Responses of ENSO and NAO to the external ...

Quaternary International xxx (2017) 1e13

Contents lists available at ScienceDirect

Quaternary International journal homepage: www.elsevier.com/locate/quaint

Responses of ENSO and NAO to the external radiative forcing during the last millennium: Results from CCSM4 and MPI-ESM-P simulations Tingting Xu a, b, *, Zhengguo Shi b, c, **, Zhisheng An b a Zhejiang “Environmental Pollution & Control” Journal Press, Environmental Science Research & Design Institute of Zhejiang Province, Hangzhou, 310007, China b State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, Xi'an, 710061, China c CAS Center for Excellence in Tibetan Plateau Earth Sciences, Beijing, 100101, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 June 2017 Received in revised form 20 December 2017 Accepted 20 December 2017 Available online xxx

~ oeSouthern Oscillation (ENSO) and North Atlantic The decadalecentennial variations of the El Nin Oscillation (NAO) are analyzed, based on the outputs of last millennium (LM) and Historical experiments (850e2005 AD) and control (CTL) experiments from two climate models Community Climate System Model, version 4 (CCSM4) and Max Planck Institute for Meteorology Paleoclimate Model (MPI-ESM-P). Variation of ENSO and NAO among the Medieval Warm Period (MWP, 1050e1150 AD), Little Ice Age (LIA, 1600e1700 AD) and 20th Century Warming (20CW, 1905e2005 AD) are focused. Significant bicentennial and multi-decadal periods are detected in the ENSO and NAO series, respectively. Both models have exerted statistically-significant contrasts of ENSO and NAO phases among the three typical ~ o-like state during the MWP, a La Nin ~ a-like state periods. The model-independent shifts of an El Nin ~ o-like state during the 20CW are captured. However, the NAO series display a during the LIA and an El Nin model-dependence. CCSM4 features a negative-phased NAO during the LIA, while MPI-ESM-P is characterized by a positive-phased NAO in the same period. The shifts of the ENSO and NAO phases are mainly due to the external radiative forcing, including the solar-volcanic (SV) forcing and greenhouse gases (GHG) forcing. © 2017 Elsevier Ltd and INQUA. All rights reserved.

Keywords: ENSO NAO Last millennium External radiative forcing CCSM4 MPI-ESM-P

1. Introductions ~ oeSouthern Oscillation (ENSO) and North Atlantic The El Nin Oscillation (NAO) are two important internal phenomena in climate system, each of which can impact hemispheric and even global climate. The ENSO originates in the tropical Pacific Ocean with ~ o) and a cold periodic oscillation between a warm phase (El Nin ~ a). An El Nin ~ o event features an increase of the eastern phase (La Nin Pacific sea surface temperature (SST), a decrease of equatorial ~a easterlies and a flattening of the thermocline slope; while a La Nin event roughly leads to the opposite effect (Rasmusson and ~ o (La Nin ~ a) is assoCarpenter, 1982; Neelin et al., 1998). The El Nin ciated with decreased (increased) summer precipitation in India

* Corresponding author. Zhejiang “Environmental Pollution & Control” Journal Press, Environmental Science Research & Design Institute of Zhejiang Province, Hangzhou, 310007, China. ** Corresponding author. State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, Xi'an, 710061, China. E-mail addresses: [email protected] (T. Xu), [email protected] (Z. Shi).

and northern Australia, and increased (decreased) winter precipitation in southeast Asia (Ropelewski and Halpert, 1987, 1989; Lau and Nath, 2000; Deser et al., 2010). The atmospheric manifesta~ o/La Nin ~ a is the Southern Oscillation (SO). It is a tion of El Nin tropical eastewest seesaw in southern Pacific sea level pressure (SLP) (Walker and Bliss, 1932). A strong (weak) SO state corre~ a (El Nin ~ o) event. The NAO is the major atmosponds to a La Nin spheric mode over North Atlantic, characterized by a bipolar seesaw between the Azores High and Icelandic Low. The positive phases of NAO are accompanied by a stronger northesouth pressure gradient and westerlies, increased temperatures and precipitation across northern Europe and decreased temperature in Greenland, southern Europe and Middle East. During the negative phases of NAO, the reverse phenomena are observed (Wallace and Gutzler, 1981; Hurrell, 1995; Hurrell and van Loon, 1997; Wanner et al., 2001; Pinto and Raible, 2012). Because of the importance in the global climate system, the variations of the ENSO and NAO have become one of the hot topics of recent climate researches. On interannual timescale, ENSO displayed 2e7 years' variation (D'Arrigo et al., 2005; Deser et al., 2006;

https://doi.org/10.1016/j.quaint.2017.12.038 1040-6182/© 2017 Elsevier Ltd and INQUA. All rights reserved.

Please cite this article in press as: Xu, T., et al., Responses of ENSO and NAO to the external radiative forcing during the last millennium: Results from CCSM4 and MPI-ESM-P simulations, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.038

2

T. Xu et al. / Quaternary International xxx (2017) 1e13

Emile-Geay et al., 2013) while on decadaleinterdecadal timescale, it shows 10e90 years' variation (Li et al., 2011, 2013; Man and Zhou, 2011; Chowdary et al., 2012). NAO owns variations among 2e3, 7e8 years on interannual timescale (Hurrell and van Loon, 1997; Cook et al., 1998; Hurrell et al., 2003) and 20, 50e70 years on decadaleinterdecadal timescale (Mann and Park, 1994; Schlesinger and ~ o-like” and Ramankutty, 1994; Mann et al., 1995). The terms “El Nin ~ a-like” are defined to denote patterns of mean change (or “La Nin trend) in the equatorial tropical Pacific, which resemble the present ~ o and La Nin ~ a events. An apparent higher eastern Pacific day El Nin SST (Fedorov and Philander, 2000) and a lower SO index (Power and Smith, 2007) in the end of 20th century have been noted. These ~ o-like response to global warming. changes indicate an El Nin ~ a-like response, with However, Cane et al. (1997) showed a La Nin the eastern equatorial Pacific cooled and the zonal SST gradient strengthened since 1900, which was supported by Karnauskas et al. (2009). Since 1980, the NAO has tended to remain in a positive phase (Hurrell, 1995; Jones et al., 1997; Moritz et al., 2002). The observational records of SLP and SST are too short to detect their long-time changes. To understand the full range of variability in ENSO and NAO during the historical period, paleoclimatic proxy records have been used. The changes in ENSO and NAO seem complicated and controversial during the last millennium among ~ o-like the proxies. Several paleoclimate records suggest an El Nin (Tierney et al., 2010; Yan et al., 2011) mean state and a persistent positive NAO mode (Trouet et al., 2009, 2012) during the Medieval Warm Period (MWP), while the situation is quite the opposite during ~ o-like conditions (Conroy et al., the Little Ice Age (LIA). The El Nin 2008; Emile-Geay et al., 2013) and positive-phased NAO (Luterbacher et al., 2001) also occur during the 20th Century Warming (20CW). In contrast, colder SSTs over central tropical Pacific during the MWP (Chen et al., 2015; Cobb et al., 2003; Mann et al., ~ a events 2005, 2009; Liu et al., 2015), and more frequent La Nin during the 20CW (Stahle et al., 1998) are delineated, claiming La ~ a-like mean states in the typical warm periods of the last milNin lennium. Evident from the subtropical North Atlantic is associated with a persistent, enhanced positive NAO pattern during the LIA (Richey et al., 2009). Olsen et al. (2012) discovered that the onset of the MWP was not associated with any notable changes in the NAO. Due to the limited proxies, it is difficult to detect the forcingresponse mechanisms of the characteristics of ENSO and NAO merely using reconstruction data. In this aspect, numerical model simulations with forced and control experiments can provide useful supplements. Thus, a coupled climate model output analysis is conducted here in order to examine the forced response of ENSO and NAO during the last millennium, as well as the sensitivity of different models to the external radiative forcing. Following questions are mainly addressed in this study: (1) How does ENSO and NAO response to the external forcing on decadal-centennial timescale? (2) Can the models detect any differences for ENSO and NAO among the MWP, LIA and 20CW? (3) What are the consistence and difference between different models? A brief description on the models and data is presented in section 2. The results are analyzed and discussed in section 3 and 4, respectively. Section 5 is a summary of this study.

Paleoclimate Model (MPI-ESM-P), which are available in Paleoclimate Modeling Intercomparison Project 3 (PMIP3). The data can be downloaded at http://esgf-data.dkrz.de/. Both models are coupled oceaneatmosphere models. CCSM4 consists of Community Atmosphere Mode, version 4 (CAM4), Community Land Model, version 4 (CLM4), Parallel Ocean Program, version 2 (POP2) and Community Ice Code, version 4 (CICE4). CAM4 is run at a uniform resolution of 1.25 in latitude by 0.9 in longitude with 26 layers in the vertical (Neale et al., 2013); CLM4 uses the same horizontal grids as CAM4; POP2 owns nominal 1 1 horizontal resolution and 60 layers in the vertical (Danabasoglu et al., 2012); CICE4 adopts the same horizontal resolution with POP2. More details about CCSM4 can be found from Landrum et al. (2013). MPI-ESM-P is composed of the atmospheric general circulation model ECHAM6 and the Max Planck Institute Ocean Model (MPI-OM). ECHAM6 is run at the horizontal resolution of 1.9 1.9 with 47 vertical levels (Stevens et al., 2013). MPI-OM employs a conformal mapping grid with a horizontal grid of 1.5 1.5 and 40 vertical levels (Marsland et al., 2003). More detail about MPI-ESM-P has been described by Jungclaus et al. (2013). The LM simulation starts at 850 AD and continues to 1850 AD, where it matches up the Historical simulation that ends in December 2005 AD. So the LM simulations in our study are from 850 AD to 2005 AD. The forcing and boundary conditions are discussed in detail by Schmidt et al. (2011). Note that the LM simulations from both models adopt the reconstruction of Vieira et al. (2011) for solar irradiation and MacFarling Meure et al. (2006) for greenhouse gases (GHG). However, the volcanic forcing in CCSM4 and MPI-ESM-P is from the reconstructions of Gao et al. (2008) and Crowley (2000), respectively. The CTL experiments of both models impose non-evolving preindustrial conditions and integrate for about 1000 years. The SST, SLP and surface air temperature (SAT) data were used in our research. 2.2. Definitions of the atmospheric oscillation indices ~ o3.4 index is calculated as the averaged SST anomalies The Nin over the central and eastern tropical Pacific region (5 Se5 N, 170 e120 W): ~ o3.4 index ¼ SST(5 Se5 N, 170 e120 W) Nin

(1)

The SO index is defined as normalized SLP anomalies between the Tahiti (20 e15 S, 148 W) and Darwin (15 e10 S, 131 E): SO index ¼ SLP(20 e15 S, 148 W)-SLP(15 e10 S, 131 E)

(2)

The NAO index is computed using normalized SLP anomalies between the Azores High (35 S, 10 We10 E) and Icelandic Low (65 N, 30 e10 W): NAO index ¼ SLP(35 S, 10 We10 E)-SLP(65 N, 30 e10 W)

(3)

All indices are focused on the DecembereJanuaryeFebruary means, as the ENSO and NAO are strongest in boreal winter. 2.3. Meteorological and reconstruction data

2. Models and data 2.1. Models and experiments description The modeling data used in this study is from the last millennium (LM), Historical and Preindustrial control (CTL) experiments by two coupled climate modelseCommunity Climate System Model, version 4 (CCSM4) and Max Planck Institute for Meteorology

The meteorological data, including the SST and SLP, are derived from the Kaplan data set (1856e2011 AD) and National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data (1948e2013 AD), respectively. In addition, a number of reconstructed atmospheric oscillation series are employed in model comparisons comprehensively



3

Fig. 1. Northern Hemispheric surface air temperature (SAT) anomaly during the last millennium derived from (a) CCSM4, (b) MPI-ESM-P and (c) reconstruction. The reconstructed Northern Hemispheric SAT is derived from Mann et al. (1999). The time series have been smoothed with a 31-year low-pass filter. The gray bars denote the three typical periods selected as Medieval Warm Period (MWP), Little Ice Age (LIA) and 20th century global warming (20CW).

considering their data accuracy and time span. (1) The Northern Hemisphere temperature series (1000e1980 AD, in annual resolution) originate from Mann et al. (1999). ~ o3.4 series (1150e1995 AD, in annual resolution) is (2) The Nin provided by Emile-Geay et al. (2013). (3) The SO series (1699e1971 AD, in annual resolution) is from Stahle and Cleaveland (1993).

(4) The NAO series (1049e1995 AD, after a 30-year cubic spline) is reconstructed by Trouet et al. (2009). 3. Results 3.1. Last millennium variations In order to validate the decadalecentennial timescale variation

~ o3.4 (black) and SO index (gray) during the last millennium derived from (a) CCSM4, (b) MPI-ESM-P and (c) reconstruction. The reconstructed Nin ~ o3.4 and SO are Fig. 2. Winter Nin derived from Emile-Geay et al. (2013) and Stahle and Cleaveland (1993), respectively. The time series have been smoothed with a 31-year low-pass filter.


4


in the forced runs, the time series of the simulated SAT averaged over the Northern Hemisphere are presented in comparison with the reconstruction (Fig. 1). All time series show a warm interval 1000e1200 AD (generally considered as the MWP), a following transition interval till 1600 AD, the 17th century cold period (generally considered as the LIA), the 18th century recovery, a cold period in the early 19th century and a rapid warming period in the 20th century (generally considered as the 20CW). Moreover, the mid-15th cooling is also captured in the simulations and reconstruction. In summary, the agreement between the simulated and reconstructed temperature add confidence to our subsequent analysis of the ENSO and NAO using outputs from the forced runs. ~ o3.4 indices display multi-centennial The simulated Nin

fluctuations during the last millennium (Fig. 2a and b). Extreme low values in 13th, 15th and early 19th century are detected, denoting ~ a-like states. After 1900 AD, a fast increasing trend occurs. La Nin ~ a-like states near The MPI-ESM-P represents other obvious La Nin 1000 AD and 1700 AD. The upward trend in 20th century is also ~ o3.4 index. Besides, the conreflected in the reconstructed Nin ~ a-like condition during the whole struction exhibits a strong La Nin 18th century, which is not exactly reproduced in the simulations (Fig. 2c). The power spectra for the simulated and reconstructed ~ o3.4 series show common periods of 200e300 years (Fig. 3a, c, Nin ~ o3.4 3e), confirming the multi-centennial oscillations in the Nin series during the last millennium. Additionally, the reconstructed ~ o3.4 series own multi-decadal periods (Fig. 3e). The SO series Nin

~ o3.4 (left panels) and SO series (right panels) derived from CCSM4 (a, b), MPI-ESM-P (c, d) and reconstruction (e, f). The dashed lines indicate the 90% Fig. 3. Power spectra of the Nin and 95% confidence levels, respectively.



~ o3.4 series, but not completely. are almost consistent with the Nin ~ o-like For example, CCSM4 simulated SO series indicates an El Nin ~ o3.4 series denotes a La Nin ~ acondition near 1800 AD, while the Nin like condition (Fig. 2a). The power spectra of the SO series pronounce 40-yr and 200-yr peaks; however, the peaks are not significant, except for the 40-yr period in the SO series from MPI-ESMP (Fig. 3b, d, 3f). For the NAO index, good phase coherence can be seen in simulations before 1100 AD and after 1800 AD (Fig. 4a and b). However, in the intervals 1100e1800 AD, the simulated NAO series seem antiphased. Especially near the years 1250 AD, 1380 AD, 1550 AD, 1650 AD and 1780 AD, the peaks of NAO series in CCSM4 coincide with the valleys in MPI-ESM-P, and vice versa. The reconstructed NAO series displays a persistent positive NAO before 1400 AD, after which the NAO becomes positive-negative alternating (Fig. 4c). The positive phase near 1080 AD and negative phase near 1450 AD are detected both in simulations and reconstruction. The multi-decadal periods of simulated NAO series are diagnosed from the power spectra. What's more, the approximate bi-centennial cycle is rather stronger in the NAO series from MPI-ESM-P than those from CCSM4 and reconstruction (Fig. 5). 3.2. Differences among the MWP, LIA and 20CW The MWP, LIA and 20CW are three typical periods during the last millennium. Based on the SAT from simulations, the MWP, LIA and 20CW are defined as the intervals with same length: 1051e1150 AD, 1601e1700 AD and 1906e2005 AD, respectively (Fig. 1a and b). The ~ o3.4 series indicate that there are an El Nin ~ o-like simulated Nin ~ a-like state in the LIA and an El Nin ~ o-like state in the MWP, a La Nin ~ o3.4 index are state in the 20CW (Fig. 2). The differences in the Nin significant among MWP, LIA and 20CW, except for the contrast between MWP and LIA from CCSM4 simulation (Table 1). Besides, ~ o-like state in the 20CW is stronger than the MWP, the El Nin ~ o-like state in the 20CW may reflect implying the stronger El Nin

5

the impact of GHG. What are the spatial patterns of ENSO during the three typical periods? To address this question, empirical orthogonal function (EOF) analysis is adopted to examine the principle modes of the SST over Pacific. The leading EOF modes of the SST from the simulations show remarkable warm anomaly in the central and eastern tropical Pacific, as well as two cold anomalies north and south of the warm area (Fig. 6a and f). However, in the LIA of CCSM4 simulation, the south cold anomaly disappears, and then in the 20CW, both of the cold anomalies vanish. The simulated ENSO patterns are similar to the pattern from Kaplan data (Fig. 6g), but the anomaly centers in simulations are relatively weaker. Besides, the explained variance in CCSM4 is close to the Kaplan data, while that in MPI-ESM-P is much lower. In the CCSM4 simulation, the warming over the tropical Pacific and cooling over the Sea of Okhotsk may appear in phase (Fig. 6a, c, 6e). During the MWP, the strong tropical Pacific warming is companied by an obvious cooling over the Sea of Okhotsk. When the tropical Pacific warming is weakened during the LIA and 20CW, the cooling over the Sea of Okhotsk is instead by the warming. In the simulation from MPI-ESM-P, the tropical warm anomaly and the cold anomaly over the Sea of Okhotsk seem anti-correlated. The 20CW produces the strongest tropical warming, the MWP second and the LIA last. But the ranking of the Sea of Okhotsk cooling is just in an inverse order (Fig. 6b, d, 6f). The SST contrast in the preindustrial period (MWP minus LIA) and industrial period (20CW minus LIA), which indicates the SST response to the solar-volcanic (SV) forcing and GHG forcing, are shown in Fig. 7. The results are consistent with Table 1. For the CCSM4 simulation, an insignificant warming for MWP minus LIA exists over the central and eastern tropical Pacific. A significant warming for 20CW minus LIA exists over the eastern tropical Pacific, with enhanced warmer on eastern tropical Pacific than western tropical Pacific, which shows a weakened westeeast SST gradient. For the MPI-ESM-P simulation, the warming is significant over eastern tropical Pacific for MWP

Fig. 4. Winter NAO index during the last millennium derived from (a) CCSM4, (b) MPI-ESM-P and (c) reconstruction. The reconstructed NAO is derived from Trouet et al. (2009). The time series have been smoothed with a 31-year low-pass filter.


6

T. Xu et al. / Quaternary International xxx (2017) 1e13 Table 1 ~ o3.4 index and NAO index among three typical periods (MWP The contrast of Nin minus LIA, 20CW minus LIA and 20CW minus MWP), and the contrast of these typical periods and control simulation (MWP minus CTL, LIA minus CTL and 20CW minus CTL). Values followed by “*” and “**” are statistically significant at 90% and 95% confidence, respectively. Index

MWP-LIA 20CW-LIA 20CW-MWP MWP-CTL LIA-CTL 20CW-CTL

~ o3.4 index Nin

NAO index

CCSM4

MPI

CCSM4

MPI

0.22 0.69** 0.47** 0.10 0.12 0.57**

0.23** 0.47** 0.24** 0.12 0.11 0.36**

0.55** 0.23 0.32 0.22 0.33* 0.10

0.70** 0.56** 0.14 0.47** 0.23 0.33**

enhanced warmer on eastern tropical Pacific than western tropical Pacific, showing a weakened westeeast SST gradient. Thus, the SV ~ o-like condiforcing and GHG forcing may both leads to an El Nin tion; however, the response of ENSO to GHG forcing is stronger than SV forcing. For the NAO series, the responses in CCSM4 and MPI-ESM-P are inconsistent (Fig. 4a and b). A most distinguished feature in CCSM4 is the strong positive-phased NAO during the LIA. Conversely, the MPI-EMS-P simulation exhibits a negative-phased NAO during the LIA. In addition, an obvious positive phase is detected during the MWP in MPI-ESM-P, matching with the reconstructed NAO series (Fig. 4c). Table 1 confirms the results mentioned above and further points out that there is a significant positive-phased anomaly in the MPI-ESM-P simulated NAO during the 20CW compared with the LIA. The SLP EOF modes from the simulations capture the main features of the mode from NCEP/NCAR reanalysis: a northesouth dipole distribution and accounting for more than 50% of the variance of the Atlantic section (Fig. 8). But discrepancies still exist between different simulations. The action centers in CCSM4 show a northeastesouthwest tilt, while the action centers in MPI-ESM-P is northwestesoutheast tilted. For the CCSM4 simulation, the south action center experiences an eastward shift during the LIA, compared with the MWP and 20CW. For the MPI-ESM-P simulation, the south action center locates most eastward during the MWP, and then it moves westward during the LIA and more westward during the 20CW. In agreement with Table 1, the SLP contrast from CCSM4 indicates a significant weakened northesouth pressure gradient for MWP minus LIA (Fig. 9a). From the MPI-ESM-P simulation, strengthened gradients emerge for MWP minus LIA and 20CW minus LIA (Fig. 9c and d). Note that the SLP gradient for MWP minus LIA is stronger than that for the 20CW minus LIA. Although the forcing causes inverse NAO phases in CCSM4 and MPI-ESM-P simulations, both of the simulated NAO seem more sensitive to the SV forcing.

3.3. Impacts of external forcing and internal variability

Fig. 5. Power spectra of the NAO series derived from (a) CCSM4, (b) MPI-ESM-P and (c) reconstruction. The dashed lines indicate the 90% and 95% confidence levels, respectively.

minus LIA, and the warming stretching westward to cover the central and eastern tropical Pacific for 20CW minus LIA, both with

To distinguish the impacts of external forcing and internal variability on the ENSO and NAO states, the LM experiments are ~ o3.4 series do compared with CTL experiments. The variation of Nin not display decadalecentennial variation from CTL experiments (Fig. 10a and c), and the power spectra do not detect any significant periods (Fig. 11a and c). In contrast, the NAO series show decadalecentennial variation (Fig. 10b and d), and the multi-decadal periods are still significant in NAO series without external forcing (Fig. 11b and d). Also notice that the strength of bi-centennial peak from MPI-ESM-P is weakened to insignificant when the external ~ o3.4 and forcing removed. Thus the bi-centennial period in the Nin



7

Fig. 6. The leading EOF mode of winter Pacific SST (40 Se60 N, 120 E60 W) derived from CCSM4 (left panels) and MPI-ESM-P (right panels) during the MWP (a, b), LIA (c, d) and 20CW (e, f) periods and from (g) Kaplan data (1857e2011). The percentages in the top-right corners denote the explained variance.

NAO series may derive from the external forcing, primarily as a response to the effective solar forcing (Liu et al., 2009), and the multi-decadal period possibly originates from the internal variability.

~ o-like anomalies Compared with the CTL experiments, the El Nin during the 20CW are significant in both models. CCSM4 show a significant anomaly of NAO only during the LIA. The anomalous NAO phases are significant during the MWP and 20CW in MPI-


8


Fig. 7. Winter SST contrast in preindustrial (MWP minus LIA, left panels) and industrial (20CW minus LIA, right panels) periods derived from CCSM4 (upper panels) and MPI-ESM-P (lower panels). Gray shadings indicate areas where the contrasts are statistically significant at 95% confidence level. The units are K.

ESM-P, with a larger anomaly occur in the MWP (Table 1). These phenomena backup our findings that the ENSO and NAO may response more sensitively to the SV forcing and GHG forcing, respectively. Table 2 shows the ratio of positive- and negative~ o3.4 index and NAO index in the three typical periods phased Nin of LM experiments, as well as the whole CTL experiments. Both LM ~ o-like states during MWP and simulations display more El Nin ~ a-like states during LIA. As for the NAO, the 20CW, and more La Nin situation widely differs. In CCSM4 LM simulation, more negativephased NAO occurs during the MWP, and more positive-phased NAO emerges during the LIA and 20CW. wherever, the LM simulation of MPI-ESM-P exert more positive-phased NAO during the MWP and 20CW, and more negative-phased NAO during the LIA. The positive- and negative-phased ENSO and NAO states occupy the similar ratios in the CTL experiments, indicating it is the external forcing that causes the different ENSO and NAO states in MWP, LIA and 20CW. 4. Discussion The global climate system has undergone substantial changes during the last millennium (Shi et al., 2016; Xu et al., 2016), and the current anthropogenic warming is continuing. These climatic changes have the potential to affect the modes of natural variability such as ENSO and NAO. Many modeling simulations are adopted to investigate the changes of ENSO and NAO during the last millennium and their physical causes. Knutson et al. (1997) carried out 1000-year CO2 sensitivity experiments on an early low-resolution coupled model GCM and suggested that the impacts of increasing CO2 on ENSO is unlikely to be distinguishable from the internal variability of the climate system. In addition, many other model

simulations showed that decadal ENSO-like variation could be generated through an internal atmosphere-ocean coupling process (Timmermann and Jin, 2002; Rodgers et al., 2004; Dewitte et al., 2007; Choi et al., 2009). The response of ENSO to natural radiative forcing changes over the past millennium was explored employing the ZebiakeCane model by Mann et al. (2005) and Community Earth System Model (CESM) by Liu et al. (2013, 2016, 2017), which had been also demonstrated by other model simulations (Robock, 2000; Li et al., 2013; Ohba et al., 2013; Maher et al., ~ a-like 2015; Pausata et al., 2015). Their outputs showed a La Nin state during the MWP compared with LIA. The shift of ENSO phases was induced by the changes in volcanic and solar radiative forcing. ~ a-like state for the MWP minus LIA pattern is However, the La Nin not reproduced in either of the models NCEP CSM1.4 and GISS-ER (Mann et al., 2009). The discrepancy was considered to derive from the inactive “thermostate” mechanism over tropical Pacific (Clement et al., 1996) in these models. In many model projections of greenhouse warming, an El Nino like SST pattern occurs (Meehl and Washington, 1996; Held and Soden, 2006; Vecchi et al., 2006). The ~ o-likeeLa Nin ~ a-likeeEl Nin ~ o-like shift during the MWP, LIA El Nin and 20CW was clearly displayed in LM experiments of FGOALS_gl model (Man and Zhou, 2011). In our research, the shift of ENSO phases agrees with Man and Zhou (2011) for both of the models CCSM4 and MPI-ESM-P, highlighting the response of ENSO to external forcing. To understand the NAO variation over the last millennium, the model ECBILT-CLIO was used to perform 25 simulations, driven by the main natural and anthropogenic forcing. Averaged NAO series over the 25 simulations exhibited quite weak changes, except during the 20CW (Goosse et al., 2005). A 2000-year control simulation with CESM showed that NAO variation could still existed



9

Fig. 8. The leading EOF mode of winter North Atlantic SLP (20 Ne80 N, 90 We40 N) derived from CCSM4 (left panels) and MPI-ESM-P (right panels) during the MWP (a, b), LIA (c, d) and 20CW (e, f) periods and from (g) NCEP/NCAR reanalysis (1949e2013). The percentages in the top-right corners denote the explained variance.

without the external radiative forcing as driving factors (Wang et al., 2015). The enhanced positive-phased NAO during the 20CW might be triggered by the increasing GHG forcing. By the models NCEP CSM1.4 and GISS-ER, Mann et al. (2009) also examined the changes of NAO and evinced a prevailing positive NAO pattern during the MWP and a negative NAO condition during the LIA. The shift of NAO phases was suggested result from the solar forcing difference. The NAO series from FGOALS_gl was generally positive during the MWP, negative during the LIA and positive during the 20CW, implying the influence of external forcing (Man

and Zhou, 2011). Our study captures different changes of NAO phases from the LM experiments of CCSM4 and MPI-ESM-P. The NAO series from MPI-ESM-P is approximately consistent with Mann et al. (2009) and Man and Zhou (2011), while that from CCSM4 opposes. However, the response of NAO to external forcing was captured by both models. LM simulation of CCSM4 did not contain solar-related ozone changes, which might weaken the response of the climate system to solar radiation and greenhouse gases (Mann et al., 2009). The volcanic forcing in CCSM4 was from the reconstructions of Gao et al. (2008), while MPI-ESM-P from


Fig. 9. The SLP contrast in preindustrial period (MWP minus LIA, left panels) and industrial (20CW minus LIA, right panels) periods derived from CCSM4 (upper panels) and MPIESM-P (lower panels). Gray shadings indicate areas where the contrasts are statistically significant at 95% confidence level. The units are hPa.

~ o3.4 (left panels) and NAO index (right panels) derived from CCSM4 (upper panels) and MPI-ESM-P (lower panels) control (CTL) experiments. The time series Fig. 10. Winter Nin have been smoothed with a 31-year low-pass filter.



11

~ o3.4 series (left panels) and NAO series (right panels) derived from CCSM4 (a, b) and MPI-ESM-P (c, d) CTL experiments. The dashed lines indicate the Fig. 11. Power spectra of Nin 90% and 95% confidence levels, respectively.

Table 2 ~ o3.4 index and NAO index in MWP, The ratio of positive- and negative-phased Nin LIA, 20CW and control simulation. The positive and negative phases are chosen based on the criterion that the index must exceed ±1 standard deviation. Index

MWPþ MWPLIAþ LIA20CWþ 20CWCTLþ CTL-

~ o3.4 index Nin

NAO index

CCSM4

MPI

CCSM4

MPI

19.0% 8.0% 14.0% 18.0% 25.0% 4.0% 17.6% 17.0%

23.0% 10.0% 10.0% 19.0% 27.0% 10.0% 15.4% 15.8%

12.0% 20.0% 18.0% 7.0% 18.0% 14.0% 15.1% 15.8%

23.0% 14.0% 12.0% 19.0% 25.0% 13.0% 16.4% 16.6%

Crowley (2000), end with the volcanic forcing strength in CCSM4 two times of that in MPI-ESM-P. The different settings between CCSM4 and MPI-ESM-P might lead to the model-dependence of NAO. However, more research should be done before drawing a definite conclusion. The reconstructions shed light on the issue why the ENSO and NAO change during the last millennium. Cobb et al. (2003) implied that the majority of ENSO variability may have arisen from the internal dynamics. However, from the reconstruction of Yan et al. (2011), the solar irradiance and mean Northern Hemisphere

temperature fluctuations was suggested to exert an impact. Reconstruction results from Adams et al. (2003), D'Arrigo et al. (2009) and McGregor et al. (2010) proved volcanic forcing effect on the ENSO system. Conroy et al. (2008) indicated the strong El ~ o-like state during the 20CW might last with the continued Nin anthropogenic warming. The shift of NAO phases was possibly initiated by changes in solar irradiance (Gray et al., 2010) and volcanic activity (Ortega et al., 2015) and amplified by Atlantic meridional overturning circulation (AMOC) (Trouet et al., 2009, 2012). A reduction in the intensity and spatial extent of the Atlantic Warm Pool was consistent with the positive phase of NAO during the LIA (Richey et al., 2009). By comparing the LM experiments with CTL experiments from CCSM4 and MPI-ESM-P, we tend to support the impact of external forcing on the phases of the ENSO and NAO, including the SV forcing and GHG forcing. The ENSO phases seem more sensitive to GHG forcing while the NAO phases may respond more robust to the SV forcing. 5. Conclusions In order to evaluate the effect of external forcing, including the SV forcing and GHG forcing, the decadalecentennial timescale variations of the ENSO and NAO, as well as the differences among MWP, LIA and 20CW are investigated. Our study is based on the outputs from LM and CTL experiments by two models CCSM4 and


12


MPI-ESM-P. Our results show that: (1) The ENSO is found to vary in an apparent period of bicentennial years and the multi-decadal variation is highlighted in the NAO series. (2) Statistically-significant contrasts of ENSO and NAO phases, as well as spatial modes, among MWP, LIA and 20CW have exerted in both models. ~ o-like, La Nin ~ a-like and El (3) Both models agree on the El Nin ~ o-like mean states during the MWP, LIA and 20CW, Nin respectively. However, the changes of NAO phases show model-dependent. The NAO series from CCSM4 is characterized by a strong negative phase during the LIA. Conversely, the positive-phased NAO dominates the LIA in MPI-ESM-P. The MWP minus LIA and 20CW minus LIA patterns indicate the impact of SV forcing and GHG forcing on the shift of ENSO and NAO phases, which is confirmed subsequently by the comparison between LM and CTL experiments. What is the mechanism of the shift of ENSO and NAO phases response to external radiative forcing? What definitely cause the model-dependence of NAO variation in different models? Further investigations will be taken to provide a potential insight into these questions. Acknowledgements The authors thank the anonymous reviewers for their valuable suggestions. This work was jointly supported by the National Natural Science Foundation of China (41690115, 41572160) and CAS 00 Light of West China" Program and the Youth Innovation Promotion Association CAS. References Adams, J.B., Mann, M.E., Ammann, C.M., 2003. Proxy evidence for an El Nino-like response to volcanic forcing. Nature 426, 274e278. Cane, M.A., Clement, A.C., Kaplan, A., Kushnir, Y., Pozdnyakov, D., Seager, R., Zebiak, S.E., Murtugudde, R., 1997. Twentieth-century sea surface temperature trends. Science 275, 957. Chen, J., Chen, F., Feng, S., Huang, W., Liu, J., Zhou, A., 2015. Hydroclimatic changes in China and surroundings during the Medieval climate anomaly and Little ice Age: spatial patterns and possible mechanisms. Quat. Sci. Rev. 107, 98e111. Choi, J., An, S.I., Dewitte, B., Hsieh, W.W., 2009. Interactive feedback between the tropical Pacific decadal oscillation and ENSO in a coupled general circulation model. J. Clim. 22, 6597e6611. Chowdary, J.S., Xie, S., Tokinaga, H., Okumura, Y.M., Kubota, H., Johnson, N., Zheng, X., 2012. Interdecadal variations in ENSO teleconnection to the Indoewestern Pacific for 1870e2007. J. Clim. 25, 1722e1744. Clement, A.C., Seager, R., Cane, M.A., Zebiak, S.E., 1996. An ocean dynamical thermostat. J. Clim. 9, 2190e2196. ~ o/Southern OscillaCobb, K.M., Charles, C.D., Cheng, H., Edwards, R.L., 2003. El Nin tion and tropical Pacific climate during the last millennium. Nature 424, 271e276. Conroy, J.L., Restrepo, A., Overpeck, J.T., Steinitzkannan, M., Cole, J.E., Bush, M.B., Colinvaux, P.A., 2008. Unprecedented recent warming of surface temperatures in the eastern tropical Pacific Ocean. Nat. Geosci. 2, 46e50. Cook, E.R., D'Arrigo, R.D., Briffa, K.R., 1998. A reconstruction of The North Atlantic oscillation tree-ring chronologies from north America and Europe. Holocene 8, 9e17. Crowley, T.J., 2000. Causes of climate change over the past 1000 years. Science 289, 270e277. Danabasoglu, G., Bates, S., Briegleb, B.P., Jayne, S.R., Jochum, M., Large, W.G., Peacock, S., Yeager, S.G., 2012. The CCSM4 ocean component. J. Clim. 25, 1361e1389. Deser, C., Capotondi, A., Saravanan, R., Phillips, A.S., 2006. Tropical Pacific and Atlantic climate variability in CCSM3. J. Clim. 19, 2451e2481. Deser, C., Alexander, M.A., Xie, S.P., Phillips, A.S., 2010. Sea surface temperature variability: patterns and mechanisms. Annu. Rev. Mater. Sci. 2, 115e143. Dewitte, B., Yeh, S.W., Moon, B.K., Cibot, C., Terray, L., 2007. Rectification of ENSO variability by interdecadal changes in the equatorial background mean state in a CGCM simulation. J. Clim. 20, 2002e2021. D'Arrigo, R., Cook, E.R., Wilson, R.J., Allan, R., Mann, M.E., 2005. On the variability of ENSO over the past six centuries. Geophys. Res. Lett. 32, L03711. D'Arrigo, R., Wilson, R., Tudhope, A., 2009. The impact of volcanic forcing on tropical

temperatures during the past four centuries. Nat. Geosci. 2, 51e56. Emile-Geay, J., Cobb, K.M., Mann, M.E., Wittenberg, A.T., 2013. Estimating central equatorial Pacific SST variability over the past millennium. Part Ⅱ: reconstructions and implications. J. Clim. 26, 2329e2352. ~ o changing? Science 288, 1997e2002. Fedorov, A.V., Philander, S.G., 2000. Is El Nin Gao, C., Robock, A., Ammann, C., 2008. Volcanic forcing of climate over the past 1500 years: an improved ice core-based index for climate models. J. Geophys. Res.: Atmos. 113, 2036e2044. Goosse, H., Renssen, H., Timmermann, A., Bradley, R.S., 2005. Internal and forced climate variability during the last millennium: a model-data comparison using ensemble simulations. Quat. Sci. Rev. 24, 1345e1360. Gray, L.J., Beer, J., Geller, M., Haigh, J.D., Lockwood, M., Matthes, K., Cubasch, U., Fleitmann, D., Harrison, G., Hood, L., Luterbacher, J., Meehl, G.A., Shindell, D., van Geel, B., White, W., 2010. Solar influences on climate. Rev. Geophys. 48, RG4001. Held, I.M., Soden, B.J., 2006. Robust responses of the hydrological cycle to global warming. J. Clim. 19, 5686e5699. Hurrell, J.W., 1995. Decadal trends in the North Atlantic Oscillation: regional temperatures and precipitation. Science 269, 676e679. Hurrell, J.W., van Loon, H., 1997. Decadal variations in climate associated with The North Atlantic oscillation. Climatic Change 36, 301e326. Hurrell, J.W., Kushnir, Y., Ottersen, G., Visbeck, M., 2003. An Overview of the North Atlantic Oscillation: Climatic Significance and Environmental Impact (AGU Geophysical Monograph, Wastington D. C., USA). Jones, P., Jonsson, T., Wheeler, D., 1997. Extension to the North Atlantic Oscillation using early instrumental pressure observations from Gibraltar and South-West Iceland. Int. J. Climatol. 17, 1433e1450. Jungclaus, J.H., Fischer, N., Haak, H., Lohmann, K., Marotzke, J., Matei, D., Mikolajewicz, U., Notz, D., von Storch, J.S., 2013. Characteristics of the ocean simulations in the Max Planck Institute Ocean Model (MPIOM) the ocean component of the MPI-Earth system model. J. Adv. Model. Earth Syst. 5, 422e446. Karnauskas, K.B., Seager, R., Kaplan, A., Kushnir, Y., Cane, M.A., 2009. Observed strengthening of the zonal sea surface temperature gradient across the equatorial Pacific Ocean. J. Clim. 22, 4316e4321. Knutson, T.R., Manabe, S., Gu, D., 1997. Simulated ENSO in a global coupled oceaneatmosphere model: multidecadal amplitude modulation and CO2 sensitivity. J. Clim. 10, 138e161. Landrum, L., Otto-Bliesner, B.L., Wahl, E.R., Conley, A., Lawrence, P.J., Rosenbloom, N., Teng, H., 2013. Last millennium climate and its variability in CCSM4. J. Clim. 26, 1085e1111. Lau, N.C., Nath, M.J., 2000. Impact of ENSO on the variability of the AsianeAustralian monsoons as simulated in GCM experiments. J. Clim. 13, 4287e4309. Li, J., Xie, S., Cook, E.R., Huang, G., D'Arrigo, R., Liu, F., Ma, J., Zheng, X., 2011. ~ o amplitude during the past millennium. Nat. Interdecadal modulation of El Nin Clim. Change 1, 114e118. Li, J., Xie, S., Cook, E., Morales, M.S., Christie, D.A., Johnson, N.C., Chen, F., D'Arrigo, R., ~ o modulations over the past seven Fowler, A.M., Gou, X., Fang, K., 2013. El Nin centuries. Nat. Clim. Change 3, 822e826. Liu, J., Wang, B., Ding, Q., Kuang, X., Soon, W., Zorita, E., 2009. Centennial variations of the global monsoon precipitation in the last millennium: results from ECHOG model. J. Clim. 22, 2356e2371. Liu, J., Wang, B., Cane, M.A., Yim, S., Lee, J., 2013. Divergent global precipitation changes induced by natural versus anthropogenic forcing. Nature 494, 656e659. Liu, F., Chai, J., Huang, G., Chen, Z., 2015. Modulation of decadal ENSO-like variation by effective solar radiation. Dynam Oceans 72, 52e61. Liu, F., Chai, J., Wang, B., Liu, J., Zhang, X., Wang, Z., 2016. Global monsoon precipitation responses to large volcanic eruptions. Sci. Rep. 2016 (6), 24331. ~ o reLiu, F., Li, J., Wang, B., Liu, J., Li, T., Gang, G., Wang, Z., 2017. Divergent El Nin sponses to volcanic eruptions at different latitudes over the past millennium. Clim. Dynam. 1e14. Luterbacher, J., Xoplaki, E., Dietrich, D., Jones, P., Davies, T., Portis, D., GonzalezRouco, J., Von Storch, H., Gyalistras, D., Casty, C., 2001. Extending North Atlantic oscillation reconstructions back to 1500. Atmos. Sci. Lett. 2, 114e124. MacFarling Meure, C., Etheridge, D., Trudinger, C., Steele, P., Langenfelds, R., van Ommen, T., Smith, A., Elkins, J., 2006. Law dome CO2, CH4 and N2O ice core records extended to 2000 years BP. Geophys. Res. Lett. 33, 70e88. Maher, N., McGregor, S., England, M.H., Gupta, A.S., 2015. Effects of volcanism on tropical variability. Geophys. Res. Lett. 42, 6024e6033. Man, W., Zhou, T., 2011. Forced response of atmospheric oscillations during the last millennium simulated by a climate system model. Chin. Sci. Bull. 56, 3042e3052. Mann, M.E., Park, J., 1994. Global-scale modes of surface temperature variability on interannual to century timescales. J. Geophys. Res. 99, 25819e25833. Mann, M.E., Park, J., Bradley, R., 1995. Global interdecadal and century-scale climate oscillations during the past five centuries. Nature 378, 266e270. Mann, M.E., Bradley, R.S., Hughes, M.K., 1999. Northern hemisphere temperatures during the past millennium: inferences, uncertainties, and limitations. Geophys. Res. Lett. 26, 759e762. Mann, M.E., Cane, M.A., Zebiak, S.E., Clement, A., 2005. Volcanic and solar forcing of the tropical Pacific over the past 1000 years. J. Clim. 18, 447e456. Mann, M.E., Zhang, Z., Rutherford, S., Bradley, R.S., Hughes, M.K., Shindell, D., Ammann, C., Faluvegi, G., Ni, F., 2009. Global signatures and dynamical origins of the Little ice Age and medieval climate anomaly. Science 326, 1256e1260. €ske, F., 2003. The Max Planck Marsland, S.J., Haak, H., Jungclaus, J.H., Latif, M., Ro


T. Xu et al. / Quaternary International xxx (2017) 1e13 Institute global ocean/sea ice model with orthogonal curvilinear coordinates. Ocean Model. 5, 91e127. McGregor, S., Timmermann, A., Timm, O., 2010. A unified proxy for ENSO and PDO variability since 1650. Clim. Past 6, 1e17. ~ o-like climate change in a model with Meehl, G.A., Washington, W.M., 1996. El Nin increased atmospheric CO2 concentrations. Nature 382, 56e60. Moritz, R.E., Bitz, C.M., Steig, E.J., 2002. Dynamics of recent climate change in the Arctic. Science 297, 1497e1502. Neale, R.B., Richter, J.H., Park, S., Lauritzen, P.H., Vavrus, S.J., Rasch, P.J., Zhang, M., 2013. The mean climate of the Community Atmosphere Model (CAM4) in forced SST and fully coupled experiments. J. Clim. 26, 5150e5168. Neelin, J.D., Battisti, D.S., Hirst, A.C., Jin, F.F., Wakata, Y., Yamagata, T., Zebiak, S.E., 1998. ENSO theory. J. Geophys. Res.: Oceans 103, 14261e14290. Ohba, M., Shiogama, H., Yokohata, T., Watanabe, M., 2013. Impact of strong tropical volcanic eruptions on ENSO simulated in a coupled GCM. J. Clim. 26, 5169e5182. Olsen, J., Anderson, N.J., Knudsen, M.F., 2012. Variability of The North Atlantic oscillation over the past 5,200 years. Nat. Geosci. 5, 808e812. Ortega, P., Lehner, F., Swingedouw, D., Masson-Delmotte, V., Raible, C.C., Casado, M., Yiou, P., 2015. A model-tested North Atlantic Oscillation reconstruction for the past millennium. Nature 523, 71e74. Pausata, F.S.R., Chafik, L., Caballero, R., Battisti, D.S., 2015. Impacts of high-latitude volcanic eruptions on ENSO and AMOC. Proc. Natl. Acad. Sci. Unit. States Am. 112, 13784e13788. Pinto, J.G., Raible, C.C., 2012. Past and recent changes in the North Atlantic oscillation. Wiley Interdiscip. Rev.: Climatic Change 3, 79e90. Power, S.B., Smith, I.N., 2007. Weakening of the Walker Circulation and apparent ~ o both reach record levels, but has ENSO really changed? dominance of El Nin Geophys. Res. Lett. 34, L18702. Rasmusson, E.M., Carpenter, T.H., 1982. Variations in tropical sea surface tempera~ o. ture and surface wind fields associated with the Southern Oscillation/El Nin Mon. Weather Rev. 110, 354e384. Richey, J.N., Poore, R.Z., Flower, B.P., Quinn, T.M., Hollander, D.J., 2009. Regionally coherent Little ice Age cooling in the Atlantic warm Pool. Geophys. Res. Lett. 36, 272e277. Robock, A., 2000. Volcanic eruptions and climate. Rev. Geophys. 38, 191e219. Rodgers, K.B., Friederichs, P., Latif, M., 2004. Tropical Pacific decadal 420 variability and its relation to decadal modulations of ENSO. J. Clim. 17, 3761e3774. Ropelewski, C.F., Halpert, M.S., 1987. Global and regional scale precipitation patterns ~ o/Southern Oscillation. Mon. Weather Rev. 115, associated with the El Nin 1606e1626. Ropelewski, C.F., Halpert, M.S., 1989. Precipitation patterns associated with the high index phase of the Southern Oscillation. J. Clim. 2, 268e284. Schlesinger, M.E., Ramankutty, N., 1994. An oscillation in the global climate system of period 65e70 years. Nature 367, 723e726. Schmidt, G.A., Jungclaus, J.H., Ammann, C.M., Bard, E., Braconnot, P., Crowley, T.J., Delaygue, G., Joos, F., Krivova, N.A., Muscheler, R., 2011. Climate forcing reconstructions for use in PMIP simulations of the last millennium (v1.0). Geosci.

13

Model Dev. (GMD) 4, 1549e1586. Shi, Z., Xu, T., Wang, H., 2016. Sensitivity of Asian climate change to radiative forcing during the last millennium in a multi-model analysis. Global Planet. Change 139, 195e210. Stahle, D.W., Cleaveland, M.K., 1993. Southern oscillation extremes reconstructed from tree rings of the Sierra Madre Occidental and southern Great Plains. J. Clim. 6, 129e140. Stahle, D.W., Cleaveland, M.K., Therrell, M.D., Gay, D., Darrigo, R., Krusic, P.J., Cook, E.R., Allan, R., Cole, J., Dunbar, R.B., 1998. Experimental dendroclimatic reconstruction of the southern oscillation. Bull. Am. Meteorol. Soc. 79, 2137e2152. Stevens, B., Giorgetta, M., Esch, M., Mauritsen, T., Crueger, T., Rast, S., Salzmann, M., Schmidt, H., Bader, J., Block, K., Brokopf, R., Fast, I., Kinne, S., Kornblueh, L., Lohmann, U., Pincus, R., Reichler, T., Roeckner, E., 2013. Atmospheric component of the MPI-m Earth system model: ECHAM6. J. Adv. Model. Earth Syst. 5, 146e172. Tierney, J.E., Oppo, D.W., Rosenthal, Y., Russell, J.M., Linsley, B.K., 2010. Coordinated hydrological regimes in the Indo-Pacific region during the past two millennia. Paleoceanography 25, 118. ~o Timmermann, A., Jin, F.F., 2002. A nonlinear mechanism for decadal El Nin amplitude changes. Geophys. Res. Lett. 29, 1003. Trouet, V., Esper, J., Graham, N.E., Baker, A., Scourse, J.D., Frank, D., 2009. Persistent positive North Atlantic oscillation mode dominated the Medieval climate anomaly. Science 324, 78e80. Trouet, V., Scourse, J.D., Raible, C.C., 2012. North Atlantic storminess and Atlantic meridional overturning circulation during the last millennium: reconciling contradictory proxy records of NAO variability. Global Planet. Change 84/85, 48e55. Vecchi, G.A., Soden, B.J., Wittenberg, A.T., Held, I.M., Leetmaa, A., Harrison, M.J., 2006. Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature 441, 73e76. Vieira, L.E.A., Solanki, S.K., Krivova, N.A., Usoskin, I.G., 2011. Evolution of the solar irradiance during the Holocene. Astron. Astrophys. 531, 576e587. Walker, G.T., Bliss, E.W., 1932. World weather Ⅴ. Q. J. Roy. Meteorol. Soc. 4, 53e84. Wallace, J.M., Gutzler, D.S., 1981. Teleconnexions in the geopotential height field during the Northern Hemisphere winter. Mon. Weather Rev. 109, 784e812. Wang, Z., Li, Y., Liu, B., Liu, J., 2015. Global climate internal variability in a 2000-year control simulation with Community Earth System Model (CESM). Chin. Geogr. Sci. 25, 263e273. Wanner, H., Bronnimann, S., Casty, C., Gyalistras, D., Luterbacher, J., Schmutz, C., Stephenson, D.B., Xoplaki, E., 2001. North Atlantic oscillation - concepts and studies. Surv. Geophys. 22, 321e381. Xu, T., Shi, Z., Wang, H., An, Z., 2016. Nonstationary impact of the winter North Atlantic Oscillation and the response of mid-latitude Eurasian climate. Theor. Appl. Climatol. 124, 1e14. Yan, H., Sun, L., Wang, Y., Huang, W., Qiu, S., Yang, C., 2011. A record of the Southern Oscillation Index for the past 2,000 years from precipitation proxies. Nat. Geosci. 4, 611e614.
