1. Introduction
The tropical Pacific Walker circulation (PWC), a large-scale zonal overturning atmospheric circulation across the tropical Pacific Ocean, named after the early twentieth-century British meteorologist Sir Gilbert Walker, is one of the most important components of the global climate system. PWC is a thermally direct circulation, driven by ascending motion and deep convection over the warm western Pacific and Maritime Continent and subsidence over the cooler eastern Pacific, with surface easterlies along the equator and westerlies in the upper troposphere (Bjerknes 1969). Its interannual variability is closely linked to those of El Niño–Southern Oscillation (ENSO), monsoonal circulation, and rainfall over adjacent continents (Philander 1990; Webster et al. 1998; Tanaka et al. 2004). The long-term changes of PWC intensity and structure are associated with precipitation and temperature changes, such as the drying of eastern Africa, intensified Northern Hemisphere summer monsoon precipitation, and global warming hiatus in recent decades (Williams and Funk 2011; Wang et al. 2012; Kosaka and Xie 2013; Liu et al. 2013; England et al. 2014; McGregor et al. 2014; Thompson et al. 2015). The understanding of the PWC change is therefore of great importance to meteorological disaster prediction, ecosystem maintenance, water resource management, and agricultural development.
Long-term changes in PWC have recently been the subject of intense debate in the climate change community (Power and Smith 2007; Power and Kociuba 2011; Meng et al. 2012; Sohn et al. 2013; Tokinaga et al. 2012a,b; DiNezio et al. 2013; L’Heureux et al. 2013; Bayr et al. 2014; McGregor et al. 2014; Sandeep et al. 2014). Long-term changes in PWC are dominated by two mechanisms: homogeneous warming (Knutson and Manabe 1995; Held and Soden 2006; Gastineau et al. 2009) and inhomogeneous warming (DiNezio et al. 2009, 2010; Tokinaga et al. 2012a; Sandeep et al. 2014). The homogeneous warming mechanism argues that the weakening of the tropical PWC under global warming is caused by the decrease of convective mass flux, which balances the slower increase of global-mean precipitation at a rate of around 1%–3% K−1, compared to the increase of atmospheric water vapor content at a rate of around 7% K−1, with constant relative humidity (Held and Soden 2006). This hydrological constraint does not require any horizontal gradients in the warming pattern and was referred to as the homogeneous warming mechanism (Knutson and Manabe 1995; Gastineau et al. 2009). The inhomogeneous warming mechanism argues that the tropical Pacific SST change patterns dominate the PWC change (Tokinaga et al. 2012a; Sandeep et al. 2014). The El Niño–like or La Niña–like pattern of the equatorial Pacific SST, which resulted from the regional differences in the cooling effect of surface evaporative cooling in response to global warming (Knutson and Manabe 1995; Xie et al. 2010) and the ocean dynamical thermostat (Clement et al. 1996; Cane et al. 1997), respectively, weaken or strengthen the zonal SST gradient and drive the weakening or strengthening of PWC (Collins et al. 2010; DiNezio et al. 2010; Solomon and Newman 2012). Meanwhile, recent enhanced Indian Ocean and Atlantic warming may also modulate Pacific climate change by driving the intensification of PWC (Williams and Funk 2011; Luo et al. 2012; McGregor et al. 2014).
There is physically consistent evidence for a reduction in the strength of PWC during the twentieth century in observations of sea level pressure (SLP), precipitation, sea surface temperature (SST), cloud, and surface wind (Tanaka et al. 2004; Vecchi et al. 2006; Zhang and Song 2006; Deser et al. 2010; Tokinaga et al. 2012b). Most coupled model simulations also indicate a slowdown of PWC over the twentieth century (Vecchi and Soden 2007; DiNezio et al. 2009; Power and Kociuba 2011; DiNezio et al. 2013; Kociuba and Power 2015). In the meantime, some studies argue for a strengthening of the twentieth-century PWC (Meng et al. 2012; Sandeep et al. 2014). As there is large uncertainty in the observed SST warming pattern and the observational SLP dataset (Deser et al. 2010; Solomon and Newman 2012; Tokinaga et al. 2012b; DiNezio et al. 2013; L’Heureux et al. 2013), there is still considerable uncertainty concerning the twentieth-century trends of the PWC. The strong internal variability of the tropical Pacific superimposes on the externally forced change of PWC and finally either promotes or cancels out the forced weakening trend of the PWC (Vecchi et al. 2006; Power and Kociuba 2011; Meng et al. 2012; DiNezio et al. 2013; Ma and Zhou 2014). For instance, the Pacific decadal oscillation (PDO) or the interdecadal Pacific oscillation (IPO) in its warm (cold) phase indicates an SST warm (cold) background condition in the central and eastern tropical Pacific, reducing (increasing) the climatological zonal SST gradient, which locally acts to weaken (enhance) PWC (Deser and Wallace 1990; Garcia and Kayano 2008; Dong and Lu 2013). The intensified PWC over the past 30 years (1979–2008) may also be caused by the decadal variation of El Niño with more frequent occurrence of central Pacific-type El Niño (Sohn et al. 2013). The recent intensified PWC is suggested to be driven by a mega–El Niño–Southern Oscillation (mega-ENSO), which is a leading mode of interannual-to-interdecadal variation of global SST (Wang et al. 2012).
In recent years, PWC change has received considerable attention as its recent intensification is closely connected with the rapid sea level rise in the western Pacific (Merrifield 2011), global warming hiatus (Kosaka and Xie 2013; England et al. 2014; Thompson et al. 2015), SLP change (L’Heureux et al. 2013), intensified Northern Hemisphere summer monsoon (Wang et al. 2012), and the reversals of multidecadal weakening of the Indonesian Throughflow and Leeuwin Current transports (Feng et al. 2011). Most previous studies mainly focused on the change of the PWC intensity and are usually based on the analyses of surface wind and SLP of only one observational or reanalysis dataset (Feng et al. 2011; Li and Ren 2012; Luo et al. 2012; de Boisséson et al. 2014; McGregor et al. 2014; England et al. 2014). As found in some previous studies (Annamalai et al. 1999; Kinter et al. 2004; Tokinaga et al. 2012b; de Boisséson et al. 2014; Schwendike et al. 2014), the observing system used for the data assimilation frequently undergoes major changes and then may induce some spurious signals of the tropical circulation changes derived from the atmospheric reanalysis products. There were observing system changes in the period after 1979 that may have been as significant as earlier changes insofar as they may have significantly altered the deep heating and, consequently, the PWC (e.g., Onogi et al. 2007; Chen et al. 2008). Up to now, less effort has been devoted to the study on how the whole three-dimensional structure of the PWC changed in the recent three decades, a period with better observations covering it and multiple reanalysis datasets available. Meanwhile, few studies have determined how reliable the changes of PWC revealed by reanalysis datasets are and how uncertain the corresponding PWC changes among different datasets are. Moreover, mechanisms causing the recent strengthened PWC remain inconclusive. In this study, we examine the robustness of PWC change characteristics, including the intensity, western edge, and spatial structure in the recent three decades (1979–2012) using multiple reanalysis datasets and independent observations of atmospheric variables that are closely linked to the changes in the PWC, based on multi-PWC indices. To further examine how the PWC responds to observed SST forcing alone and understand the mechanism of PWC change, we further analyze the PWC change in the Atmospheric Model Intercomparison Project (AMIP) experiments. We show evidences that, in terms of the changes in the three-dimensional structure, the PWC has strengthened and shifted westward during the recent three decades. The intensification and westward shift of PWC are robust in all seven sets of reanalysis data, with an intensified rate of 15% decade−1 (range from 8% to 18% decade−1) and westward-shifting rate of 3.7° longitude decade−1 (range from 2.6° to 4.7° longitude decade−1). Analyses of 26 model simulations from phase 5 of the Coupled Model Intercomparison Project (CMIP5), used for the AMIP experiments, found that the models successfully reproduced the observed strengthening and westward shift of PWC, which is indicative of a dominant forcing of the La Niña–like SST anomalies to the observed PWC change.
The remainder of the paper is organized as follows. The observational datasets, seven sets of reanalysis data, models, and analysis methods are described in section 2. Section 3 makes a comparison of changes in PWC characteristics in seven sets of reanalysis data. Simulated features of PWC in 26 CMIP5 models are analyzed in section 4. Finally, a summary is given in section 5.
2. Data and method
a. Data description
The seven reanalysis datasets used in this study are the National Oceanic and Atmospheric Administration (NOAA)–Cooperative Institute for Research in Environmental Sciences (CIRES) Twentieth Century Reanalysis (20CR; Compo et al. 2011; http://www.esrl.noaa.gov/psd/data/gridded/data.20thC_ReanV2.html), the European Centre for Medium-Range Weather Forecast (ECMWF) interim reanalysis (ERAIM; Dee et al. 2011; http://apps.ecmwf.int/archive-catalogue/?class=ei), the Japanese 25-year (JRA-25; Onogi et al. 2007; http://jra.kishou.go.jp/JRA-25/index_en.html) and 55-year Reanalysis Projects (JRA-55; Ebita et al. 2011; http://jra.kishou.go.jp/JRA-55/index_en.html), the Modern-Era Retrospective Analysis for Research and Applications (MERRA; Rienecker et al. 2011; http://gmao.gsfc.nasa.gov/merra/), the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research reanalysis (NCEP-1; Kalnay et al. 1996; http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html), and the NCEP–U.S. Department of Energy reanalysis (NCEP-2; Kanamitsu et al. 2002; http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis2.html). These datasets are the currently state-of-the-art reanalyses, and their characteristics are summarized in Table 1. It is shown that these reanalyses differ in model system, data assimilation method, satellite data processing, and model resolution. They cover different time periods but overlap in our interest analysis period 1979–2012. Multiple reanalyses are applied to avoid data dependence of the results. Results of individual reanalysis datasets are derived from their own original spatial resolutions. The multiple reanalysis ensemble mean (MRE) is calculated after individual reanalyses are uniformly interpolated onto a common 2.5° × 2.5° grid format using bilinear interpolation.
Reanalysis datasets used in this study and their detailed information.
To examine whether the precipitation and SLP change are coherent with the change of PWC and climate impacts of the recent PWC change, the monthly Global Precipitation Climatology Project (GPCP), version 2.2, combined precipitation dataset (Huffman et al. 2009), the monthly mean SLP data from the Hadley Centre Sea Level Pressure dataset, version 2 (HadSLP2; Allan and Ansell 2006), and surface temperature from the Goddard Institute for Space Studies (GISS) Temperature Analysis (GISTEMP; Hansen et al. 2010) are used. To study how SST variations affect PWC, the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST) is used here to describe global SST variation patterns and extract the decadal-scale signals from its long record (Rayner et al. 2003).
To examine the robustness of recent PWC change and drive mechanism, the AMIP experiments, using the monthly mean observed SST to force the atmospheric general circulation model (AGCM), from 26 climate models (listed in Table 2) participating in CMIP5, are analyzed. Detailed descriptions about the models and experiments can be seen online (http://cmip-pcmdi.llnl.gov/cmip5/). Only the first member of each model, with the same period from 1979 to 2008, is used for analysis. All model data are regridded onto a common 2.5° × 2.5° grid using bilinear interpolation.
Details of 26 CMIP5 AGCMs used in the study. Horizontal resolution shows the number of grid points in the meridional by zonal directions. (Expansions of acronyms are available at http://www.ametsoc.org/PubsAcronymList.)
b. PWC indices
Three kinds of PWC index are used:
The first is the equatorial Pacific east–west SLP gradient (dSLP), defined by Vecchi et al. (2006), the difference of the area-averaged SLP between the Tahiti region (5°S–5°N, 160°–80°W) and the Darwin region (5°S–5°N, 80°–160°E), as a proxy of PWC intensity. The value dSLP offers a good measure of the equatorial trade wind in both observations and models (Vecchi et al. 2006; Vecchi and Soden 2007; Zhang and Song 2006), but dSLP is not a direct measure of the atmospheric circulation.
The second index is the surface zonal wind (Us) averaged over the equatorial Pacific (5°S–5°N, 150°E –150°W), as defined by Luo et al. (2012) and McGregor et al. (2014). The Pacific trade wind is only the surface part of PWC.
The third index is based on zonal mass streamfunction (Yu and Zwiers 2010; Bayr et al. 2014).
The divergent component of the zonal wind is obtained by solving the Poisson equation globally for the potential function with divergence as the forcing term and then calculating the divergent wind. The zonal mass flux streamfunction is computed subsequently by vertically integrating uD meridionally averaged between 5°S and 5°N, from top level downward. Thus, the calculation of Ψ depends on accurate analyses of uD, which may be highly model dependent because the divergent circulation in reanalyses is strongly influenced by the model heating, especially in the upper-tropospheric wind, in data-sparse regions such as are common in the tropics (Annamalai et al. 1999; Schwendike et al. 2014). For example, Kumar et al. (1999), using the divergent circulation deduced from the velocity potential at 200 hPa based on the NCEP-1 reanalysis as an indicator variable of PWC, found a southeastward shift in the PWC anomalies associated with ENSO events during 1958–97 that may lead to a reduced subsidence over the Indian region, thus causing the weakening relationship between the Indian monsoon and ENSO. However, comparisons between reanalysis products and independent observations suggested that the abrupt shifts of divergent circulation in NCEP-1 were probably spurious artifacts of the treatment of convection in models used to produce reanalyses or the changes in the observing system since the late 1970s (Kinter et al. 2004). Moreover, the observing system used for the data assimilation frequently undergoes major changes that may create spurious signals in atmospheric reanalyses (de Boisséson et al. 2014).
In addition, although the atmospheric circulation in a reanalysis dataset is the best estimate of real atmospheric circulation since the zonal and meridional wind are directly assimilated from the observational data (Kalnay et al. 1996), there may be some differences in divergent circulation described by different reanalysis datasets, because the atmospheric forecast model, assimilation algorithm, vertical and horizontal resolution, satellite data processing (Table 1), and convective parameterizations are not the same among different reanalysis dataset. This has led to the differences of the global monsoon precipitation (Lin et al. 2014) and the atmospheric water vapor transport for summer precipitation over the Tibetan Plateau (Feng and Zhou 2012) derived from different reanalysis datasets. To ensure the robustness and reliability of results, we focus on the better-observed period since the late 1970s; the changes of PWC derived from multiple reanalysis datasets in the period of 1979–2012 were examined and compared.
The latitudinal interval from 5°S to 5°N is chosen because the region is typically used to define the Walker circulation (Cane et al. 1997; Vecchi et al. 2006; DiNezio et al. 2013; Bayr et al. 2014), and the result is insensitive to reasonable variation of the latitudinal interval averaged (Oort and Yienger 1996). To make the results comparable to previous studies, we follow the definition. We have noticed the possible influence of the latitudinal interval. In a recent paper, Schwendike et al. (2014) compared results based on latitudinal interval averaged over 5°S–5°N, 10°S–10°N, and 35°S–10°N when they used vertical mass flux to define PWC and also found the essential characteristics of the PWC basically remain the same among the different latitude average. We have also examined the Walker circulation changes based on the 10°S–10°N average; the conclusion remains the same.
The zonal mass streamfunction in Eq. (1) not only can isolate the zonal convective flow from original three-dimensional tropical atmospheric circulation, which includes meridional and zonal components—namely, the Hadley and Walker circulation (Schwendike et al. 2014)—but also proves a more intuitive image of PWC (Fig. 1). The vertically averaged Ψ of all levels averaged over the western and central Pacific (150°E–150°W) is built as another PWC intensity index (STRF). We define the PWC western edge by the zero line of Ψ on the westward side of the international date line, averaged between 400 and 600 hPa. The results are not sensitive to the pressure levels chosen to determine the PWC intensity and western edge.
Mean state (1979–2012) of the tropical PWC in (a) 20CR, (b) ERAIM, (c) JRA-25, (d) JRA-55, (e) MERRA, (f) NCEP-1, (g) NCEP-2, and (h) the multiple reanalysis ensemble mean. Shading and contours are for zonal mass streamfunction (1011 kg s−1) along the equatorial Pacific (5°S–5°N). Vectors are the composite of pressure velocity (ω × −50; Pa s−1) and zonal divergent wind (m s−1).
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
c. Statistical significance evaluation of trends
As in many previous studies (Luo et al. 2012; Sohn et al. 2013; Kosaka and Xie 2013; L’Heureux et al. 2013; McGregor et al. 2014; Sandeep et al. 2014), trends of all quantities presented in this study were calculated using least-squares linear regressions, and their level of significance was estimated using a two-tailed Student’s t test (null hypothesis of zero linear trend) with an effective degree of freedom (Bretherton et al. 1999).
3. Comparison of seven sets of reanalysis data
a. Long-term mean PWC characteristics
To better understand the changing character of the PWC during the recent three decades (1979–2012), we first discuss the long-term climatology of the PWC defined using the zonal mass streamfunction. Meanwhile, differences of the climatological PWC characteristics between different reanalysis datasets are compared.
The long-term (1979–2012) mean annual zonal mass streamfunction along the equatorial Pacific (5°S–5°N) and corresponding zonal divergent winds and vertical winds derived from seven reanalysis datasets and a multiple reanalysis ensemble mean are shown in Fig. 1. The results above 100 hPa are not displayed, because the atmospheric mass flux above 100 hPa is negligible. In terms of Ψ, the zonal atmospheric circulation is depicted by alternating negative and positive cells. The PWC is the strongest cell located east of the Maritime Continent, with positive values indicative of clockwise rotation, consistent with those indicated by composite vectors of zonal divergent winds and vertical winds. The center of the PWC cell is located in the middle troposphere (~500 hPa) and fairly close to the equatorial central Pacific (~160°W). The relatively denser lines of constant streamfunction over the region around 150°E and the region around 120°W, separately, are coherent with the stronger rising motion over the Maritime Continent and western Pacific and sinking motion over the eastern Pacific. Combined with the results of the zonal divergent circulation, Ψ depicts visually the whole structure of the PWC, which is characterized by an ascent center over the Maritime Continent and western Pacific, westerlies in the upper troposphere, a strong decent in the eastern Pacific, and surface easterlies, resulting in an enclosed cell.
PWC structure features captured by seven reanalyses are similar, with the pattern correlation of the zonal mass streamfunction along the equatorial Pacific between individual reanalysis and a multiple reanalysis ensemble mean larger than 0.96. Nevertheless, slight differences are notable between different reanalyses. For example, the PWC is the strongest in 20CR and the weakest in MERRA, with the zonal mass flux Ψ of the circulation center larger than 5.0 × 1011 kg s−1 and close to 4.0 × 1011 kg s−1, respectively. The western edge of the PWC (zero line of Ψ over the western Pacific) indicated by four reanalyses (ERAIM, JRA-25, JRA-55, and MERRA) with relatively higher horizontal resolution is shifted farther westward than that of three reanalyses (20CR, NCEP-1, and NCEP-2) with lower horizontal resolution.
b. Recent changes in PWC characteristics
As introduced in section 2, although the SLP gradient between the western and eastern Pacific is not a direct measure of the PWC strength, it is a simple estimate of the PWC strength. The equatorial Pacific trade winds are only the surface parts of the PWC, but it is the main bridge of the air–sea interactions over the tropical Pacific. Thus, in this section, we carry out an integrated study of the changes in PWC using the zonal mass streamfunction, which captures PWC structure features well, combined with the results of SLP and surface trade.
The long-term trends of Ψ along the equatorial Pacific during 1979–2012 are shown in Fig. 2. In all reanalyses and multiple reanalysis ensemble mean, Ψ trends (shading) show a quite similar spatial pattern with respect to the long-term mean (contour), with a westward movement of the maximum positive Ψ trend center, compared to the climatological center of the PWC cell. Both the positive trend of Ψ on the west side of the PWC cell and the negative trend on the right side of the PWC cell are statistically significant at the 5% level, indicating an intensification and westward shift of the PWC in recent decades. The trends of the zonal mass streamfunction over the equatorial exhibit a similar pattern among seven reanalysis datasets, with the pattern correlation coefficients between individual reanalysis dataset and multiple reanalysis ensemble mean all larger than 0.71. Discrepancies are still obvious between the seven reanalysis products, despite the large similarities of trend spatial pattern. For instance, the positive trends of the zonal mass streamfunction have a broader coverage in 20CR, ERAIM, JRA-25, JRA-55, and MERRA in comparison to those in NCEP-1 and NCEP-2. The strong negative trends of Ψ almost control all levels of the eastern Pacific in 20CR, ERAIM, JRA-25, JRA-55, and MERRA, with the maximum center located in the middle troposphere (~500 hPa). In NCEP-1 and NCEP-2, the strongest negative trends dominate the low levels of the eastern Pacific and extend westward as far as the international date line, accompanied by a significant positive trend in the upper troposphere. A similar discrepancy of NCEP-1 and NCEP-2 with other reanlyses is also found for the Hadley circulation (Mitas and Clement 2005).
Linear trends of the annual zonal mass streamfunction (shading; 1011 kg s−1 decade−1) along the equatorial Pacific (5°S–5°N) from 1979 to 2012 in (a) 20CR, (b) ERAIM, (c) JRA-25, (d) JRA-55, (e) MERRA, (f) NCEP-1, (g) NCEP-2, and (h) multiple reanalysis ensemble mean, with trends that are statistically significant at the 5% level dotted. Contours denote the long-term mean of the zonal mass streamfunction.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
From above analyses of the changes of the zonal mass streamfunction, the strengthening and westward shift of the PWC structure on the zonal vertical cross section are robust among seven reanalysis products. Then what are the characteristics of the changes from the PWC on the ground surface? How is their significance and robustness between different reanalysis products? Are the PWC changing characteristics derived from reanalysis datasets coherent with other independent climate variables? To answer these questions, we examine the change of SLP and surface winds over the tropical Pacific in seven reanalyses in Fig. 3. The linear trends of the observational SLP and precipitation are also shown in Fig. 3. In the observations, SLP decreases significantly over the tropical Indian Ocean and western Pacific and increases significantly over the tropical central and eastern Pacific, while precipitation shows a moistening tendency over the central Indian Ocean and western Pacific and a drying tendency over the central Pacific (Fig. 3a), strongly supporting the intensification and westward shift of the PWC indicated by reanalysis data. Seven reanalyses show similar SLP trend patterns to those of the observations. Consistent with the significant decreasing of SLP over the western Pacific and increasing over the eastern Pacific, the enhancement of the tropical Pacific trade wind is evident in all seven reanalyses (Figs. 3b–h).
Linear trends of SLP (shading; hPa decade−1) and surface winds (vectors; m s−1 decade−1) from 1979 to 2012 in (a) HadSLP2, (b) 20CR, (c) ERAIM, (d) JRA-25, (e) JRA-55, (f) MERRA, (g) NCEP-1, and (h) NCEP-2. The black dotted areas indicate that the SLP trends are statistically significant at the 5% level. Vectors are plotted only for regions with surface zonal wind trends that are statistically significant at the 5% level. Contours in (a) represent the positive (green), zero (orange), and negative (purple) trend of observed precipitation (mm day−1 decade−1).
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
To get a clearer picture of the changes in the PWC structure, the long-term trends and long-term mean of the mass-weighted vertically averaged zonal streamfunction, vertical wind at 500 hPa, and surface zonal wind over the equatorial Pacific are shown in Fig. 4. No matter the vertical average of the zonal mass exchange flux Ψ or for 500-hPa vertical wind and surface zonal wind, in all reanalyses, the shapes of the linear trends are similar with respect to the long-term mean. Compared with the climatological zonal distribution, centers of the trends shift westward. For instance, in terms of the climate mean state, both the maximum center of the vertically averaged zonal streamfunction and the strongest surface zonal winds are located in regions of the central Pacific around 150°W (Figs. 4a,c), and the strongest ascending motions are concentrated on the western Pacific (Fig. 4b). In terms of long-term change, the largest increasing of the vertically averaged Ψ and strongest strengthening of surface zonal winds in the reanalyses all occur in regions west of the international date line, indicating a clear picture of the intensification and westward shift of the PWC.
Long-term mean (dashed lines) and linear trends (solid lines) during 1979–2012 of (a) vertical average of zonal mass streamfunction, (b) pressure velocity at 500 hPa, and (c) surface zonal wind along the equatorial Pacific (averaged for 5°S–5°N) derived from MRE. Light pink (blue) shading denotes the range of the linear trend (mean state) in data from seven reanalyses. The left (right) axis is the linear trend (mean state).
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
Considering the PWC is a zonal asymmetric tropical Pacific circulation and is driven by the SST difference along the equatorial Pacific, created by the continental interruption of major oceanic circulations over the Maritime Continent (Bjerknes 1969; Philander 1990), and as a consequence of the seasonal variation in the energy exchange between the atmosphere and ocean, PWC has a seasonal cycle (Yu et al. 2012; Schwendike et al. 2014); thus, we further examine the seasonality of PWC and its corresponding changes. As shown in Fig. 5 for the climatology, the PWC strength indicated by the maximum of the vertically averaged zonal mass flux is relatively stronger in boreal summer and winter than those in spring and fall, and the corresponding maximum center migrates zonally during the course of the year. The PWC western edge indicated by the zero line of the vertically averaged zonal mass flux over the western equatorial Pacific exhibits an obvious seasonal migration with the shifting from west of 150°E in boreal summer [June–August (JJA)] to the east of 160°E in winter [December–February (DJF)]. For the long-term change during 1979–2012, the strengthening and westward shifting of the PWC is found in all seasons. So the robust strengthening and westward shifting of the PWC in the recent three decades is not seasonally dependent.
Long-term mean (contours) and linear trends (shading) during 1979–2012 of the monthly means of vertically averaged zonal mass flux over the equatorial Pacific plane in (a) 20CR, (b) ERAIM, (c) JRA-25, (d) JRA-55, (e) MERRA, (f) NCEP-1, (g) NCEP-2, and (h) multiple reanalysis ensemble mean. Trends statistically significant at the 5% level are dotted in green.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
To quantitatively measure the changes of PWC, the time series of the smoothed annual PWC intensity indices based on the zonal streamfunction, SLP gradient, and the equatorial Pacific trade, after using a 3-yr running mean, are displayed in Fig. 6. All three PWC intensity indices show a high degree of consistency in the multiyear variations and a similarly intensifying trend among seven reanalysis datasets. The interannual and multiyear variation of PWC is dominated by ENSO, as evidenced by close coherent variation (Table 3). We further compare the PWC western edge index derived from different reanalysis data in Fig. 6d. A coherent variation among different reanalysis data, and between the western edge index and Niño-3.4 index, is also evident. The PWC intensity indices are significantly and negatively correlated to the Niño-3.4 index with an absolute correlation coefficient |r| > 0.81, while the PWC western edge index is positively correlated to the Niño-3.4 index with a correlation coefficient of up to 0.81 for the multiple reanalysis ensemble mean (Table 3). Thus, the PWC exhibits a weakening (strengthening) and an eastward (westward) shift during El Niño (La Niña) events, accompanied by the corresponding eastward (westward) shift of the centers of large-scale convection and precipitation (Philander 1990). All reanalysis data are highly consistent in revealing the above relationship.
Time series (1979–2012) of the smoothed annual anomalies of the tropical PWC intensity indices based on (a) STRF (1011 kg s−1), (b) dSLP (hPa), (c) Us (easterly is positive; m s−1), and (d) western edge (° lon), derived by applying 3-yr running mean to the annual anomalies. Different colored lines represent different reanalysis data denoted by the colored character strings at the top of (a).
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
Correlation coefficients of the annual PWC intensity and western edge indices with the Niño-3.4 index computed using the observational dataset ERSST.v3b. All correlations are statistically significant at the 1% level.
Time series of the PWC intensity indices and western edge (Fig. 6) show an intensification and westward shift of the PWC cell since 1979. The quantitative measure of these changes is shown in Fig. 7 and Table 4. For STRF defined by the zonal mass streamfunction, all reanalyses except for NCEP-2 exhibit a strengthening tendency at a rate of larger than 0.28 × 1011 kg s−1 decade−1 (or 13.45% decade−1), and all are statistically significant at the 1% level. In NCEP-2, STRF also shows an increasing trend of 0.14 × 1011 kg s−1 decade−1 (around 8% decade−1) but is not statistically significant at the 10% level. dSLP shows a consistent increasing trend among seven reanalyses, with the largest trend of 0.45 hPa decade−1 in NCEP-1 and the smallest trend of 0.15 hPa decade−1 in MERRA. All reanalyses show a significant enhancement of trade wind with a trend range from 0.27 m s−1 decade −1 in NCEP-2 to 0.59 m s−1 decade −1 in 20CR.
Linear trends of three PWC intensity indices and the western edge during 1979–2012. (a) STRF (1011 kg s−1 decade−1), (b) dSLP (hPa decade−1), (c) Us × −1 (m s−1 decade−1), and (d) western edge (° lon decade−1). The trends that passed the 99% (95%) statistical confidence level are filled with black dots (horizontal lines). Bars filled with vertical lines indicate that the trends are not statistically significant at the 5% level.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
The linear trends of the PWC indices during 1979–2012 in seven reanalysis datasets: STRF (1011 kg s−1 decade−1), dSLP (hPa decade −1), Us × −1 (m s−1 decade −1), and western edge (° lon decade −1). Trends that are statistically significant at the 5% are in italic and 1% levels are in boldface. The numbers in parentheses represent the rate of the linear trend in percentage (% decade −1).
The westward shift of PWC is evident in all reanalysis datasets (Fig. 7d), with a trend for the western edge of 4.68°, 4.48°, and 3.58° longitude decade−1 in JRA-55, NCEP-1, and NCEP-2, respectively; all are statistically significant at the 1% level. MERRA and 20CR also show a westward shift trend of 3.58° and 3.08° longitude decade−1, which is statistically significant at the 5% and 10% levels, respectively. The westward shift trends in ERAIM and JRA-25 are smaller than in the other five reanalyses and are not statistically significant at the 10% level.
c. Changes of precipitation and surface temperature associated with the PWC change
To reveal the precipitation and temperature changes associated with the intensification and westward shift of the PWC, following Thompson et al. (2000) and Zhou et al. (2008), the observed trends of precipitation and surface temperature are partitioned into linearly congruent and linearly independent components with respect to the PWC intensity and western edge index. The component of the precipitation trend that is linearly congruent with the PWC intensity (western edge) is estimated at each grid point by regressing the precipitation time series onto the time series of the PWC intensity (western edge) index and then multiplying the linear trends of the PWC intensity (western edge) index. As in Zhou et al. (2008), an f test is used to test the statistical significance of the linearly congruent component.
The linear trends of precipitation and surface temperature during 1979–2012, and the corresponding components congruent with the PWC intensity and western edge are shown in Fig. 8. In precipitation and temperature fields, PWC intensity and western-edge-congruent components highly resemble that of total trends. The impacts of the PWC intensity and western edge changes on the surface climate are similar. A wetter Indo-Pacific warm pool and drier central and eastern tropical Pacific stand out; a drying northern Atlantic and moistening equatorial Atlantic is also evident. The PWC changes are associated with a cooling central and eastern Pacific but warming western Pacific. Compared to the westward shift of PWC, the strengthening of PWC has stronger impacts on changes of precipitation and surface temperature. For instance, the moistening of the Indo-Pacific warm pool and drying of the central and eastern tropical Pacific in intensity-congruent components are stronger than that in western-edge-congruent components. The cooling in the central and eastern Pacific in intensity-congruent components and the warming in the western Pacific are also stronger than that in the western-edge-congruent components. Because the chosen linear analysis method is deficient over the eastern equatorial Pacific, the cooling and drying over the eastern equatorial Pacific are statistically significant in the linearly congruent component, but the results based on the total change are not statistically significant.
Linear trends in (left) precipitation (mm day−1 decade−1) and (right) surface temperature (K decade−1) during 1979–2012 of (a),(b) annual trends, (c),(d) the components linearly congruent with the multiple reanalysis mean STRF, and (e),(f) components linearly congruent with the western edge. Areas exceeding the confidence limit of 5% using an f test are dotted.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
4. CMIP5 simulations
In the observations, during 1979–2012, the SST trend pattern displays an obviously eastern Pacific cooling and a significant warming over the western Pacific (Fig. 9a), with a clear strengthening of the tropical Pacific SST gradient. Meanwhile, the tropical Atlantic and Indian Ocean also show a significant warming but with relatively smaller magnitudes. The La Niña–like SST change pattern is physically consistent with the changes of SLP and surface wind displayed in Fig. 3. The PWC is thought to be a thermally divergent zonal atmospheric circulation over the tropical Pacific the equatorial Pacific SST gradient is one dominant driving force of the PWC strength (Philander 1990). For the long-term change of PWC, an increased (decreased) equatorial Pacific zonal SST gradient is key to the enhancement (reduction) of the PWC (Liu et al. 2013; Meng et al. 2012; Tokinaga et al. 2012a; Sandeep et al. 2014). Meng et al. (2012) used the history of observed SST, which exhibits a zonally asymmetric evolution with La Niña–like pattern since 1870 over the equatorial Indo-Pacific sector, to force an AGCM. They suggested that the Pacific SST gradient drove the PWC variability, and the La Niña–like SST change induced the strengthening of the PWC over the twentieth century. Based on the results of a suite of SST-forced AGCM experiments conducted by Tokinaga et al. (2012a,b), it was suggested that the SST warming pattern, resembling El Niño with a reduced zonal gradient in the tropical Indo-Pacific sector, is the main cause of the weakened PWC over the past six decades (1950–2009). To further confirm the driving effect of the observational La Niña–like SST change pattern with increased zonal gradient in the tropical Indo-Pacific sector on recent PWC strengthening and westward shift, we examine the changes of PWC in CMIP5 simulations, which were performed using the monthly mean observed SST to force the atmosphere models.
(a) Linear trends of SST (shading; K decade−1) during 1979–2012, derived from ERSST.v3b. The black dotted areas indicate that the corresponding trends are statistically significant at the 5% level. (b) The climatological distribution of the zonal mass steam function (1011 kg s−1) along the equatorial Pacific in the CMIP5 MME. (c) Taylor diagram illustrating the relative mean squared difference (radial coordinate, the ratio of spatial standard deviation of the individual reanalysis dataset and models against the MRE) and the spatial pattern correlation between model and MRE (angular coordinate). Different letters and numbers denote different reanalyses and models, respectively.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
Before examining the changes of the PWC in CMIP5 simulations, we first examine the performance of CMIP5 AGCMs in reproducing the climate mean state and interannual variability of PWC. The climatological vertical structures (Fig. 9b) of the tropical PWC in the CMIP5 multimodel ensemble mean (MME) are similar to those of the reanalysis dataset (Fig. 1). All CMIP5 models reasonably reproduced the mean state of the PWC structure, as evidenced by the spatial pattern correlation coefficients (PCCs) of Ψ between models and MRE larger than 0.88 and the relative squared difference smaller than 1.25 (Fig. 9c). MME shows better performance in reproducing the PWC mean state with highest PCC with MRE and smaller relative squared difference compared to most models.
The time series of the PWC intensity and western edge index derived from CMIP5 models are compared to MRE, which is regarded as the observed PWC intensity and western edge index, in Figs. 10a and 10b. The interannual variation of the PWC intensity and western edge are well reproduced by CMIP5 models (Figs. 10c and 10d), with the correlation coefficients of the PWC index between CMIP5 models and MRE larger than 0.87 and 0.7 for intensity and western edge index, respectively; all are statistically significant at the 0.1% level. Meanwhile, MME shows the best PWC interannual variation, with higher r than any individual model. Overall, CMIP5 models capture the mean state and interannual variation of the PWC reasonably well. Thus, simulations of CMIP5 models can be used to further study the changes of PWC.
The temporal evolution of PWC (a) intensity (1011 kg s−1) and (b) western edge (° lon) index in CMIP5 MME (red line) and range of individual model (pink shading). The MRE PWC intensity and western edge indices are also shown in (a) and (b), respectively. Correlation coefficients of PWC (c) intensity and (d) western edge index between CMIP5 models and MRE. Letters and numbers denote different reanalyses and models, respectively, as shown in Fig. 9c.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
In response to the observed La Niña–like SST forcing, a strengthening and westward shift trend in PWC is apparent in trends of both Ψ and SLP (Figs. 11a,b), with decreasing SLP over the Maritime Continent and extending westward to the Atlantic and prominent increasing SLP over the eastern tropical Pacific. The positive trends of Ψ extend eastward to the central Pacific from the Maritime Continent (nearly from 135°E to 135°W), and negative trends dominate the zonal mass flux change over the eastern Pacific. Physically consistent with changes in SLP and Ψ, surface easterlies enhance significantly over the central Pacific, while precipitation increases over the equatorial Indian Ocean and the Maritime Continent and decreases over the tropical central Pacific.
Linear trends (shading) during 1979–2008 of (a) zonal mass streamfunction along the equatorial Pacific (1011 kg s−1 decade−1; long-term mean is overlaid using contours), and (b) SLP (hPa decade−1) in CMIP5 MME; trends that are statistically significant at the 5% level are dotted. Precipitation trends are overlaid by contours in (b) for positive (green), zero (orange), and negative (purple) trends. In (b), 1000-hPa horizontal wind trends (vectors) that are statistically significant at the 5% level are also shown. (c) Scatterplot of linear trends of the western edge and intensity of the tropical PWC. The black solid line in (c) denotes the least-squares linear fit of the trends based on reanalysis data and CMIP5 simulations. Numbers denote different models, as shown in Fig. 9c.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
The AMIP experiments from CMIP5 reproduce the PWC structure changes derived from reanalysis datasets and independent observations quite well, indicative of a driver of the recent La Niña–like SST change pattern to the strengthening and westward shift of PWC. The driving forcing of the recent La Niña–like SST change on the strengthened PWC is supported by the evidence provided by previous studies, which suggested the current global-warming hiatus is tied specially to the equatorial Pacific La Niña–like decadal cooling through intensifying PWC (Meehl et al. 2011, 2013; Kosaka and Xie 2013; England et al. 2014). But most CMIP5 simulations have a less westward shift and a weaker intensification of PWC than that in all reanalyses except for NCEP-1 (Fig. 11c). This difference might be due to neglecting air–sea feedback (Zhou et al. 2008) and deficiencies in the downward mixing of momentum through the boundary in AGCM (McGregor et al. 2014). The coherent changes of intensified and westward shift of PWC seen in the reanalysis dataset are also evident in CMIP5 models. As shown in Fig. 11c, the PWC intensifying trends are negatively correlated with the PWC westward shift trends, with the correlation coefficient up to −0.52, which is statistically significant at the 1% level.
5. Summary and discussion
a. Summary
In this study, we analyzed the recent trend of the PWC structure using multiple reanalysis datasets and observations of climate variables associated with PWC, based on multi-PWC indices. At the same time, to advance our understanding of SST forcing to the recent change of PWC during 1979–2012, we examined the PWC change simulated by AMIP experiments of 26 CMIP5 atmospheric models. The main findings are summarized as follows.
The zonal mass streamfunction intuitionally depicts the zonal–vertical structure of the enclosed PWC cell very well, with ascending motion over the Maritime Continent and tropical western Pacific, descending motion over the eastern Pacific, westerlies in the upper troposphere, and surface easterlies. The equatorial Pacific zonal mass streamfunction produces an extremely similar mean state PWC structure among seven sets of reanalysis products, with the pattern correlation coefficient between individual reanalysis dataset and multiple reanalysis ensemble mean being larger than 0.95. During the recent three decades (1979–2012), the linear trends of Ψ along the equatorial Pacific derived from seven sets of reanalysis data all exhibited a significant strengthening and westward shifting trends of the PWC. All reanalyses exhibited significant increasing trend of SLP over the western Pacific and decreasing trend of SLP over the eastern Pacific and thus enhanced trade winds, which is physically consistent with the change of the zonal mass streamfunction. The observed SLP trends resemble that derived from the reanalyses, providing more evidence of PWC strengthening and westward shift.
Quantitatively, the PWC intensity and western edge defined using the zonal mass streamfunction show a robust strengthening and westward shift trend among different reanalysis datasets over the past three decades, with a significant trend of 15.08% decade−1 and 3.70° longitude decade−1, respectively, in the ensemble mean of the seven reanalysis datasets, and with the strongest (weakest) intensification of 0.38 × 1011 kg s−1 decade−1 (0.14 × 1011 kg s−1 decade−1) in 20CR (NCEP-2) and the largest (smallest) westward shift −4.68° longitude decade−1 (−2.55° longitude decade−1) in JRA-55 (JRA-25). The PWC intensity index based on the SLP gradient and Pacific trade winds also revealed a robust PWC enhancement in all reanalysis datasets. The recent PWC intensification and westward shift contribute greatly to the observed moistening over the Indo-Pacific warm pool and drying and cooling over the central and eastern tropical Pacific.
The CMIP5 models reasonably reproduced the mean state and interannual variation of PWC quite well. More important, in response to the recently observed anomalous SST forcing, which resembles a La Niña–like state, the CMIP5 models reasonably reproduced the recently observed PWC strengthening and westward shift trends, indicating dominant forcing of the La Niña–like SST anomalies to the observed PWC change.
b. Discussion
Based on multiple reanalysis datasets and 26 CMIP5 simulations, it is found that there is a robust strengthening and westward shift of the PWC during 1979–2012, which is different from the results revealed in some previous studies (Knutson and Manabe 1995; Held and Soden 2006; Vecchi et al. 2006; Power and Smith 2007; Power and Kociuba 2011) that have stated an observed and simulated twentieth-century weakening of PWC as a result of global warming. One of the possible reasons for this difference is the PWC change during 1979–2012 has a lower signal-to-noise ratio due to the shorter time period and a stronger influence of natural variability such as the Pacific decadal oscillation (PDO) or interdecadal Pacific oscillation (IPO). Both global warming induced by increasing greenhouse gas forcing and internally generated variability [such as a mega–El Niño–Southern Oscillation (ENSO), PDO, or IPO] contributed to the observed weakening of the twentieth-century PWC (Collins et al. 2010; Power and Smith 2007; Power and Kociuba 2011; Dong and Lu 2013; Wang et al. 2013; Bayr et al. 2014; Sandeep et al. 2014). As pointed out in Vecchi et al. (2006), the strong decadal variability complicates the detection of a relatively small forced PWC change even in multidecadal records; to detect the forced change of the PWC, a record longer than 100–120 yr is required. It is also suggested that, for the detection periods shorter than 100 yr, internal variability is likely to play a more important role in the changes of PWC (DiNezio et al. 2013; Ma and Zhou 2014; Sohn et al. 2013; Wang et al. 2013). DelSole et al. (2011) also found that trends of SST have very high confidence intervals when evaluated over 16- or 32-yr periods, making the detection of trends above noise in climate time series very difficult. A similar conclusion was drawn by Harrison and Chiodi (2015): because of the strong multidecadal variability in ENSO superimposed on the century-scale variability, the centennial-scale (or longer) trend associated with ENSO cannot be distinguished from zero with 95% confidence. Thus, the recent three decades (1979–2012) may not be long enough to isolate the weakening trend of the PWC induced by global warming from the strong internal variability.
Perhaps because of the low signal-to-noise ratio, the recently robust strengthening and westward shift of PWC deduced from the 1979–2012 period are inconsistent with some previous studies that revealed a weakening PWC during the twentieth century (Vecchi et al. 2006; Power and Smith 2007; Power and Kociuba 2011). Insofar as surface pressure and other surface observations can sufficiently constrain the three-dimensional circulation [as has been argued by Compo et al. (2011)], the 20CR made use of only surface observations, which is a relatively homogeneous set of observations in the reanalysis that extends back to 1900; thus, in order to determine how much confidence can be placed in the PWC changes deduced from the 1979–2012 period, the 20CR was used to extend the PWC analysis farther back in time. The changes of PWC characteristics during 1900–2012 derived from the 20CR dataset are shown in Fig. 12. A significant strengthening and westward shift of the twentieth-century PWC is revealed, characterized by significantly positive trends of zonal mass streamfunction Ψ over the western Pacific and significantly negative trends of Ψ over the eastern Pacific (Fig. 12a), accompanied by the increasing (decreasing) SLP over the central and eastern Pacific (Indo-Pacific) and the enhanced trade winds over the western Pacific, physically consistent with the La Niña–like SST change (Fig. 12b). In terms of the time evolution of PWC indices (Fig. 12c), STRFs, dSLP, and surface zonal wind (Us) multiplied by −1 show a significantly increasing trend of 0.68 × 1011 kg s−1 century−1, 1.26 hPa century −1, and 1.34 m s−1 century −1, respectively, and the western edge also shows a westward shift with a trend of 6.38° longitude century −1, consistent with the results of Meng et al. (2012), who also suggested a strengthened PWC during the twentieth century.
Changes of PWC during 1900–2012 derived from the 20CR. Linear trends of (a) zonal mass streamfunction along the equatorial Pacific (shading, 1011 kg s−1 century−1; long-term mean is overlaid using contours), (b) SLP [contours, hPa century−1; contours are for positive (green), zero (orange), and negative (purple) trends] and surface winds (vectors, m s−1 century−1). Trends of SST derived from HadISST are also shown in (b) using shading, and trends that are statistically significant at the 5% level are dotted. (c) Time series of 11-yr smoothed standardized anomalies of annual PWC indices defined as STRF (black line), western edge (red), dSLP (green), and Us (blue), respectively.
Citation: Journal of Climate 29, 9; 10.1175/JCLI-D-15-0398.1
In addition, there is still considerable uncertainty concerning the twentieth-century trends of PWC. The observed weakened PWC over the twentieth century is based on the observations and reconstructions of SLP (Vecchi et al. 2006; Power and Smith 2007). However, the bias in the SLP observations and reconstructions is large before the 1950s because of sparse in situ measurements (DiNezio et al. 2013; L’Heureux et al. 2013). Multidecadal trends evaluated since 1950 reveal a significant strengthening of the PWC (L’Heureux et al. 2013). SST trend patterns are the key drivers of the PWC change (DiNezio et al. 2010; Meng et al. 2012; Tokinaga et al. 2012a), but there are large discrepancies in the SST trends during the twentieth century over the Indo-Pacific in the different observational datasets (Deser et al. 2010; Solomon and Newman 2012; Tokinaga et al. 2012b) leading to the change of the twentieth-century PWC in response to global warming that is actually uncertain.
For the results of climate models, the simulated SST responses to global warming are different among models: although a number of coupled models simulate an El Niño–like SST response (Vecchi and Soden 2007; Yu and Boer 2002; DiNezio et al. 2009), some models exhibit a La Niña–like pattern (Clement et al. 1996; DiNezio et al. 2010), and a number of models show a more or less homogeneous warming in the tropics with little of either pattern (DiNezio et al. 2009; Collins et al. 2010), resulting in a large spread of the twentieth-century PWC change (Power and Kociuba 2011; DiNezio et al. 2013). The twentieth-century trends in the PWC have been hard to characterize from observations and model simulations.
The zonal thermal contrast over the tropical Pacific is one direct driving force of the zonally asymmetric atmospheric overturning circulation PWC. CMIP5 simulations forced by the observed SST with a change pattern resembling La Niña reproduced reasonably the recent robust strengthening and westward shifting of the PWC revealed by multiple reanalyses and independent observations. This strongly suggests that the recently intensified and westward-shifting PWC arise from the increased zonal Pacific SST gradient induced by the La Niña–like SST change. However, whether the recent La Niña–like decadal SST trend is internal or forced is still unclear. There are some studies that have concluded the recent cooling of the tropical Pacific is probably due to natural internal variability rather than a forced response (Kosaka and Xie 2013). As a result of frequent occurrences of central Pacific-type El Niño in recent decades, the eastern tropical Pacific and western Pacific have undergone cooling and warming, respectively (Sohn et al. 2013). The shift of IPO or PDO from its positive phase to negative phase and a mega–El Niño–Southern Oscillation probably contributed the eastern tropical Pacific cooling (Meehl et al. 2013; Wang et al. 2013; England et al. 2014; Thompson et al. 2015). There are other studies that have suggested forcing from global warming may contribute to the recent La Niña–like SST trend pattern resulting from ocean dynamics (Li and Ren 2012). Dong and Zhou (2014) suggested that the recent SST cooling in the eastern tropical Pacific has resulted from a competition between the global warming mode, IPO mode, and Atlantic multidecadal oscillation (AMO) mode.
Meanwhile, Indian Ocean and Atlantic warming could play an important role in modulating the Pacific climate change. For example, Luo et al. (2012) point out that the enhanced tropical Indian Ocean warming in recent decades is likely to have contributed to the La Niña–like state through the Pacific Ocean–atmosphere interactions and favors a stronger PWC. In addition, the AMO has had a substantial influence on the changes of the Pacific climate (Wang et al. 2013; McGregor et al. 2014). The stronger warming of the tropical Atlantic Ocean compared to the global SST mean is associated with a La Niña–like mean state change in the tropical Pacific (Kucharski et al. 2016). The Atlantic warming can modify the atmospheric circulation through an atmospheric bridge mechanism, leading to a high pressure anomaly in the central Pacific. The easterly wind anomalies associated with this high pressure anomaly to the west then trigger oceanic Kelvin waves that travel to the east and lead eventually to La Niña–type conditions in the tropical Pacific region (Kucharski et al. 2015). Therefore, recent Indian Ocean and Atlantic warming may further indirectly promote the strengthening and westward shifting of the PWC through modulating the Pacific climate. The contribution percentage of individual basin SST variability to the recent intensification and westward shifting of PWC is a very interesting scientific question and needs to be studied in the future.
Acknowledgments
This work is supported by the National Natural Science Foundation of China (Grants 41330423 and 41420104006) and the R&D Special Fund for Public Welfare Industry (meteorology) (GYHY201506012).
REFERENCES
Allan, R., and T. Ansell, 2006: A new globally complete monthly historical gridded mean sea level pressure dataset (HadSLP2): 1850–2004. J. Climate, 19, 5816–5842, doi:10.1175/JCLI3937.1.
Annamalai, H., J. M. Slingo, K. R. Sperber, and K. Hodges, 1999: The mean evolution and variability of the Asian summer monsoon: Comparison of ECMWF and NCEP–NCAR reanalyses. Mon. Wea. Rev., 127, 1157–1186, doi:10.1175/1520-0493(1999)127<1157:TMEAVO>2.0.CO;2.
Bayr, T., D. Dommenget, T. Martin, and S. B. Power, 2014: The eastward shift of the Walker Circulation in response to global warming and its relationship to ENSO variability. Climate Dyn., 43, 2747–2763, doi:10.1007/s00382-014-2091-y.
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.
Bretherton, C. S., M. Widmann, V. P. Dymnikov, J. M. Wallace, and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990–2009, doi:10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.
Cane, M. A., A. C. Clement, A. Kaplan, Y. Kushnir, D. Pozdnyakon, R. Seager, S. E. Zebiak, and R. Murtugudde, 1997: Twentieth-century sea surface temperature trends. Science, 275, 957–960, doi:10.1126/science.275.5302.957.
Chen, J., A. D. Del Genio, B. E. Carlson, and M. G. Bosilovich, 2008: The spatiotemporal structure of twentieth-century climate variations in observations and reanalyses. Part II: Pacific pan-decadal variability. J. Climate, 21, 2634–2650, doi:10.1175/2007JCLI2012.1.
Clement, A. C., R. Seager, M. A. Cane, and S. E. Zebiak, 1996: An ocean dynamical thermostat. J. Climate, 9, 2190–2196, doi:10.1175/1520-0442(1996)009<2190:AODT>2.0.CO;2.
Collins, M., and Coauthors, 2010: The impact of global warming on the tropical Pacific Ocean and El Niño. Nat. Geosci., 3, 391–397, doi:10.1038/ngeo868.
Compo, G. P., and Coauthors, 2011: The Twentieth Century Reanalysis project. Quart. J. Roy. Meteor. Soc., 137, 1–28, doi:10.1002/qj.776.
de Boisséson, E., M. A. Balmaseda, S. Abdalla, E. Källén, and P. A. E. M. Janssen, 2014: How robust is the recent strengthening of the tropical Pacific trade winds? Geophys. Res. Lett., 41, 4398–4405, doi:10.1002/2014GL060257.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
DelSole, T., M. K. Tippett, and J. Shukla, 2011: A significant component of unforced multidecadal variability in the recent acceleration of global warming. J. Climate, 24, 909–926, doi:10.1175/2010JCLI3659.1.
Deser, C., and J. M. Wallace, 1990: Large-scale atmospheric circulation features of warm and cold episodes in the tropical Pacific. J. Climate, 3, 1254–1281, doi:10.1175/1520-0442(1990)003<1254:LSACFO>2.0.CO;2.
Deser, C., A. S. Phillip, and M. A. Alexander, 2010: Twentieth century tropical sea surface temperature trends revisited. Geophys. Res. Lett., 37, L10701, doi:10.1029/2010GL043321.
DiNezio, P. N., A. C. Clement, G. A. Vecchi, B. J. Solden, B. P. Kirtman, and S. K. Lee, 2009: Climate response of the equatorial Pacific to global warming. J. Climate, 22, 4873–4892, doi:10.1175/2009JCLI2982.1.
DiNezio, P. N., A. C. Clement, and G. A. Vecchi, 2010: Reconciling differing views of tropical Pacific climate change. Eos, Trans. Amer. Geophys. Union, 91, 141–142, doi:10.1029/2010EO160001.
DiNezio, P. N., G. A. Vecchi, and A. C. Clement, 2013: Detectability of changes in the Walker circulation in response to global warming. J. Climate, 26, 4038–4048, doi:10.1175/JCLI-D-12-00531.1.
Dong, B. W., and R. Y. Lu, 2013: Interdecadal enhancement of the Walker circulation over the tropical Pacific in the late 1990s. Adv. Atmos. Sci., 30, 247–262, doi:10.1007/s00376-012-2069-9.
Dong, L., and T. Zhou, 2014: The formation of the recent cooling in the eastern tropical Pacific Ocean and the associated climate impacts: A competition of global warming, IPO, and AMO. J. Geophys. Res., 119, 11 272–11 287, doi:10.1002/2013JD021395.
Ebita, A., and Coauthors, 2011: The Japanese 55-year Reanalysis “JRA-55”: An interim report. SOLA, 7, 149–152, doi:10.2151/sola.2011-038.
England, M. H., and Coauthors, 2014: Recent intensification of wind-driven circulation in the Pacific and the ongoing warming hiatus. Nat. Climate Change, 4, 222–227, doi:10.1038/nclimate2106.
Feng, L., and T. Zhou, 2012: Water vapor transport for summer precipitation over the Tibetan Plateau: Multidata set analysis. J. Geophys. Res., 117, D20114, doi:10.1029/2011JD017012.
Feng, M., C. Böning, A. Biastoch, E. Behrens, E. Weller, and Y. Masumoto, 2011: The reversal of the multi‐decadal trends of the equatorial Pacific easterly winds, and the Indonesian Throughflow and Leeuwin Current transports. Geophys. Res. Lett., 38, L11604, doi:10.1029/2011GL047291.
Garcia, S. R., and M. T. Kayano, 2008: Climatological aspects of Hadley, Walker and monsoon circulations in two phases of the Pacific Decadal Oscillation. Theor. Appl. Climatol., 91, 117–127, doi:10.1007/s00704-007-0301-9.
Gastineau, G., L. Li, and H. L. Treut, 2009: The Hadley and Walker circulation changes in global warming conditions described by idealized atmospheric simulations. J. Climate, 22, 3993–4013, doi:10.1175/2009JCLI2794.1.
Hansen, J., R. Ruedy, M. Sato, and K. Lo, 2010: Global surface temperature change. Rev. Geophys., 48, RG4004, doi:10.1029/2010RG000345.
Harrison, D. E., and A. M. Chiodi, 2015: Multi-decadal variability and trends in the El Niño–Southern Oscillation and tropical Pacific fisheries implications. Deep-Sea Res. II, 113, 9–21, doi:10.1016/j.dsr2.2013.12.020.
Held, I. M., and B. J. Soden, 2006: Robust responses of the hydrological cycle to global warming. J. Climate, 19, 5686–5699, doi:10.1175/JCLI3990.1.
Huffman, G. J., R. F. Adler, D. T. Bolvin, and G. Gu, 2009: Improving the global precipitation record: GPCP version 2.1. Geophys. Res. Lett., 36, L17808, doi:10.1029/2009GL040000.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Kanamitsu, M., W. Ebisuzaki, J. Woollen, S.-K. Yang, J. J. Hnilo, M. Fiorino, and G. L. Potter, 2002: NCEP–DOE AMIP-II Reanalysis (R-2). Bull. Amer. Meteor. Soc., 83, 1631–1643, doi:10.1175/BAMS-83-11-1631.
Kinter, J. L., III, M. J. Fennessy, V. Krishnamurthy, and L. Marx, 2004: An evaluation of the apparent interdecadal shift in the tropical divergent circulation in the NCEP–NCAR reanalysis. J. Climate, 17, 349–361, doi:10.1175/1520-0442(2004)017<0349:AEOTAI>2.0.CO;2.
Knutson, T. R., and S. Manabe, 1995: Time-mean response over the tropical Pacific to increase CO2 in a coupled ocean–atmosphere model. J. Climate, 8, 2181–2199, doi:10.1175/1520-0442(1995)008<2181:TMROTT>2.0.CO;2.
Kociuba, G., and S. B. Power, 2015: Inability of CMIP5 models to simulate recent strengthening of the Walker circulation: Implications for projections. J. Climate, 28, 20–35, doi:10.1175/JCLI-D-13-00752.1.
Kosaka, Y., and S. P. Xie, 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403–407, doi:10.1038/nature12534.
Kucharski, F., F. S. Syed, A. Burhan, I. Farah, and A. Gohar, 2015: Tropical Atlantic influence on Pacific variability and mean state in the twentieth century in observations and CMIP5. Climate Dyn., 44, 881–896, doi:10.1007/s00382-014-2228-z.
Kucharski, F., and Coauthors, 2016: Atlantic forcing of Pacific decadal variability. Climate Dyn., 46, 2337–2351, doi:10.1007/s00382-015-2705-z.
Kumar, K. K., B. Rajagopalan, and M. A. Cane, 1999: On the weakening relationship between the Indian monsoon and ENSO. Science, 284, 2156–2159, doi:10.1126/science.284.5423.2156.
L’Heureux, M. L., S. Lee, and B. Lyon, 2013: Recent multidecadal strengthening of the Walker circulation across the tropical Pacific. Nat. Climate Change, 3, 571–576, doi:10.1038/nclimate1840.
Li, G., and B. Ren, 2012: Evidence for strengthening of the tropical Pacific Ocean surface wind speed during 1979–2001. Theor. Appl. Climatol., 107, 59–72, doi:10.1007/s00704-011-0463-3.
Lin, R., T. Zhou, and Q. Yun, 2014: Evaluation of global monsoon precipitation changes based on five reanalysis datasets. J. Climate, 27, 1271–1289, doi:10.1175/JCLI-D-13-00215.1.
Liu, J., B. Wang, M. A. Cane, S. Y. Yim, and J. Y. Lee, 2013: Divergent global precipitation changes induced by natural versus anthropogenic forcing. Nature, 493, 656–659, doi:10.1038/nature11784.
Luo, J. J., W. Sasaki, and Y. Masumoto, 2012: Indian Ocean warming modulates Pacific climate change. Proc. Natl. Acad. Sci. USA, 109, 18 701–18 706, doi:10.1073/pnas.1210239109.
Ma, S., and T. Zhou, 2014: Changes of the tropical Pacific Walker circulation simulated by two versions of FGOALS model. Sci. China Earth Sci., 57, 2165–2180, doi:10.1007/s11430-014-4902-8.
McGregor, S., A. Timmermann, M. F. Stuecker, M. H. England, M. Merrifield, F.-F. Jin, and Y. Chikamoto, 2014: Recent Walker circulation strengthening and Pacific cooling amplified by Atlantic warming. Nat. Climate Change, 4, 888–892, doi:10.1038/nclimate2330.
Meehl, G. A., J. M. Arblaster, J. T. Fasullo, A. Hu, and K. E. Trenberth, 2011: Model-based evidence of deep-ocean heat uptake during surface-temperature hiatus periods. Nat. Climate Change, 1, 360–364, doi:10.1038/nclimate1229.
Meehl, G. A., A. Hu, J. M. Arblaster, J. Fasullo, and K. E. Trenberth, 2013: Externally forced and internally generated decadal climate variability associated with the interdecadal Pacific oscillation. J. Climate, 26, 7298–7310, doi:10.1175/JCLI-D-12-00548.1.
Meng, Q., M. Latif, W. Park, N. S. Keenlyside, V. A. Semenov, and T. Martin, 2012: Twentieth century Walker circulation change: Data analysis and model experiments. Climate Dyn., 38, 1757–1773, doi:10.1007/s00382-011-1047-8.
Merrifield, M. A., 2011: A shift in western tropical Pacific sea level trends during the 1990s. J. Climate, 24, 4126–4138, doi:10.1175/2011JCLI3932.1.
Mitas, C. M., and A. Clement, 2005: Has the Hadley cell been strengthening in recent decades? Geophys. Res. Lett., 32, L03809, doi:10.1029/2004GL021765.
Onogi, K., and Coauthors, 2007: The JRA-25 Reanalysis. J. Meteor. Soc. Japan, 85, 369–432, doi:10.2151/jmsj.85.369.
Oort, A. H., and J. J. Yienger, 1996: Observed interannual variability in the Hadley circulation and its connection to ENSO. J. Climate, 9, 2751–2767, doi:10.1175/1520-0442(1996)009<2751:OIVITH>2.0.CO;2.
Philander, S. G., 1990: El Niño, La Niña, and the Southern Oscillation. Academic Press, 293 pp.
Power, S. B., and I. N. Smith, 2007: Weakening of the Walker circulation and apparent dominance of El Niño both reach record levels, but has ENSO really changed? Geophys. Res. Lett., 34, L18702, doi:10.1029/2007GL030854.
Power, S. B., and G. Kociuba, 2011: What caused the observed twentieth-century weakening of the Walker circulation? J. Climate, 24, 6501–6514, doi:10.1175/2011JCLI4101.1.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, doi:10.1029/2002JD002670.
Rienecker, M. M., and Coauthors, 2011: MERRA: NASA’s Modern-Era Retrospective Analysis for Research and Applications. J. Climate, 24, 3624–3648, doi:10.1175/JCLI-D-11-00015.1.
Sandeep, S., F. Stordal, P. D. Sardeshmukh, and G. P. Compo, 2014: Pacific Walker Circulation variability in coupled and uncoupled climate models. Climate Dyn., 43, 103–117, doi:10.1007/s00382-014-2135-3.
Schwendike, J., P. Govekar, M. J. Reeder, R. Wardle, G. J. Berry, and C. Jakob, 2014: Local partitioning of the overturning circulation in the tropics and the connection to the Hadley and Walker circulations. J. Geophys. Res., 119, 1322–1339, doi:10.1002/2013JD020742.
Sohn, B. J., S. W. Yeh, J. Schmetz, and H. J. Song, 2013: Observational evidences of Walker circulation change over the last 30 years contrasting with GCM results. Climate Dyn., 40, 1721–1732, doi:10.1007/s00382-012-1484-z.
Solomon, A., and M. Newman, 2012: Reconciling disparate 20th century Indo-Pacific ocean temperature trends in the instrumental record. Nat. Climate Change, 2, 691–699, doi:10.1038/nclimate1591.
Tanaka, H. L., N. Ishizaki, and A. Kitoh, 2004: Trend and interannual variability of Walker, monsoon and Hadley circulations defined by velocity potential in the upper troposphere. Tellus, 56A, 250–269, doi:10.1111/j.1600-0870.2004.00049.x.
Thompson, D., J. Wallace, and G. Hegerl, 2000: Annular modes in the extratropical circulation. Part II: Trends. J. Climate, 13, 1018–1036, doi:10.1175/1520-0442(2000)013<1018:AMITEC>2.0.CO;2.
Thompson, D. M., J. E. Cole, G. T. Shen, A. W. Tudhope, and G. A. Meehl, 2015: Early twentieth-century warming linked to tropical Pacific wind strength. Nat. Geosci., 8, 117–121, doi:10.1038/ngeo2321.
Tokinaga, H., S. P. Xie, C. Deser, Y. Kosaka, and Y. M. Okumura, 2012a: Slowdown of the Walker circulation driven by tropical Indo-Pacific warming. Nature, 491, 439–443, doi:10.1038/nature11576.
Tokinaga, H., S. P. Xie, and A. Timmermann, 2012b: Regional patterns of tropical Indo-Pacific climate change: Evidence of the Walker circulation weakening. J. Climate, 25, 1689–1710, doi:10.1175/JCLI-D-11-00263.1.
Vecchi, G. A., and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20, 4316–4340, doi:10.1175/JCLI4258.1.
Vecchi, G. A., B. J. Soden, A. T. Wittenberg, I. A. Held, A. Leetma, and M. J. Harrison, 2006: Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature, 441, 73–76, doi:10.1038/nature04744.
Wang, B., J. Liu, H. J. Kim, P. J. Webster, and S. Y. Yim, 2012: Recent change of the global monsoon precipitation (1979–2008). Climate Dyn., 39, 1123–1135, doi:10.1007/s00382-011-1266-z.
Wang, B., J. Liu, H. J. Kim, P. J. Webster, S. Y. Yim, and B. Xiang, 2013: Northern Hemisphere summer monsoon intensified by mega-El Niño/Southern Oscillation and Atlantic multidecadal oscillation. Proc. Natl. Acad. Sci. USA, 110, 5347–5352, doi:10.1073/pnas.1219405110.
Webster, P. J., V. O. Magana, T. N. Palmer, J. Shukla, R. A. Tomas, M. Yanai, and T. Yasunari, 1998: Monsoons: Processes, predictability, and the prospects for prediction. J. Geophys. Res., 103, 14 451–14 510, doi:10.1029/97JC02719.
Williams, A. P., and C. Funk, 2011: A westward extension of the warm pool leads to a westward extension of the Walker circulation, drying eastern Africa. Climate Dyn., 37, 2417–2435, doi:10.1007/s00382-010-0984-y.
Xie, S. P., C. Deser, G. A. Vecchi, J. Ma, H. Teng, and A. T. Wittenberg, 2010: Global warming pattern formation: Sea surface temperature and rainfall. J. Climate, 23, 966–986, doi:10.1175/2009JCLI3329.1.
Yu, B., and G. J. Boer, 2002: The roles of radiation and dynamical processes in the El Niño-like response to global warming. Climate Dyn., 19, 539–553, doi:10.1007/s00382-002-0244-x.
Yu, B., and F. W. Zwiers, 2010: Changes in equatorial atmospheric zonal circulations in recent decades. Geophys. Res. Lett., 37, L05701, doi:10.1029/2009GL042071.
Yu, B., F. W. Zwiers, G. J. Boer, and M. F. Ting, 2012: Structure and variances of equatorial zonal circulation in a multimodel ensemble. Climate Dyn., 39, 2403–2419, doi:10.1007/s00382-012-1372-6.
Zhang, M., and H. Song, 2006: Evidence of deceleration of atmospheric vertical overturning circulation over the tropical Pacific. Geophys. Res. Lett., 33, L12701, doi:10.1029/2006GL025942.
Zhou, T., R. Yu, H. Li, and B. Wang, 2008: Ocean forcing to changes in global monsoon precipitation over the recent half-century. J. Climate, 21, 3833–3852, doi:10.1175/2008JCLI2067.1.
