Influence of Atlantic meridional overturning circulation on the East Asian winter monsoon
大西洋经向翻转环流对东亚冬季风的影响

Youbin Sun $^{1 ⋆}$ , Steven C. Clemens $^{2}$ , Carrie Morrill $^{3}$ , Xiaopei Lin $^{4}$ , Xulong Wang $^{1}$ and Zhisheng An $^{1}$
Youbin Sun $^{1 ⋆}$ ， Steven C. Clemens $^{2}$ ， Carrie Morrill $^{3}$ ， Xiaopei Lin $^{4}$ ， Xulong Wang $^{1}$ 和 Zhisheng An $^{1}$

Abstract 抽象

The last glacial period was characterized by abrupt, millennial-scale climate change. These climate fluctuations are particularly pronounced in records of the East Asian monsoon system $^{1 - 8}$ , and seem to be linked to changes in North Atlantic circulation. Here we present records of grain size variations from the northwestern Chinese Loess Plateau, dated using optically stimulated luminescence. We reconstruct changes in the strength of the East Asian winter monsoon over the past 60,000 years and find reconstructed millennial-scale variations that are broadly correlated with temperature variations over Greenland, suggesting a common forcing. We investigate the effect of a slow-down of Atlantic meridional overturning circulation on the monsoon system using a coupled climate model simulation with added freshwater flux into the northern North Atlantic, and find a strengthening winter monsoon circulation over the regions that supply dust to the Loess Plateau and a reduction in summer monsoon precipitation over East Asia. We conclude that Atlantic meridional overturning circulation is a driver of abrupt change in the East Asian winter and summer monsoon systems, and that the northern westerlies play a role in transmitting this signal from the North Atlantic to the Asian monsoon regions.
上一个冰河期的特点是突然的、千年尺度的气候变化。这些气候波动在东亚季风系统的 $^{1 - 8}$ 记录中尤为明显，并且似乎与北大西洋环流的变化有关。在这里，我们展示了中国西北黄土高原粒度变化的记录，使用光刺激发光进行测年。我们重建了过去 60,000 年东亚冬季季风强度的变化，并发现了重建的千年尺度变化，这些变化与格陵兰岛的温度变化广泛相关，表明存在共同的强迫。我们使用耦合气候模型模拟和增加的淡水通量进入北大西洋北部，研究了大西洋经向翻转环流减慢对季风系统的影响，发现为黄土高原提供沙尘的地区冬季季风环流加强，东亚夏季季风降水减少。我们得出结论，大西洋经向翻转环流是东亚冬季和夏季季风系统突变的驱动因素，并且北部西风带在将这一信号从北大西洋传递到亚洲季风区方面发挥了作用。

Interactions between the tropics and high latitudes play a key role in the inter- and intra-hemispheric coupling of abrupt climate change. Greenland temperature, as an indication of highlatitude climate

^{9, 10}

, is largely dependent on the Atlantic meridional overturning circulation (AMOC; refs 11,12). By contrast, the Asian summer monsoon, a low-latitude process, is dynamically linked to the migration of the Intertropical Convergence Zone

^{13 - 15}

. Similarity of glacial millennial-scale climate variability recorded by Chinese speleothems

^{4}

and Greenland ice cores

^{9, 10}

implies a plausible influence of high-latitude northern hemisphere climate on East Asian monsoon circulation

^{16, 17}

, although tropical processes have also been implicated as possible drivers of millennial-scale climate change

^{18}

. Changes in the westerly jet and the Siberian-Mongolian High were previously considered as a mechanism linking abrupt climate changes in the North Atlantic and East Asia

^{1 - 5, 7, 19 - 21}

. The remote impacts of the North Atlantic climate change on the wind and precipitation variability in East Asia can be further assessed by integrating palaeoclimate proxy data and model simulations.
热带和高纬度地区之间的相互作用在突然气候变化的半球间和半球内耦合中起着关键作用。格陵兰岛温度作为高纬度气候

^{9, 10}

的指标，在很大程度上取决于大西洋经向翻转环流（AMOC;参考文献 11,12）。相比之下，亚洲夏季季风是一个低纬度过程，与热带辐合带

^{13 - 15}

的迁移动态相关。中国洞穴

^{4}

岩和格陵兰冰芯记录的冰川千年尺度气候变率的相似性

^{9, 10}

意味着北半球高纬度气候对东亚季风环流

^{16, 17}

的可能影响，尽管热带过程也被认为是千年尺度气候变化

^{18}

的可能驱动因素.西风急流和西伯利亚-蒙古高压的变化以前被认为是连接北大西洋和东亚气候变化突变的机制

^{1 - 5, 7, 19 - 21}

。通过整合古气候代理数据和模式模拟，可以进一步评估北大西洋气候变化对东亚风和降水变化的远程影响。

Unlike the African and Indian monsoons, the East Asian monsoon contains a unique mid-latitude component (a strong winter monsoon), which is characterized by prevailing lowlevel northwesterly winds associated with the Siberian-Mongolian

H_{i g h}^{22}

. The winter monsoon encompasses a larger meridional
与非洲和印度季风不同，东亚季风包含独特的中纬度成分（强烈的冬季季风），其特征是与西伯利亚-蒙古相关的盛行低层西北风

H_{i g h}^{22}

。冬季季风包括一个较大的经向
domain than the summer monsoon and can act as a medium to transmit climate signals from the high-latitude North Atlantic to the mid- and low-latitude regions

^{1, 3, 6, 7}

. Thus, reconstruction of past winter monsoon fluctuations is key to deciphering the hemispheric coupling between abrupt climate changes in the North Atlantic and East Asia. Here we report closely spaced optically stimulated luminescence (OSL) ages for two loess sequences in the northwestern Chinese Loess Plateau, yielding an independent chronology for assessing rapid winter monsoon shifts and a robust assessment of the coupling between Greenland temperature and East Asian monsoon variability. Furthermore, an integration of loess, speleothem and ice-core records with model simulations tests hypotheses about the dynamical causes of rapid climate changes.
域，可以作为介质将气候信号从北大西洋高纬度地区传输到中低纬度地区

^{1, 3, 6, 7}

。因此，重建过去的冬季季风波动是破译北大西洋和东亚气候变化突变之间的半球耦合的关键。在这里，我们报告了中国西北黄土高原两个黄土序列的紧密间隔的光刺激发光（OSL）年龄，产生了一个独立的年表，用于评估冬季季风的快速变化，并对格陵兰温度与东亚季风变化之间的耦合进行了稳健的评估。此外，黄土、洞穴岩和冰芯记录与模式模拟相结合，检验了关于快速气候变化动力学原因的假设。

Loess sequences at Jingyuan

({36.35}^{\circ} N, {104.6}^{\circ} E, 2, 210 m

above sea level) and Gulang

({37.49}^{\circ} N, {102.88}^{\circ} E, 2, 400 m

above sea level) are situated in the depocentre of modern dust storms

^{23}

, at distances of 80 km and 10 km , respectively, from the modern Tengger Desert margin (Fig. 1). Mean annual precipitation over the past 55 years is 233 mm at Jingyuan and 164 mm at Gulang, with

55 %

of the annual precipitation falling during the summer season. Two

20 - m

sampling pits were excavated at Jingyuan and Gulang. Forty OSL ages were used to generate an independent loess chronology, indicating that the

20 - m

loess sequences at Jingyuan and Gulang accumulated over the past 48 and 60 kyr , respectively (Methods, Supplementary Note S1, Tables S1 and S2, and Figs S1-S5). High sedimentation rates

(10 - 130 cm {kyr}^{- 1}

at Jingyuan and

9 - 63 cm {kyr}^{- 1}

at Gulang; Supplementary Fig. S6) and weak pedogenesis make these two loess sequences sensitive recorders of rapid monsoon changes.
静源

({36.35}^{\circ} N, {104.6}^{\circ} E, 2, 210 m

海拔的黄土层序）和古浪

({37.49}^{\circ} N, {102.88}^{\circ} E, 2, 400 m

海拔）位于现代沙尘暴的中心

^{23}

，距离现代腾格里沙漠边缘分别 80 公里和 10 公里（图 1）。过去 55 年，景源的年平均降水量为 233 毫米，鼓浪的年平均降水量为 164 毫米，

55 %

其中年降水量在夏季。在景源和鼓浪挖掘了两个

20 - m

采样坑。利用 40 个 OSL 年龄生成独立的黄土年代学，表明景源和鼓浪的

20 - m

黄土序列分别在过去 48 和 60 kyr 中积累（方法，补充说明 S1，表 S1 和 S2，以及图 S1-S5）。Jingyuan 和

9 - 63 cm {kyr}^{- 1}

Gulang 的沉积速率

(10 - 130 cm {kyr}^{- 1}

高;补充图 S6）和弱的土壤生成使这两个黄土序列成为季风快速变化的敏感记录器。

We argue that the grain size of the loess samples at Gulang and Jingyuan mainly reflects changes in winter monsoon strength

^{24}

(Supplementary Note S2). The Gulang and Jingyuan grain size records indicate distinct millennial-scale and glacial-interglacial variability (Fig. 2). Gulang grain size shows higher amplitude and more abrupt transitions, yielding a well-resolved record of millennial-scale winter monsoon fluctuations. Relatively high accumulation rates and coarser grain size characterize the Jingyuan section (Supplementary Fig. S6), a result of its proximity to the Yellow River (a likely local dust source). Significant dust contribution from this local source may attenuate the abrupt grain size changes recorded in the Jingyuan loess section. Our OSL-dated records indicate, with certain exceptions, that rapid grain size oscillations are aligned with Heinrich and Dansgaard-Oeschger (DO) events recorded in the North Greenland ice core (NGRIP; refs

10, 25, 26

) and Chinese speleothems

^{4, 27, 28}

(Fig. 2).
我们认为，鼓浪和景源黄土样品的粒度主要反映了冬季季风强度

^{24}

的变化（补充说明 S2）。鼓浪和靖远的粒度记录表明了明显的千年尺度和冰期间冰期的变化（图 2）。鼓浪颗粒尺寸显示出更高的振幅和更突然的转变，从而产生了千年尺度冬季风波动的清晰记录。相对较高的积累速率和较粗的晶粒尺寸是景源剖面的特征（补充图 S6），这是由于它靠近黄河（可能是当地的尘埃源）的结果。来自这种局部来源的显著灰尘贡献可能会减弱景源黄土剖面记录的突然颗粒尺寸变化。我们的 OSL 测年记录表明，除了某些例外，快速粒度振荡与北格陵兰冰芯（NGRIP;refs

10, 25, 26

）和中国洞穴中记录的 Heinrich 和 Dansgaard-Oeschger （DO）事件一致

^{4, 27, 28}

（图 2）。

Figure 1 | Site location and atmospheric circulation. Major atmospheric circulation regimes in East Asia and locations of Gulang and Jingyuan loess profiles as well as Hulu and Wulu caves. ISM-Indian summer monsoon, EASM-East Asian summer monsoon, EAWM-East Asian winter monsoon, WJ-Westerly jet.
图 1 |场地位置和大气环流。东亚的主要大气环流状况以及 Gulang 和 Jingyuan 黄土剖面以及 Hulu 和 Wulu 洞穴的位置。ISM-印度夏季季风，EASM-东亚夏季季风，EAWM-东亚冬季季风，WJ-西风急流。

Coarsening of mean grain size around 16, 24, 30, 39, 48 and 55 kyr indicates that the North Atlantic Heinrich events

(H_{1} - H_{5 a})

are associated with strong winter monsoon circulation in the Chinese Loess Plateau, although the timing and amplitude of coarsening grain size around 24 and 30 kyr at the Gulang section are slightly mismatched relative to the

H_{2}

and

H_{3}

events. Many of the millennial-scale events (DO 7-17, 34-60 kyr) are well aligned among the loess, speleothem and ice-core records, showing similar amplitudes and durations as well (Fig. 2). A remarkable grain size event is present between DO 14 and 15 , which is absent in the Hulu speleothem record but distinct in the NGRIP (ref. 10) and Wulu records

^{28}

. The validity of the OSL chronology is strongly supported by the good correlation between loess grain size and radiometrically dated speleothem

δ^{18} O

across events from DO 7 to DO 17. The correlative nature of these DO
平均粒径在 16、24、30、39、48 和 55 kyr 左右的粗化表明北大西洋海因里希事件

(H_{1} - H_{5 a})

与中国黄土高原强烈的冬季季风环流有关，尽管鼓浪剖面 24 和 30 kyr 附近粗化粒径的时间和幅度相对于

H_{2}

和

H_{3}

略有不匹配事件。许多千年尺度的事件（DO 7-17、34-60 kyr）在黄土、洞穴和冰芯记录之间很好地对齐，也显示出相似的振幅和持续时间（图 2）。在 DO 14 和 15 之间存在一个显著的粒度事件，这在 Hulu 洞穴岩记录中不存在，但在 NGRIP（参考文献 10）和 Wulu 记录

^{28}

中却不同。从 DO 7 到 DO 17 的事件

δ^{18} O

中黄土粒度与辐射测年洞穴之间的良好相关性有力地支持了 OSL 年代学的有效性。这些 DO 的相关性
events in the ice-core, loess and cave records (all on independent chronologies) directly links Heinrich- and DO-scale variability in the Loess Plateau to that in Greenland and eastern China over the interval 34-60 kyr.
冰芯、黄土和洞穴记录中的事件（均在独立的年代学上）直接将黄土高原的海因里希和溶解尺度变化与格陵兰岛和中国东部在 34-60 kyr 范围内的变化联系起来。

In contrast, DO events in the interval

20 - 34 kyr

are not as well correlated among the three records, indicating different amplitudes, numbers of resolved events and event durations. Within this interval the Gulang grain size record indicates thirteen distinct events (labelled ‘a’ to ’

m

'). Neither the cave nor the ice-core records indicate thirteen events within this interval. Three exceptionally strong events (

a, b

and c ) between 20 and 23 kyr in the Gulang record have no counterparts in the NGRIP or cave records. However, these large amplitude oscillations are clearly expressed as heavy-isotope events in Guliya ice-core

δ^{18} O

on the Tibetan Plateau

^{8}

. The OSL age model suggests that events

d, i, l

and

m

are most proximal to DO 2, 4, 5 and 6 in the cave records

^{4, 27}

, although we acknowledge that other correlations are possible within the OSL age error (for example, grain size event k to DO 5 , and events 1 and

m

to DO 6; Fig. 2).
相比之下，区间

20 - 34 kyr

内的 DO 事件在三个记录之间没有很好的相关性，表明不同的幅度、已解决事件的数量和事件持续时间。在此区间内，Gulang 粒度记录表明了 13 个不同的事件（标记为 'a' 到 '

m

'）。洞穴和冰芯记录都没有表明在此间隔内发生了 13 次事件。在 Gulang 记录中，在 20 到 23 kyr 之间的三个异常强烈的事件（

a, b

和 c ）在 NGRIP 或 cave 记录中没有对应物。然而，这些大振幅振荡在青藏高原的古里雅冰芯

δ^{18} O

中清楚地表现为重同位素事件

^{8}

。OSL 年龄模型表明，事件

d, i, l

和

m

最接近洞穴记录

^{4, 27}

中的 DO 2、4、5 和 6 ，尽管我们承认在 OSL 年龄误差中可能存在其他相关性（例如，粒度事件 k 到 DO 5 ，事件 1 和

m

DO 6;图 2）。

Clear differences in resolution exist between the loess grain size, ice-core and speleothem records. For example, the Gulang grain size data show three distinct fine-scale oscillations within DO 8, four within DO 12 and six within DO 14. This centennial-scale variability is not clearly expressed in either the speleothem or ice-core records. Replication of these structures in proximal Gulang sites will be required to test for reproducibility. Different expression of these abrupt events in NGRIP

δ^{18} O

, loess grain size, and speleothem

δ^{18} O

records indicates that these palaeo-proxies document different aspects (temperature, wind and precipitation) of the climate response in three different regions (Greenland, Chinese Loess Plateau and eastern China). Similar regional differences also exist within the Chinese Loess Plateau, possibly due either to age model discrepancies, or to the resolution and continuity of the loess sequences

^{1 - 3, 29, 30}

.
黄土颗粒尺寸、冰芯和洞穴记录之间存在明显的分辨率差异。例如，Gulang 粒度数据显示 DO 8 内有 3 个不同的精细尺度振荡，DO 12 中有 4 个，DO 14 中有 6 个。这种百年尺度的变化在洞穴岩或冰芯记录中都没有明确表达。需要在近端 Gulang 部位复制这些结构以测试可重复性。这些突变事件在 NGRIP

δ^{18} O

、黄土粒度和洞穴

δ^{18} O

记录中的不同表达表明，这些古替代物记录了三个不同地区（格陵兰岛、中国黄土高原和中国东部）气候响应的不同方面（温度、风和降水）。中国黄土高原也存在类似的区域差异，可能是由于年龄模型的差异，或者是由于黄土序列的分辨率和连续性

^{1 - 3, 29, 30}

。

The DO-scale variability common to Greenland ice cores, loess and cave records, particularly within the interval

34 - 60 kyr

, is plausibly linked to rapid AMOC oscillations

^{11, 12}

. Both geological
格陵兰冰芯、黄土和洞穴记录常见的 DO 尺度变化，尤其是在间隔

34 - 60 kyr

内，可能与快速 AMOC 振荡有关

^{11, 12}

。既是地质学的

Figure 2 | Monsoonal and North Atlantic climate records. Comparison of Gulang (red) and Jingyuan (black) mean grain size with NGRIP

δ^{18} O

(blue) and

{Ca}^{2 +}

(orange) records

^{10, 25, 26}

and Hulu(purple)/Wulu(green) speleothems

^{4, 27, 28}

. Sky-blue bars denote the Heinrich-like events identified in the three records. Black numbers (

7 - 17, 34 - 60 kyr

) denote well aligned DO events identified in the three records. Over the interval

20 - 34 kyr

the Gulang grain size record indicates thirteen distinct events (labelled ‘a’ to ’

m

'). OSL ages and errors (

1 σ

error bars) of two loess sequences are marked on the grain size curves. The

^{230}

Th ages and errors (

2 σ

error bars) are shown on the speleothem records. Age errors for the NGRIP ice core are the maximum counting errors on the Greenland Ice Core Chronology 2005 (GICCO5) timescale.
图 2 |季风和北大西洋气候记录。鼓浪（红色）和晶元（黑色）平均晶粒尺寸与 NGRIP

δ^{18} O

（蓝色）和

{Ca}^{2 +}

（橙色）记录

^{10, 25, 26}

以及 Hulu（紫色）/Wulu（绿色）洞穴的比较

^{4, 27, 28}

。天蓝色条表示在三条记录中确定的类似海因里希的事件。黑色数字（

7 - 17, 34 - 60 kyr

）表示在三条记录中识别出的对齐良好的 DO 事件。在这段时间内

20 - 34 kyr

，Gulang 粒度记录表明了 13 个不同的事件（标记为 'a' 到 '

m

'）。两个黄土序列的 OSL 年龄和误差（

1 σ

误差线）标记在粒度曲线上。

^{230}

Th ages 和错误（

2 σ

error bars）显示在洞穴记录上。NGRIP 冰芯的年龄误差是 2005 年格陵兰冰芯年表（GICCO5）时间尺度上的最大计数误差。
evidence and modelling results suggest that rapid climate change in the North Atlantic region (for example, Heinrich and DO events) can result in variability in the strength and position of the westerly jet, influencing northern hemisphere climate, including East Asia

^{1 - 3, 7, 19 - 21}

. Considering a North Atlantic origin of abrupt climate change, one would expect a distinct response in monsoonal wind and precipitation variability in East Asia. Monsoonal proxy results do indicate that significant strengthening of the winter monsoon and weakening of the summer monsoon are associated with strong cooling in the North Atlantic. North Atlantic water-hosing experiments using the Community Climate System Model version 3 (CCSM3) under Last Glacial Maximum (LGM) boundary conditions also suggest a coupled response of the winter and summer monsoons to reductions of the AMOC (Methods and Supplementary Note S3).
证据和建模结果表明，北大西洋地区的快速气候变化（例如，海因里希和 DO 事件）会导致西风急流的强度和位置发生变化，从而影响北半球气候，包括东亚

^{1 - 3, 7, 19 - 21}

。考虑到气候变化的起源于北大西洋，人们可以预期东亚的季风风和降水变化会做出明显的反应。季风代理结果确实表明，冬季季风的显著增强和夏季季风的减弱与北大西洋的强烈降温有关。在末次冰盛期（LGM）边界条件下使用社区气候系统模型第 3 版（CCSM3）的北大西洋水管实验也表明，冬季和夏季季风对 AMOC 减少的耦合响应（方法和补充说明 S3）。

In the water-hosing experiments, the AMOC weakens from a control mean of 16.8 Sv to a mean of 5.6 Sv (here Sv stands for Sverdrup;

1 Sv = 10^{6} m^{3} s^{- 1}

) after 50 years (Supplementary Note S3 and Fig. S7). Climate in most parts of the Northern Hemisphere, including Greenland, responds in a matter of decades or less. As AMOC slows, global northward ocean heat transport is reduced at all latitudes, including a

\sim 60 %

reduction north of

40^{\circ} N

. This is partially, but not completely, offset by an increase in atmospheric heat transport. Consequently, temperatures in the northern hemisphere decrease, with the largest cooling during the winter associated with areas of sea-ice expansion in both the North Pacific and North Atlantic (Supplementary Note S3 and Fig. S8). The increase in the meridional (latitudinal) temperature gradient leads to stronger westerly winds in the mid-latitudes and strengthened winter wind speed above the northwestern Chinese Loess Plateau and the major Asian dust source regions (Fig. 3a). Both surface winter wind stress, which is key for dust entrainment and transportation, as well as 750 mb wind speed (westerly), which is important for long-range dust transport, significantly increase in the hosing experiment.
在水管实验中，AMOC 从 16.8 Sv 的对照平均值减弱到 5.6 Sv 的平均值（这里 Sv 代表 Sverdrup;

1 Sv = 10^{6} m^{3} s^{- 1}

）（补充说明 S3 和图 S7）。北半球大部分地区（包括格陵兰岛）的气候在几十年或更短的时间内就会做出反应。随着 AMOC 的减慢，全球向北的海洋热传输在所有纬度上都减少了，包括以北的

\sim 60 %

40^{\circ} N

减少。这部分但并非完全被大气热传递的增加所抵消。因此，北半球的气温下降，冬季最大的降温与北太平洋和北大西洋的海冰扩张区域有关（补充注 S3 和图 S8）。经向（纬度）温度梯度的增加导致中纬度地区出现更强的西风，中国西北部黄土高原和亚洲主要沙尘源地区上空的冬季风速增强（图 3a）。在软管实验中，对尘埃夹带和传输至关重要的地表冬季风应力，以及对远距离尘埃传输很重要的 750 mb 风速（西风）都显著增加。

Following hosing, there is a decrease in the summer precipitation in East Asia (Fig. 3b), which today is associated with moisture convergence along the Meiyu front

^{22}

. Cyclonic circulation anomalies associated with a weakening of the subtropical high probably reduced southwesterly moisture transport and moisture convergence to the Meiyu front. Furthermore, an overall reduction in specific humidity caused by colder temperatures dries the northern hemisphere atmosphere (Supplementary Table S3). Many of our modelled summer circulation anomalies are consistent with results from a pre-industrial hosing experiment

^{17}

, indicating that similar atmospheric and oceanic responses to hosing were probably in operation during the last glacial. In that simulation, cooling in the North Atlantic associated with the reduction of the AMOC causes a southward shift of the Intertropical Convergence Zone in both the Atlantic and the Pacific, a strengthening of the Walker circulation in the northern tropical Pacific, and a weakening of the East Asian summer monsoon

^{17}

.
软管后，东亚夏季降水减少（图 3b），今天这与梅雨锋面

^{22}

的水汽会聚有关。与副热带高压减弱相关的气旋环流异常可能减少了西南水汽输送和水汽向梅雨锋的收敛。此外，由于温度较低而导致的比湿度总体降低，使北半球大气干燥（补充表 S3）。我们建模的许多夏季环流异常与工业化前软管实验

^{17}

的结果一致，表明大气和海洋对软管的类似反应可能在末次冰期期间起作用。在该模拟中，与 AMOC 减少相关的北大西洋冷却导致大西洋和太平洋的热带辐合带向南移动，热带太平洋北部沃克环流加强，东亚夏季季风

^{17}

减弱。

A robust comparison of loess grain size with Chinese speleothem and Greenland ice-core records indicates that Greenland temperature and EAM variability are coupled on millennial timescales during the interval

34 - 60 kyr

. North Atlantic water-hosing experiments suggest a dynamical response of the monsoonal wind and precipitation in East Asia to a reduction of the AMOC. Freshwater forcing results in an increased latitudinal temperature gradient and strong winter winds throughout much of the atmosphere in the Northern Hemisphere mid-latitudes. Meanwhile, the North Pacific subtropical high weakens following the hosing and the resulting strong cyclonic circulation anomaly leads to reduced convergence along the Meiyu front in East Asia and decreased summer precipitation surrounding Hulu Cave. Integration of these modelling results
黄土颗粒大小与中国洞穴和格陵兰冰芯记录的有力比较表明，格陵兰温度和 EAM 变化在间隔期间在千年时间尺度上耦合

34 - 60 kyr

。北大西洋吸水实验表明，东亚季风风和降水对 AMOC 减少的动力响应。淡水强迫导致北半球中纬度地区大部分大气中的纬度温度梯度增加和强冬季风。同时，北太平洋副热带高压在软管后减弱，由此产生的强气旋环流异常导致东亚梅雨锋面的辐合减少，葫芦洞周围夏季降水减少。整合这些建模结果

Figure 3 | CCSM3 simulated surface wind stress and precipitation anomalies between the water-hosing experiment and the glacial control simulation. a, Boreal winter (December-February) surface wind stress (

{Nm}^{- 2}

, colour), sea-level pressure ( hPa , lines) and 750 mb wind vectors (

m s^{- 1}

b

, Boreal summer (July-August) precipitation (

mm d^{- 1}

, colour), sea-level pressure (hPa, lines) and wind vectors from the lowest atmospheric model level (

m s^{- 1}

). Stippling shows areas of statistically-significant differences in surface wind stress and precipitation at the

95 %

level according to a Student’s

t

-test. Black dots show locations of Gulang, Jingyuan and Hulu discussed in text.
图 3 |CCSM3 模拟了水管试验和冰川控制模拟之间的地表风应力和降水异常。a，北方冬季（12 月至 2 月）表面风应力（

{Nm}^{- 2}

，颜色）、海平面压力（ hPa ，线）和 750 mb 风矢量（

m s^{- 1}

）。

b

、北方夏季（7 月至 8 月）降水（

mm d^{- 1}

，颜色）、海平面压力（hPa，线）和最低大气模式水平（）的风矢量（

m s^{- 1}

）。点画显示根据 Student

t

检验，在该

95 %

级别上表面风应力和降水具有统计学显著性差异的区域。黑点表示文本中讨论的 Gulang、Jingyuan 和 Hulu 的位置。
and our palaeoclimate data reveals that reduction of the AMOC probably played a key role in driving abrupt monsoon changes in East Asia.
我们的古气候数据显示，AMOC 的减少可能在驱动东亚季风的突然变化中发挥了关键作用。

Methods 方法

We took 2000 powder samples at

2 - cm

intervals from these two

20 - m

loess pits. Grain size distribution of the samples was measured using a Malvern 2000 laser instrument after removal of organic matter and carbonate. Samples for OSL dating were collected in stainless tubes hammered into a freshly-cleaned section face, and then wrapped in aluminium foil until processing under subdued red light in the luminescence dating laboratory of the Institute of Earth Environment, Chinese Academy of Sciences. Material from both ends of the sample tubes was excluded from equivalent dose (ED) estimation and was used for radioisotope measurements to obtain the environmental dose rate. Samples were pretreated with

30 % HCl

and

10 % H_{2} O_{2}

to remove the carbonates and organic matter, respectively. Then the
我们每

2 - cm

隔一段时间从这两个

20 - m

黄土坑中采集了 2000 个粉末样本。去除有机物和碳酸盐后，使用 Malvern 2000 激光仪器测量样品的粒度分布。用于 OSL 测年的样品收集在不锈钢管中，锤打到新清洁的切片面上，然后用铝箔包裹，直到在中国科学院地球环境研究所的发光测年实验室在柔和的红光下进行处理。样品管两端的材料被排除在等效剂量（ED）估计之外，并用于放射性同位素测量以获得环境剂量率。样品分别用

30 % HCl

和

10 % H_{2} O_{2}

预处理以去除碳酸盐和有机物。然后

$^{1}$ State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, Xi’an 710075, China, $^{2}$ Department of Geology, Brown University, Providence, Rhode Island 02912-1846, USA, $^{3}$ Cooperative Institute for Research in Environmental Sciences (CIRES),
$^{1}$ 中国科学院地球环境研究所，黄土与第四纪地质国家重点实验室 710075 习， $^{2}$ 布朗大学地质系，美国罗德岛州普罗维登斯 02912-1846， $^{3}$ 环境科学合作研究所，
University of Colorado at Boulder, and NOAA’s National Climatic Data Center, Boulder, Colorado 80305-3328, USA, $^{4}$ Physical Oceanography Laboratory, Ocean University of China, Qingdao 266100, China. *e-mail: sunyb@ieecas.cn.
科罗拉多大学博尔德分校和 NOAA 国家气候数据中心，美国科罗拉多州博尔德 80305-3328， $^{4}$ 中国海洋大学物理海洋学实验室，中国青岛 266100。*电子邮件：sunyb@ieecas.cn。

Influence of Atlantic meridional overturning circulation on the East Asian winter monsoon 大西洋经向翻转环流对东亚冬季风的影响

Abstract 抽象

Methods 方法

Influence of Atlantic meridional overturning circulation on the East Asian winter monsoon
大西洋经向翻转环流对东亚冬季风的影响