMs) are needed for further exploration of the characteristics and dynamics of the PIOAM in the tropical Pacific-Indian Ocean. Using an OGCM developed by (Jin et al., 1999) and (Li, 2005), (Wu and Li, 2009) investigated the three-dimensional thermal and dynamic structures of the associated mode and its mechanism of evolution. Their subsequent numerical experiments indicated that the Indonesian Throughflow plays an important role in the formation of the PIOAM, especially in the subsurface (Wu et al., 2010). However, the spatial resolution of their OGCM was relatively low and so could not identify the narrow channels in the Indonesian Sea. Thus, a relatively high-resolution OGCM is required to further improve the simulation of the PIOAM (Wu et al., 2010). In the present study, a state-of-the-art eddy-resolving OGCM, LICOM2.0, is used to examine the properties of the surface and subsurface PIOAM, with an emphasis on the evolution of the subsurface PIOAM and its relationship to the surface PIOAM.

    2. Model and data

    LICOM2.0 is the newest version of the fourth-generation OGCM developed by the State Key Laboratory of Numerical Modeling for Atmospheric Science and Geophysical Fluid Dynamics at the Institute of Atmospheric Physics, Chinese Academy of Sciences. It is also the oceanic component of FGOALS2.0, which participated in CMIP5. The model was forced by the normal forcing data derived from CORE (Griffies et al., 2009), as detailed in Table 1. More information on LICOM2.0 can be found in (Liu et al., 2012).

    With respect to the standard edition of LICOM2.0 (Liu et al., 2012), for this study we made some improvements to the model according to (Yu et al., 2012), as follows: (1) The horizontal resolution was increased to an eddy-resolving 1/10° and the number of vertical layers to 55. In the upper 300 m, there were 36 uneven layers with a mean thickness of less than 10 m, and the depth of the first layer was 5 m. (2) To exclude the Arctic Ocean, the model domain was set to 66°N-79°S. (3) Biharmonic viscosity and diffusivity schemes were adopted in the horizontal direction in the momentum and thermohaline equations, respectively, while the parameterization of mesoscale eddies was turned off in the thermohaline equations. (4) The methods of barotropic and baroclinic decomposition were improved.

    In our simulation, LICOM2.0 was spun-up for 500 years from zero velocity and initialized from the observed temperature and salinity obtained from WOA05, repeating the daily-corrected Normal Year Forcing data from (Large and Yeager, 2004) as the forcing condition. Then, according to the parameterization schemes of Liu et al. (2014a, 2014b), the model was forced using the forcing data in Table 1 and integrated for 60 years. The last 50 years of output, from January 1958 to December 2007, were used for the analysis and discussion below.

    Table 1. The forcing data of LICOM2.0.
    Resolu- Time
    Variable Dataset Period tion interval
    Air temperature at 2 m CORE v2 1948-2007 T62 Daily
    Relative humidity at2 m CORE v2 1948-2007 T62 Daily
    Sea level pressure CORE v2 1948-2007 T62 Daily
    Wind at 10 m CORE v2 1948-2007 T62 Daily
    Downward shortwaveradiation CORE v2 1948-2007 T62 Daily
    Downward longwaveradiation CORE v2 1948-2007 T62 Daily
    Precipitation CORE v2 1948-2007 T62 Daily
    Runoff CORE v2 1948-2007 1° Annual
    Ice density NSIDC 1979-2006 1° Monthly
    SST and SSS WOA05 Before 2005 1° Monthly

    Table 1. The forcing data of LICOM2.0.

    To validate the model results, we used two datasets——one observational (ERSST.v3b) and one reanalysis (SODA v2.2.4). The ERSST.v3b data were obtained from the National Oceanic and Atmospheric Administration with a horizontal resolution of 2°× 2° and were constructed using ICOADS SST data and improved statistical methods (Smith et al., 2008). The SODA v2.2.4 ocean reanalysis data (Carton and Giese, 2008) were provided by the University of Maryland with a horizontal resolution of 0.5°× 0.5° and 40 levels in the vertical direction.

    3. Model verification

    Firstly, the error associated with simulated variables, such as SST, thermocline depth and heat content, was analyzed to examine the results of the numerical simulation. Here, the "error" is the difference between the model data and the ERSST or SODA data, as these two datasets are relatively realistic in their representation of ocean conditions (Smith et al., 2008; Carton and Giese, 2008).

    3.1. SST

    Figures 1a and b show the January climatological SST simulated by LICOM2.0 and that obtained from ERSST, respectively. The results show that they are essentially consistent in their spatial pattern, although the warm pool SST in the simulation is significantly higher than that in the observations, as is the SST at the East African coast. On the other hand, although the spatial pattern of the SST standard deviation in the simulation (Fig. 1c) is similar to that in the ERSST data (Fig. 1d), the values of the former are generally larger than those of the latter. One possible cause of this is that the relatively high resolution of the model may contain some dynamic and thermodynamic processes that are undetectable in the observational data. Another possibility is that the model systematically overestimates the SST variability in the tropical Pacific-Indian Ocean.

    Fig.1. The (a, b) January climatological SST and (c, d) SST standard deviation (units: °C) in (a, c) LICOM2.0 and (b, d) ERSST in the tropical Pacific-Indian Ocean.

    Fig.2. Distribution of the SST trend [units: °C (50 yr)-1] in (a) LICOM and (b) ERSST in the tropical Pacific-Indian Ocean, and the time series of (c) the ño3 index and (d) the IOBMI calculated using the original LICOM2.0 (solid line) and ERSST (dotted line) data.

    We also examined the SST trend in the LICOM2.0 simulation, as shown in Fig. 2a. Compared with the result based on the ERSST data (Fig. 2b), the simulated SST trend is obviously higher in the central and eastern Pacific but markedly lower in the warm pool region and Indian Ocean. Thus, the ño3 region (5°S-5°N, 150°-90°W) and tropical Indian Ocean basin region (20°S-20°N, 40°-110°E) were selected to further compare the SST trend in the simulation and observational data. Figure 2c shows the time series of the original ño3 index in the LICOM2.0 simulation and the ERSST data, respectively. In LICOM2.0, the index shows a warming trend [up to 0.92°C (50 yr)-1], whereas in ERSST this warming trend is relatively weaker [only 0.51°C (50 yr)-1]. However, the warming trend of the Indian Ocean basin mode index (IOBMI) in the simulation is not as obvious as it is in the observational data [0.27°C (50 yr)-1 versus 0.55 °C (50 yr)-1. These results again indicate that LICOM2.0 overestimates the warming trend in the central and eastern Pacific but underestimates the warming trend in the tropical Indian Ocean. These deficiencies may stem from systematic biases in the model. Thus, we removed all the trends in the model and observational data hereinafter.

    Fig.3. Time series of (a) the ño3.4 index and (b) the DMI calculated using the detrended LICOM2.0 (solid line) and ERSST (dotted line) data.

    Figure 3 shows the ño3.4 index and Dipole Mode Index (DMI) calculated separately using the simulated SST and that in the ERSST observations. The correlation coefficient between the LICOM2.0 ño3.4 index and the ERSST ño3.4 index reaches 0.94 (exceeding the 99% significance level; Fig. 3a), indicating that LICOM2.0 can simulate the ENSO variability very well. Meanwhile, the correlation coefficient between the two DMI time series is about 0.71 (exceeding the 99% significance level; Fig. 3b), showing that the model also performs well in simulating the main SST variability in the Indian Ocean, but not as well as in Pacific. This may be because the processes of air-sea interaction are more complex in the Indian Ocean than in the Pacific, which is hard for a single OGCM to simulate. Regardless, considering the performance of LICOM2.0 in simulating ENSO and the IOD, we had no reason to doubt that the surface PIOAM could be satisfactorily reproduced in this model.

    3.2. Thermocline depth

    The thermocline depth was calculated using the vertical gradient method by separately applying the LICOM2.0 and SODA data. The results showed that the climatology of the thermocline depth in LICOM2.0 (Fig. 4a) is consistent with that of SODA (Fig. 4b) in most regions, but is much deeper than the latter in the equatorial northwestern Pacific and in the ITCZ domain (5°-15°N, 150°-110°W), leading to the thermocline ridge along 10°N in the northeastern Pacific being poorly defined in the model (Fig. 4a). By comparing the amplitude (standard deviation) of thermocline depth in these two datasets, we found that the spatial pattern in the model (Fig. 4c) is generally similar to that in the reanalysis data (Fig. 4d). The major differences are that the oscillation of the thermocline in the equatorial eastern Pacific and eastern Indian Ocean are larger in LICOM2.0 than they are in SODA, whereas the reverse is true for the tropical northeastern Pacific, equatorial western Pacific, and western Indian Ocean (Figs. 4c and d). This may indicate that the oceanic long waves in LICOM2.0 are stronger than those observed in the eastern part of the ocean basin, but weaker in the western part.

    The simulated upper ocean heat content is also consistent with the observations (figures not shown). These results indicate that LICOM2.0 is also capable of accurately simulating the tropical Pacific-Indian Ocean variability in the subsurface zone. Thus, we were confident that the model could be used to examine the PIOAM both at the sea surface and in the thermocline, and in considering their relationship with each other. These points will be discussed in the next section.

    Fig.4. (a, b) January climatological thermocline depth (units: m) in (a) LICOM2.0 and (b) SODA, and (c, d) thermocline depth standard deviation (units: m) in (c) LICOM2.0 and (d) ERSST, in the tropical Pacific-Indian Ocean.

    4. Characteristics of the PIOAM in LICOM2.0

    We began by analyzing the SSTA in LICOM2.0 using the EOF method and comparing it with the same in the ERSST data. The spatial pattern of EOF1 presents as a tripole mode; that is, the SSTAs are positive across most of the tropical Indian Ocean and central eastern Pacific, whereas they are negative across the equatorial southeastern Indian Ocean and in the tropical western Pacific (Fig. 5a), which is consistent with the pattern seen in the ERSST data (Fig. 5b). The first mode in the model and in the ERSST data explains 51.4% and 49.8% of the variance of the SSTA, respectively, indicating that this SSTA mode is typical. The correlation coefficients between the corresponding PC1 in LICOM2.0 and in the ERSST data reaches 0.94 (exceeding the 99% significance level). Thus, the major mode of SSTA variation in the tropical Pacific-Indian Ocean, the surface PIOAM, is well represented in the LICOM2.0 simulation.

    Fig.5. EOF-1 of tropical Pacific-Indian Ocean SSTA derived from (a) LICOM2.0 and (b) ERSST; (c) temporal variation of the corresponding PC1 in LICOM2.0 (solid line) and ERSST (dotted line).

    Fig.6. Time series of the SAMI (a) and the composed SAMI (b) in LICOM and ERSST, respectively.

    Following the definition proposed by (Yang and Li, 2005), the SSTA associated mode index (SAMI) was calculated from the LICOM2.0 and from the ERSST data. Their time series can be seen in Fig. 6a. Generally, the SAMI in the model is consistent with that in ERSST (with correlation coefficients above 0.8 and exceeding the 99% significance level), although it is obviously higher or lower than observed in several years. Positive SAMI years were selected when the SAMI was larger than +1.0 standard deviation in three consecutive months, and vice versa. The composite analysis from these data (Fig. 6b) shows that positive SAMI in the model tends to develop rapidly in late spring and summer, peaks in autumn, and decays in the following spring. In the year prior to positive SAMI years, weak negative SAMI often occurs. However, the negative PIOAM in the model seems to grow slowly from the previous autumn, peaks in the early part of the concurrent autumn, and can then last through to the following spring. Notably, the evolution of positive SAMI in the model is very close to that in the observational data (Fig 6b), whereas negative SAMI in the model develops 6-9 months earlier and decays 2-3 months later than observed. Moreover, the amplitude of negative SAMI in the model is larger than that in the observational data. In short, the simulated negative PIOAM is more durable and stronger than observed. This may indicate that there are some problems related to the ability of LICOM2.0 to simulate the negative PIOAM and the amplitude asymmetry of PIOAM.

    The next step was to investigate the performance of LICOM2.0 in reproducing the subsurface PIOAM. In section 3, we showed that there are systematic "errors" in the climatology and standard deviation of the thermocline depth in LICOM2.0, indicating the temperature variation in the simulated thermocline does not accurately represent the upper ocean thermal anomaly seen in the observations. However, the climatology and standard deviation of the upper ocean heat content anomaly (HCA) are close to those in the reanalysis data; consequently, we used the HCA of the upper 300 m (HCA300) to examine the subsurface ocean variability in LICOM2.0. The EOF1 of HCA300 in LICOM2.0 explains 32.6% of the variance. In this mode, the HCA300 in both the Pacific and Indian oceans shows a significant dipole pattern, while the HCA300 in the eastern Indian Ocean has the same sign as that of the western Pacific (Fig. 7a). Compared to the surface PIOAM (Fig. 5a), the pattern exhibits clearer physical significance and more associated features between the Pacific and Indian oceans. Thus, it forms an HCA300 associated mode in the tropical Pacific-Indian Ocean that is similar to the TOTA mode referred to by (Li et al., 2013). The PC1 of the simulated HCA300 (Fig. 7b) shows stronger amplitude asymmetry than that of the simulated SSTA (Fig. 5c). These results indicate that the subsurface PIOAM is also significant in LICOM2.0. However, there are still some differences between the HCA300 associated mode in LICOM2.0 and that in the SODA data; e.g., the variability in the model is often stronger than that seen in the reanalysis data, especially in the ño3 region and eastern Indian Ocean. This may indicate a need to further improve the representation of the dynamic and thermodynamic processes in these regions.

    Fig.7. The (a) EOF-1 of HCA300 in the tropical Pacific-Indian Ocean derived from LICOM2.0 and (b) the temporal variation of the corresponding PC1.

    Fig.8. (a) Time series of the TAMI in LICOM2.0. (b) Wavelet power spectrum of TAMI, the yellow shading denotes where the value exceed 95% confidence level.(c) Global wavelet spectrum of TAMI, the dotted line represent the 95% confidence level.

    The subsurface associated mode index, referred to as TAMI by (Li et al., 2013), was calculated using the HCA300 from LICOM2.0. From the simulated TAMI series (Fig. 8a), it is apparent that the amplitude of positive SAMI is clearly larger than that of negative SAMI, indicating the subsurface PIOAM also has a significant feature of amplitude asymmetry. Wavelet analysis shows that the TAMI mainly has 2-4-yr interannual variability and a 10-14-yr interdecadal periodicity (Figs. 8b and c). Composite analysis shows that the subsurface PIOAM usually develops in summer, peaks in late autumn, and ends in early spring, and the amplitude of its warm phase is larger than that of the cold phase (figure not shown). These characteristics are similar to the findings of (Li et al., 2013), again indicating that the PIOAM is a robust mode in the subsurface of the Pacific-Indian Ocean and that LICOM2.0 can accurately describe the co-variations of temperature in the tropical Pacific and Indian oceans.

    Figure 9 shows that the maximum correlation coefficient between TAMI and SAMI in LICOM2.0 is 0.75 (exceeding the 99% significance level), at a lag of 0 months. This indicates that there is close relationship between the temperature variations at the surface and in the subsurface zone of the Pacific-Indian Ocean. The zero lag in the maximum correlation between TAMI and SAMI may imply that the exchange of momentum and heat between the surface and subsurface of the Pacific-Indian Ocean occurs rapidly, and the progress of both is likely to be forced by the wind anomaly.

    Fig.9. Time series of (a) TAMI, and Hovmöller diagrams of HCA300 (color scale bar; units: °C) propagating along (b) the equatorial zone (2.5°S-2.5°N) in the Pacific-Indian Ocean, (c) the zonal belt off the equator in the Northern Hemisphere (10°-15°N), and (d) the zonal belt off the equator in the Southern Hemisphere (10°-15°S). Positive/negative values denote an anomalously warm/cold upper ocean.

    Using ocean reanalysis data, (Li et al., 2013) revealed that the propagation of the SOTA results in the development and transition of the subsurface PIOAM. It can also affect the surface, and leads to the evolution of the surface PIOAM.

    Fig.10. Time-lag correlation between TAMI and SAMI in LICOM2.0.

    We further explored whether LICOM2.0 can reproduce these processes and relationships. The evolution of TAMI (Fig. 10a), combined with the longitude-time profile of HCA300 along key latitude belts (Fig. 10b-d), shows that the upper ocean heat content in the equatorial western Pacific and eastern Indian Ocean is significantly warmer than normal before the PIOAM develops into its warm phase. Then, the warming signals propagate along the equator (Fig. 10b) and westwards along the zonal belt of 10°-15°S in the southern Indian Ocean (Fig. 10d). Thus, the equatorial eastern Pacific and western Indian Ocean become regions of warming, thereby contributing to the development of a positive PIOAM. After the positive PIOAM reaches its peak, the warming signals in the eastern Pacific begin to propagate westwards along 10°-15°N (Fig. 10c). Meanwhile, the equatorial western Pacific and eastern Indian Ocean have been gradually occupied by anomalously cold water, due to upwelling Kelvin and Rossby waves, respectively. The anomalous cold signals in the equatorial western Pacific then propagate eastwards along the equator, while some of the warming signals propagate eastwards along the equator in the Indian Ocean (Fig. 10b), leading to decay of the PIOAM. When the warming signals return to the western Pacific and eastern Indian oceans, the PIOAM transitions to a negative phase. In the development of a negative PIOAM, these processes are opposite to those during a positive PIOAM. Clearly, the evolution of the PIOAM is closely related to the propagation of the ocean HCA (i.e., the SOTA) in LICOM2.0, with a specific passage of the SOTA signals in the Pacific and in the Indian Ocean. These results support the findings of (Li et al., 2013) and further confirm the important role of the SOTA in the evolution of the PIOAM. Meanwhile, there is a suggestion that LICOM2.0 can describe the subsurface dynamic and thermodynamic processes well, such as oceanic long waves.

    5. Discussion and conclusions

    Following improvements including the vertical resolution, horizontal range, parameterization schemes, and barotropic and baroclinic split, LICOM2.0 was integrated with forcing data from CORE. The outputted monthly ocean temperature data for the period 1958-2007 were analyzed and compared with those of the SODA and ERSST datasets. The results showed that the climatology and variability of temperature in the tropical Pacific-Indian Ocean are reproduced well by LICOM2.0, as are the basic features and evolution of the PIOAM.

    EOF analysis of the surface and subsurface temperature anomalies showed that the co-variations of the Indian and Pacific oceans (i.e., the associated modes) are captured by LICOM2.0. The surface and subsurface PIOTM indexes (SAMI and TAMI, respectively) calculated from the model data are consistent with those based on the SODA data, and the high correlation between these two associated modes is also reproduced by LICOM2.0. This indicated that the model can reliably simulate the upper ocean dynamic and thermodynamic processes in the tropical Pacific-Indian Ocean.

    The zonal propagations of the HCA300 in LICOM2.0 showed that the evolution of the PIOAM is closely related to the propagations of HCA300 along the equator and in the off-equatorial zonal belt. In the Pacific, the anomaly signals propagate mainly eastwards along the equator and westwards along 10°-15°N, but the westward propagation is not present, or is very weak, in the South Pacific. However, in the Indian Ocean the signals propagate mainly westwards along 10°-15°S and then partially return to the east along the equator, while the westward propagation in the North Indian Ocean is rather weak. These features are similar to the propagation signals of the SOTA based on SODA data, as reported by (Li et al., 2013).

    In summary, the results obtained from the LICOM2.0 model further confirm that the PIOAM is the principle pattern of the tropical Pacific-Indian Ocean temperature anomaly in both the surface and subsurface zones. As seen in the observational data, the evolution of the PIOAM is related to the development and propagation of the SOTA, and there are also close relationships between the variations of the surface and subsurface PIOAM.

    Based on the SODA ocean reanalysis data and NCEP atmospheric circulation reanalysis data, the physical mechanisms of the PIOAM were initially inferred by (Li, 2015). The general framework is that the anomalous atmospheric circulation (often the Walker circulation) incites the surface wind stress and sea level anomalies, exciting anomalous currents, oceanic long waves, and upwelling. This can further lead to surface and subsurface ocean temperature anomalies through a series of thermodynamic-dynamic processes. These anomalous heat conditions of the ocean then change the atmospheric circulation and SST through evaporation, precipitation, sensible heat fluxes, and latent heat fluxes, which can eventually lead to the system returning to normal or even reversing. In these processes, the coupled Walker circulation above the Pacific and Indian oceans plays an important bridging role, while evaporation, precipitation, shortwave radiation and the reflection of oceanic long waves at the ocean boundary play a role in regulation and feedback.

    However, this framework is based only on data analysis and deductions, combined with an OGCM simulation, which limits its capacity to reveal the mechanisms involved in the evolution of the PIOAM in more depth. Thus, an ocean-atmosphere coupled model is needed to explore the key factors and physical processes associated with the development and evolution of the PIOAM, supported by a series of sensitivity experiments; e.g., by turning the Indonesian Throughflow on or off or by modulating the parameters of air-sea coupling.

    Acknowledgements. Two anonymous reviewers provided careful comments on the submitted manuscript, which helped improve the article, and for which we are grateful. The authors also sincerely thank Professor Hailong LIU for his help with LICOM2.0. This work was supported by the National Basic Research Program of China (Grant No. 2013CB956203), the National Natural Science Foundation of China (Grant Nos. 41490642 and 41575062), and the Open Fund of LASG.

    References

    [1] Annamalai H., S. P. Xie, J. P. McCreary, and R. Murtugudde, 2005: Impact of Indian ocean sea surface temperature on developing El Niño.J. Climate,18,302-319,doi: 10.1175/JCLI-3268.1.
    Prior to the 1976-77 climate shift (1950-76), sea surface temperature (SST) anomalies in the tropical Indian Ocean consisted of a basinwide warming during boreal fall of the developing phase of most El Ni os, whereas after the shift (1977-99) they had an east-west asymmetry -- a consequence of El Ni o being associated with the Indian Ocean Dipole/Zonal mode. In this study, the possible impact of these contrasting SST patterns on the ongoing El Ni o is investigated, using atmospheric reanalysis products and solutions to both an atmospheric general circulation model (AGCM) and a simple atmospheric model (LBM), with the latter used to identify basic processes. Specifically, analyses of reanalysis products during the El Ni o onset indicate that after the climate shift a low-level anticyclone over the South China Sea was shifted into the Bay of Bengal and that equatorial westerly anomalies in the Pacific Ocean were considerably stronger. The present study focuses on determining influence of Indian Ocean SST on these changes.A suite of AGCM experiments, each consisting of a 10-member ensemble, is carried out to assess the relative importance of remote (Pacific) versus local (Indian Ocean) SST anomalies in determining precipitation anomalies over the equatorial Indian Ocean. Solutions indicate that both local and remote SST anomalies are necessary for realistic simulations, with convection in the tropical west Pacific and the subsequent development of the South China Sea anticyclone being particularly sensitive to Indian Ocean SST anomalies. Prior to the climate shift, the basinwide Indian Ocean SST anomalies generate an atmospheric Kelvin wave associated with easterly flow over the equatorial west-central Pacific, thereby weakening the westerly anomalies associated with the developing El Ni o. In contrast, after the shift, the east-west contrast in Indian Ocean SST anomalies does not generate a significant Kelvin wave response, and there is little effect on the El Ni o-induced westerlies. The Linear Baroclinic Model (LBM) solutions confirm the AGCM's results.
    DOI:10.1175/JCLI-3268.1      URL     [Cited within:2]
    [2] Bjerknes J., 1969: Atmospheric teleconnections from the equatorial Pacific.Mon. Wea. Rev.,97,163-172,doi: 10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2.
    DOI:10.1175/1520-0493(1969)097&lt;0163:ATFTEP&gt;2.3.CO;2      URL     [Cited within:1]
    [3] Carton J. A., B. S. Giese, 2008: A reanalysis of ocean climate using simple ocean data assimilation (SODA).Mon. Wea. Rev.,136,2999-3017,doi: 10.1175/2007MWR1978.1.
    DOI:10.1175/2007MWR1978.1      URL     [Cited within:2]
    [4] Chao J. P., Q. C. Chao, and L. Liu, 2005: The ENSO events in the tropical Pacific and Dipole events in the Indian Ocean.Acta Meteologica Sinica,63,594-602,doi: 10.3321/j.issn:0577-6619.2005.05.005. (in Chinese with English abstract)
    正A depth map (close to that of the thermocline as defined by 20℃) of climatically maximum seatemperature anomaly was created at the subsurface of the tropical Pacific and Indian Ocean, based on which the evolving sea-temperature anomaly at this depth map from 1960 to 2000 was statistically analyzed. It is noted that the evolving sea temperature anomaly at this depth map can be better analyzed than the evolving sea surface one. For example, during the ENSO event in the tropical Pacific, the seatemperature anomaly signals travel counter-clockwise within the range of 10°S-10°N, and while moving, the signals change in intensity or even type. If Dipole is used in the tropical Indian Ocean for analyzing the depth map of maximum sea-temperature anomaly, the sea-temperature anomalies of the eastern and western Indian Oceans would be negatively correlated in statistical sense (Dipole in real physical sense), which is unlike the sea surface temperature anomaly based analysis which demonstrates that the inter-annual positive and negative changes only occur on the gradients of the western and eastern temperature anomalies. Further analysis shows that the development of ENSO and Dipole has a time lag features statistically, with the sea-temperature anomaly in the eastern equatorial Pacific changing earlier (by three months or so). And the linkage between these two changes is a pair of coupled evolving Walker circulations that move reversely in the equatorial Pacific and Indian Oceans.
    DOI:10.1007/s10409-004-0010-x      URL     [Cited within:2]
    [5] Griffies, S. M., Coauthors, 2009: Coordinated ocean-ice reference experiments (COREs).Ocean Modelling,26,1-46,doi: 10.1016/j.ocemod.2008.08.007.
    The atmospheric forcing from [Large, W., Yeager, S., 2004. Diurnal to decadal global forcing for ocean and sea-ice models: the data sets and flux climatologies. NCAR Technical Note: NCAR/TN-460+STR. CGD Division of the National Center for Atmospheric Research] was developed for coupled-ocean and sea ice models. We found it to be suitable for our purposes, even though its evaluation originally focussed more on the ocean than on the sea-ice. Simulations with this atmospheric forcing are presented from seven global ocean-ice models using the CORE-I design (repeating annual cycle of atmospheric forcing for 500 years). These simulations test the hypothesis that global ocean-ice models run under the same atmospheric state produce qualitatively similar simulations. The validity of this hypothesis is shown to depend on the chosen diagnostic. The CORE simulations provide feedback to the fidelity of the atmospheric forcing and model configuration, with identification of biases promoting avenues for forcing dataset and/or model development.
    DOI:10.1016/j.ocemod.2008.08.007      URL     [Cited within:1]
    [6] Izumo, T., Coauthors, 2010: Influence of the state of the Indian Ocean Dipole on the following year's El Niño.Nature Geoscience,3,168-172,doi: 10.1038/ngeo760.
    El Ni09o-Southern Oscillation (ENSO) consists of irregular episodes of warm El Ni09o and cold La Ni09a conditions in the tropical Pacific Ocean(1), with significant global socio-economic and environmental impacts(1). Nevertheless, forecasting ENSO at lead times longer than a few months remains a challenge(2,3). Like the Pacific Ocean, the Indian Ocean also shows interannual climate fluctuations, which are known as the Indian Ocean Dipole(4,5). Positive phases of the Indian Ocean Dipole tend to co-occur with El Ni09o, and negative phases with La Ni09a(6-9). Here we show using a simple forecast model that in addition to this link, a negative phase of the Indian Ocean Dipole anomaly is an efficient predictor of El Ni09o 14 months before its peak, and similarly, a positive phase in the Indian Ocean Dipole often precedes La Ni09a. Observations and model analyses suggest that the Indian Ocean Dipole modulates the strength of the Walker circulation in autumn. The quick demise of the Indian Ocean Dipole anomaly in November-December then induces a sudden collapse of anomalous zonal winds over the Pacific Ocean, which leads to the development of El Ni09o/La Ni09a. Our study suggests that improvements in the observing system in the Indian Ocean region and better simulations of its interannual climate variability will benefit ENSO forecasts.
    DOI:10.1038/ngeo760      URL     [Cited within:1]
    [7] Jin X. Z., X. H. Zhang, and T. J. Zhou, 1999: Fundamental framework and experiments of the third generation of IAP/LASG world ocean general circulation model.Adv. Atoms. Sci.,16,197-215,doi: 10.1007/BF02973082.
    A new generation of the IAP / LASG world ocean general circulation model is designed and presented based on the previous 20-layer model, with enhanced spatial resolutions and improved parameterizations. The model uses a triangular-truncated spectral horizontal grid system with its zonal wave number of 63 (T63) to match its atmospheric counterpart of a T63 spectral atmosphere general circulation model in a planned coupled ocean-atmosphere system. There are 30 layers in vertical direction, of which 20 layers are located above 1000 m for better depicting the permanent thermocline. As previous ocean models developed in IAP / LASG, a free surface (rather than 渞igid-lid approximation) is included in this model. Compared with the 20-layer model, some more detailed physical parameterizations are considered, including the along / cross isopycnal mixing scheme adapted from the Gent-MacWilliams scheme.The model is spun up from a motionless state. Initial conditions for temperature and salinity are taken from the three-dimensional distributions of Levitus annual mean observation. A preliminary analysis of the first 1000-year integration of a control experiment shows some encouraging improvements compared with the twenty-layer model, particularly in the simulations of permanent thermocline, thermohaline circulation, meridional heat transport, etc. resulted mainly from using the isopycnal mixing scheme. However, the use of isopycnal mixing scheme does not significantly improve the simulated equatorial thermocline. A series of numerical experiments show that the most important contribution to the improvement of equatorial thermocline and the associated equatorial under current comes from reducing horizontal viscosity in the equatorial regions. It is found that reducing the horizontal viscosity in the equatorial Atlantic Ocean may slightly weaken the overturning rate of North Atlantic Deep Water.
    DOI:10.1007/BF02973082      URL     [Cited within:1]
    [8] Ju J. H., L. L. Chen, and C. Y. Li, 2004: The preliminary research of Pacific-Indian Ocean sea surface temperature anomaly mode and the definition of its index.Journal of Tropical Meteorology,20,617-624,doi: 10.3969/j.issn.1004-4965.2004.06.001. (in Chinese with English abstract)
    With the application of the empirical orthogonal function (EOF) analysis to analyze the first EOF mode, we find that the sea surface temperature in the mid-western part of the Indian Ocean and mid-eastern part of equatorial Pacific and that in the western Pacific are inverse during the four seasons of a year. We also find that the variance of the first EOF mode of every season is more than 33 %. This shows that this kind of spatial distribution of the sea surface temperature is stable. We named this pattern the Pacific-Indian Ocean sea surface temperature anomaly mode. Through the careful analysis and comparison, we defined the index of this mode.
    URL     [Cited within:1]
    [9] Large W. G., S. Yeager, 2004: Diurnal to decadal global forcing for ocean and sea-ice models: The data sets and flux climatologies. NCAR/TN-460+STR,1-105, doi: 10.5065/D6KK98Q6.
    DOI:10.5065/D6KK98Q6      URL     [Cited within:1]
    [10] Li C. Y., M. Q. Mu, 1999: El Niño occurrence and sub-surface ocean temperature anomalies in the pacific warm pool.Chinese Journal of Atmospheric Sciences,23,513-521,doi: 10.3878/j.issn.1006-9895.1999.05.01. (in Chinese with English abstract)
    [Cited within:1]
    [11] Li C. Y., M. Q. Mu, 2000: Relationship between East Asian winter monsoon,warm pool situation and ENSO cycle. Chinese Science Bulletin, 45, 1448-1455, doi: 10.1007/BF02898885.
    Based on the observational data analyses and numerical simulations with the air-sea coupled model (CGCM), a new perspective on the occurrence mechanism of ENSO is advanced in this paper. The continuous strong (weak) East Asian winter monsoon will lead to continuous westerly (easterly) wind anomalies over the equatorial western Pacific region. The anomalous equatorial westerly (easterly) winds can cause eastward propagation of the subsurface ocean temperature anomalies (SOTA) in the warm pool region, the positive (negative) SOTA have been in the warm pool region for quite a long time. The eastward propagating of positive (negative) SOTA along the thermocline will lead to positive (negative) SSTA in the equatorial eastern Pacific and the occurrence of El Ni(n)o (La Ni(n)a) event. After the occurrence of ENSO, the winter monsoon in East Asia will be weak (strong) due to the influence of El Ni(n)o (La Ni(n)a).
    DOI:10.1007/BF02898885      URL     [Cited within:1]
    [12] Li C. Y., M. Q. Mu, 2001: The influence of the Indian Ocean dipole on atmospheric circulation and climate. Adv. Atmos. Sci., 18, 831- 843.
    [Cited within:1]
    [13] Li D. H., 2005: Establishment and application of oceanic general circulation model. PhD. dissertation, PLA University of Science and Technology, 200 pp. (in Chinese with English abstract)
    [Cited within:1]
    [14] Li X., 2015: The joint evolution of subsurface ocean temperature in tropical Indo-Pacific and its climate impacts. PhD dissertation, PLA University of Science and Technology, 165 pp. (in Chinese with English abstract)
    [Cited within:1]
    [15] Li X., C. Y. Li, Y. K. Tan, R. Zhang, and G. Li, 2013: Tropical Pacific-Indian Ocean thermocline temperature associated anomaly mode and its evolvement.Chinese Journal of Geophysics,56,3270-3284,doi: 10.6038/cjg20131005. (in Chinese with English abstract)
    The Pacific-Indian Ocean Thermocline Temperature Anomaly Mode (PITM) was brought forward and its corresponding index(PITMI) was defined by analyzing the relation of thermocline temperature between Pacific and India Ocean using SODA reanalysis data and sea surface height data from satellite altimeter. The results show that PITMI has quasi-biennial, 4-year periods and interdecadal period of about 11-12 years. Moreover, it also shows seasonal phase-locked and amplitude asymmetric characteristics. Further studies show that the evolution of PITM is closely related with the evolution and spread of Thermocline Ocean Temperature Anomaly (TOTA). In the Pacific, TOTA moves eastward from the equatorial western Pacific to the equatorial eastern Pacific along the equator (5 degrees S-5 degrees N), turning north there, then turning west at 10 degrees N-14 degrees N to the western Pacific and then propagates to the equatorial western Pacific; while in the Indian Ocean, the major path appears as westward propagation along 8 degrees S-12 degrees S and eastward propagation along the equator(1.25 degrees S-1. 25 degrees N). Composite analyses of NCEP/NCAR wind reanalysis data show that the evolution progress of PITM has a tight relationship with atmosphere circulation, especially the zonal circulation (Walker circulation) over the tropical Pacific and Indian Ocean. Positive phase of PITM is corresponding with clockwise Walker circulation over the equatorial India Ocean and anticlockwise Walker circulation over the equatorial Pacific while the negative phase has reverse situation. Moreover, the propagation and evolution of TOTA also has a clear correlation with that of zonal wind anomaly at 850 hPa.
    DOI:10.6038/cjg20131005      URL     [Cited within:7]
    [16] Lian T., D. K. Chen, Y. M. Tang, and B. G. Jin, 2014: A theoretical investigation of the tropical Indo-Pacific tripole mode.Science China Earth Sciences,57,174-188,doi: 10.1007/s11430-013-4762-7.
    The El Ni o-Southern Oscillation (ENSO) phenomenon in the tropical Pacific has been a focus of ocean and climate studies in the last few decades. Recently, the short-term climate variability in the tropical Indian Ocean has attracted increasingly more attention, especially with the proposition of the Indian Ocean Dipole (IOD) mode. However, these phenomena are often studied separately without much consideration of their interaction. Observations reveal a striking out-of-phase relationship between zonal gradients of sea surface height anomaly (SSHA) and sea surface temperature anomaly (SSTA) in the tropical Indian and Pacific Oceans. Since the two oceans share the ascending branch of the Walker cells over the warm pool, the variation within one of them will affect the other. The accompanied zonal surface wind anomalies are always opposite over the two basins, thus producing a tripole structure with opposite zonal gradients of SSHA/SSTA in the two oceans. This mode of variability has been referred to as Indo-Pacific Tripole (IPT). Based on observational data analyses and a simple ocean-atmosphere coupled model, this study tries to identify the characteristics and physical mechanism of IPT with a particular emphasis on the relationships among ENSO, IOD, and IPT. The model includes the basic oceanic and atmospheric variables and the feedbacks between them, and takes into account the inter-basin connection through an atmospheric bridge, thus providing a valuable framework for further research on the short-term tropical climate variability.
    DOI:10.1007/s11430-013-4762-7      URL     [Cited within:1]
    [17] Liu H. L., P. F. Lin, Y. Q. Yu, and X. H. Zhang, 2012: The baseline evaluation of LASG/IAP Climate system Ocean Model (LICOM) version 2.Acta Meteologica Sinica,26,318-329,doi: 10.1007/s13351-012-0305-y.
    The baseline performance of the latest version (version 2) of an intermediate resolution, stand-alone climate oceanic general circulation model, called LASG/IAP (State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics/Institute of Atmospheric Physics) Climate system Ocean Model (LICOM), has been evaluated against the observation by using the main metrics from Griffies et al. in 2009. In general, the errors of LICOM2 in the water properties and in the circulation are comparable with the models of Coordinated Ocean-ice Reference Experiments (COREs). Some common biases are still evident in the present version, such as the cold bias in the eastern Pacific cold tongue, the warm biases off the east coast of the basins, the weak poleward heat transport in the Atlantic, and the relatively large biases in the Arctic Ocean. A unique systematic bias occurs in LICOM2 over the Southern Ocean, compared with CORE models. It seems that this bias may be related to the sea ice process around the Antarctic continent.
    DOI:10.1007/s13351-012-0305-y      URL     [Cited within:2]
    [18] Liu H. L., Y. Q. Yu, P. F Lin., and F. C. Wang, 2014a: High-resolution LICOM. Flexible Global Ocean-Atmosphere-Land System Model, T. Zhou et al., Eds., Springer-Verlag,Berlin Heidelberg, 321- 331.
    [Cited within:]
    [19] Liu, H. L, P. F. Lin, Y. Q. Yu, F. C. Wang, X. Y. Liu, X. H. Zhang, 2014b: LASG/IAP climate system ocean model version 2: LICOM2. Flexible Global Ocean-Atmosphere-Land System Model, T. Zhou et al., Eds., Springer-Verlag,Berlin Heidelberg, 15- 26.
    [Cited within:]
    [20] Qian W. H., Y. F. Zhu, and J. Y. Liang, 2004: Potential contribution of maximum subsurface temperature anomalies to the climate variability.International Journal of Climatology,24,193-212,doi: 10.1002/joc.986.
    Abstract On the interannual time scale, sea-surface temperature anomalies (SSTAs) that are concerned with climate variability at global and regional scales have been widely investigated in previous studies. Through the analysis of the monthly 46-year (1955–2000) expendable bathythermograph data, we show that subsurface temperature anomalies (STAs) can directly affect the SSTAs in the major air–sea interaction regions. Along the equatorial Pacific, four important features for STAs have been characterized. (1) The STAs and SSTAs are well correlated in the eastern equatorial Pacific (EEP) due to the fact that the thermocline anomalies have only to be mixed with the surface over a very short distance. (2) The STAs are always stronger than SSTAs at any location. (3) In the time between El Ni09o and La Ni09a, and vice versa , the STAs propagate eastward along the thermocline without mixing with SSTAs in the central Pacific. (4) An El Ni09o or La Ni09a can develop only when the maximum positive or the maximum negative STA propagates to the EEP. Inside and outside the tropical basins the STA was more centred on the thermocline than the 20°C isotherm. These features inform us that the maximum STAs (MSTAs) from each vertical STA profile can be used to indicate the anomalous wave-propagation signal or thermocline variations in the worldwide oceans. This analysis implies that the MSTA is also a potential factor controlling climate variability and is a better indicator than SSTA, because MSTAs memorize the change in air–sea interaction signals and represent a huge deposit of energy in the upper ocean. The correlations between SSTAs and MSTAs with a coefficient of more than 0.60 are predominantly located in the EEP, the northern North Pacific, the southern subtropical Indian Ocean, and the northern North Atlantic Ocean. These correlations are discussed from case and statistical analyses. The leading pattern of SSTAs and MSTAs in the tropical Pacific, Atlantic and Indian Oceans are decomposed using empirical orthogonal functions (EOFs) and their EOF patterns might be able to explain the difference between MSTAs and SSTAs. Their periods of mode series are performed from a global wavelet spectral analysis. Copyright 08 2004 Royal Meteorological Society
    DOI:10.1002/joc.986      URL     [Cited within:1]
    [21] Saji N. H., B. N. Goswami, P. N. Vinayachand ran, and T. Yamagata, 1999: A dipole mode in the tropical Indian Ocean. Nature, 401, 360- 363.
    Abstract For the tropical Pacific and Atlantic oceans, internal modes of variability that lead to climatic oscillations have been recognized, but in the Indian Ocean region a similar ocean-atmosphere interaction causing interannual climate variability has not yet been found. Here we report an analysis of observational data over the past 40 years, showing a dipole mode in the Indian Ocean: a pattern of internal variability with anomalously low sea surface temperatures off Sumatra and high sea surface temperatures in the western Indian Ocean, with accompanying wind and precipitation anomalies. The spatio-temporal links between sea surface temperatures and winds reveal a strong coupling through the precipitation field and ocean dynamics. This air-sea interaction process is unique and inherent in the Indian Ocean, and is shown to be independent of the El Ni脙卤o/Southern Oscillation. The discovery of this dipole mode that accounts for about 12% of the sea surface temperature variability in the Indian Ocean--and, in its active years, also causes severe rainfall in eastern Africa and droughts in Indonesia--brightens the prospects for a long-term forecast of rainfall anomalies in the affected countries.
    DOI:10.1038/43854      PMID:16862108      URL     [Cited within:]
    [22] Smith T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore, 2008: Improvements to NOAA's historical merged land-ocean surface temperature analysis (1880-2006).J. Climate,21,2283-2296,doi: 10.1175/2007JCLI2100.1.
    DOI:10.1175/2007JCLI2100.1      URL     [Cited within:2]
    [23] Ueda H., J. Matsumoto, 2000: A possible triggering process of east-west asymmetric anomalies over the Indian Ocean in relation to 1997/98 El Niño.J. Meteor. Soc. Japan,78,803-818,doi: 10.2151/jmsj1965.78.6_803.
    DOI:10.2151/jmsj1965.78.6_803      URL     [Cited within:]
    [24] Webster P. J., A. M. Moore, J. P. Loschnigg, and R. R. Leben, 1999: Coupled ocean-atmosphere dynamics in the Indian Ocean during 1997-98.Nature,401,356-360,doi: 10.1038/43848.
    DOI:10.1038/43848      URL     [Cited within:]
    [25] Wu H. Y., C. Y. Li, 2009: Numerical simulation of the tropical Pacific-Indian Ocean associated temperature anomaly mode.Climatic and Environmental Research,14(6),567-586,doi: 10.3878/j.issn.1006-9585.2009.06.01. (in Chinese with English abstract)
    Three-dimensional thermal and dynamic structures of the Pacific - Indian Ocean associated temperature anomaly mode and the forming and evolvement mechanism of the associated mode are studied in terms of numerical modeling and composite analyses.The sea temperature anomalies in the equatorial western Indian Ocean and the equatorial eastern Pacific are reversed with that in the equatorial eastern Indian Ocean - western Pacific.This associated mode has annual,interannual,and interdecadal variability.The composite analyses of modeling results disclose that the circulation anomaly caused by the anomalous surface wind stress is one of the main dynamical reasons and the vertical transport is important to the subsurface associated sea temperature mode.In addition,anomalous net short wave radiation flux and latent heat flux are the main thermodynamic reasons to the associated mode forming. From the occurrence associated to its crest,the anomalous surface advection and subsurface vertical transportation spur its evolution,while anomalous net short wave radiation flux,latent heat flux,sensible heat flux,and subsurface advection restrain its further evolution.
    URL     [Cited within:2]
    [26] Wu H. Y., C. Y. Li, and M. Zhang, 2010: The preliminary numerical research of effects of ITF on tropical Pacific-Indian Ocean associated temperature anomaly mode.Journal of Tropical Meteorology,26(5),513-520,doi: 10.3969/j.issn.1004-4965.2010.05.001. (in Chinese with English abstract)
    Geographical condition in Indonesia sea area is the passage of tropical Pacific and Indian.In order to find out the influence of the Indonesian Throughflow(ITF) on tropical Pacific-Indian Ocean associated temperature anomaly mode,experiments of opening or closing the ITF in an Ocean Global Circulation Model(OGCM) are done.The results suggest that ITF is important in the simulations of the sea temperature and currents in tropical Pacific and Indian Ocean.This paper is also shown that for the associated temperature anomaly mode,the ITF is playing subordinate factor while the external forcing is the primary on the sea surface,however,it plays certain role in the subsurface.
    DOI:10.1080/09500340.2010.529951      URL     [Cited within:2]
    [27] Wu S., Q. Y. Liu, and R. J. Hu, 2005: The main coupled mode of SSW and SST in the tropical Pacific South China Sea-Tropical Indian Ocean on interannual time scale.Periodical of Ocean University of China,35,521-526,doi: 10.3969/j.issn.1672-5174.2005.04.001. (in Chinese with English abstract)
    Based on the COADS data, the coupled mode of SSWA (sea surface wind anomaly) and SSTA for the whole tropical Pacific-the South China Sea-the tropical Indian Ocean is studied, using various methods such as correlation analysis, Empirical Orthogonal Function (EOF) analysis, Singular Value Decomposition (SVD) and wavelet analysis. It is shown that there exists a dominant mode of SSWA and SSTA when considering the tropical Pacific, the Indian Ocean and the South China Sea as a whole. Through conducting Singular Value Decomposition (SVD), the main coupled mode for such a big area is found and discussed. In the Pacific the main mode is the ENSO cycle, a unanimous variance is found to be the dominant mode in the Tropical Indian Ocean, and the South China Sea has a basin-scale homogeneity.
    DOI:10.1360/gs050302      URL     [Cited within:1]
    [28] Yan H. M., J. H. Ju, and Z. N. Xiao, 2001: The variable characteristics analysis of SSTA over the Indian ocean during the two phases of ENSO cycle.Journal of Nanjing Institute of Meteorology,24,242-249,doi: 10.3969/j.issn.1674-7097.2001.02.014. (in Chinese with English abstract)
    By analyzing the variable characteristics of SST in the Indian Ocean during different phases of ENSO cycle,it is found that the correlation of the variation of SST between in the Indian Ocean and in eastern equatorial Pacific is significant,and the variation of SST in the Indian Ocean is an important component of ENSO cycle.The SST pattern in the Indian Ocean is warm in the western part and cold in the eastern part when SST in the eastern equatorial Pacific is warm,and v.v.Furthmore,the results showed that the most significant area of difference of SSTA between in the Western and Eastern Indian Ocean is in 0锝25℃S.It's interannual and seasonal change is significant.The zonal difference of SST variation in the Indian Ocean during summer monsoon season is more significant than that during winter monsoon season,and the difference during cold phases is larger than that during warm phases.The correlation coefficient is 0.55 between Dipole Mode Index in the Indian Ocean and SSTA in the eastern equatorial Pacific.
    URL     [Cited within:1]
    [29] Yang H., C. Y. Li, 2005: Effect of the tropical Pacific-Indian ocean temperature anomaly mode on the south Asia high.Chinese Journal of Atmospheric Sciences,29,99-110,doi: 10.3878/j.issn.1006-9895.2005.01.12. (in Chinese with English abstract)
    The east-west difference of tropical Indian Ocean temperature anomaly (dipole) is closely related to the ENSO in the Pacific Ocean through the Walker circulation and other processes. Thus the ENSO and Indian dipole should be regarded as an air-sea coupled system in the tropical Pacific and Indian Ocean. The tropical Pacific-Indian Ocean temperature anomaly mode is presented,and its impact on the South Asia high is studied. The positive phase of the tropical Pacific-Indian Ocean temperature anomaly mode (positive SSTAs in the western Indian Ocean and eastern Pacific, negative SSTAs in the eastern Indian Ocean-western Pacific) is favorable for a weaker South Asia high with southeastwards shift. The negative phase (reverse SSTAs of the positive phase) contributes to a stronger South Asia high with northwestwards shift. The physical mechanisms that the tropical Pacific-Indian Ocean temperature anomaly mode influences the South Asia high are proposed. 1) The tropical Pacific-Indian Ocean temperature anomaly mode greatly influences the Asian summer monsoon, resulting in the rainfall anomaly. The analyses of the vertically integrated heat sources and moisture sinks, and precipitation reveal that the release of latent heat of condensation is the primary heat source. When the tropical Pacific-Indian Ocean temperature anomaly mode is in a positive phase, the heat source over the Tibetan Plateau is smaller than normal with less precipitation there. However, when the tropical Pacific-Indian Ocean temperature anomaly mode is in a negative phase, more abundant precipitation appears,indicating larger heat source there. The negative (positive) heat source anomaly maintains the negative (positive) ascent anomaly over the Tibetan Plateau, which provide a key mechanism for the distinct change of intensity of the South Asia high. The positive (negative) phase is favorable for the stronger (weaker) South China Sea summer monsoon and the intertropical convergence zone. Thus positive (negative) heat source anomaly is found from the South China Sea to the western Pacific via the Philippines, resulting in an anticyclonic (cyclonic) circulation to the southeast of the Tibetan Plateau at the high level. So the South Asia high is located in the southeast (northwest) position. 2) The tropical Pacific-Indian Ocean temperature anomaly mode greatly influences the zonal vertical (Walker) circulation. The positive (negative) phase drives the westerly (easterly) current anomaly at the high level over the Indian Ocean, contributing to the weaker (stronger) easterlies in the south part of South Asia high. 3) The positive (negative) phase drives a clear local Hadley cell anomaly with ascent (descent) over the western Indian Ocean and descent (ascent) over the Iran Plateau, resulting in the weaker (stronger) South Asia high with southeastwards (northwestwards) shift.
    DOI:10.1111/j.1745-7254.2005.00209.x      URL     [Cited within:3]
    [30] Yang H., X. L. Jia, and C. Y. Li, 2006: The tropical Pacific-Indian Ocean temperature anomaly mode and its effect.Chinese Science Bulletin,51,2878-1884,doi: 10.1007/s11434-006-2199-5.
    DOI:10.1007/s11434-006-2199-5      URL     [Cited within:1]
    [31] Yu L. S., M. M. Rienecker, 1999: Mechanisms for the Indian Ocean warming during the 1997-98 El Niño.Geophys. Res. Lett.,26,735-738,doi: 10.1029/1999GL900072.
    This study examines primary mechanisms that gave rise to the basin-wide variations of the sea surface temperature (SST) in the Indian Ocean during the 1997-98 El Ni o by using multi-source data sets. The evolution of some key atmosphere-ocean variables indicated that the SST variability in the Indian Ocean was largely attributable to the ENSO impact on the large-scale atmospheric circulation. During June-December 1997, when the El Ni o in the Pacific was maturing, the Indian Ocean experienced the reversal of the Walker Circulation and the prolonged equatorward displacement of the southeast trades. The resultant changes in surface wind influenced the SST through the following means. In the equatorial region, the easterly winds associated with the reversed Walker Circulation forced equatorial Kelvin/Rossby waves, which then affected the equatorial ocean heat balance (mainly through upwelling/downwelling) and led to the reversal of the zonal SST gradient in the fall of 1997. The negative SST anomalies in the east and positive anomalies in the west in turn helped maintain and prolong the equatorial easterlies, a clear indication of coupled atmosphere-ocean interactions in operation. Outside of the equatorial waveguide, changes of latent heat flux induced by wind speed variations played a major role in the broad-scale warming. The effect was most significant during the summer/fall of 1997 when the southeasterly trade winds weakened considerably, leading to a dramatic reduction of latent heat release and subsequently a rapid surface warming in the southern ocean.
    DOI:10.1029/1999GL900072      URL     [Cited within:2]
    [32] Yu Y. Q., H. L. Liu, and P. F. Lin, 2012: A quasi-global 1/10 eddy-resolving ocean general circulation model and its preliminary results. Chinese Science Bulletin,57, 3908-3916, doi: 10.1007/s11434-012-5234-8.
    在这研究,一伪全球(排除北极海洋) 解决旋涡的海洋将军发行量模型(OGCM ) 基于 LASG/IAP 气候系统海洋模型(LICOM2.0 ) 的最近的版本被建立。水平决定和垂直分辨率被增加到 1/10 吗??
    DOI:10.1007/s11434-012-5234-8      URL     [Cited within:1]
    [33] Yuan D. L., 2005: Role of the Kelvin and Rossby waves in the seasonal cycle of the equatorial Pacific Ocean circulation. J. Geophys. Res., 110,C04004, doi: 10.1029/2004JC002344.
    The dynamics of the seasonal cycle of the equatorial Pacific Ocean circulation are investigated in a hindcast study generated by the Poseidon quasi-isopycnal general ocean circulation model forced by wind stress from a Special Sensor Microwave Imager product and heat flux from an atmospheric mixed layer model. The simulated annual and semi-annual oscillations are validated by the Tropical Atmosphere-Ocean mooring measurements and the altimetry sea levels on the equator, based on which the roles of the equatorial Kelvin and Rossby waves in the seasonal cycle dynamics are investigated. The Kelvin and Rossby waves are extracted from the hindcast in such a way as to treat the model zonal nonlinear momentum terms and wind stress as forcing terms so that the validity of the linear theory on wave propagation and reflection can be examined. It is found that the reflections of the first baroclinic mode waves at the Pacific western and eastern boundaries are in good agreement with the linear theory. In comparison, the reflections of the second and higher baroclinic mode waves are significantly different from the linear theory, suggesting that the reflections are potentially nonlinear. The hindcast has simulated the westward propagation of the dominant annual oscillations of the seasonal sea level in the central basin and the eastward propagation of the dominant semi-annual oscillations of the seasonal sea level in the western and eastern basins successfully. The westward propagation of the annual oscillations is primarily associated with the annual Rossby waves forced by the annual winds in the eastern-to-central equatorial Pacific, the upwelling phase of which forces progressive springtime reversal of the surface zonal currents from the east to the west. However, the internal dynamics of the ocean alter the amplitudes and phases of the springtime reversal significantly. The results of this study have shown that the internal dynamics of the ocean are an important source of the semi-annual oscillations in the equatorial Pacific Ocean. The semi-annual harmonics of the Rossby waves are boosted by the internal dynamics and eventually dominate the seasonal cycle in the far western Pacific. The reflections of the Rossby waves at the western boundary produce semi-annual dominant Kelvin waves, which force significant semi-annual oscillations of sea level in the eastern basin.
    DOI:10.1029/2004JC002344      URL     [Cited within:1]
    [34] Yuan, D. L., Coauthors, 2011: Forcing of the Indian Ocean Dipole on the interannual variations of the tropical Pacific Ocean: Roles of the Indonesian Throughflow.J. Climate,24,3593-3608,doi: 10.1175/2011JCLI3649.1.
    Controlled numerical experiments using ocean-only and ocean-atmosphere coupled general circulation models show that interannual sea level depression in the eastern Indian Ocean during the Indian Ocean dipole (IOD) events forces enhanced Indonesian Throughflow (ITF) to transport warm water from the upper-equatorial Pacific Ocean to the Indian Ocean. The enhanced transport produces elevation of the thermocline and cold subsurface temperature anomalies in the western equatorial Pacific Ocean, which propagate to the eastern equatorial Pacific to induce significant coupled evolution of the tropical Pacific oceanic and atmospheric circulation. Analyses suggest that the IOD-forced ITF transport anomalies are about the same amplitudes as those induced by the Pacific ENSO. Results of the coupled model experiments suggest that the anomalies induced by the IOD persist in the equatorial Pacific until the year following the IOD event, suggesting the importance of the oceanic channel in modulating the interannual climate variations of the tropical Pacific Ocean at the time lag beyond one year.
    DOI:10.1175/2011JCLI3649.1      URL     [Cited within:2]
    [35] Yuan D. L., H. Zhou, and X. Zhao, 2013: Interannual climate variability over the tropical Pacific Ocean induced by the Indian Ocean Dipole through the Indonesian Throughflow.J. Climate,26,2845-2861,doi: 10.1175/JCLI-D-12-00117.1.
    The authors' previous dynamical study has suggested a link between the Indian and Pacific Ocean interannual climate variations through the transport variations of the Indonesian Throughflow. In this study, the consistency of this oceanic channel link with observations is investigated using correlation analyses of observed ocean temperature, sea surface height, and surface wind data. The analyses show significant lag correlations between the sea surface temperature anomalies (SSTA) in the southeastern tropical Indian Ocean in fall and those in the eastern Pacific cold tongue in the following summer through fall seasons, suggesting potential predictability of ENSO events beyond the period of 1 yr. The dynamics of this teleconnection seem not through the atmospheric bridge, because the wind anomalies in the far western equatorial Pacific in fall have insignificant correlations with the cold tongue anomalies at time lags beyond one season. Correlation analyses between the sea surface height anomalies (SSHA) in the southeastern tropical Indian Ocean and those over the Indo-Pacific basin suggest eastward propagation of the upwelling anomalies from the Indian Ocean into the equatorial Pacific Ocean through the Indonesian Seas. Correlations in the subsurface temperature in the equatorial vertical section of the Pacific Ocean confirm the propagation. In spite of the limitation of the short time series of observations available, the study seems to suggest that the ocean channel connection between the two basins is important for the evolution and predictability of ENSO.
    DOI:10.1175/JCLI-D-12-00117.1      URL     [Cited within:1]
    Downloaded    
    RichHTML Viewed    
    Abstract Viewed    

    Share
    Export

    Related articles:
    Key words
    ocean general circulation model
    numerical simulation
    tropical Pacific-Indian Ocean associated mode
    subsurface ocean temperature anomaly

    Authors
    Xin LI
    Chongyin LI