Quantitative reconstruction of the East Asian Winter Monsoon evolution over the past 100 years: Evidence from high-resolution sedimentary records of the inner continental shelf of the East China Sea

Abstract

The East Asian Monsoon (EAM) is a regional factor affecting the East Asian climate and the oceanographic processes of the marginal seas along the Western Pacific. Finding proxies for the EAM intensity and reconstructing the interannual and interdecadal variations of the EAM using high-resolution records are necessary to improve our understanding of the EAM’s role in the global climate system and for predicting climate change. In this paper, high-resolution sedimentary records of sedimentary core C0702 obtained from the inner continental shelf of the East China Sea were comprehensively analyzed using a laser particle size analyzer, an Itrax^TM core scanner, and a ²¹⁰Pb and ¹³⁷Cs radionuclide analyzer to explore potential proxies for the East Asian Winter Monsoon (EAWM). By combining the obtained results with instrumental observations of the EAM, we established a quantitative formula for the EAWM, which enables to reconstruct the evolution trend of the EAWM during the past 130 years. The sensitive grain-size component F2, with a grain-size range of 14.35–230 µm, and principal component PC1 of the sedimentary deposits of the East China Sea inner shelf can be used as EAWM proxies. The evolution of EAWM in 1880–1950 could be roughly divided into two stages: a weak EAWM period from 1882 to 1900 and a strong EAWM period from 1900 to 1945. This study improves our understanding of the variations in the EAWM on interannual and interdecadal temporal scales.

Keywords

East Asian Monsoon inner shelf of East China Sea proxy quantitative reconstruction sedimentary record

Introduction

The East Asian Monsoon (EAM) is an important component of the global climate system that directly affects the terrestrial and marine climate of the East Asian region (Wang et al., 2005). Previous studies have obtained good results in analyzing the EAM intensity variations in loess accumulation, stalagmites, and cave deposits (An, 2000; Cheng et al., 2016; Dong et al., 2010; Sun et al., 2012). However, because loess and stalagmite have low accumulation rates, such deposits can only provide information on the long-term evolution of the monsoon; they are unable to provide high-resolution records of monsoon variability on interannual or interdecadal scales.

Marine sediments can contain continuous high-resolution and relatively complete records of paleoclimate and paleoenvironmental evolution and have been widely used for paleoclimate and paleoenvironmental inferences. In particular, a series of important achievements has been made in the reconstruction of the East Asian Winter Monsoon (EAWM; Qiao et al., 2010; Sagawa et al., 2014; Zheng et al., 2014; Zhou et al., 2007; Zong et al., 2006). These studies generally used proxies such as sensitive grain-size components, elemental ratios, and the Mg/Ca sea surface temperature (SST) and δ¹⁸O contents of planktonic foraminifera to quantitatively reconstruct the variation of the EAWM in the Holocene, revealing its evolution patterns at the millennial and centennial scales (Hu et al., 2012; Liu et al., 2010b; Xiang et al., 2006; Yu et al., 2012). Most of this research is based on radiometric dating (¹⁴C dating), which means that the accuracy of the temporal resolution can only reach a few decades. Since the beginning of the Industrial Revolution, the global climate has shown a significant warming trend (IPCC, 2013). However, under the influence of the EAM, the SST along the eastern margin of China has shown an overall cooling trend over the past three centuries (He et al., 2014; Kong et al., 2015), revealing the complexity and diversity of the regional climate response under global climate change conditions. Revealing the variation trends of the EAM at a sub-centennial scale requires accurate dating at an interannual or interdecadal scale, which is where the radiocarbon dating falls short. Another challenge is that instrumental data only cover a period of about 100 years, leaving out the most significant period of global climate change since the Industrial Revolution. Although some scholars have attempted to reconstruct the intensity variations of the EAM during this period, the shortage of suitable archives and proxies largely limits these studies (He et al., 2014; Li et al., 2015).

The inner shelf mud deposits of the East China Sea have been forming since the postglacial sea-level high stand (Li et al., 2014; Liu et al., 2010a). Large amounts of terrigenous materials were transported into the sea by the Yangtze River and were then moved southward under the combined action of tides, waves, and coastal currents to be deposited along the coasts of Zhejiang and Fujian, with small amounts of biogenic sediments added. As a result, this region has a high sediment accumulation rate, which can reach 1–3 cm/yr (Qiao et al., 2017). On the contrary, the flux of the terrigenous material entering the sea from the Yangtze River, the marine biomass supply, or the variations in intensity and direction of nearshore currents are all restricted by the EAM, which is evident from seasonal fluctuations during summer and winter (Hu et al., 2011; Liu et al., 2007). These seasonal variations are reflected by the deposits of centimeter-thick varves, which can be used for precise dating of muddy sediments in this area (Fan et al., 2011; Zhang et al., 2018). However, the inner shelf of the East China Sea is also affected by frequent typhoons (Li et al., 2012), the deposits of which are formed in the muddy sediments (Tian et al., 2019). The development of these event deposits interrupts the formation of high-resolution sedimentary records, making accurate results difficult to obtain.

In this paper, we analyzed the sedimentary characteristics of a sediment core collected in the inner continental shelf of the East China Sea using a laser-diffraction particle size analyzer and a non-destructive and fast-data-acquisition Itrax^TM core scanner. After identifying and eliminating the typhoon event deposits, the sediment core was precisely dated based on the analysis of radioactive nuclides ²¹⁰Pb and ¹³⁷Cs and sediment varves. Next, we explored the use of sensitive proxies and constructed quantitative formulas to reconstruct the EAWM evolution trends during the past 130 years, thereby providing a new means for better understanding of the interannual and interdecadal EAM variability based on high-resolution records.

Study area

The East China Sea is a typical extensive epi-continental sea. Bordering the Eurasian continent to the west, the sea receives large amounts of terrigenous debris from mainland China. On the east side, the East China Sea connects to the Pacific Ocean through the Okinawa Trough and is restricted by the ocean’s western boundary current. The terrigenous input entering the East China Sea from mainland China amounts to 0.9–1.0 × 10⁹ t/yr. The Yangtze River and the Yellow River are its main contributors (Deng et al., 2006), each accounting for 34–50% and 2–16% (Su and Huh, 2002) of the total terrigenous input, respectively. The nearshore current system includes the Yangtze River diluted water, the Zhejiang–Fujian Coastal Current, and the Taiwan Warm Current, making this marginal sea one of the most complex dynamic sedimentary environments in the world (Hu et al., 2011).

The sedimentation of the East China Sea is strongly affected by the EAM. Under its influence, the water and sediment input from the Yangtze River, the supply of marine biogenic materials, and the intensity and direction of the nearshore current systems show significant seasonal variations (Hu et al., 2011; Wang et al., 2005). During the summer monsoon, the sediment flux from the rivers into the sea is high, whereas the coastal currents are weak. The sediments from the Yangtze River are first deposited in the estuary area, with a small amount transported into the open sea and along the coast. During winter, under the action of strong waves, sediments from the Yangtze estuary and coastal areas are reworked and resuspended, and about 30% of the suspended material is transported by the Zhejiang–Fujian Coastal Current and deposited in the inner shelf along the coast (Liu et al., 2007; Xiao et al., 2006).

The abundant supply of sediment coupled with its suitable dynamic environment makes the continental shelf of the East China Sea an important sediment sink. Since the sea-level high stand after the Last Glacial Period, multiple mud deposit centers have been formed, including the Yangtze River estuary mud area, the Zhejiang–Fujian coastal mud wedge, and the Southwestern Jeju Island mud area (Figure 1; Liu et al., 2007; Yang and Liu, 2007). These mud areas are characterized by a high sediment accumulation rate and good continuity of the sedimentary strata, making them good archives of high-resolution paleoclimate and paleoenvironmental records.

Figure 1.

Study area and sampling location. The left figure is from Google Earth. The circulation system and muddy areas (dark gray) in the right figure are modified from Liu et al. (2007) and Hu et al. (2011). The solid arrows in the figure indicate the Kuroshio Current (KC), Taiwan Warm Current (TWWC), Zhejiang–Fujian Coastal Current (ZFCC), Yellow Sea Coastal Current (YSCC), Yellow Sea Warm Current (YSWC), Shandong Coastal Current (SCC), and Liaonan Coastal Current (LCC).

Materials and analysis

Sampling method

Sediment core C0702 (122°27.176′E, 29°12.527′N) was sampled from a mud wedge area located in the northern part of Zhejiang–Fujian, with a water depth of 40 m. The core sample was obtained in 2009 by R/V Dong Fang Hong 2 using a gravity corer. The core was scanned and analyzed using an Itrax core scanner. After being x-rayed in a lab, a 177.5-cm sample was cut into 0.25 cm segments for grain-size analysis. At the same time, the left half of the sedimentary core was cut into 0.5 cm segments for radiometric and elemental analysis. The sub-samples were precisely cut based on the sediment laminae distribution shown in an x-ray grayscale image. All samples were placed in sampling bags and were stored at room temperature for later use.

Grain-size analysis

Samples of uniform thickness were obtained by sediment coring with a customized square polyvinyl chloride (PVC) pipe 200 cm × 2 cm × 5.5 cm in length, width, and height, respectively, and were analyzed using x-ray imagery. Next, the x-rayed sub-core was sampled for grain analysis by selecting 1 g samples at 0.25 cm intervals. The grain-size analysis procedure was as follows. In the pre-treatment for grain-size analysis, 5 mL of 30% H₂O₂ was added to 0.5 g samples for 24 h to remove organic matter. Next, a sufficient amount of (NaPO₃)₆ dispersing agent was added, and the samples were ultrasonically dispersed for 0.5 h. The samples were analyzed by a laser particle size analyzer (Mastersizer 2000 Malvern Instruments Ltd, UK) with a measuring range of 15.6 to −1Φ (0.02–2000 µm), a resolution of 0.01Φ, and an analytical error of ±2%.

Radionuclide analysis

About 5–10 g of sediment was first dried and homogenized. The activity of ²¹⁰Pb and ¹³⁷Cs in the sediments was determined by a gamma spectrum analysis system (EG&G ORTEC, Oak Ridge, Tennessee, US). The test was conducted at Nanjing Institute of Geography and Limnology, Chinese Academy of Science. The standard samples of ¹³⁷Cs and ²²⁶Ra were provided by the China Institute of Atomic Energy. The standard sample of ²¹⁰Pb was provided by the University of Liverpool, UK.

Geochemical analysis

Itrax core scanner enables non-destructive fast-data-acquisition from sediment cores. It can acquire elemental geochemistry data at sub-millimeter resolution (Cuven et al., 2011) and is thus a powerful tool for the study of high-resolution sedimentary records. In this study, after cutting the core longitudinally into two sub-cores, the Itrax core scanner (at 100 s exposure time with a Mo-tube) was used to acquire the geochemical data by scanning the sub-core at 1 mm intervals. The experimental conditions included current at 30 mA and voltage at 50 kV. Because the geochemical data acquired by the Itrax core scanner are susceptible to compaction (Cuven et al., 2011), we used a x-ray fluorescence (XRF) elemental analyzer (SPECTRO XEPOS) to effectively extract and analyze the element content of the core sediments. The Itrax core scanner acquired intensive signals from Ca, Fe, K, Ti, Si, Sr, Cl, Rb, and Zr. Comparative analysis showed that the signal intensity of these elements has a good linear relationship with the element content determined from the XRF analysis (Zhang et al., 2013). Therefore, the most effective elements from the Itrax core scanning results, specifically Ca, Fe, K, Ti, Si, Sr, Cl, Rb, and Zr, were selected for further analysis.

Statistical analysis

The statistical method of Ensemble Empirical Mode Decomposition (EEMD), which is internally consistent and suitable for fragment data analysis and has been widely applied in the Earth sciences to process nonlinear and nonstationary data (Wu and Huang, 2009), was used to process the grain size and geochemical data in this paper. This method decomposes the raw data into a series of Intrinsic Mode Functions and the residual trend, which can be illustrated as

X (t) = H (t) + L (t) + R e s

where $X (t)$ is the time series of the grain size or geochemical data from the core C0702, $H (t)$ is the high-frequency component, $L (t)$ is the low frequency component, and $R e s$ is the residual trend. Here, the EEMD method is used to remove the residual trend of proxies for the EAWM.

Analysis results

Sediment lithology and dating

The core C0702 sediment is composed mainly of silt and clay, which together account for about 98% of the sample material; the sand content is low, at about 2%. The contents of sand and silt fluctuate along the vertical direction, and a significant increase in their content can occasionally be observed in individual laminae. In our previous study, we found that there were visible sediment varves in core C0702, which were composed of normal sedimentary layers and event sedimentary layers (Zhang et al., 2018), and we found that these coarse-grained sedimentary layers and typhoon activities were well correlated. For example, the event layers E1–E5 recorded in core C0702 correspond to strong typhoons passing by the area in 1983, 1981, 1974, 1961, and 1956, respectively. A conceptual model to illustrate the sedimentary process of the formation of typhoon event layer is as follows: coastal sediment is eroded and resuspended during the intense typhoon period, transported by an outward current in flume form, and deposited on the inner shelf (Tian et al., 2019). We identified 15 event deposits in the sediment core, and the thickness of event layers varies from 0.5 to 5.5 cm with an average thickness of 2.05 cm (Figure 2) based on the three-sigma rule of thumb method. Briefly, first box plot quartile analysis was used to determine extreme outliers, then the average $\bar{X}$ was calculated, and standard deviation $D$ was calculated after the removal of outliers; finally, $\bar{X} + 3 D$ was taken as the background sedimentary threshold (Ercolani et al., 2015). These storm event layers cover the underlying deposits without causing erosion-induced depositional discontinuities (Tian et al., 2019). In addition, the normal sedimentary layers of the core sample are well preserved and show no signs of erosion.

Figure 2.

Identification parameters of event deposits in core C0702. (a) Change curve of the x-ray and gray value with depth (white line), (b) grain composition of the core, (c) change curve of average grain size with depth, and (d)–(h) variations of Si, Ca, Fe, K, and Ti with depth based on the Itrax^TM core scanner. Fifteen event deposits were identified, and the thickness of E1–E15 was 0.75, 1.25, 3.25, 2.5, 5.5, 2.75, 1, 2.75, 0.5, 3, 0.75, 1.75, 0.5, 1.5, and 3 cm, respectively.

To accurately determine the chronological framework of the sediment core, we first removed the event layers in the core. Then, we obtained the sedimentation rate of the core based on the constant initial concentration (CIC) model of excess ²¹⁰Pb (Sun et al., 2018). The mean sediment accumulation rates of core C0702 were 0.9 and 1.2 cm/yr, at 0–20.5 cm and 20.5–80 cm, respectively (Figure 3). At the same time, by using three ¹³⁷Cs temporal markers corresponding to 1986, 1963, and 1954, respectively (Palinkas and Nittrouer, 2007), we had calculated sedimentary rates of the core by 0.9 cm/yr at 0–20.5 cm and 1.1 cm/yr at 20.5–52 cm. The sedimentary rates from excess ²¹⁰Pb were consistent with the results from ¹³⁷Cs temporal markers. Therefore, the sedimentary rates calculated from excess ²¹⁰Pb were used to date the core in this paper. Furthermore, our previous studies have reported that core C0702 has well-defined varves; there was only a 2- to 5-year difference between the chronologies based on varve counting and radiometric dating (Zhang et al., 2018). Based on the above analysis results, the base of core C0702 was calculated to be about 127 years (1882–2009), and the mean sedimentation rate was 1.1 cm/yr (Figure 2). These results are consistent with the findings of other scholars for this region (Qiao et al., 2017).

Figure 3.

Intensity variations with depth of radioactive nuclides ²¹⁰Pb and ¹³⁷Cs of sample core C0702.

Sensitive grain-size components

The average grain size of core C0702 varied from 1.9 to 39.4 µm, with a mean value of 5.6 µm. In some layers, the sediment grain size displayed significantly higher values. The changes in sensitive grain-size components of the sediments are more significant than those of the mean size grains, which means that sensitive grain-size components can better represent the regional sedimentary hydrodynamic environment (Fan et al., 2011; Sun et al., 2003; Xiao et al., 2006). In this study, the grain-size standard deviation method was used to determine the sensitive grain-size components of the core. The following three types of components were identified in the core: <14.35 µm (Fraction 1 (F1)), 14.35–230 µm (Fraction 2 (F2)), and >230 µm (Fraction 3 (F3); Figure 4b). Among them, F1 is the most important part of the core sediment, accounting for 9–84% of the total content and a mean of 65.6%. F2 is a secondary component, accounting for 15–91% and a mean of 33.8%; F2 shows significant negative correlation with F1. F3 has the lowest content. The F3 content is relatively stable throughout the entire core with a mean value of 0.6% except at depths of 21–23 cm, where its content is relatively high.

Figure 4.

(a) Change curve of sediment grain size versus standard deviation of core C0702. (b) Changes in sensitive grain-size component content with depth.

High-resolution geochemical records

Intensity data of 27 elements, including Mg, Al, and Si, were obtained by using the Itrax scanning system (Table 1). The scan results show large variations in element intensity, with the high value exceeding n × 10⁴ S⁻¹ and the low value falling below 10² S⁻¹. Eight elements – Ca, Fe, K, Ti, Si, Rb, Sr, and Zr – had very strong signal-to-noise ratios. Combining these results with those of conventional XRF spectrometry revealed that these elements account for more than 85% of the total element content of the core, which means that they can represent the overall variation trend of the core (Zhang et al., 2013). To reduce the number of variables, principal component analysis (PCA) was applied to the elements with high signal-to-noise ratios (Table 2, Figure 5). As shown in Table 2 and Figure 5, principal component 1 (PC1) and principal component 2 (PC2) are the two primary components of the core, accounting for 49% and 17% of the total variance, respectively. Moreover, PC1 and PC2 satisfy the requirements of the cumulative variables, which means that they can represent the overall variation trend (Boyer-Villemaire et al., 2013). PC1 is represented mainly by Si, Ca, Sr, and Zr, all of which are positively correlated and represents the coarse particle components of the terrigenous sediment. PC2 is represented mainly by Fe, K, and Ti, all of which are positively correlated. PC2 represents the fine particle components of the terrigenous sediment.

Table 1.

Mean values of core C0702 elements based on Itrax^TM scan results (unit: cps).

Mg	Al	Si	P	S	Cl	Ar	K	Ca
6.55	350.04	4594.25	0.33	144.36	4105.36	1560.46	23,342.82	46,419.3
Ti	V	Fe	Co	Ni	Cu	Zn	Ga	Ge
17,937.2	259.14	39,521.92	144.01	79.79	35.5	167.95	29.73	21.61
As	Br	Rb	Sr	Y	Zr	Mo	Ba	Ta
22.13	132.12	343.99	323.89	45.25	281.63	189.87	374.27	52.22
W	Pb	Mn
138.84	24.73	–

Table 2.

Principal component analysis of core C0702: elements and principal component (PC1 and PC2) loadings (the bold values denote elements with significant positive correlation).

Loadings	C0702 (n = 1724)
Loadings	PC1 (49%)	PC2 (17%)
Ca	0.9	–0.01
Fe	–0.73	0.52
K	–0.39	0.73
Ti	0.11	0.96
Si	0.9	–0.15
Rb	–0.21	0.09
Sr	0.58	–0.12
Zr	0.67	–0.23

Figure 5.

The principal component analysis results of eight elements – Ca, Fe, K, Ti, Si, Rb, Sr, and Zr, which were obtained using the Itrax^TM scanning system.

Discussion

Sediment transportation mechanisms along the inner shelf of the East China Sea

Because the grain size of marine sediment preserves abundant information on the sediment transportation and deposition (Weltje and Prins, 2003), the grain-size composition can be used to understand the form of sediment transport and to further deduce the corresponding hydrodynamic conditions and their driving factors (Hu et al., 2012; Xiao et al., 2006). The cumulative probability curves of the sediment grain size of core C0702 can be divided into three types: a one-segment pattern, a two-segment pattern, and a three-segment pattern. The one-segment pattern occurred most often for finer sediments, whereas the two-segment pattern occurred primarily for more coarse sediments. The three-segment pattern occurred for individual lamina (Figure 6).

Figure 6.

Cumulative probability curves of core C0702 sediment grain size: (a) one-segment, (b) two-segment, and (c) three-segment.

The results of C–M diagram of the sediment grain size show that the C and M values of L1 show synchronous changes, which is roughly parallel to the C–M line, and this represents graded suspension; the C value of L2 changes significantly and the M value of L2 is lower than L1. In addition, the sample points are more closely related, representing homogeneous suspension; compared with L1 and L2, the sample points of L3 are more dispersed, characterized by little change in M value, and M value is within the range of M value of homogeneous suspension. We considered that the whole population belongs to fine particles and a small amount of coarse particles contained in the sediments, which represents biological debris.

F1 is the finest part of the core sediment, which corresponds to the suspended loading of cumulative probability curves; furthermore, the C–M diagram showed F1 belongs to a homogeneous suspension. The grain-size interval of F2 corresponds to the rolling loading (two-stage type) and suspended loading (one-stage type) of cumulative probability curves. However, the C–M diagram showed F2 belongs to the graded suspension, indicating this part of the sediment in the water column is varied. Furthermore, the cumulative probability curves of these grain sizes have a large slope and good sorting, reflecting a high hydrodynamic energy. F3 is the coarsest sediment, and the grain-size interval corresponds to the rolling loading of cumulative probability curves. It showed a typical mode of rolling loading of cumulative probability curves, but on the C–M diagram it appears as a wider strip-shaped region that is nearly parallel to the C-axis, which is significantly different from the rolling loading caused by the tractive current. Consider the high primary production in this area (Wang et al., 2015), which includes diatoms, dinoflagellates, and benthic organisms, and various biological fragments in the core sediments. F3 is probably produced by the biological processes and cannot reflect the hydrodynamic conditions of this region (Figure 7).

Figure 7.

C–M diagram of core C0702 sediment grain size. Area L1 represents graduated suspension, Area L2 represents homogeneous suspension, and Area L3 represents bioclastics.

The sedimentation process of the inner continental shelf of the East China Sea shows significant seasonal variation (Fan et al., 2011). In the summer, large amounts of terrigenous material are transported primarily by the Yangtze River and its tributaries; the wave energy is low, and the Zhejiang–Fujian Coastal Current weakens or even disappears completely. As a result, most of the sediment transported by the river is deposited in the estuary and in the adjacent waters, with only small amounts of sediment transported away from the estuary by means of homogeneous suspension to the muddy areas of the inner continental shelf (Hu et al., 2011). During this period, the hydrodynamic environment is conducive to the transportation and deposition of F1 sediments. In the winter, small amounts of terrigenous material are transported into the sea by the Yangtze River and its tributaries; strong winter storms occur; the wave energy is high; and the Zhejiang–Fujian Coastal Current strengthens. As a result, the sediments deposited in the summer become resuspended and are transported and deposited in the inner shelf along the coast of Zhejiang and Fujian by means of graded suspension (Liu et al., 2007). During this season, the hydrodynamic environment is conducive to the deposition of F2 sediments.

The principal element content of sediment is restricted by the grain-size control rate. PC1 is characterized mainly by Si, Ca, Sr, and Zr and primarily represents the coarse-grain components of terrigenous material, which are conducive to deposition in winter. PC2 is characterized mainly by Fe, K, and Ti and primarily represents the fine grain components of the terrigenous material, which tend to be deposited in summer.

Reconstruction of EAM intensity variations

Based on the above analysis, both the sensitive grain-size component F2 and the principal component PC1 of the sediment core C0702 are restricted by the EAWM. F2 and PC1 are the products of sediment deposition as a function of the hydrodynamic environment under the influence of the winter monsoon. In theory, a relationship should exist between the F2 or PC1 components and the EAWM. With respect to the chronological framework of the core, because of the existence of three temporal markers, which can be readily identified, the dating results are more accurate from 1950 to 2009. Therefore, we selected the F2 and PC1 components beginning in 1950 and compared them with the East Asian Winter Monsoon Index (EAWMI), from He and Wang (2012), which is composed of the weighted average of the normalized strength index of the East Asian Rapids (EAJ) radial shear, the Siberian high pressure, and the East Asian trough. The analysis revealed that a significant correlation exists between the F2 or PC1 components and EAWMI, that is, high values of F2 and PC1 correspond to high values of EAWMI and vice versa (Figure 8a–c). After applying data homogenization and moving average processing to the initial data, we found a good linear correlation between the EAWMI and F2, PC1 (Figure 8d and e), respectively, expressed by

Y_{EAWM} = 0.495 \times X_{F 2} + 0.304; r = 0.46

(1)

Y_{EAWM} = 0.831 \times X_{PC1} + 0.3; r = 0.55

(2)

Figure 8.

Comparative analysis and correlation variations between the sensitive grain-size component F2, principal component PC1, and corrected components F2′ and PC1′ of core C0702 and the EAWMI. (a) Variation in the sensitive grain-size component F2 (red solid line) and the corrected F2′ (blue solid line) of core C0702 from 1950 to 2008. (b) Variation in principal component PC1 (red solid line) and the corrected PC1′ of core C0702 from 1950 to 2008. (c) East Asian Winter Monsoon Index (EAWMI), based on data obtained from He and Wang (2012). In the figure, the gray solid line represents the initial data, the red and blue solid lines in (a) and (b) represent the five-point moving average results, and the red solid line in (c) represents the three-point moving average results. The intersection points of the 12 black dotted lines and the moving average curves are used as data points for the linear analysis in Figure 8. (d) and (e) Linear correlation between the sensitive grain-size component F2, principal component PC1, and the EAWMI; black dots represent the initial data, and red dots represent the corrected data. The data for linear analysis were extracted from the intersection of black dotted lines and the moving average curves in the left-hand image.

It should be noted that at certain times, this correlation is not very obvious, which may be caused by the large-scale artificial dams in the Yangtze River Basin and other long-term climatic factors (e.g. Pacific Decadal Oscillation, Arctic Oscillation) (Li et al., 2015). In addition, under the same hydrodynamic environment conditions, individual grain-size components interfere with each other, which affects the intrinsic relationship between F2, PC1, and EAWMI. To better understand the relationship between F2, PC1, and EAWMI, the ratio of F2/(F1 + F2) was first used to reflect more clearly the correlation between the grain-size components and to eliminate the effects of inherent physical factors (Weltje and Tjallingii, 2008). Next, any trend components of F2 and PC1 were removed using EEMD. As a result, we were able to eliminate, to a certain extent, the effects of long-term climatic factors and human activity, primarily referring to the artificial dam construction. On this basis, the corrected sensitive grain-size component F2′ and principal component PC1′ were obtained. After applying the correction, the correlation coefficient, R, between F2, PC1, and EAWMI increased significantly. New linear relationships between the corrected core components F2′, PC1′, and EAWMI were established as follows (Figure 8d and e):

Y_{EAWM} = 0.757 \times X_{F 2'} + 0.13; r = 0.69

(3)

Y_{EAWM} = 0.888 \times X_{F 2} + 0.217; r = 0.70

(4)

As shown in Figure 8, there is a good correlation between EAWM intensity based on measurement (REF) and that based on our estimates. However, there was an interval of inconsistency during the period 1960–1970. The sediment discharge from Changjiang decreased by ~25% between mid-1960s and early 2000. Annual sediment loads at Huang-jiagang station has decreased dramatically, from ~150 mt/yr in early 1960s to ~5 mt/yr in late 1960s (Yang et al., 2006). This may be the main reason for the difference between the two values.

Based on the linear relationship between the corrected sensitive grain-size component F2′ and principal component PC1′ and the EAWMI, this study established the intensity variation curve of the EAWM starting from 1882, as shown in Figure 9a and b. The interdecadal variations of the reconstructed EAWM from 1900 to 1950 are consistent with the fluctuation range of Siberian High index (SHI) during the same period. As shown in Figure 9, the period 1882–1945 can be divided into two stages: a weak EAWM period from 1882 to 1900 and a strong EAWM period from 1900 to 1945. These findings are consistent, to a certain extent, with the results of a previous study on the reconstruction of EAWM intensity variations using a millennial scale based on the comparison of the diatomic ratio of the diatoms A. granulate/C. stelligera in lakes; the strong and the weak stages of the EAWM in both studies are in good agreement (Wang et al., 2012). Good agreements were found between our results and temperature records based on tree-ring (Figure 9e), from Changbai Mountain, which reflected February–April temperature of Northeast China in this area (Zhu et al., 2009). The reconstructed EAWM intensity shows negative correlations with the temperature records based on tree-ring. It is also possible to compare our reconstructed EAWM intensity with a composite oxygen isotopic record of cave calcite based on a stalagmite from the Wanxiang Cave in northern China, which reflected the precipitation variability in northern China (Tan et al., 2011). The reconstructed EAWM intensity was correlated with the monsoon precipitation variations based on stalagmite and showed an increase in the EAWM matching with a decreased trend of the precipitation recorded in the stalagmite (Figure 9f). Compared with the SHI, which directly reflects climatic variations, the fluctuation range is consistent, but the fluctuation period and phase are not in complete agreement (Panagiotopoulos et al., 2005).

Figure 9.

Comparative analysis of the reconstructed EAWMI and proxy indicators based on the calibrated F2′ and PC1′ of core C0702. (a) and (b) Reconstructed EAWMI (dotted line) based on F2′ and PC1′. Solid lines represent the EAWMI reconstructed on the basis of the 1950–2009 instrumental data in Zhou et al. (2007) and He and Wang (2012). (c) Siberian High index (SHI) data from Panagiotopoulos et al. (2005). (d) Lake diatom A. granulate and C. stelligera content ratio AG/CS, cited from Wang et al. (2012). (e) Temperature records based on tree-ring from Changbai Mountain of China, cited from Zhu et al. (2009). (f) The composite oxygen isotopic record of cave calcite based on stalagmite from Wanxiang cave of China, cited from Tan et al. (2011).

The lack of comparable results from previous studies creates limitations in evaluating the accuracy of this EAWM reconstruction. However, the present reconstruction method has the following advantages. First, the quantitative proxies of the EAWM were determined on the basis of statistical correlation between measured observations and sedimentological indicators. Second, the inner shelf mud deposits have high sediment accumulation rates, which are conducive for radiometric dating and the identification of event deposits, which enables more accurate dating. As a result, it has been possible to obtain high-resolution sedimentary records of the EAWM variations on interannual or interdecadal scales.

Conclusion

The grain size of the inner shelf mud deposits of the East China Sea reflects the sediment transport conditions and the hydrodynamic environment. More specifically, the sensitive grain-size component F2, with a grain-size interval of 14.35–230 µm, shows good correspondence with the intensity of the EAM and can therefore be used as a proxy of the EAWM. PC1, which is represented mainly by Si, Ca, Sr and Zr, is restricted by the coarse grain component of the sediment. PC1 also shows a good correspondence with the EAWM. After eliminating the event deposits and long-term climatic trend factors, we established quantitative formulas for reconstruction of the EAWM based on F2 and PC1. Based on the established formulas, we reconstructed the evolutionary history of the EAM from 1880 to 1950, which can be roughly divided into the following two stages: a weak EAWM period from 1882 to 1900 and a strong EAWM period from 1900 to 1945. This study presents a new method for studying the EAM at a high temporal resolution, which fills a gap in the existing research on the use of interannual and interdecadal event scales to study monsoon variations and provides a basis for further research.

Footnotes

Acknowledgements

The authors thank the crew of the R/V Dong Fang Hong 2 for kindly assisting in sedimentary core sampling on the cruise in May 2009. Finally, we are grateful to the associate editor Prof. Viv Jones for her great effort on editing a previous version of this manuscript and to the two anonymous reviewers for their very constructive comments.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the National Natural Science Foundation of China under contract Nos 41530966 and 41376055, the Marine Geological Survey Program of China Geological Survey (DD20190819), and National Key Basic Research and Development Program of China (2018YFC0310003 and 2017YFC0307704).

ORCID iD

Dejiang Fan

References

(2000) The history and variability of the East Asian paleomonsoon climate. Quaternary Science Reviews 19: 171–187.

Boyer-Villemaire

St-Onge

Bernatchez

, et al. (2013) High-resolution multiproxy records of sedimentological changes induced by dams in the Sept-Îles area (Gulf of St. Lawrence, Canada). Marine Geology 338: 17–29.

Cheng

Edwards

Sinha

, et al. (2016) The Asian monsoon over the past 640,000 years and ice age terminations. Nature 534(7609): 640–646.

Cuven

Francus

Lamoureux

(2011) Mid to Late-Holocene hydroclimatic and geochemical records from the varved sediments of East Lake, Cape Bounty, Canadian High Arctic. Quaternary Science Reviews 30(19–20): 2651–2665.

Deng

Zhang

(2006) Recent sediment accumulation and carbon burial in the East China Sea. Global Biogeochemical Cycles 20(3): 466–480.

Dong

Wang

Cheng

, et al. (2010) A high-resolution stalagmite record of the Holocene East Asian monsoon from Mt Shennongjia, central China. The Holocene 20(2): 257–264.

Ercolani

Muller

Collins

, et al. (2015) Intense southwest Florida hurricane landfalls over the past 1000 years. Quaternary Science Reviews 126: 17–25.

Fan

Sun

, et al. (2011) Annual lamination and its sedimentary implications in the Yangtze River delta inferred from High-resolution biogenic silica and sensitive grain-size records. Continental Shelf Research 31(2): 129–137.

Wang

(2012) An integrated East Asian winter monsoon index and its interannual variability. Chinese Journal of Atmospheric Sciences 36(3): 523–538 (in Chinese, with English Summary).

10.

Zhou

Liu

, et al. (2014) Weakened Yellow Sea Warm Current over the last 2–3 centuries. Quaternary International 349(349): 252–256.

11.

Yang

Zhao

, et al. (2012) Grain size records reveal variability of the East Asian Winter Monsoon since the Middle Holocene in the Central Yellow Sea mud area, China. Science China Earth Sciences 55(10): 1656–1668.

12.

Lin

Shi

, et al. (2011) The role of shelf mud depositional process and large river inputs on the fate of organochlorine pesticides in sediments of the Yellow and East China seas. Geophysical Research Letters 38(3): 246–258.

13.

IPCC (2013) Climate Change 2013: The Physical Science Basis (Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change). Cambridge; New York: Cambridge University Press.

14.

Kong

Tse

Jia

, et al. (2015) Cooling trend over the past 4 centuries in northeastern Hong Kong waters as revealed by alkenone-derived SST records. Journal of Asian Earth Sciences 114: 497–503.

15.

Yao

Bianchi

, et al. (2015) Historical reconstruction of organic carbon inputs to the East China Sea inner-shelf: Implications for anthropogenic activities and regional climate variability. The Holocene 25(12): 1869–1881.

16.

Liu

, et al. (2014) Sedimentary system response to the global sea level change in the East China Seas since the last glacial maximum. Earth-Science Reviews 139: 390–405.

17.

Wang

Qiao

, et al. (2012) The impact of typhoon Morakot on the modern sedimentary environment of the mud deposition center off the Zhejiang–Fujian coast, China. Continental Shelf Research 37: 92–100.

18.

Liu

Saito

Kong

, et al. (2010a) Sedimentary record of environmental evolution off the Yangtze River estuary, East China Sea, during the last ~13,000 years, with special reference to the influence of the Yellow River on the Yangtze River delta during the last 600 years. Quaternary Science Reviews 29(17): 2424–2438.

19.

Liu

, et al. (2007) Flux and fate of Yangtze River sediment delivered to the East China Sea. Geomorphology 85(3): 208–224.

20.

Liu

Shi

Liu

, et al. (2010b) Records of the East Asian winter monsoon from the mud area on the inner shelf of the East China Sea since the Mid-Holocene. Chinese Science Bulletin 55: 2306–2314.

21.

Palinkas

Nittrouer

(2007) Modern sediment accumulation on the Po shelf, Adriatic Sea. Continental Shelf Research 27: 489–505.

22.

Panagiotopoulos

Shahgedanova

Hannachi

, et al. (2005) Observed trends and teleconnections of the Siberian High: A recently declining center of action. Journal of Climate 18(9): 1411–1422.

23.

Qiao

Shi

Wang

, et al. (2017) Sediment accumulation and budget in the Bohai Sea, Yellow Sea and East China Sea. Marine Geology 390: 270–281.

24.

Qiao

Yang

Liu

, et al. (2010) Records of Late-Holocene East Asian winter monsoon in the East China Sea: Key grain-size component of quartz versus bulk sediments. Quaternary International 230(1–2): 106–114.

25.

Sagawa

Kuwae

Tsuruoka

, et al. (2014) Solar forcing of centennial-scale East Asian winter monsoon variability in the Mid- to Late-Holocene. Earth and Planetary Science Letters 395(3): 124–135.

26.

Huh

(2002) ²¹⁰Pb, ¹³⁷Cs and ^239,240Pu in East China Sea sediments: Sources, pathways and budgets of sediments and radionuclides. Marine Geology 183(1–4): 163–178.

27.

Sun

Fan

Tian

, et al. (2018) Normalization of excess ²¹⁰Pb with grain size in the sediment cores from the Yangtze River Estuary and adjacent areas: Implications for sedimentary processes. The Holocene 28(4): 545–557.

28.

Sun

Gao

(2003) Preliminary analysis of grain-size populations with environmentally sensitive terrigenous components in marginal sea setting. Chinese Science Bulletin 48(2): 184–187.

29.

Sun

Clemens

Morrill

, et al. (2012) Influence of Atlantic meridional overturning circulation on the East Asian winter monsoon. Nature Geoscience 5: 46–49.

30.

Tan

Cai

, et al. (2011) Centennial- to decadal-scale monsoon precipitation variability in the semi-humid region, northern China during the last 1860 years: Records from stalagmites in Huangye Cave. The Holocene 21(2): 287–296.

31.

Tian

Fan

Zhang

, et al. (2019) Event deposits of intense typhoon in the muddy wedge of East China Sea over the past 150 years. Marine Geology 410: 109–121.

32.

Wang

Fan

, et al. (2015) The variation of clay minerals contents on the inner shelf of the East China Sea in the last hundred years and its environment implication. Acta Oceanologica Sinica 37(5): 87–100 (in Chinese, with English Summary).

33.

Wang

, et al. (2012) The East Asian winter monsoon over the last 15,000 years: Its links to high-latitudes and tropical climate systems and complex correlation to the summer monsoon. Quaternary Science Reviews 32(1): 131–142.

34.

Wang

Clemens

Beaufort

, et al. (2005) Evolution and variability of the Asian monsoon system: State of the art and outstanding issues. Quaternary Science Reviews 24(5–6): 595–629.

35.

Weltje

Prins

(2003) Muddled or mixed? Inferring palaeoclimate from size distributions of deep-sea clastics. Sedimentary Geology 162: 39–62.

36.

Weltje

Tjallingii

(2008) Calibration of XRF core scanners for quantitative geochemical logging of sediment cores: Theory and application. Earth and Planetary Science Letters 274(3): 423–438.

37.

Huang

(2009) Ensemble empirical mode decomposition: A noise-assisted data analysis method. Advances in Adaptive Data Analysis 1(1): 1–41.

38.

Xiang

Yang

Saito

, et al. (2006) East Asia winter monsoon changes inferred from environmentally sensitive grain-size component records during the last 2300 years in mud area southwest off Cheju Island, ECS. Science in China (Series D) 49(6): 604–614.

39.

Xiao

Liu

, et al. (2006) Coherence between solar activity and the East Asian winter monsoon variability in the past 8000 years from Yangtze River-derived mud in the East China Sea. Palaeogeography, Palaeoclimatology, Palaeoecology 237: 293–304.

40.

Yang

Liu

(2007) A unique Yellow River-derived distal subaqueous delta in the Yellow Sea. Marine Geology 240(1–4): 169–176.

41.

Yang

Wang

Sation

, et al. (2006) Dam impacts on the Changjiang (Yangtze) River sediment discharge to the sea: The past 55 years and after the Three Gorges Dam. Water Resources Research 42(4): W04407.

42.

Zong

Lloyd

, et al. (2012) Mid-Holocene variability of the East Asian monsoon based on bulk organic δ13C and C/N records from the Pearl River estuary, southern China. The Holocene 22(6): 705–715.

43.

Zhang

Fan

Wang

, et al. (2013) The calibration and quality evaluation of elemental analysis results of marine sediment measured by an X-ray fluorescence core scanner. Acta Oceanologica Sinica 35(6): 86–95 (in Chinese, with English Summary).

44.

Zhang

Fan

, et al. (2018) Sedimentary laminae in muddy inner continental shelf sediments of the East China Sea: Formation and implications for geochronology. Quaternary International 464: 343–351.

45.

Zheng

Wan

, et al. (2014) ITCZ and ENSO pacing on East Asian winter monsoon variation during the Holocene: Sedimentological evidence from the Okinawa Trough. Journal of Geophysical Research Oceans 119(7): 4410–4429.

46.

Zhou

Wang

Zhou

, et al. (2007) Interdecadal variability of the relationship between the East Asian winter monsoon and ENSO. Meteorology and Atmospheric Physics 98(3–4): 283–293.

47.

Zhu

Fang

Shao

, et al. (2009) Tree ring-based February–April temperature reconstruction for Changbai Mountain in Northeast China and its implication for East Asian winter monsoon. Climate of the past 5(4): 661–666.

48.

Zong

Lloyd

Leng

, et al. (2006) Reconstruction of Holocene monsoon history from the Pearl River Estuary, southern China, using diatoms and carbon isotope ratios. The Holocene 16(2): 251–263.