1. Introduction
The East Asian winter monsoon (EAWM) is one of the most active climate systems in the Northern Hemisphere during boreal winter (Lau and Li 1984; Chen and Sun 1999; He et al. 2007). The EAWM has broad impacts on the East Asian weather and climate (Tao and Chen 1987). A strong EAWM can lead to intense snowfall and cold surge activity in the winter (Guo 1994; Fan 2009; Wang et al. 2011) and is associated with a subsequent dust climate during the following spring over China (Wang et al. 2003). In addition, the EAWM can contribute to summer drought/flooding over eastern China (Sun and Sun 1995; Shi and Zhu 1996; Yan et al. 2003). Hence, studies on the EAWM are among the keys to understanding the mechanisms of East Asian climate variation (Wang and Fan 2013).
Early studies of the EAWM focused on the synoptic scale, especially for the East Asian cold wave activity and associated atmospheric circulation (e.g., Tao 1957, 1959). More recently, variabilities in the EAWM at intraseasonal and interdecadal time scales received more attention. Some studies have reported the quasi-2-week and quasi-40-day oscillations of the observed EAWM (Pan and Zhou 1985; Yang and Zhu 1990; Tang and Wang 1994). On the interannual time scale, Mu and Li (1999) noted that the EAWM has quasi-biennial and 3–5-yr cycles, and El Niño–Southern Oscillation (ENSO) signals are correlated with the interannual anomaly of the EAWM. Huang et al. (2012) also found a quasi-4-yr oscillation in the EAWM. Many studies indicate that ENSO plays a vital role in regulating the interannual variability of the EAWM (Wang et al. 2000; Zhou et al. 2007a; Wang and He 2012; He and Wang 2013). The mature phase of an El Niño event is usually accompanied by a weaker EAWM, while a La Niña event is often associated with a stronger EAWM (Wang et al. 2000). Recently, however, Wang and He (2012) found the close relationship between ENSO and EAWM weakens after the 1970s. This is partly caused by suppressed ENSO-associated tropical Indo–western Pacific sea surface temperature (SST) variability, reduced EAWM interannual variability, and northward-retreating EAWM signals. The EAWM intensity is also regulated by the Arctic Oscillation (AO) on the interannual time scale (Gong et al. 2001; Wu and Wang 2002). Furthermore, the Arctic amplification and sea ice loss may affect the EAWM (Wang and Liu 2016; Zhou 2017).
Both anthropogenic forcings [e.g., greenhouse gases (GHGs) and anthropogenic aerosols (AAs)] and natural forcings (e.g., volcanic aerosols and solar variability) can influence the East Asian climate (e.g., the EAWM). Hori and Ueda (2006) simulated a weakened EAWM under a global warming scenario. Ding et al. (2007) also suggested a weakened EAWM at the end of the twenty-first century due to anthropogenic global warming. Some modeling studies have shown that the influence of AAs contributed to the weakening of the East Asian summer monsoon in the late 1970s (e.g., Wang et al. 2013; Song et al. 2014). A recent study also indicates that the AAs tend to intensify the northern mode of the EAWM (Jiang et al. 2017). Additionally, previous modeling studies suggested that natural external forcings play an important role in regulating the evolution of ENSO (Wang et al. 2018), the Pacific decadal oscillation (PDO; Wang et al. 2012), and the Atlantic multidecadal oscillation (AMO; Otterå et al. 2010; Zanchettin et al. 2012), which could have potential impacts on the East Asian climate (e.g., Zhu et al. 2015; Miao et al. 2016, 2018). For instance, Miao et al. (2018) found that the simulated EAWM is strengthened when the solar forcing is stronger on a multidecadal time scale through regulating decadal variability of the North Atlantic SST. Overall, the external forcings play an important role in the East Asian winter climate. Thus in the last decades, how natural and increasing anthropogenic external forcings have affected the EAWM requires more attention.
Since the mid-1980s, the EAWM has experienced an interdecadal weakening (Xu et al. 1999; Kang et al. 2006; He and Wang 2012; Wang and Fan 2013). The mechanisms of the interdecadal variation in the EAWM are complex (Ding et al. 2014). This is likely related to the internal variability in the climate system and modulated by external forcings (Ding et al. 2014). The interdecadal variations in the EAWM are associated with dominant modes of atmospheric circulation variability (Ao and Sun 2016). He and Wang (2012) revealed the out-of-phase relationship between the AO and EAWM during the late twentieth century and inferred that the interdecadal weakening of the EAWM in the mid-1980s could be caused by the significantly enhanced AO. In addition, some recent studies suggest that the interdecadal variations in the EAWM are also related to the PDO (Yang et al. 2004; Zhou et al. 2007b; Wang et al. 2008; Ding et al. 2014). For instance, Wang et al. (2008) found the interdecadal modulation of the PDO on the impact of ENSO on the EAWM. Ding et al. (2014) noted that when the PDO is in a positive (negative) phase, the EAWM is usually weaker (stronger). In addition, the interdecadal variations in the EAWM are also affected by the AMO (Li and Bates 2007; Li et al. 2009; Wang et al. 2009).
However, the influences from external forcings on the interdecadal weakening of the EAWM during the mid-1980s are still unclear. Many previous studies used ensembles of coupled models to analyze climatic responses to external forcings and to attribute internally generated and externally forced changes (Deser et al. 2012a; Marvel et al. 2015; Deser et al. 2016; Meehl et al. 2016; Soden and Chung 2017). Thus, in this study, we investigate the relative roles of anthropogenic forcings and natural forcings in the interdecadal weakening of the EAWM during the mid-1980s using a suite of ensembles of various coupled models from phase 5 of the Coupled Model Intercomparison Project (CMIP5), which are driven by different forcing agent combinations. The model, data, and methods are described in section 2. In section 3, we examine the observed and simulated interdecadal climate changes over East Asia, as well as the associated physical processes. The conclusions and some discussion are provided in section 4.
2. Model, data, and methods
We analyze multimodel results from CMIP5 (see Table 1) in this study, including historical, historicalGHG, historicalNat, historicalMisc and piControl experiments (Taylor et al. 2012). The historical simulations (named ALL) are forced by both anthropogenic forcings (e.g., GHGs and AAs) and natural forcings (e.g., volcanic aerosols and solar variability). The historicalGHG (named GHG) and historicalNat (named NAT) simulations are the same as ALL, except that they are only forced by well-mixed GHG changes or natural agents (i.e., volcanic aerosols and solar variability), respectively. The simulations forced by only AAs (named AA) in the historicalMisc experiment are also used in this study. These simulations cover the period of 1850–2005. The piControl simulations are long-term control runs with imposed non-evolving preindustrial conditions (i.e., anthropogenic and natural forcings are set to preindustrial levels).
Details of 17 CMIP5 models used in this study.
In this study, the Climate Research Unit (CRU) dataset (Mitchell and Jones 2005), Hadley Centre sea level pressure dataset (HadSLP2; Allan and Ansell 2006), Hadley Centre monthly sea surface temperature dataset (HadISST; Rayner et al. 2003), and National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis data (Kalnay et al. 1996) are also used to evaluate the performance of the model results and to investigate the observed interdecadal climate changes over East Asia.
For the statistical analysis, we focus on the linear trends in the boreal winter (December–February). For example, the winter of 1987 refers to the average of December 1986–February 1987. In addition, the Mann–Kendall test (Mann 1945) is used to evaluate the significance of the linear trends. Pearson’s linear correlation coefficient is also used to describe the correlation between the variables.
The tropospheric meridional temperature gradient (MTG) is calculated as 
The multimodel ensemble mean (MME) method is used in this study. The averages are first computed over the number of realizations available for each model and then computed over the number of models.
The key elements of the EAWM system include the upper-tropospheric jet stream, midtropospheric EAT, and low-level monsoon circulation (Jhun and Lee 2004). In the following, to investigate the relative contributions of different external forcings to the interdecadal weakening of the EAWM during the mid-1980s, we mainly analyze the linear trends of the observed and simulated East Asian winter climate during the period of 1967–2004 and make model–observation comparisons.
3. Results
a. Climatology evaluation of the model
First, we examine whether the models can reliably reproduce the East Asian winter climatology because this ability is a primary factor in judging whether the model simulation of the interdecadal variation is credible (Sun and Ding 2008). Taylor diagrams (Taylor 2001) are used here to assess the performances of the 17 models for the typical region of East Asia (20°–50°N, 100°–145°E). A similar analysis has been used in some previous studies that evaluated the performances of the model (e.g., Zhu and Wang 2010; Gong et al. 2014; Yan et al. 2014; Hao et al. 2016).
The diagrams include information on the pattern correlations and the ratio of the standard deviation between the model and observations. The better the models agree with the observations, the nearer the results will be to the reference point. As shown in Fig. 1, the simulated surface air temperature (SAT) has correlations greater than 0.95 with the observations, and the normalized standard deviations fall into the range of 0.75–1.25 for most models. This indicates that most models simulate a realistic distribution of winter SAT over East Asia and capture the amplitude of the observed signal. Similarly, the models generally show good performances in Z500. Though the simulated U300 also has good correlations with the observations, some models overestimate the amplitude of the observed signal. In contrast, for SLP, most models underestimate the amplitude of the observed signal. Additionally, some models show lower correlations with the observations, and the intermodel spread is larger than that of the other variables. Note that a noticeable improvement in the simulated results can be seen from the individual models to the MME. In this study, 17 MME results are used to represent the MME of the 17 CMIP5 results.
Taylor diagrams of winter climatology for SAT (°C; triangles), SLP (hPa; circles), Z500 (m; plus signs), and U300 (m s−1; stars). The evaluation is for East Asia (20°–50°N, 100°–145°E) during 1967–2004.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
In summary, the CMIP5 models differ in their abilities to reproduce the observed East Asian winter climatology. Nevertheless, most models perform well, especially in simulating the SAT, Z500, and U300. In addition, the 17 MME results agree with the observations better than most of the CMIP5 models.
b. The observed and simulated decadal changes in the EAWM
As illustrated in Fig. 2, the EAWM undergoes a significant interdecadal weakening during the mid-1980s in the observations. The intensity of the EAWM is usually stronger before the mid-1980s but weaker thereafter. The observed linear trend of the EAWM index is −1.2 per 40 years (p = 0.03) during the period of 1967–2004. This interdecadal weakening of the EAWM can be evident from changes in the Siberian high, East Asian trough, and almost every member of the EAWM system (He and Wang 2012). The 17 MME results can reproduce the observed interdecadal weakening well. After the mid-1980s, the simulated EAWM is significantly weaker. The linear trend of the simulated EAWM index is −1.7 per 40 years (p < 0.01), which is larger than the observed trend. The following analyses are thus based on the 17 MME results. Details of the ensemble simulations for the 17 models are listed in Table 2. Because some models do not contain single forcing simulations, we analyze the ensemble mean of nine models (9 MME) for the AA simulations. The ensemble members are large enough to detect the forced changes in the EAWM (Deser et al. 2012b).
(left) Observed and (right) simulated EAWM index (bars). The observation is based on the NCEP–NCAR reanalysis data and the HadSLP2 data. For the modeling, the index is based on the MME of ALL simulations for each CMIP5 model. Their least squares linear trends during 1967–2004 are plotted as solid lines in each panel.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
Simulations of the 17 CMIP5 models in this study. The numbers in the table are ensemble simulations for each model. Note that the historicalMisc (AA) simulations of GISS-E2-H(R) used here are based on concentrations and not emissions.
c. Changes in the EAT and SST
The EAT is an important component of the EAWM system. The deepened EAT is favorable for cold air spilling south, causing cold SAT anomalies over East Asia. Therefore, a strengthened (weakened) EAT can lead to a stronger (weaker) EAWM (Fig. 3a; Cui and Sun 1999). Figures 3b–f show the linear trends of Z500 during 1967–2004. In the observations, positive trends of Z500 are evident over mid- and low-latitude Asian regions, and the maximum values appear over northeast Asia (Fig. 3b). In contrast, negative trends can be seen over the Ural Mountains and Bering Sea. This means that the EAT weakens during the period of 1967–2004. Consistent with the observations, significant positive trends of Z500 can be found over the Asian continent in the ALL ensemble (Fig. 3c). Although the positive trends are weaker, they still indicate a shallower EAT. The ALL ensemble captures the interdecadal weakening of the EAT. In the GHG ensemble, the positive trends of Z500 are stronger. Higher values can be found over northeastern China and southern Japan (Fig. 3d). This suggests a weakened EAT in the GHG ensemble, which is similar to the observations and ALL ensemble. In contrast, the AA ensemble simulates negative trends for Z500 over the mid- and low-latitude regions during the period of 1967–2004 (Fig. 3e). For the NAT ensemble, no significant trends can be found over the East Asian region. Overall, the GHGs probably play an important role in weakening the EAT during the period of 1967–2004 in the CMIP5 models analyzed. Furthermore, the linear trends of the EAT index are examined here (Fig. 4a). The ALL ensemble accounts for 55% of the observed trend of the EAT index. The GHG plays an important role and accounts for 67% of the observed trend, while the AA plays an opposite role and accounts for 19%. The NAT ensemble contributes less (~5%).
(a) The climatological winter 500-hPa geopotential height (m) during 1967–2004. The linear trend of winter 500-hPa geopotential height [shaded; m (40 yr)−1] during 1967–2004 in (b) NCEP–NCAR and (c) ALL, (d) GHG, (e) AA, and (f) NAT from MME. In (b)–(f), dots denote regions where the statistical significance level is above 95%, and lines represent climatological winter 500-hPa geopotential height during 1967–2004.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
The linear trends of the observed and simulated (a) EAT index [m (40 yr)−1], (b) EAJS index [m s−1 (40 yr)−1], and (c) Siberian high index [hPa (40 yr)−1] during 1967–2004. The number represents the ratio of simulation with individual forcings to observations. The error bars represent the standard deviation of intermodel variability.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
A recent study indicates that anomalous SSTs over the southwestern North Pacific play an important role in the decadal variability of the EAT through the regulation of the air–sea interactions over the North Pacific (Sun et al. 2016). Sun et al. (2016) note that the EAT index and the North Pacific SST index are highly correlated on a decadal time scale in observations. Here, we examine a simulated relationship between these two indices in the CMIP5. For the different forcing ensembles, significant correlations between the EAT index and the North Pacific SST index are well reproduced (figure not shown). Therefore, the CMIP5 results also suggest that the North Pacific SST does play an important role in regulating the EAT variability.
As shown in Fig. 5a, a significant SST warming trend can be found over the southwestern North Pacific in the observation data. In the different forcing ensembles, the observed positive trends of SST over the southwestern North Pacific are only reproduced by the ALL and GHG ensembles (Figs. 5b,c). This suggests that when forced by the increased GHG concentrations, the ALL ensemble reproduces the southwestern North Pacific warming trend, which most likely leads to the weakening trend of the EAT (Fig. 3). In contrast, the cooling induced by the AAs over this region probably plays an opposite role in modulating the EAT (Fig. 5d). The linear trend of the SST in the NAT ensemble is not significant over this region (Fig. 5e).
The linear trend of winter SST [shaded; °C (40 yr)−1] during 1967–2004 in the (a) HadISST and (b) ALL, (c) GHG, (d) AA, and (e) NAT from MME. Dots denote regions where the statistical significance level is above 95%.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
We further investigate the surface heat flux budget in the GHG ensemble (Fig. 6). It can be seen that the sensible heat flux as well as the surface net longwave and shortwave radiation favor the warming trends of the southwestern North Pacific SST, with the sensible heat flux playing a dominant role. In contrast, the latent heat flux favors cooling SST trends over this region. Similarly, we analyze the surface heat budget in the AA ensemble (figure not shown). We find that the decreased surface net shortwave radiation can cause the cooling trends of the southwestern North Pacific SST in the AA ensemble. This confirms that AAs play an opposite role in regulating the variability of the EAT, compared with the GHGs. Overall, the observed weakening of the EAT is probably attributed to an anomalous SST warming trend in the southwestern North Pacific, which is mainly caused by increased GHG concentrations.
The linear trend of winter (a) surface net longwave radiation, (b) surface net shortwave radiation, (c) sensible heat flux, and (d) latent heat flux [W m−2 (40 yr)−1] during 1967–2004 in the GHG from MME. Positive values mean downward. Dots denote regions where the statistical significance level is above 95%.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
d. Changes in the East Asian jet stream
Jhun and Lee (2004) indicate that the meridional shear of U300, which is associated with the change in the EAJS, reflects the intensity of the EAWM. We thus examine the trends of U300 during the period of 1967–2004. In the observation data, the climatological core of U300 is usually located along ~30°N (Fig. 7a). During the period of 1967–2004, negative U300 trends are evident over the entrance of the EAJS core region, whereas positive trends are located over Siberia (Fig. 7b). This indicates a weakened and poleward-shifted EAJS, which contributes to the weakening of the EAWM. Compared to the observations, similar negative and positive trends of U300 can be found in corresponding regions in the ALL ensemble (Fig. 7c). The correlation coefficient of the spatial pattern between the ALL ensemble and observations is 0.45. Although the trends are weaker, they still indicate a weakened meridional shear of the EAJS. Thus, the ALL ensemble captures the interdecadal weakening of the EAJS. For the NAT ensemble, negative and positive trends of U300 are evident in the EAJS core region and to the north, respectively. The correlation coefficient between the NAT ensemble and observations is 0.34 (0.72 between the NAT and ALL ensembles). This suggests that natural external forcings contribute to the weakening of the EAJS. In contrast, an opposite trend pattern can be found in the AA ensemble. Based on the observed and simulated trends of the EAJS index, the ALL ensemble accounts for 16% of the observed trends of the EAJS index (Fig. 4b). The NAT plays a dominant role and accounts for 20%, while the AA plays an opposite role and accounts for 15%. However, the discrepancies are large among the models, and the internal variability of the jet stream may be an important source (Deser et al. 2016). In addition, the larger model biases in simulating U300 could have some contribution to these discrepancies (Fig. 1). Changes in GHGs have few effects (~7%) on regulating the meridional shear of the EAJS during this period. The ALL ensemble contributions (16%) are nearly equal to the summation of NAT (20%), GHG (7%), and AA (−15%) ensembles.
(a) The climatological winter 300-hPa zonal wind (m s−1) during 1967–2004. The linear trend of winter 300-hPa zonal wind [shaded; m s−1 (40 yr)−1] during 1967–2004 in (b) NCEP–NCAR and (c) ALL, (d) GHG, (e) AA, and (f) NAT from MME. In (b)–(f), dots denote regions where the statistical significance level is above 95%, and lines represent climatological winter 300-hPa zonal wind during 1967–2004. In (c)–(f), the numbers in the top right are the correlation coefficients between the trends in NCEP–NCAR and the simulations.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
The variability of EAJS is closely associated with changes in the tropospheric MTG. As shown in Fig. 8a, negative trends in the MTG are evident at the entrance of the EAJS core region, while positive trends are located over Siberia. These are the same as the distribution of the U300 trends. This suggests that changes in the MTG weaken U300 in the EAJS core region, whereas they strengthen U300 over Siberia (Fig. 7a) through the thermal wind relationship. The observed interdecadal changes in the MTG are reproduced by the ALL and NAT ensembles, except for smaller magnitudes (Fig. 8). This further confirms that the natural external forcings probably play an important role in causing interdecadal weakening of the EAJS during this period through the regulation of the tropospheric MTG.
The linear trend of the winter equator-to-pole temperature gradient averaged from 850 to 300 hPa [shaded; 10−7 °C m−1 (40 yr)−1] during 1967–2004 in (a) NCEP–NCAR and (b) ALL, (c) GHG, (d) AA, and (e) NAT from MME. Dots denote regions where the statistical significance level is above 95%.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
e. Changes in the lower-tropospheric circulation
In winter, the Siberian high controls the lower-tropospheric circulation over the entire Asian continent (Fig. 9a). In addition, a strengthened Siberian high causes a stronger EAWM circulation (Gong et al. 2001). We therefore examine the trends of the Siberian high during the period of 1967–2004. In the observations, negative trends in the SLP can be seen over the northern Asian continent, suggesting that the Siberian high undergoes a decadal weakening during this period (Fig. 9b), consistent with the trend of the observed Siberian high index (Fig. 4c). Compared with the observations, the negative trends in the SLP can be found over the Siberian region in the ALL ensemble (Fig. 9c), whereas positive trends are evident in the southern part of East Asia. Although the magnitude of the SLP trend is smaller, it still indicates a weakened Siberian high during this period. The ALL ensemble successfully simulates this interdecadal weakening of the Siberian high. For the GHG, AA, and NAT ensembles, however, the linear trends are not significant. This means that the GHGs, anthropogenic aerosols, and natural forcings may work together to weaken the Siberian high. The ALL ensemble accounts for 25% of the observed trends in the Siberian high index (Fig. 4c). The GHG, AA, and NAT ensembles account for 4%, −2%, and 5%, respectively.
(a) The climatological winter SLP (hPa) during 1967–2004. The linear trend of winter SLP [shaded; hPa (40 yr)−1] during 1967–2004 in (b) HadSLP2 and (c) ALL, (d) GHG, (e) AA, and (f) NAT from MME. In (b)–(f), dots denote regions where the statistical significance level is above 95%.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
f. Potential role of the internal variability
Our analysis is based on the MME, and thus, the internal variability is averaged out to a large extent. Partly for this reason, the MME shows better performance between the observed and simulated patterns, compared to the individual model (Fig. 1). Simulated internal variabilities of the East Asian climate in each model are possibly different from the observations, which can contribute to the discrepancies between the observed patterns and ALL ensembles. In fact, the internal variability of the EAWM also plays an important role in the interdecadal change in the EAWM, which was noted by Ding et al. (2014). To investigate the relative roles of internal variability in weakening the EAWM, we further analyze the piControl simulations for the 17 coupled models. A nearly 500-yr-long simulation is available for each model [except for MIROC-ESM-CHEM (250 years)].
Figure 10 shows the linear trend ranges of the EAT index, the EAJS index, and the Siberian high index in the piControl simulations for the 17 models. This is performed by breaking the time series into 38-yr running segments (the same length as the analyzed period in the observation) and then calculating the linear trends of each segment. The probability density distributions are then calculated. As shown in Fig. 10a, the observed linear trends in the EAT index are much larger than the simulated trends for most models in the piControl simulations. Only five models reproduce the same linear trends under some extreme conditions. The statistics show that the same/larger trends as the observations in the EAT index occur only 0.4 times per 500 years in the piControl simulations (Table 3). This means that external forcings most likely play an essential role in this observed interdecadal change in the EAT, which is not caused solely by the internal variability of the EAT. For the EAJS and Siberian high indices, the observed linear trends can be caused by the internal variability alone in most models (Figs. 10b,c). The statistics show that the same/larger trends as the observations in the EAJS and Siberian high indices occur 8.5 and 3.8 times per 500 years in the piControl simulations, respectively. The internal variability from high-latitude systems (i.e., EAJS and Siberian high) can contribute to the observed interdecadal weakening of the EAWM, particularly for changes in the higher troposphere. However, it is difficult to quantitatively distinguish its relative contribution from the external forcings based on the present analysis, especially in a nonlinear climate system.
Density distributions of linear trends of the (a) EAT index [m (40 yr)−1], (b) EAJS index [m s−1 (40 yr)−1], and (c) Siberian high index [hPa (40 yr)−1] in the piControl simulations. Color curves correspond to different CMIP5 models. The linear trends in the observations and the ALL from 17 MMEs are shown as solid black lines and dashed black lines, respectively.
Citation: Journal of Climate 31, 21; 10.1175/JCLI-D-17-0868.1
Number of 38-yr periods when the linear trends of the indices (EAT index, EAJS index, and Siberian high index) in the piControl simulations are as large as those in the observations [(500 yr)−1]. Only independent 38-yr periods are selected here. “Independent” means that the gaps between two periods are longer than 37 years.
4. Summary and discussion
This study investigates the relative contributions of anthropogenic forcings and natural external forcings to the interdecadal weakening of the EAWM during the mid-1980s using outputs from the CMIP5. The results show that the weakening of the EAWM is well reproduced in the 17 MME results. The comparisons of the separate forcing ensembles and observations reveal that in the upper troposphere, the weakened and poleward-shifted EAJS in the ALL ensemble is mainly caused by the natural external forcing through the regulation of the MTG over the East Asian region. In the midtroposphere, the increased GHG forcing plays an important role in weakening the EAT. Increased longwave and shortwave radiation and downward sensible heat flux–induced warming trends in the North Pacific SSTs, which probably act as a bridge to link the external forcings and EAT on a decadal time scale. In the lower troposphere, both anthropogenic and natural forcings may contribute to the weakening of the Siberian high during this period. Overall, the GHGs and natural forcings play more important roles in causing the interdecadal weakening of the EAWM during the mid-1980s.
The results based on the piControl simulations indicate that the same/larger trends as the observations in the EAJS and Siberian high indices occur 8.5 and 3.8 times per 500 years, respectively. The internal variability could contribute to the observed interdecadal weakening of the EAWM during the 1980s, especially for the upper and lower troposphere.
Our analysis further shows that some discrepancies exist between the observed patterns and ALL ensembles. Both the models’ biases and internal variability can contribute to these discrepancies. Additionally, the coupled models show best performances in simulating Z500 climatology (Fig. 1). Thus, results based on the responses of the EAT to external forcings are more robust.
A recent study suggests that the EAWM recovered from its weak episode and reamplified in the mid-2000s, and the internal variability in the climate system plays an important role during this process (Wang and Chen 2014). The external forcings may also have important effects on the changes in the EAWM in the mid-2000s, which needs further investigation.
Acknowledgments
We thank three anonymous reviewers and the editor for their valuable comments and suggestions, which helped to improve the quality of this paper significantly. This research was supported by the National Natural Science Foundation of China (41575086 and 41661144005) and the CAS–PKU Joint Research Program. We thank the IPCC for providing the CMIP5 datasets (http://www.ipcc-data.org/sim/gcm_monthly/AR5/Reference-Archive.html).
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, https://doi.org/10.1175/JCLI3937.1.
Ao, J., and J. Q. Sun, 2016: Decadal change in factors affecting winter precipitation over eastern China. Climate Dyn., 46, 111–121, https://doi.org/10.1007/s00382-015-2572-7.
Chen, J., and S. Sun, 1999: East Asian winter monsoon anomaly and variation of global circulation. Part I: A comparison study on strong and weak winter monsoons (in Chinese). Chin. J. Atmos. Sci., 23, 101–111.
Cui, X., and Z. Sun, 1999: East Asian winter monsoon index and its variation analysis (in Chinese). J. Nanjing Inst. Meteor., 22, 321–325.
Deser, C., R. Knutti, S. Solomon, and A. S. Phillips, 2012a: Communication of the role of natural variability in future North American climate. Nat. Climate Change, 2, 775–779, https://doi.org/10.1038/nclimate1562.
Deser, C., A. S. Phillips, V. Bourdette, and H. Teng, 2012b: Uncertainty in climate change projections: The role of internal variability. Climate Dyn., 38, 527–546, https://doi.org/10.1007/s00382-010-0977-x.
Deser, C., L. Terray, and A. S. Phillips, 2016: Forced and internal components of winter air temperature trends over North America during the past 50 years: Mechanisms and implications. J. Climate, 29, 2237–2258, https://doi.org/10.1175/JCLI-D-15-0304.1.
Ding, Y. H., G. Y. Ren, Z. C. Zhao, Y. Xu, Y. Luo, Q. P. Li, and J. Zhang, 2007: Detection, causes and projection of climate change over China: An overview of recent progress. Adv. Atmos. Sci., 24, 954–971, https://doi.org/10.1007/s00376-007-0954-4.
Ding, Y. H., and Coauthors, 2014: Interdecadal variability of the East Asian winter monsoon and its possible links to global climate change. J. Meteor. Res., 28, 693–713, https://doi.org/10.1007/s13351-014-4046-y.
Fan, K., 2009: Predicting winter surface air temperature in Northeast China. Atmos. Ocean. Sci. Lett., 2, 14–17, https://doi.org/10.1080/16742834.2009.11446770.
Gong, D. Y., S. W. Wang, and J. H. Zhu, 2001: East Asian winter monsoon and Arctic Oscillation. Geophys. Res. Lett., 28, 2073–2076, https://doi.org/10.1029/2000GL012311.
Gong, H. A., L. Wang, W. Chen, R. G. Wu, K. Wei, and X. F. Cui, 2014: The climatology and interannual variability of the East Asian winter monsoon in CMIP5 models. J. Climate, 27, 1659–1678, https://doi.org/10.1175/JCLI-D-13-00039.1.
Guo, Q., 1994: Relationship between the variations of East Asian winter monsoon and temperature anomalies in China (in Chinese). Quart. J. Appl. Meteor., 5, 218–225.
Hao, X., F. Li, J. Sun, H. Wang, and S. He, 2016: Assessment of the response of the East Asian winter monsoon to ENSO-like SSTAs in three U.S. CLIVAR Project models. Int. J. Climatol., 36, 847–866, https://doi.org/10.1002/joc.4388.
He, J. H., J. H. Ju, Z. P. Wen, J. M. Lu, and Q. H. Jin, 2007: A review of recent advances in research on Asian monsoon in China. Adv. Atmos. Sci., 24, 972–992, https://doi.org/10.1007/s00376-007-0972-2.
He, S. P., and H. J. Wang, 2012: An integrated East Asian winter monsoon index and its interannual variability (in Chinese). Chin. J. Atmos. Sci., 36, 523–538.
He, S. P., and H. J. Wang, 2013: Oscillating relationship between the East Asian winter monsoon and ENSO. J. Climate, 26, 9819–9838, https://doi.org/10.1175/JCLI-D-13-00174.1.
Hori, M. E., and H. Ueda, 2006: Impact of global warming on the East Asian winter monsoon as revealed by nine coupled atmosphere-ocean GCMs. Geophys. Res. Lett., 33, L03713, https://doi.org/10.1029/2005GL024961.
Huang, R. H., J. L. Chen, L. Wang, and Z. D. Lin, 2012: Characteristics, processes, and causes of the spatio-temporal variabilities of the East Asian monsoon system. Adv. Atmos. Sci., 29, 910–942, https://doi.org/10.1007/s00376-012-2015-x.
Jhun, J.-G., and E.-J. Lee, 2004: A new East Asian winter monsoon index and associated characteristics of the winter monsoon. J. Climate, 17, 711–726, https://doi.org/10.1175/1520-0442(2004)017<0711:ANEAWM>2.0.CO;2.
Jiang, Y., and Coauthors, 2017: Anthropogenic aerosol effects on East Asian winter monsoon: The role of black carbon-induced Tibetan Plateau warming. J. Geophys. Res. Atmos., 122, 5883–5902, https://doi.org/10.1002/2016JD026237.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–472, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Kang, L., W. Chen, and K. Wei, 2006: The interdecadal variation of winter temperature in China and its relation to the anomalies in atmospheric general circulation (in Chinese). Climatic Environ. Res., 11, 330–339.
Lau, K.-M., and M.-T. Li, 1984: The monsoon of East Asia and its global associations—A survey. Bull. Amer. Meteor. Soc., 65, 114–125, https://doi.org/10.1175/1520-0477(1984)065<0114:TMOEAA>2.0.CO;2.
Li, S., and G. T. Bates, 2007: Influence of the Atlantic multidecadal oscillation on the winter climate of East China. Adv. Atmos. Sci., 24, 126–135, https://doi.org/10.1007/s00376-007-0126-6.
Li, S., Y. Wang, and Y. Gao, 2009: A review of the researches on the Atlantic multidecadal oscillation (AMO) and its climate influence (in Chinese). Trans. Atmos. Sci., 32, 458–465.
Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245–259, https://doi.org/10.2307/1907187.
Marvel, K., A. S. Gavin, S. Drew, B. Céline, N. L. Allegra, N. Larissa, and T. Kostas, 2015: Do responses to different anthropogenic forcings add linearly in climate models? Environ. Res. Lett., 10, 104010, https://doi.org/10.1088/1748-9326/10/10/104010.
Meehl, G. A., A. Hu, B. D. Santer, and S.-P. Xie, 2016: Contribution of the interdecadal Pacific oscillation to twentieth-century global surface temperature trends. Nat. Climate Change, 6, 1005–1008, https://doi.org/10.1038/nclimate3107.
Miao, J. P., T. Wang, Y. L. Zhu, J. Z. Min, H. J. Wang, and D. Guo, 2016: Response of the East Asian winter monsoon to strong tropical volcanic eruptions. J. Climate, 29, 5041–5057, https://doi.org/10.1175/JCLI-D-15-0600.1.
Miao, J. P., T. Wang, H. J. Wang, and Y. Q. Gao, 2018: Influence of low-frequency solar forcing on the East Asian winter monsoon based on HadCM3 and observations. Adv. Atmos. Sci., 35, 1205–1215, https://doi.org/10.1007/s00376-018-7229-0.
Mitchell, T. D., and P. D. Jones, 2005: An improved method of constructing a database of monthly climate observations and associated high-resolution grids. Int. J. Climatol., 25, 693–712, https://doi.org/10.1002/joc.1181.
Mu, M., and C. Li, 1999: ENSO signals in the interannual variability of East-Asian winter monsoon. Part I: Observed data analyses (in Chinese). Chin. J. Atmos. Sci., 23, 276–285.
Otterå, O. H., M. Bentsen, H. Drange, and L. L. Suo, 2010: External forcing as a metronome for Atlantic multidecadal variability. Nat. Geosci., 3, 688–694, https://doi.org/10.1038/ngeo955.
Pan, H.-L., and F.-X. Zhou, 1985: The 10–20 day tropical-midlatitude interactions during the winter monsoon season. J. Meteor. Soc. Japan, 63, 829–844, https://doi.org/10.2151/jmsj1965.63.5_829.
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, https://doi.org/10.1029/2002JD002670.
Shi, N., and Q. Zhu, 1996: Anomalous East Asia winter monsoon intensity and its relation to summer 500 hPa atmospheric circulation and climate anomaly in China (in Chinese). J. Trop. Meteor., 12, 26–33.
Soden, B., and E.-S. Chung, 2017: The large-scale dynamical response of clouds to aerosol forcing. J. Climate, 30, 8783–8794, https://doi.org/10.1175/JCLI-D-17-0050.1.
Song, F., T. Zhou, and Y. Qian, 2014: Responses of East Asian summer monsoon to natural and anthropogenic forcings in the 17 latest CMIP5 models. Geophys. Res. Lett., 41, 596–603, https://doi.org/10.1002/2013GL058705.
Sun, B., and C. Li, 1997: Relationship between the disturbances of East Asian trough and tropical convective activities in boreal winter (in Chinese). Chin. Sci. Bull., 42, 500–503.
Sun, J. Q., S. Wu, and J. Ao, 2016: Role of the North Pacific sea surface temperature in the East Asian winter monsoon decadal variability. Climate Dyn., 46, 3793–3805, https://doi.org/10.1007/s00382-015-2805-9.
Sun, S., and B. Sun, 1995: The relationship between the anomalous winter monsoon circulation over East Asia and summer drought/flooding in the Yangtze and Huaihe River valley (in Chinese). Acta Meteor. Sin., 53, 440–450.
Sun, Y., and Y. H. Ding, 2008: An assessment on the performance of IPCC AR4 climate models in simulating interdecadal variations of the East Asian summer monsoon. Acta Meteor. Sin., 22, 472–488.
Tang, D., and J. Wang, 1994: Three-dimension vertical circulation LFO features of eastern Asian winter monsoon (in Chinese). J. Nanjing Inst. Meteor., 17, 351–355.
Tao, S. Y., 1957: A synoptic and aerological study on a cold wave in the Far East during the period of the break down of the blocking situation over Euroasia and Atlantic (in Chinese). Acta Meteor. Sin., 28, 63–74.
Tao, S. Y., 1959: Study on East Asian cold waves in China during recent 10 years (1949–1959) (in Chinese). Acta Meteor. Sin., 30, 226–230.
Tao, S. Y., and L. X. Chen, 1987: A review of recent research on the East Asian monsoon in China. Monsoon Meteorology, C. P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 60–92.
Taylor, K. E., 2001: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res., 106, 7183–7192, https://doi.org/10.1029/2000JD900719.
Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: An overview of CMIP5 and the experiment design. Bull. Amer. Meteor. Soc., 93, 485–498, https://doi.org/10.1175/BAMS-D-11-00094.1.
Wang, B., R. Wu, and X. Fu, 2000: Pacific–East Asian teleconnection: How does ENSO affect East Asian climate? J. Climate, 13, 1517–1536, https://doi.org/10.1175/1520-0442(2000)013<1517:PEATHD>2.0.CO;2.
Wang, H., and S. He, 2012: Weakening relationship between East Asian winter monsoon and ENSO after mid-1970s. Chin. Sci. Bull., 57, 3535–3540, https://doi.org/10.1007/s11434-012-5285-x.
Wang, H., and K. Fan, 2013: Recent changes in the East Asian monsoon (in Chinese). Chin. J. Atmos. Sci., 37, 313–318.
Wang, H., X. Lang, G. Zhou, and D. Kang, 2003: A preliminary report of the model prediction on the forthcoming winter and spring dust climate over China (in Chinese). Chin. J. Atmos. Sci., 27, 136–140.
Wang, H., E. Yu, and S. Yang, 2011: An exceptionally heavy snowfall in Northeast China: Large-scale circulation anomalies and hindcast of the NCAR WRF Model. Meteor. Atmos. Phys., 113, 11–25, https://doi.org/10.1007/s00703-011-0147-7.
Wang, L., and W. Chen, 2014: The East Asian winter monsoon: Re-amplification in the mid-2000s. Chin. Sci. Bull., 59, 430–436, https://doi.org/10.1007/s11434-013-0029-0.
Wang, L., W. Chen, and R. Huang, 2008: Interdecadal modulation of PDO on the impact of ENSO on the East Asian winter monsoon. Geophys. Res. Lett., 35, L20702, https://doi.org/10.1029/2008GL035287.
Wang, S.-Y., and J. Liu, 2016: Delving into the relationship between autumn Arctic sea ice and central–eastern Eurasian winter climate. Atmos. Ocean. Sci. Lett., 9, 366–374, https://doi.org/10.1080/16742834.2016.1207482.
Wang, T., O. H. Otterå, Y. Gao, and H. Wang, 2012: The response of the North Pacific decadal variability to strong tropical volcanic eruptions. Climate Dyn., 39, 2917–2936, https://doi.org/10.1007/s00382-012-1373-5.
Wang, T., H. J. Wang, O. H. Otterå, Y. Q. Gao, L. L. Suo, T. Furevik, and L. Yu, 2013: Anthropogenic agent implicated as a prime driver of shift in precipitation in eastern China in the late 1970s. Atmos. Chem. Phys., 13, 12 433–12 450, https://doi.org/10.5194/acp-13-12433-2013.
Wang, T., D. Guo, Y. Q. Gao, H. J. Wang, F. Zheng, Y. L. Zhu, J. P. Miao, and Y. Y. Hu, 2018: Modulation of ENSO evolution by strong tropical volcanic eruptions. Climate Dyn., 51, 2433–2453, https://doi.org/10.1007/s00382-017-4021-2.
Wang, Y., S. Li, and D. Luo, 2009: Seasonal response of Asian monsoonal climate to the Atlantic multidecadal oscillation. J. Geophys. Res., 114, D02112, https://doi.org/10.1029/2008JD010929.
Wu, B., and J. Wang, 2002: Winter Arctic Oscillation, Siberian high and East Asian winter monsoon. Geophys. Res. Lett., 29, 1897, https://doi.org/10.1029/2002GL015373.
Xu, J., Q. Zhu, and T. Zhou, 1999: Sudden and periodic changes of East Asian winter monsoon in the past century (in Chinese). Quart. J. Appl. Meteor., 10, 1–8.
Yan, H., W. Duan, and Z. Xiao, 2003: A study on relation between East Asian winter monsoon and climate change during raining season in China (in Chinese). J. Trop. Meteor., 19, 367–376.
Yan, Q., H. Wang, O. M. Johannessen, and Z. Zhang, 2014: Greenland ice sheet contribution to future global sea level rise based on CMIP5 models. Adv. Atmos. Sci., 31, 8–16, https://doi.org/10.1007/s00376-013-3002-6.
Yang, S., and Q. Zhu, 1990: Oscillation and its relation to cold air activities in Asian winter (in Chinese). J. Nanjing Inst. Meteor., 13, 339–347.
Yang, X., Y. Zhu, Q. Xie, X. Ren, and G. Xu, 2004: Advances in studies of Pacific decadal oscillation (in Chinese). Chin. J. Atmos. Sci., 28, 979–992.
Zanchettin, D., C. Timmreck, H.-F. Graf, A. Rubino, S. Lorenz, K. Lohmann, K. Krüger, and J. H. Jungclaus, 2012: Bi-decadal variability excited in the coupled ocean–atmosphere system by strong tropical volcanic eruptions. Climate Dyn., 39, 419–444, https://doi.org/10.1007/s00382-011-1167-1.
Zhou, W., 2017: Impact of Arctic amplification on East Asian winter climate. Atmos. Ocean. Sci. Lett., 10, 385–388, https://doi.org/10.1080/16742834.2017.1350093.
Zhou, W., X. Wang, T. J. Zhou, C. Y. Li, and J. C. L. Chan, 2007a: Interdecadal variability of the relationship between the East Asian winter monsoon and ENSO. Meteor. Atmos. Phys., 98, 283–293, https://doi.org/10.1007/s00703-007-0263-6.
Zhou, W., C. Y. Li, and X. Wang, 2007b: Possible connection between Pacific oceanic interdecadal pathway and East Asian winter monsoon. Geophys. Res. Lett., 34, L01701, https://doi.org/10.1029/2006GL027809.
Zhu, Y. L., and H. J. Wang, 2010: The Arctic and Antarctic Oscillations in the IPCC AR4 coupled models. J. Meteor. Res., 24, 176–188.
Zhu, Y. L., H. J. Wang, J. H. Ma, T. Wang, and J. Q. Sun, 2015: Contribution of the phase transition of Pacific decadal oscillation to the late 1990s’ shift in East China summer rainfall. J. Geophys. Res. Atmos., 120, 8817–8827, https://doi.org/10.1002/2015JD023545.