Introduction
Global warming is expected to alter the behavior of extreme hydroclimate events, amplifying their magnitude and frequency (Giorgi et al., Reference Giorgi, Im, Coppola, Diffenbaugh, Gao, Mariotti and Shi2011; Trenberth et al., Reference Trenberth, Dai, Van Der Schrier, Jones, Barichivich, Briffa and Sheffield2014). It is essential to understand natural climatic variability and the factors controlling it to test the possible effects of anthropogenic influences and to anticipate future hydroclimate changes. East Asia is an optimal region for testing the responses of climate factors, as they are characterized by marked changes in the seasonal wind direction driven by temperature differences between the low-latitude Pacific Ocean and high-latitude Eurasian continent (Fig. 1), which lead to a warm wet East Asian summer monsoon flow (EASM) and cold dry East Asian winter monsoon flow (EAWM). Both monsoonal winds are under the strong influence of seasonal shifts in the Intertropical Convergence Zone (ITCZ) (An et al., Reference An, Poter, Kutzbach, Xihao, Suming, Xiaodong, Xiaoqiang and Weijian2000; Dykoski et al., Reference Dykoski, Edwords, Cheng, Yuan, Cai, Zhang, Lin, Qing, An and Revenaugh2005; Wang et al., Reference Wang, Cheng, Edwards, He, Kong, An, Wu, Kelly, Dykoski and Li2005; Yancheva et al., Reference Yancheva, Nowaczyk, Mingram, Dulski, Schettler, Negendank, Liu, Sigman, Peterson and Haug2007; Jo et al., Reference Jo, Woo, Yi, Yang, Lim, Wang, Cheng and Edwards2014; Chen et al., Reference Chen, Xu, Chen, Birks, Liu and Zhang2015). An interesting feature of seasonal ITCZ shift is antiphasing between the Northern and Southern Hemispheres. Under northward (southward) movement of the ITCZ, summer monsoonal precipitation increases (decreases) in the Northern Hemisphere but decreases (increases) in the Southern Hemisphere, suggesting synchronous changes in global monsoons (Haug et al., Reference Haug, Hughen, Sigman, Peterson and Röhl2001; X. Wang et al., Reference Wang, Auler, Edwards, Cheng, Ito, Wang, Kong and Solheid2007; Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009; Wanner et al., Reference Wanner, Solomina, Grosjean, Ritz and Jetel2011; Jo et al., Reference Jo, Woo, Yi, Yang, Lim, Wang, Cheng and Edwards2014; P.X. Wang et al., Reference Wang, Wang, Cheng, Fasullo, Guo, Kiefer and Liu2014, Reference Wang, Wang, Cheng, Fasullo, Guo, Kiefer and Liu2017; Tan et al., Reference Tan, Shen, Löwemark, Edwards, Chawchai, Cai and Wang2019; Chawchai et al., Reference Chawchai, Tan, Löwemark, Wang, Yu, Chung, Mii, Liu, Wohlfarth and Shen2021; Lim et al., Reference Lim, Yi and Kim2023a; Fig. 1).
Figure 1. Map showing global monsoon areas and the present seasonal shift in the intertropical convergence zone (ITCZ) (modified from Google Maps). Circle indicates the East Asian summer monsoon (EASM) sites, triangle indicates the Indian summer monsoon (ISM) sites, and square indicates the South American summer monsoon (SASM) region. The locations of the geological records of EASM over the Hadong area (HD), Korea (this study) and China (DC, Dongge Cave [Dykoski et al., Reference Dykoski, Edwords, Cheng, Yuan, Cai, Zhang, Lin, Qing, An and Revenaugh2005]; GL, Gonghai Lake [Chen et al., Reference Chen, Xu, Chen, Birks, Liu and Zhang2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018]), ISM (Qunf Cave [Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003] and Tianmen Cave [Cai et al., Reference Cai, Zhang, Cheng, An, Edwards, Wang, Tan, Liang, Wang and Kelly2012]), and SASM (Padre Cave, Brazil; Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009) are also shown.
Regarding high-latitude influences, past studies suggest a strong link between summer rainfall in East Asia and the westerly jet in terms of combined influences of high/low-latitude forcing (Xiao et al., Reference Xiao, Nakamura, Lu and Zhang2002, Reference Xiao, Xu, Nakamura, Yang, Liang and Inouchi2004, Reference Xiao, Zhang, Fan, Wen, Zhai, Tian and Jiang2018; Nagashima et al., Reference Nagashima, Tada and Toyoda2013; Chen et al., Reference Chen, Xu, Chen, Birks, Liu and Zhang2015; Fig. 2A). The westerly jet over the Tibetan Plateau plays a critical role in establishing atmospheric conditions conducive to convection by facilitating the development of the EASM rainfall frontal system (Schiemann et al., Reference Schiemann, Lüthi and Schär2009). During winter and early spring, the westerly jet lies south of the Himalayas, but it shifts north of the Tibetan Plateau in late spring as the Hadley cell circulation intensifies. In recent studies, the meridional jet position was found to control the intraseasonal stages of the EASM, such as spring, pre-Meiyu, Meiyu, and post-Meiyu periods (Chiang et al., Reference Chiang, Fung, Wu, Cai, Edman, Liu, Day, Bhattacharya, Mondal and Labrousse2015, Reference Chiang, Swenson and Kong2017; Day et al., Reference Day, Fung and Risi2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018). Regarding the possible relationship between the westerly jet position and EASM over China during the Mid-Holocene, Chen et al. (Reference Chen, Zhang, Huang, Lu, Zhang and Chen2021) revealed that the northward shift of the westerly jet was driven by increased meridional air temperature differences over China, and this shift caused more precipitation in north China through the enhanced southerly moisture flow from south China, leading to a northwesterly shift of the northern boundary of the EASM.
Figure 2. Simplified climatic system around the East Asia region and study sites (modified from Google Maps, Naver Maps). (A) Schematic map showing East Asian and Indian monsoon areas and the present seasonal shift in the westerly jet over the Tibetan Plateau and Asian continent. The East Asian summer monsoon (EASM; this study), Indian summer monsoon (ISM; Qunf Cave; Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003), and atmospheric temperature data from Greenland (NGRIP ice cores) are indicated. (B) Study area (JG-04) on the southern coast of Korea and the comparison site (QJS60; Park et al., Reference Park, Lim, Kim, Shin and Lim2023). (C) Changma front affecting both study sites, QJS60 and JG-04, during 2019 (modified from a satellite image of GEO-KOMPSAT-2A, August 7, 2020). (D) Study area (JG-04) in Hadong Province and Gwangyang Bay (core STP17-14; Lim et al., Reference Lim, Yi and Kim2023a), southern coast of Korea. (E) Study area and coring site (JG-04), located in a former shallow bay (paleo-Hadong Bay), now part of a reclaimed land area.
Furthermore, it is important to test high-latitude and low-latitude influences on past EASM variability simultaneously. During the Early Holocene, which had greater summer insolation (Kutzbach, Reference Kutzbach1981) and remnant glaciers in the Northern Hemisphere (Owen, Reference Owen2009; Renssen et al., Reference Renssen, Seppä, Heiri, Roche, Goosse and Fichefet2009), the midlatitude EASM could respond dynamically to both high- and low-latitude forcing climate factors. An 8.2 ka cooling event may provide such chance to test their influences at decadal to centennial timescales, because the cooling event, which initiated in the North Atlantic region, has been recorded as signals characterized by decreases in precipitation and temperature over Asia (Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009; Hong et al., Reference Hong, Hong, Lin, Shibata, Zhu, Leng and Wang2009; Xu et al., Reference Xu, Xiao, Li, Tian and Nakagawa2010; Wu et al., Reference Wu, Wang, Cheng, Kong and Liu2012; Liu and Hu, Reference Liu and Hu2016; Song et al., Reference Song, Li, Lu, Mao, Saito, Yi and Lim2017; Park et al., Reference Park, Park, Yi, Kim, Lee and Jin2018). Liu et al. (Reference Liu, Liao, Qiu, Yang, Feng, Allan, Can, Long and Xu2020) suggested that cooling of the North Atlantic region was an essential component of the centennial timescale variation in rainfall over the Chinese Loess Plateau during the Holocene. Moreover, the decadal to centennial variation in the Indian summer monsoon (ISM) responded to changes in Himalayan snow cover and North Atlantic sea-surface temperature, including the 8.2 ka cooling event (Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003, Reference Fleitmann, Burns, Mangini, Mudelsee, Kramers, Villa and Neff2007).
Regarding this 8.2 ka event and its influence on EASM variability over Korea, a previous study has tested the time series of the high-resolution Zr/Ti ratios (a proxy for past freshwater input) reconstructed from a paleo-bay site in Buan area, western coast of Korea (Park et al., Reference Park, Lim, Kim, Shin and Lim2023), and discussed possible linkages to regional and global monsoonal changes in terms of global teleconnection. These results need to be tested with data from other coastal areas in South Korea, because it is likely that these past hydroclimate variabilities were interconnected through the seasonal march of EASM rainbands over Korea. In this study, we tested hydroclimatic implications for high-resolution elemental ratios (e.g., titanium/aluminum [Ti/Al] and zirconium/aluminum [Zr/Al]) in the sedimentary cores in the Hadong area by comparing the elemental ratios with various geochemical data (e.g., total sulfur [TS%] and C/S ratios) in terms of selection for site-specific (or local) hydroclimate or environmental indicators in combination with bedrocks and geomorphological features. Then, we reconstructed the summer monsoonal rainfall change in the paleo-bay site of the Hadong area, southern Korea, and tested possible seasonal march of EASM based on changes in Changma rain front–controlled elemental ratios (e.g., Ti/Al and Zr/Al). Using these high-resolution EASM proxies from the two sites (Hadong and Buan), we explored multi-decadal- to centennial-scale associations among the global summer monsoons over the Northern Hemisphere, including the EASM, ISM, and North Atlantic region for 8.6–7.8 ka, in terms of global atmospheric circulation teleconnection during the Early to Mid-Holocene. We also tested the effects of high-latitude climate forcing on the EASM compared with the atmospheric temperature over Greenland, tracing possible influence of the westerly jet position.
Study area and methods
Study area and meteorological information
In the Korean Peninsula, most of the monsoon rainfall, more than 70% of the annual precipitation, is concentrated in summer (June to September) along the Changma (also called “Meiyu” in China and “Baiu” in Japan) rainband, which is located at the front of the EASM (Fig. 2A and C). The present summer precipitation in the study area and surrounding areas shows decadal variation (Fig. 3A and B). The precipitation at Jinju (Hadong) and Buan was highly correlated from 1973 to 2020 (r = 0.68; Fig. 3B). The precipitation records from the two areas are directly linked to the Changma front shift over South Korea, a part of the EASM system. Of note, periods of decreased rainfall (e.g., 1995, 2015) in the Hadong and Buan areas follow a pseudo–20 year cycle.
Figure 3. (A) Annual summer precipitation records for the study area (Jinju; Figure. 2C and 2D) and Buan (core QJS6; Figure. 2B) from 1972 to 2020. (B) Correlation of summer precipitation (3 year average) between Jinju and Buan, with a correlation coefficient of R = 0.69.
Sampling and physical and geochemical analyses
The low-lying areas near the present coastal regions in the western and southern parts of the Korean Peninsula (Fig. 2B) were mainly tidal flats or bay environments in the Holocene (Fig. 2D and E). Following their reclamation, these areas are now the sites of industrial complexes and agricultural land. The study area (Hadong), a small paleo-bay with an area of ∼4 × 4 km, is located in the lower reach of the Seomjin River, which drains into Gwangyang Bay and forms an estuarine environment including tidal flats and sandbars (Fig. 2D). Hadong is a low-elevation area that is vulnerable to heavy rainfall-driven flooding events during summer; therefore, past hydrological extreme events, including the EASM, could be recorded in the paleo-bay, which had been separate from the main river channel, as a partly stagnant water–dominant bay (Fig 2E).
The bedrock at Hadong is mainly granite gneiss, which is composed of quartz, feldspar, magnetite, muscovite, and zircon (Kim and Kang, Reference Kim and Kang1965). The Hadong core site (JG-04; 35°3′58.52″N, 127°45′27.88″E; elevation, 2.43 m) was a tidal flat in 1920 and has since been gradually reclaimed. In 2020, a 31-m-long sedimentary core (JG-04) was drilled using a standard rotary piston sampler with a 50 mm inner diameter in a reclaimed paddy field at the site. As seen in Figure 4A, the recovery rate of the entire core was greater than 90%, except for intervals of 25∼31 m and surface layer (0∼1 m). The core was subsampled at 5 cm intervals after photographs were taken (Fig. 4A) and visual observations were made to describe its lithological features.
Figure 4. (A) Photograph of core JG-04 (0∼31 m) from Hadong Province, Korea. (B) Age–depth model for core JG-04. Depth (m) converted to age (cal yr BP) by using an age–depth model calculated with CLAM software (Blaauw, Reference Blaauw2010). The model was generated using linear interpolation between dated levels (type = 1), weighted by calibrated probabilities (prob. = 0.95), and excluding three identified outliers. Reversed ages (marked in red) were not used in the model.
For radiocarbon dating, plant fragments were sampled from core JG-04 (n = 20). The pretreatment for 14C dating followed reported methods (e.g., Kigoshi et al., Reference Kigoshi, Suzuki and Shiraki1980; Kretschmer et al., Reference Kretschmer, Anton, Bergmann, Finckh, Kowalzik, Klein and Leigart1997; Kim et al., Reference Kim, Lim, Yu, Park, Lee, Hong and Park2021). The plant fragments were subjected to acid (0.5 M HCl), alkali (0.5 M NaOH), and acid (0.5 M HCl) treatment to remove potential contamination. After graphitization of the pretreated samples, radiocarbon dating was performed using the accelerator mass spectrometry facility at the Korea Institute of Geoscience and Mineral Resources (KIGAM). Radiocarbon (14C) ages were converted into calibrated ages (cal yr BP) using OxCal 4.4 and IntCal20 (Bronk Ramsey, Reference Bronk Ramsey2009a, Reference Bronk Ramsey2009b; Reimer et al., Reference Reimer, Austin, Bard, Bayliss, Blackwell, Bronk Ramsey and Butzin2020). Based on the 14C dates, we established age–depth models using CLAM software (Blaauw, Reference Blaauw2010; Fig. 4B).
For grain-size analysis, we obtained ∼0.5 g subsamples from core JG-04 (n = 176) and pretreated them with H2O2 to remove organic matter and then with HCl to remove carbonate; then the subsamples were washed three times with distilled water. Sediment grain sizes were measured using a Mastersizer 3000 (Malvern Panalytical, Malvern, UK) at KIGAM.
To obtain high-resolution information on elements in the sediment core (e.g., Al, Si, K, Rb, Zr, Ti, and S), we used the XRF Core Scanner (Avaatech B.V., Alkmaar, The Netherlands) at KIGAM at 10 kV and 0.8 mA with no filter and at 30 kV and 1.0 mA with a thick lead filter (exposure time: 10 s; down-slit size: 1.2 cm). All elements were measured at 5 mm intervals. To test for similarities between the high-resolution XRF core data from the sediment core, we calculated Pearson’s correlation coefficient.
To trace possible organic sources, we measured the total organic carbon (TOC) and total nitrogen (TN) contents in HCl-pretreated subsamples (JG-04; n = 176) using a C/N/S elemental analyzer (vario MICRO cube; Elementar, Langenselbold, Germany) at KIGAM. Finally, we calculated the TOC/TN (C/N) and TOC/TS (C/S) ratios for each sample depth.
Results
Radiocarbon dating and lithological features
As shown in Table 1 and Figure 4B, the radiocarbon ages of plant fragments (n = 16) revealed that core JG-04 was deposited between 9840 cal yr BP (24.25 m) and 7140 cal yr BP (7.48 m) in ascending order, except for three slight reversal ages. The upper layer between 5.5 and 2.72 m was deposited at 350–100 cal yr BP, as shown by the radiocarbon ages (n = 4). The age jump between 7000–400 cal yr BP suggests a significant hiatus between the two sedimentary layers. We mainly focused on the lower part of the cores before the hiatus. The core can be subdivided into five sedimentary units (Fig. 5): Unit 1, yellowish, gravelly, coarse sand layer overlying weathered bedrock, at 29.7–28.3 m; Unit 2, gray gravelly sand at 28.3–25 m; Unit 3, gray silty mud with no clear layer, at 25–6.55 m; Unit 4, gray sandy mud layer with partial cross bedding, at 6.55–1.9 m; and Unit 5, grayish to yellowish mud layer with reworked weathered sandy deposits formed by reclamation.
Figure 5. Lithological features and results of age dating, grain-size, and geochemical analyses of core JG-04 from Hadong Province, southern coast of Korea. (A) Radiocarbon dating results. (B) Median grain size. (C) C/N ratios. (D) C/S ratios. (E) Total sulfur (TS) content. (F) Sulfur intensity measured by X-ray fluorescence core scanning (XRF-CS). cps, counts per second.
Table 1. Results of radiocarbon dating for core JG-04, Hadong area, Korea.
a Calibrated with radiocarbon calibration program (OxCal http://c14.arch.ox.ac.uk/embed.php?File=oxcal.html).
b Measured by using the accelerator mass spectrometry facility at the Korea Institute of Geoscience and Mineral Resources (KIGAM) (n = 20).
c Dates assumed as outliers (three points; see Fig. 4B).
We converted the core JG-04 depths (m) into ages (cal yr BP) based on sedimentation rates using age–depth models (Fig. 4B) calculated with CLAM (Blaauw, Reference Blaauw2010), with linear interpolation between dated points. Note that there was a hiatus in the dating ages in core JG-04 between Units 3 and 4 (Fig. 5A). To discuss a reliable high-resolution sedimentary environment and its physical and geochemical features, we focus on Unit 3 with the higher recovery rate of greater 95%, corresponding to a well-dated, lithologically stable period.
Grain-size and elemental analyses
As shown in Figure 5B, the gray silt of Unit 3 was fine to medium silt with coarsening upward, with a median grain size of 10–30 µm. Both the C/N and C/S ratios fluctuated significantly in the lower part of Unit 3 (Fig. 5C and D). Throughout Unit 3, C/N varied from 10 to 150 and C/S from 0.5 to 5. Especially, this unit can be divided into three subunits. Unit 3a is characterized by increasing TS% (Fig. 5E) and decreasing C/S ratios (Fig. 5D), indicating increasing seawater influence. High-amplitude fluctuation in the C/N ratios indicates a roughly decreasing trend, indicating an increase in aquatic plant influence (Fig. 5C). Unit 3b shows relatively stable TS% and C/N and C/S ratios, indicating strong aquatic plant and seawater influences. Then Unit 3c shows a significant swing in C/N ratios coupled with increased median grain size.
XRF core scanning (XRF-CS) results
Figure 6 shows the depth profiles of median grain size (µm) in Unit 3 and 4 and semiquantitative element ratios, that is, Ti/Al and Zr/Al in Unit 3 of core JG-04. The depth profiles of elemental ratios and median grain sizes show partly similar variability. For example, in Unit 3c the larger median grain sizes correlated with increased elemental ratios (e.g., Zr/Ti, Zr/Al and Ti/Al) and decreased S intensity (counts per second [cps]). Compared with other elemental ratios, Ti/Al ratios seem to be negatively correlated with S intensity. This similar pattern can be supported by correlation test. As shown in Figure 7, the correlations between the elemental ratios (Ti/Al and Zr/Al) and S intensity in Unit 3b are high (e.g., R = 0.59), and this result indicates these elemental ratios were controlled by common factors.
Figure 6. Comparison of multi-elemental information with lithological features, age dating, and grain-size results in core JG-04 from Hadong Province, southern coast of Korea. Depth profiles of semiquantitatively determined element ratios (Zr/Ti, Zr/Al, Zr/K, Zr/Rb, and Sr/Ti) measured by X-ray fluorescence core scanning (XRF-CS).
Figure 7. Cross plots and correlation coefficients between elemental ratios (Ti/Al and Zr/Al) and S intensity (cps, counts per second) in lithological Unit3b of sedimentary core JG-04, Hadong Province, southern coast of Korea.
Discussion
Early Holocene sea-level change and coastal evolution in Hadong area
Previous studies have suggested that the current coastal morphological features in Korea are the result of local or regional responses to past sea-level changes since the last glacial maximum and deglacial periods (Bloom and Park, Reference Bloom and Park1985; Chough et al., Reference Chough, Lee, Chun and Shinn2004; Song et al., Reference Song, Yi, Yu, Nahm, Lee, Lim and Kim2018; Lim et al., Reference Lim, Lee, Hong, Park, Lee and Yi2019, Reference Lim, Yi and Kim2023a; Park et al., Reference Park, Lim, Kim, Shin and Lim2023; Yang et al., Reference Yang, Han, Yoon, Kim, Choi, Shin, Kim, Jung, Park and Jun2023). These studies clearly suggest the Holocene transgression in various regions, but few studies have examined the spatial response along main river channels and their sedimentary environmental changes in terms of sea level and their water depth due to the lack of a suitable index and inter-compatible sedimentary records with well-controlled ages.
Clearly, there were significant sea-level changes of up to 20 m during the Holocene (Tanigawa et al., Reference Tanigawa, Hyodo and Sato2013; Lambeck et al., Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014), and the study area and surrounding areas might have been influenced by this sea-level transgression. There are few reconstruction records of the Holocene sea-level changes for the southern coast of Korea, including our study area. We estimated past changes in water depth at the coring site based on reconstructed changes in sea level from the Yellow Sea and East China Sea (Lambeck et al., Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014 and references therein). Based on these estimated water depths, the present tidal range (∼3 m), and geochemical information, past changes in the coastal environment in the Hadong area were reconstructed and compared with recently reconstructed coastal information for the northern part of Gwangyang Bay (Lim et al., Reference Lim, Yi and Kim2023a) in terms of spatial changes in the coast.
It is worth noting that the Early to Mid-Holocene coastal sedimentary record corresponds to a time when sea level had changed the setting from intertidal to subtidal, back to intertidal, and finally to supratidal. These changes must have had an immense effect on the elemental and grain-size composition of the sediment. Even small shifts of a wandering river mouth could have a strong influence on the sediment delivered. In addition, compaction and subsidence could affect water depth and the architecture of the deposits. It is difficult to consider all these issues in detail with one sedimentary core. Thus, we will focus on the relatively stable bay environment with enough depth but separate from main flowing systems. The sedimentary records will be tested by comparison with other sedimentary cores from the same areas.
As shown in Figure 8, the water depth at core sites in Hadong (core JG-04) and Gwangyang (core STP17–14; Lim et al., Reference Lim, Yi and Kim2023a) should have been controlled by both changes in sea level and the sediment accumulation rate (e.g., Bird et al., Reference Bird, Fifield, Teh, Chang, Shirlaw and Lambeck2007). The past water depth at the sites may be approximated by subtracting the elevation of the sediment–water interface in the core from the reconstructed sea-level elevation at that time (Lambeck et al., Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014 and references therein). As shown in Figure 8A, the reconstructed sea level at ∼9600 cal yr BP was −18 m and the sediment–water interface elevation in core JG-04 was −24 m, suggesting a water depth of ∼6 m in the area. Interestingly, a similar water depth was found for core STP17-14. These water depths increased with an increase in sea level. At 8200 cal yr BP, the reconstructed sea level was around −2 m, and the resultant water depths at sites JG04 and STP17-14 were 10 and 15 m, respectively. This increase in sea level and the resultant water depth increases along the present Seomjin River suggest that the two sites were coastal bay environments with water depths exceeding 10 m with Early Holocene transgression. These deepwater bay environments seem to have persisted until 7000 years ago.
Figure 8. (A) Comparison between the reconstructed past sea-level change (Lambeck et al., Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014 and references therein) and elevation–age curves in Hadong area (JG-04; this study), Gwangyang Bay (STP17-14; Lim et al., 2023), and Buan area (QJS60; Park et al., Reference Park, Lim, Kim, Shin and Lim2023). (B) S intensity (cps, count per second) in the Hadong area (JG-04, this study). (C and D) Zr/Al and Ti/Al ratios in the Hadong area (JG-04; this study). (E and F) Comparison of total sulfur (TS) content in core JG-04 with C/S ratios from cores JG-04 (this study) and STP17-14 (Lim et al., 2023) on the southern coast of Korea during the Early Holocene. Past water depths at the coring sites were estimated based on reconstructed sea levels from the Yellow Sea and East China Sea (Lambeck et al., Reference Lambeck, Rouby, Purcell, Sun and Sambridge2014 and references therein).
These assumed changes in the coastal environment can be tested using the geochemical dataset (e.g., S intensity, TS%, and C/S ratios). The ratio of organic carbon to pyrite sulfur (C/S ratio; Fig. 8F) is thought to be affected by the salinity at the time of deposition, as diagenetic pyrite forms more readily in marine sediments than in freshwater ones due to the relatively high availability of dissolved sulfate in seawater (Berner, Reference Berner1984; Berner and Raiswell, Reference Berner and Raiswell1984; Morse and Berner, Reference Morse and Berner1995). The C/S ratios are 17–34 at salinities <1‰, whereas they are 1.4–1.8 at salinities of 19–21‰ (Berner and Raiswell, Reference Berner and Raiswell1984 and references therein). Furthermore, freshwater sediments have high C/S values (>10) and marine sediments have low values (0.5–4) (Berner, Reference Berner1984; Berner and Raiswell, Reference Berner and Raiswell1984; Woolfe et al., Reference Woolfe, Dale and Brunskill1995 and references therein). Compared with these values, deviations toward higher C/S ratios may be interpreted as a freshwater environment, while deviations toward lower C/S ratios suggest burial under a euxinic environment (Berner and Raiswell, Reference Berner and Raiswell1984).
While the water depth (∼6 m) at the two sites was similar at 9500 cal yr BP, the C/S ratios were <3 in core STP17-14, indicating a marine environment, where those in core 20JG04 were 3∼5, showing a somewhat brackish environment (Fig. 8F). This difference is natural, because the increase in sea level and resultant transgression started earlier in the river mouth area (STP17-14) than in areas in the lower reach of the Seomjin River (e.g., core JG04).
After 9200 cal yr BP, the C/S ratios in JG04 fell to 2 and remained below 3, suggesting a stable bay environment (paleo-Hadong Bay). Interestingly, the fluctuations of the C/S ratios at the two sites were similar (Fig. 8F). For example, peaks and troughs (denoted by a–f in Fig. 8F) reveal roughly simultaneous variability, indicating common peaks at 7850 cal yr BP. It is worthy to note that this similar fluctuation was not clear at 9000–8300 cal yr BP due to very low sedimentation rate in core STP17-14. But during this period, the C/S ratios, TS%, and S intensity show high-resolution records in the core JG-04. This suggests that the sedimentary environments in paleo-Hadong Bay at 9100–7000 cal yr BP were stable enough to archive their depositional environments under a water depth of ∼10 m, and this can be supported by the similar pattern in TS% and related S intensity measured by XRF-CS and narrow varying range around 2 in C/S ratios.
In addition to sulfur-based evidence, the environmental features of the paleo-Hadong Bay at 9100–7000 cal yr BP can be tested using C/N ratios of the sediments (Lamb et al., Reference Lamb, Wilson and Leng2006 and references therein; Park et al., Reference Park, Lim, Kim, Shin and Lim2023). The C/N ratio has been used as a proxy to trace sources of organic matter in lacustrine and coastal regions (Meyers, Reference Meyers1994; Sampei and Matsumoto, Reference Sampei and Matsumoto2001; Lamb et al., Reference Lamb, Wilson and Leng2006 and references therein). It has been suggested that the C/N ratios of terrestrial plants are >20, whereas the C/N ratios of aquatic plants are 4–10 (Meyers, Reference Meyers1994). As shown in Figure 5C, C/N ratios in Unit3a and Unit3c show increased terrestrial plant input, but the intervening Unit3b shows relatively low and stable C/N ratios, indicating increased aquatic plant input. This could indicate that there had been enough water depth in the paleo-bay to provide aquatic input and dilute terrestrial input, although the absolute amounts of aquatic and terrestrial organic matter could not be determined.
Furthermore, changes in sea level may influence C/N ratios in the coastal areas by controlling the distance to the coastline. Several studies showed an ∼11 m increase in sea level during the period 8640–7850 cal yr BP (e.g., Xiong et al., Reference Xiong, Zong, Li, Long, Huang and Fu2020; Xu et al., Reference Xu, Meng, Yuan, Teng, Xin and Sun2020). This increase in sea level and resultant increase in water depth in the study site, as well as the greater distance from the coring site to the coastline, could influence the significant changes in C/N ratios among Unit3a, Unit3bm and Unit3c. Based on the estimated water depth shown in Figure 8, these units correspond to shallow (relatively high C/N ratios), deep (relatively low C/N ratios), and shallow (relatively high C/N ratios) bay environments, respectively, suggesting a long-term aquatic influence and water mass changes.
Consequently, these changes in the elevation–age curves, estimated water depth, and C/S and C/N ratios in the sedimentary cores revealed clear coastal evolution in the Hadong area during the Early to Mid-Holocene. This suggests that the terrestrial element information (e.g., Zr/Al and Zr/Ti), especially in the Unit3b corresponding to stable paleo-bay environment, can be used to trace the responsible transport processes in the Seomjin River.
Multi-decadal to centennial hydrological changes on the South Korea during the Early to Mid-Holocene
To obtain the decadal-to-centennial variation in paleo-hydroclimate changes, it is necessary to reconstruct high-resolution proxies. Recently, high-resolution multi-elemental data obtained from nondestructive, continuous X-ray fluorescence core scanners have been widely used to reconstruct past environmental and hydroclimatic changes at high resolution (Haug et al., Reference Haug, Hughen, Sigman, Peterson and Röhl2001; Blanchet et al., Reference Blanchet, Thouveny, Vidal, Ledu, Tachikawa, Bard and Beaufort2007; Nizou et al., Reference Nizou, Hanebuth and Vogt2011; Wang et al., Reference Wang, Zheng, Xie, Fan, Yang, Zhao and Wang2011; Betrand et al., Reference Bertrand, Hughen, Giosan, Croudace and Rothwell2015; Lim et al., Reference Lim, Yang, Lee, Hong and Um2015, Reference Lim, Lee, Hong, Park, Lee and Yi2019, Reference Lim, Um, Yi and Jun2022, Reference Lim, Kim, Park and San Ahn2023b; Rothwell and Croudace, Reference Rothwell, Croudace, Croudace and Rothwell2015; Longman et al., Reference Longman, Veres and Wennrich2019; Park et al., Reference Park, Lim, Kim, Shin and Lim2023). But it is noteworthy that these elemental records have been influenced by different transportation and geomorphological features in study areas, suggesting that these high-resolution elemental records should be considered as site-specific (or local) hydroclimate or environmental indicators (Lim et al., Reference Lim, Lee, Hong, Park, Lee and Yi2019).
In general, Zr and Ti indicate coarse particles, because they are normally enriched in the silt–sand fraction (Wang et al., Reference Wang, Zheng, Xie, Fan, Yang, Zhao and Wang2011; Rothwell and Croudace, Reference Rothwell, Croudace, Croudace and Rothwell2015). However, the element intensity measured using an XRF core scanner can be affected by water-pore size, grain size, and surface roughness (Böning et al., Reference Böning, Bard and Rose2007; Tjallingii et al., Reference Tjallingii, Röhl, Kölling and Bickert2007; Löwemark et al., Reference Löwemark, Chen, Yang, Kylander, Yu, Hsu, Lee, Song and Jarvis2011); therefore, element ratios (Rothwell et al., Reference Rothwell, Hoogakker, Thomson, Croudace and Frenz2006) and log-element ratios (Rothwell et al., Reference Rothwell, Hoogakker, Thomson, Croudace and Frenz2006; Weltje and Tjallingii, Reference Weltje and Tjallingii2008) have been widely used. Generally, Ti/Al and Zr/Al are used as indicators of terrestrial origin flooding (Lückge et al., Reference Lückge, Doose-Rolinski, Khan, Schulz and Von Rad2001; Aniceto et al., Reference Aniceto, Moreira-Turcq, Cordeiro, Quintana, Fraizy and Turcq2014; Croudace et al., Reference Croudace and Rothwell2015 and references therein; Lim et al., Reference Lim, Lee, Hong, Park, Lee and Yi2019; Park et al., Reference Park, Lim, Kim, Shin and Lim2023).
To identify the environmental implication of high-resolution elemental ratios (e.g., Ti/Al and Zr/Al; Fig. 8C and D) in core JG-04, we compared the elemental ratios with the S intensity (Fig. 8B) in addition to low-resolution TS% (Fig. 8E) and C/S (Fig. 8F) ratios measured at 10 cm intervals. As shown in Figure 8, the time series of Ti/Al and Zr/Al ratios are negatively correlated with S intensity, which correlates with TS%. As discussed earlier, the comparison between C/S ratios in the two cores (JG-04 and STP17-14) indicates long-term coastal environmental changes, suggesting a stable bay environment (paleo-Hadong Bay) during the period of 8600–7800 cal yr BP (Unit 3b) under an ∼10 m water depth. Especially, during this period, the inverse relationship between S intensity and Ti/Al ratios (Figs. 7A and 8) could be interpreted in terms of dilution effect by terrestrial material deposition in the bay. And this may have been caused by increased terrestrial input. Considering the high sedimentation rate during this period, the responsible transporting media should have been increased rainfall; this increased freshwater input may have caused significant changes in salinity and/or anoxic conditions in the bay, as shown in S intensity and TS%. Thus, it is likely that the time series of the Ti/Al and Zr/Al ratios in the paleo-Hadong Bay indicate past freshwater (rainfall) input variability along the Seomjin River.
A similar interpretation has been presented in previous studies. For example, Wang et al. (Reference Wang, Zheng, Xie, Fan, Yang, Zhao and Wang2011) reconstructed past flood events in the Yangtze River basin using Zr/Rb ratios from sedimentary cores sampled in the subaqueous Yangtze River delta. They concluded that Yangtze River flooding events caused increased discharge of coarse-grained minerals into the East China Sea, leading to higher Zr/Rb ratios. Lim et al. (Reference Lim, Lee, Hong, Park, Lee and Yi2019) traced past hydrological variability on the southern coast of Korea using a time series of Ti/Al ratios recovered from coastal sediments. To demonstrate that this elemental ratio could be used as a hydrological indicator, they compared Ti/Al ratios with δ13C values of TOC in the coastal sediments, which are sensitive to past freshwater input. Furthermore, Zr in the coastal sediments has recently been used to trace hydroclimate changes in coastal areas. According to Park et al. (Reference Park, Lim, Kim, Shin and Lim2023), the Zr/Ti ratio can be used as a high-resolution terrestrial origin grain-size proxy and paleo-flooding indicator for the west coast of Korea.
If the terrestrial input change in paleo-Hadong Bay can be described using the high-resolution (5 mm interval) Ti/Al and Zr/Al records of core JG-04, these elemental ratio changes should be related to past regional hydroclimate changes. To test this assumption, we compared the Ti/Al ratio records of core JG-04 with the hydrological variation reconstructed in the west coast (Park et al., Reference Park, Lim, Kim, Shin and Lim2023), particularly for the period 8.6–7.8 ka BP (Fig. 9) in terms of simultaneous responses to past summer monsoonal precipitation changes over South Korea as shown by present meteorological comparisons (Fig. 3).
Figure 9. Comparison of past freshwater input variability in the past coastal regions in South Korea representing East Asian summer monsoon (EASM) with other climatic indices. (A) Atmospheric temperature index based on δ1⁸O values from Greenland ice cores (NGRIP members, 2004). Numbers (1∼17) represent lower-temperature (cooling) events, including the 8.2 ka cooling event. (B) Freshwater input variability represented by the Ti/Al ratio time series from cores JG-04 (southern coast in Korea; this study). Each square with an error range indicates dating points. (C) Freshwater input variability in core QJS60 (western coast in Korea; Park et al., Reference Park, Lim, Kim, Shin and Lim2023). Each square with an error range indicates dating points. (D) Indian summer monsoon (ISM) index from detrended δ1⁸O values of stalagmites in Tianmen Cave (Cai et al., Reference Cai, Zhang, Cheng, An, Edwards, Wang, Tan, Liang, Wang and Kelly2012). (E) ISM variability from detrended δ1⁸O values of stalagmites in Qunf Cave, Oman (Fleitmann et al., Reference Fleitmann, Burns, Mangini, Mudelsee, Kramers, Villa and Neff2007). (F) South America summer monsoon (SASM) variability coupled with Intertropical Convergence Zone (ITCZ) shift from detrended δ1⁸O values of stalagmites in Padre Cave, Brazil (Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009). “f” and “d” indicate possible flooding and drought periods, respectively.
As shown in Figure 9, the variability in the Zr/Ti ratios of cores QJS60 from the western coastal area (Fig. 9C) and the Ti/Al ratios of cores JG-04 of the southern coastal area (Fig. 9B) can be tested based on the values of mean and control lines (or standard deviation). Park et al. (Reference Park, Lim, Kim, Shin and Lim2023) suggested that abnormalities in the Zr/Ti ratios of core QJS60 that were higher or lower than these control lines indicate significant increases or decreases in freshwater input, namely, respective periods of flooding or drought in the Buan area. Similar hydroclimate fluctuations are found in the Ti/Al ratios in core JG-04. For example, common remarkable peaks are found at f1, f3, f4, f5, f7, f10, f11, and f14, suggesting simultaneously intensified summer rainfall at those times in the Hadong and Buan areas. Similarly, periods of decreased Ti/Al ratios centered at ca. 8400 (d5), 8200 (d7, d8), 8020 (d9), 7890 (d16), and 7820 (d17) cal yr BP suggest that drought-dominant or weakened flooding periods occurred in both areas. Despite differences between the age controls, the observed similar variability of the elemental ratios in the two areas suggests that simultaneous multi-decadal changes in the hydrological climate occurred in the western and southern parts of Korean Peninsula during the Early Holocene as in the present (Fig. 3).
As mentioned earlier, heavy rainfall that causes flooding on the Korean Peninsula mainly occurs in summer and is mainly affected by the intensified EASM (Qian et al., Reference Qian, Kang and Lee2002; Ha et al., Reference Ha, Heo, Lee, Yun and Jhun2012; Fig. 2C). The decadal to centennial hydroclimate variation shared by Buan and Hadong may represent variability in the EASM over South Korea during the Early Holocene.
Centennial-scale linkage between global summer monsoon intensity and ITCZ movement during the Early Holocene
The change in orbital scale in the EASM during the Holocene has been attributed to the northward/southward shift of the ITCZ through the rainfall front position, coupled with insolation changes (Dykoski et al., Reference Dykoski, Edwords, Cheng, Yuan, Cai, Zhang, Lin, Qing, An and Revenaugh2005; Yancheva et al., Reference Yancheva, Nowaczyk, Mingram, Dulski, Schettler, Negendank, Liu, Sigman, Peterson and Haug2007; Wang et al., Reference Wang, Wang, Cheng, Fasullo, Guo, Kiefer and Liu2017). Han et al. (Reference Han, Yu, Yan, Yan, Tao, Zhang, Wang and Chen2019) found that decadal-scale hydrological variability in the South China Sea from 1851 to 2010 CE was modulated by meridional ITCZ migration. Because of the meridional ITCZ shift (Fig. 1), the regional monsoon regions are strongly tied to an interhemispheric antiphase relationship called the global monsoon system (Wang and Ding, Reference Wang and Ding2008; Cheng et al., Reference Cheng, Sinha, Wang, Cruz and Edwards2012; Jo et al., Reference Jo, Woo, Yi, Yang, Lim, Wang, Cheng and Edwards2014).
It is noteworthy that there have been a number of studies on the EASM changes using various climatic proxies over China, covering low-latitude monsoonal to high-latitude arid areas. A previous study (Park et al., Reference Park, Lim, Kim, Shin and Lim2023) tested possible linkage between freshwater input variability in Buan area and the oxygen isotope (δ18O) values from stalagmites in various caves in China, which have been used to trace summer monsoon changes (e.g., Wang et al., Reference Wang, Cheng, Edwards, He, Kong, An, Wu, Kelly, Dykoski and Li2005; Dykoski et al., Reference Dykoski, Edwords, Cheng, Yuan, Cai, Zhang, Lin, Qing, An and Revenaugh2005; Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009; Dong et al., Reference Dong, Wang, Cheng, Hardt, Edwards, Kong and Wu2010, Reference Dong, Shen, Kong, Wang and Jiang2015). But there was no clear relationship. Park et al. (Reference Park, Lim, Kim, Shin and Lim2023) suggested that the difference may reflect different responses to regional and global changes in the atmospheric–oceanic circulation, in addition to distinct age control intervals. Interestingly, they found strong correlation between precipitation change over South Korea and the pollen-based precipitation change from Gonghai Lake in north China (Chen et al., Reference Chen, Xu, Chen, Birks, Liu and Zhang2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018). This suggests the reconstructed precipitation signals based on the Changma rainband over South Korea (Fig. 2C) can be used to trace past EASM variability. Thus, this study can provide a chance to elaborate the climatic meaning of decadal to centennial freshwater input variability found in the Hadong area (this study) and Buan area (Park et al., Reference Park, Lim, Kim, Shin and Lim2023) in the South Korea.
By comparing a high-resolution EASM index from the South Korea with an ISM index from caves in Tianmen (Cai et al., Reference Cai, Zhang, Cheng, An, Edwards, Wang, Tan, Liang, Wang and Kelly2012) and Qunf (Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003), this study examined the correlation between decadal-scale EASM and ITCZ shifts during the Early Holocene from a global monsoon perspective (Fig. 9). Despite differences due to different age controls, the numbered peaks and troughs in the EASM index seem to correspond to those of the δ18O values of Tianmen and Qunf Caves. The prominent flooding (f3∼f14) events recorded in the EASM index are well correlated with the intensified ISM δ18O values of Tianmen Cave. We found common decreases in summer rainfall (e.g., d6, d7, d8, d9, and d12) in the Hadong area and Qunf Cave (Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003), suggesting an in-phase relationship. These synchronous reductions in rainfall in East Asia, Oman (Qunf Cave), and the southern part of the Tibetan Plateau (Tianmen Cave) clearly demonstrate strong teleconnection between the regions through systematic, coeval ITCZ shifts in the same direction, whereas South American summer monsoon rainfalls intensified (Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009; Fig. 9F), showing an antiphase relationship with the EASM during the same periods. This antiphase relationship between the data presented here and Southern Hemisphere monsoon changes supports spatial global monsoon changes driven by ITCZ shifts (Haug et al., Reference Haug, Hughen, Sigman, Peterson and Röhl2001; Fleitmann et al., Reference Fleitmann, Burns, Mudelsee, Neff, Kramers, Mangini and Matter2003; Cheng et al., Reference Cheng, Fleitmann, Edwards, Wang, Cruz and Auler2009).
Interestingly, the significant decrease in both EASM and ISM corresponds well with the decreasing δ18O in the North Greenland Ice Core Project (NGRIP) ice core (NGRIP members, 2004; Fig. 9A), suggesting strong teleconnection. According to several previous studies, Northern Hemisphere atmospheric temperature and sea ice coverage are strongly linked to ITCZ movement via the atmospheric circulation system (Haug et al., Reference Haug, Hughen, Sigman, Peterson and Röhl2001; Chiang and Bitz, Reference Chiang and Bitz2005; Jo et al., Reference Jo, Woo, Yi, Yang, Lim, Wang, Cheng and Edwards2014; Liu et al., Reference Liu, Liao, Qiu, Yang, Feng, Allan, Can, Long and Xu2020). For example, Chiang, and Bitz (Reference Chiang and Bitz2005) suggested that the extent of northern high-latitude land ice and sea ice recorded in the Greenland ice core influenced the marine tropical climate. Liu et al. (Reference Liu, Liao, Qiu, Yang, Feng, Allan, Can, Long and Xu2020) also inferred a strong relationship between Northern Hemisphere sea ice and the ITCZ during 1998–2012. The significant decrease in δ18O in the NGRIP ice core to the lower control line corresponded to weakening of the EASM and ISM around 8200 cal yr BP (NGRIP members, 2004; Fig. 9A). Moreover, the multi-decadal low-temperature periods, indicated as points 2 to 11 (Fig. 9A), suggest that high-latitude climate changes were strongly connected to the global monsoon region through the meridional ITCZ movements during the Early Holocene.
Previous studies have suggested that high-latitude climate change can influence the climate in downwind areas, including East Asia, via the winter monsoon and westerly jet (Xiao et al., Reference Xiao, Xu, Nakamura, Yang, Liang and Inouchi2004; Lim and Matsumoto, Reference Lim and Matsumoto2006, Reference Lim and Matsumoto2008; Nagashima et al., Reference Nagashima, Tada, Matsui, Irino, Tani and Toyoda2007, Reference Nagashima, Tada and Toyoda2013, Reference Nagashima, Addison, Irino, Omori, Yoshimura and Harada2022; Wen et al., Reference Wen, Xiao, Fan, Zhang and Yamagata2017; Cho et al., Reference Cho, Lim, Kim and San Ahn2022). The position of the westerly jet is a critical determinant of the onset of the summer monsoon, particularly over East Asia, due to its interaction with the Asian monsoon system and the westerly jet. At present, a change in the westerly jet position can precede the onset of the summer monsoon, which is linked to the activities of warm-moist and cold air masses (Nagashima et al., Reference Nagashima, Tada and Toyoda2013; Chiang et al., Reference Chiang, Fung, Wu, Cai, Edman, Liu, Day, Bhattacharya, Mondal and Labrousse2015; Day et al., Reference Day, Fung and Risi2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018). During the spring, the westerly jet, which is located south of the Tibetan Plateau in winter, begins its northward migration. This shift is closely tied to changes in the intensity and position of the EASM rainbands (also called the Changma). The timing of this northward jump of the westerly jet is crucial; when it occurs in early May, it typically precedes the onset of the summer monsoon over regions such as the South China Sea (Li and Pan, Reference Li and Pan2006; Chiang et al., Reference Chiang, Fung, Wu, Cai, Edman, Liu, Day, Bhattacharya, Mondal and Labrousse2015, Reference Chiang, Swenson and Kong2017; Day et al., Reference Day, Fung and Risi2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018). For example, when the westerlies are situated over the southern part of Tibet during early summer (May to early June), heavy rainfall tends to be concentrated in southern China (Chiang et al., Reference Chiang, Fung, Wu, Cai, Edman, Liu, Day, Bhattacharya, Mondal and Labrousse2015; Zhang et al., Reference Zhang, Griffiths, Chiang, Kong, Wu, Atwood, Huang, Cheng, Ning and Xie2018). However, due to the latitudinal difference, the Changma front does not extend to southern Korea during this period. Instead, from late June to July, the westerlies shift northward, allowing the Changma front to develop over southern-central Korea. These characteristics suggest that the position of the westerlies can influence both the onset and duration of the Changma period, ultimately affecting the intensity of the EASM over Korea.
As shown in Figure 9, the 8.2 ka cooling event can provide a rare chance to examine regional responses to global climatic event and their possible mechanisms. The 8.2 ka cooling event occurred because of the drainage of a final outburst from glacial Lakes Agassiz and Ojibway during the terminal demise of the Laurentide Ice Sheet. This event was characterized by a considerable decrease in temperature (1–2°) over Greenland (NGRIP members, 2004). In a previous study on the western coast of Korea (Park et al., Reference Park, Lim, Kim, Shin and Lim2023), we tested the possible influence of the 8.2 ka cooling event on hydroclimate over the western part of Korea, suggesting that the change in the sedimentary elemental ratios from the coastal sediments can be used as a freshwater input change. As shown in Figure 9, the global monsoon system responded to this cooling event in different ways. EASM and ISM significantly weakened, while the South American summer monsoon (SASM) intensified, showing climatic synchroneity in the global atmospheric–oceanic circulation changes. Including the 8.2 ka cooling event, several natural cooling events (e.g., numbers 2–5 and 8–11) correlated well with weakening of the EASM and ISM (Fig. 9).
It has been suggested that during the cold period, with increased ice volume in the high-latitude areas, the Siberian High (or winter monsoon) intensified and was displaced southward via downstream atmospheric cooling, showing antiphasing with EASM strength (Kang et al., Reference Kang, Du, Wang, Dong, Wang, Wang, Qiang and Song2020). Especially, the EAWM variation during the Holocene has been influenced by meridional temperature gradients between high and low latitudes, feedbacks from the Arctic sea ice and Eurasian snow cover, anomalous freshening of the North Atlantic, and solar activity (Lim et al., 2005; Chen et al., Reference Chen, Xu, Chen, Birks, Liu and Zhang2015; Wen et al., Reference Wen, Xiao, Fan, Zhang and Yamagata2017; Cho et al., Reference Cho, Lim, Kim and San Ahn2022; Zhou et al., Reference Zhou, Li, Shi, Sha, Lei and An2023). During the intensified EAWM, the westerly jet may have remained south of the Tibetan Plateau for prolonged periods of prevailing cooler climate (Xiao et al., Reference Xiao, Nakamura, Lu and Zhang2002, Reference Xiao, Xu, Nakamura, Yang, Liang and Inouchi2004; Lim and Matsumoto, Reference Lim and Matsumoto2006, Reference Lim and Matsumoto2008). Consequently, the relatively prolonged southward positioning of the westerly jet could have delayed the onset of the EASM, resulting in relatively weakened summer precipitation in the western and southern parts of Korea. Thus, it is likely that the warm or cool climatic conditions in high latitudes have primarily influenced the position of the Siberian High and the westerly jet, influencing the extent of the strong or weak EASM and ISM.
Our results provide clear evidence that the position of the ITCZ strongly affects both the EASM and ISM and shifts systematically over decadal to centennial timescales. Moreover, the high-latitude cooling climate is also directly connected to the EASM and ISM variability via the southward migration or duration of the westerly jet position over the Tibetan Plateau. These results emphasize the urgent need for further study of the factors controlling the interaction between the westerly jet and ITCZ and the physical mechanisms linking them in terms of the global monsoon system over various timescales to prepare for anticipated global warming.
Conclusions
To examine summer hydroclimate variability and its behavior along the western and southern coasts of Korea during the Early Holocene, we reconstructed multi-decadal- to centennial-scale variations in freshwater input using sedimentary cores from southern Korea and compared these results with records from western Korea. To reconstruct summer rainfall intensity, we analyzed grain size, TS, C/S ratios, terrestrial elemental ratios (e.g., Ti/Al, Zr/Ti), and sulfur intensity, which reflect changes in terrestrial element input and associated EASM variations.
Our results reveal synchronous summer rainfall variability between the southern and western parts of Korea, strongly suggesting coordinated EASM variability over these regions during the Early Holocene. This study highlights that weakened EASM and ISM intensities, along with a strong SASM at approximately 8400, 8200, and 7800 cal yr BP, can be attributed to global monsoon behavior influenced by ITCZ shifts and high-latitude climate changes. Furthermore, significant weakening of the EASM and ISM could be influenced by the prolonged southward positioning of the westerly jet along the southern edge of the Tibetan Plateau during the cooling events (e.g., 8.2 ka event) in high-latitude regions. This prolonged westerly flow over the southern Tibetan Plateau hindered the development of the Changma front over Korea, similar to present-day early summer conditions, and consequently resulted in a weaker EASM. The multi-decadal- to centennial-scale change in EASM intensity described by the terrestrial element ratios may represent past coupled ITCZ–westerly jet influences over East Asia.
Acknowledgments
This research was supported by a Basic Research Project of the Korea Institute of Geoscience and Mineral Resources (GP2022–006 [24–3111–3], GP2022–005 [24–3807], GP2025–027 [25–3225]).
Open access
