INTRODUCTION
One of the most important discoveries in the study of paleoclimate changes was the recognition in the early 1990s that the climate during the last glacial period was not always cold and that a series of millennial-scale climate warming events occurred, which are called Dansgaard-Oeschger (D/O) events (Johnsen et al., Reference Johnsen, Clausen, Dansgaard, Fuhrer, Gundestrup, Hammer, Iversen, Jouzel, Stauffer and Steffensen1992; Bond et al., Reference Bond, Broecker, Johnsen, McManus, Labeyrie, Jouzel and Bonani1993; Dansgaard et al., Reference Dansgaard, Johnsen, Clausen, Dahl-Jensen, Gundestrup, Hammer and Hvidberg1993). The D/O events were abrupt changes from stadials to interstadials with durations of hundreds to thousands of years (Thomas et al., Reference Thomas, Wolff, Mulvaney, Johnsen, Steffensen and Arrowsmith2009). Leuschner and Sirocko (Reference Leuschner and Sirocko2000) and Voelker et al. (Reference Voelker2002) summarized the geographic distribution of the D/O events on a global scale as recorded in different geologic records and indicated that the D/O events occurred worldwide.
The global footprint of the D/O events suggests that the earth system is integrated and that several interrelations and interactions occur between the ocean-atmosphere-continent systems, the Northern and Southern Hemispheres, and high and low latitudes (Bond et al., Reference Bond, Broecker, Johnsen, McManus, Labeyrie, Jouzel and Bonani1993; Blunier and Brook, Reference Blunier and Brook2001; Stocker and Johnsen, Reference Stocker and Johnsen2003; Cheng et al., Reference Cheng, Sinha, Cruz, Wang, Edwards, d’Horta, Ribas, Vuille, Stott and Auler2013b). However, there are differences between the D/O events recorded in geologic archives in different latitude regions. These differences might be attributed to regional characteristics of physical environment, uncertainties in the age models, and/or resolution of records (Zhao et al., Reference Zhao, Wang, Edwards, Cheng and Liu2010). D/O events have been used as important tie points in research on the linkages between climate changes in different regions.
In this study, we reconstructed the evolution of the Asian summer monsoon (ASM) between 49.1 and 59.5 ka using a stalagmite record (JYFK7) from Yangkou Cave in Chongqing, southwest China. The improved resolution of the δ18O series and the high-precision chronology of this record provide a good opportunity to study in the detail variability of the ASM during this time period. By comparing the JFYK7 record with other records from northern China (Duan et al., Reference Duan, Cheng, Tan and Edwards2016) and polar regions in the Northern Hemisphere, we investigated the response of the ASM to D/O events 13–17 during this period and assessed the different response of the ASM in northern and southern China.
GEOGRAPHIC SETTING
Stalagmite JFYK7 was collected in Yangkou Cave (29°1.68′N, 107°11.25′E; elevation: 2140 m; length: 2245 m), which is located on Jinfo Mountain (Fig. 1), Chongqing City, southwest China (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014). Yangkou Cave is located near the top of Jinfo Mountain, developed in Permian limestone stratum with overlying bedrock as thick as 50–100 m (Zhang et al., Reference Zhang, Zhu, Han, Zhang and Fang1998). It is a gallery-type cave, and its width and height are generally 15–20 m (with the maximum being 64 m) and 8–12 m (with the maximum being 27 m), respectively (Zhang et al., Reference Zhang, Zhu, Han, Zhang and Fang1998). Stalagmite JFYK7 was sampled in the south branch gallery of the cave, about 800 m and 400 m away from the entrance and the south exit, respectively (Fig. 1c).
Figure 1 (a) Locations of Yangkou Cave and other sites mentioned in this study (circles) (Modified from Cheng et al., 2012). Stalagmite JFYK7 (Yangkou Cave: 29°1.68′N, 107°11.25′E) (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014), stalagmite XL-1 (Xinglong Cave: 40°29.21′N, 117°29.27′E) (Duan et al., Reference Duan, Cheng, Tan and Edwards2016), stalagmite MSL (Hulu Cave: 32°3.43′N, 119°2.44′E) (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001), and North Greenland Ice Core Project (NGRIP) ice core (Greenland: 75°1.00′N, 42°32.00′W) (North Greenland Ice Core project members, 2004). The current seasonal migrations of the Intertropical Convergence Zone (ITCZ; dashed lines) are denoted in red (July) and blue (January) (Lutgens and Tarbuck, Reference Lutgens and Tarbuck2001). The prevailing surface winds are shown by green arrows, which indicate the directions of the Indian summer monsoon (ISM) and East Asian summer monsoon (EASM) (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014). (b) Detail of the location of Yangkou Cave in Chongqing, China. (c) Sketch map of Yangkou Cave (after Zhang et al., Reference Zhang, Zhu, Han, Zhang and Fang1998). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
The region has a subtropical and humid monsoon climate, which is mild and rainy. The average annual temperature is 14.5°C, and the average annual precipitation is 1435 mm. From April to October, the monthly rainfall is greater than 100 mm, and the total rainfall in this period accounts for more than 80% of the annual rainfall (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014; Han et al., Reference Han, Li, Cheng, Edwards, Shen, Li and Huang2016).
The top of Jinfo Mountain is surrounded by cliffs, without any surface river. All the rain and snowmelt is discharged through the karst fissures and sinkholes. Accordingly, the drip waters in Yangkou Cave are fed by local precipitation. A multiyear monitoring project was carried out in Yangkou Cave. Compared with the δ18O of local precipitation ranging from −1.44‰ to −16.10‰, the δ18O of drip waters (sampled from six sites) in Yangkou Cave falls in the range from −7.05‰ to −9.96‰, obviously compressed by the mixing effect of overlying host rocks. The monitoring results show that no obvious evaporation occurred when precipitation infiltrated into the cave to form drips, even though drip waters in different monitoring sites lag behind local precipitation to different extents considering the difference in geologic characteristics, such as the thickness of overlying stratum and the development and connection of fissures. The δ18O values of the drip water reflected the isotopic composition of the local precipitation (Wang et al., Reference Wang, Li, Yuan and Li2014). Unfortunately, no active stalagmite was collected in Yangkou Cave, nor any carbonate deposited on the glass plates placed under the drip waters in past monitoring years. Instead, some scraped powders from the surface of rocks under the drip waters can be taken as the active deposits, and the average δ18O value of these deposits is −6.98‰ (Vienna Pee Dee belemnite [V-PDB]), close to the calculated value (−6.70‰) of deposits deposited at isotopic equilibrium fractionation, with the average δ18O of drip waters being −8.80‰ (Vienna standard mean ocean water) and average cave temperature being 7.9°C (Li, T.-Y., unpublished data). In spite of this, the δ18O values of the local precipitation showed clear seasonal variations that were strongly influenced by the summer monsoon, with light values in summer and heavier values in winter (Wang et al., Reference Wang, Li, Yuan and Li2014). Consequently, the δ18O values of the stalagmite can be used to assess the variability of the summer monsoon; low stalagmite δ18O values denote a stronger summer monsoon and vice versa (Cheng et al., Reference Cheng, Edwards, Broecker, Denton, Kong, Wang, Zhang and Wang2009; Liu et al., Reference Liu, Wen, Brady, Otto-Bliesner, Yu, Lu and Cheng2013; Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014; Zhou et al., Reference Zhou, Zhao, Feng, Chen, Mi, Shen and He2014; Han et al., Reference Han, Li, Cheng, Edwards, Shen, Li and Huang2016).
ANALYTIC METHODS
Stalagmite JFYK7 has a cylindrical shape and is approximately 125 mm and 40 mm in diameter at the bottom and top, respectively. The stalagmite was cut in half along the growth axis. The vertical section is dark brown, and growth banding is clearly visible. Measured along the growth axis, stalagmite JFYK7 is 555 mm long, and the section between 193 mm and 399 mm is discussed in this study. Fifteen subsamples were drilled for 230Th dating from the polished surface of JFYK7 using carbide dental burrs with a diameter of 1 mm. The subsamples were 230Th dated by multicollector inductively coupled plasma mass spectrometry in the Isotope Laboratories of the University of Minnesota and Xi’an Jiaotong University. The age errors are given in two standard deviations (2σ) (Edwards, Reference Edwards1988; Shen et al., Reference Shen, Wu, Cheng, Edwards, Hsieh, Gallet and Chang2012; Cheng et al., Reference Cheng, Edwards, Shen, Polyak, Asmerom, Woodhead, Hellstrom, Wang, Kong and Spötl2013a). The decay constants of 230Th, 234U, and 238U are 9.1705×10−6/yr (Cheng et al., Reference Cheng, Edwards, Shen, Polyak, Asmerom, Woodhead, Hellstrom, Wang, Kong and Spötl2013a), 2.82206×10−6/yr (Cheng et al., Reference Cheng, Edwards, Shen, Polyak, Asmerom, Woodhead, Hellstrom, Wang, Kong and Spötl2013a), and 1.55125×10−6/yr (Jaffey et al., Reference Jaffey, Flynn, Glendenin, Bentley and Essling1971), respectively. The age correction for the initial 230Th was performed using the average crustal 230Th/232Th ratio of 4.4±2.2×10−6 (Taylor and McLennan, Reference Taylor and McLennan1995).
The δ18O and δ13C analyses of JFYK7 were performed in the Geochemistry and Isotope Laboratory of Southwest University, China. A total of 413 subsamples were drilled for isotope analysis along the growth axis using a 0.5-mm-diameter drill bit and a sampling interval of 0.5 mm. The analyses were performed using a Delta V Plus mass spectrometer, which was combined with a Kiel IV Carbonate Device. Every seven samples were bracketed with one work standard (SWU1). The isotopic results are given with respect to the V-PDB standard with a 1-sigma external error <±0.1‰ for δ18O and <±0.06‰ for δ13C (Li et al., Reference Li, Shen, Li, Li, Chang, Song and Yuan2011).
The spectral analysis was performed on the δ18O records of JFYK7 using the program REDFIT 37. This software allows the variance of a time series to be separated into contributions associated with different time scales, which is useful for better understanding the origins of the variability recorded in a time series (Schulz and Mudelsee, Reference Schulz and Mudelsee2002).
RESULTS
230Th dating
The dating results are shown in Table 1. Because the concentration of 238U in the stalagmites from Yangkou Cave is approximately 10 ppm (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014; Han et al., Reference Han, Li, Cheng, Edwards, Shen, Li and Huang2016), the dating precision for JFYK7 is high, and all of the age errors are less than ±200 yr. The 230Th/232Th ratio is another important factor that affects the dating precision (Shen et al., Reference Shen, Wu, Cheng, Edwards, Hsieh, Gallet and Chang2012). The 230Th /232Th ratio is as high as 27,000 to 970,000x10−6 (Table 1). Thus, it is not necessary to correct the results using the initial 230Th, and the age errors are less than 0.4%. In this study, all ages are in the stratigraphic order, and a linear interpolation method was used to establish an age-depth model (Fig. 2). The timing for the section of JFYK7 discussed in this study (193 to 399 mm in depth) is in the period of 49–59 ka, and the average resolution for the δ18O record is 24 yr.
Figure 2 Age-depth model of stalagmite JFYK7 based on linear interpolation.
Table 1 Inductively coupled plasma mass spectrometry dating results of stalagmite JFYK7.
Notes: 230=9.1705×10−6/yr; λ234=2.82206×10−6/yr; λ238=1.55125×10−10/yr.
δ234U=([234U/238U]activity −1)×1000.
δ234Uinitial was calculated based on 230Th age (T) (i.e., δ234Uinitial=δ234Umeasured×eλ234 ×T).
Corrected 230Th ages assume the initial 230Th/232Th atomic ratio of 4.4±2.2×10−6. Those are the values for a material at secular equilibrium, with the bulk earth 232Th/238U value of 3.8. The errors are arbitrarily assumed to be 50%.
Carbon and oxygen isotopes
Previous work on stalagmite JFYK7 indicated that it was deposited under isotope equilibrium fractionation conditions and that the variation of δ18O was mainly controlled by climatic factors (Han et al., Reference Han, Li, Cheng, Edwards, Shen, Li and Huang2016). Although both the δ18O and δ13C values shows large changes at century to millennial time scales, the details show several significant differences (Fig. 3). During the period from 49.1 to 59.5 ka, the δ18O values of JFYK7 fluctuate between −9.8‰ and −7.0‰ with an average of −8.4‰. The maximum and minimum δ18O values occurred at approximately 55.0 and 52.3 ka, respectively (Fig. 3).
Figure 3 (color online) The δ18O and δ13C records of stalagmite JFYK7.
Spectral analysis
Spectra of paleoclimatic time series frequently show a continuous decrease of spectral amplitude with increasing frequency (“red noise”) (Schulz and Mudelsee, Reference Schulz and Mudelsee2002). The red noise test showed that periods of 144 yr and 125 yr are statistically significant periodicities in the δ18O data at the 95%–99% confidence level (Fig. 4). The 144 yr period reflects the century-scale period in the interstadials during D/O events 13–17, which is approximately consistent with the 148 yr cycle of 14C recorded by tree rings (Stuiver and Braziunas, Reference Stuiver and Braziunas1993). The 125 yr period reflects the solar cycle as has been observed in other paleoclimate records (Stuiver and Braziunas, Reference Stuiver and Braziunas1989; Castagnoli et al., Reference Castagnoli, Bonino, Provenzale and Serio1990; Agnihotri et al., Reference Agnihotri, Dutta, Bhushan and Somayajulu2002). In addition, there are statistically significant periodicities at the 99% confidence level of 133, 81, 73, 67, 62, 57, and 53 yr (Fig. 4). Based on the resolution of the δ18O record (~24 yr), most of the cycles, such as 73, 67, 62, 57, and 53 yr, may simply be because of random fluctuations. Only the 133 yr and 81 yr cycles can be trusted. The 133 yr period is known from 14C data of tree-ring records (Damon and Sonnett, Reference Damon and Sonnett1991). The 81 yr cycle is consistent with the solar centennial cycles (70–100 yr, the Gleissberg cycle) and was also identified in Chinese stalagmites (Dykoski et al., Reference Dykoski, Edwards, Cheng, Yuan, Cai, Zhang, Lin, Qing, An and Revenaugh2005).
Figure 4 Spectral analysis (Schulz and Mudelsee, 2002) (green solid line) of the δ18O records of stalagmite JFYK7 from Yangkou Cave. GXX is the spectrum of the data, and 95%-Chi (pink short dashed curve) and 99%-Chi (red long dashed curve) are the 95% and 99% confidence levels, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
DISCUSSION
Difference between northern and southern Chinese stalagmite records
A recent stalagmite record from northern China, which covers the change of the ASM during D/O 14 and 15, shows abrupt onsets of D/O 14 and 15.2 and strongly resembles those of the North Greenland Ice Core Project (NGRIP) record (Fig. 5b; Fig. 2 in Duan et al., Reference Duan, Cheng, Tan and Edwards2016). The abrupt transition into D/O 14 is significantly different from the relatively gradual change in our record and MSL, which is from Hulu Cave in eastern China (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001) (Fig. 5c and d, Fig. 6). In spite of the resolution difference in δ18O and chorology, the duration for the transition of D/O 14 in the records of NGRIP (Svensson et al., Reference Svensson, Andersen, Bigler, Clausen, Dahl-Jensen, Davies and Johnsen2008), XL-1 (Duan et al., Reference Duan, Cheng, Tan and Edwards2016), JFYK7 (this study), and MSL (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001) is 40, 110, 360, and 400 yr, respectively (Fig. 6). The transition of D/O 14 recorded in northern China is obviously faster than that in southern China.
Figure 5 Comparison between the stalagmite δ18O record from Yangkou Cave and other records. (a) δ18O of North Greenland Ice Core Project (NGRIP; Svensson et al., Reference Svensson, Andersen, Bigler, Clausen, Dahl-Jensen, Davies and Johnsen2008). (b) δ18O of stalagmite XL-1 from Xinglong Cave (Duan et al., Reference Duan, Cheng, Tan and Edwards2016). (c) δ18O of stalagmite JFYK7 from Yangkou Cave in southwestern China (this study). (d) δ18O of stalagmite MSL from Hulu Cave in Nanjing, China (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001). We suggest that the initial signature of Dansgaard/Oeschger 14 and 17 in the MSL record should be shifted to 54.3 and ~59 ka, which is presented by the arabic numerals 14 and 17 in red, respectively. The gray curves denote the variation of summer insolation (Berger, Reference Berger1978; Berger and Loutre, Reference Berger and Loutre1991; Paillard et al., Reference Paillard, Labeyrie and Yiou1996; Laskar et al., Reference Laskar, Robutel, Joutel, Gastineau, Correia and Levrard2004). 230Th dates and age uncertainties for the stalagmite records XL-1, JFYK7, and MSL are shown in the corresponding colors. The gray dashed lines with an arrow denote the trend of changes in different records. V-PDB, Vienna Pee Dee belemnite; V-SMOW, Vienna standard mean ocean water. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Figure 6 (color online) Expanded comparison for the transition of Dansgaard/Oeschger 14 between the stalagmite δ18O record from Yangkou Cave and other records. (a) δ18O of North Greenland Ice Core Project (NGRIP; Svensson et al., Reference Svensson, Andersen, Bigler, Clausen, Dahl-Jensen, Davies and Johnsen2008). (b) δ18O of stalagmite XL-1 from Xinglong Cave (Duan et al., Reference Duan, Cheng, Tan and Edwards2016). (c) δ18O of stalagmite JFYK7 from Yangkou Cave in southwestern China (this study). (d) δ18O of stalagmite MSL from Hulu Cave in Nanjing, China (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001).
We speculate that this discrepancy between northern and southern China may have been caused by the different responses to climate changes in the high northern latitudes. The westerlies have been proposed as the linkage for rapid atmospheric teleconnection between the North Atlantic and ASM regions (Porter and An, Reference Porter and An1995), which mainly influence the middle latitudes of the Northern Hemisphere (Railsback et al., Reference Railsback, Xiao, Liang, Akers, Brook, Dennis, Lanier, Tan, Cheng and Edwards2014). The Yangkou and Hulu Caves located in southern China and meteoric observations confirmed that during the summer months, the moisture from the Indian Ocean, the South China Sea, and the western Pacific affects southern China (Ding et al., Reference Ding, Li and Liu2004). Thus, climate change in southern China is not only influenced by the high northern latitudes, but also may be modulated by changes in the tropical oceans (Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014; Zhao et al., Reference Zhao, Wang, Edwards, Cheng, Liu, Kong and Ning2016). In other words, the differences in climate change patterns between northern and southern China presented in stalagmites may be attributed to the differences of the influence on these two regions from climate change in the northern high latitudes, the tropical Indian Ocean, and the Pacific Ocean.
Climate events D/O 13–17 recorded in JFYK7 are similar to those in the MSL stalagmite record, especially the gradual decreases of the ASM at the end of D/O 16 and at the onset of D/O 14 (Fig. 5c and d). This similarity suggests synchronous changes of the ASM in eastern and southwestern China from 49.1 to 59.5 ka.
Although the overall changes in the JFYK7 and MSL records are in broad agreement, there are several small-scale differences (Fig. 5c and d). First, in the JFYK7 record, the start and end of D/O 17 occurred at ~59.3±0.2 and ~58.4±0.2 ka BP, respectively, and the start and end of D/O 16 occurred at ~58.4±0.2 and ~56.0±0.2 ka, respectively; the duration of D/O 16, therefore, was approximately 2.4 ka. However, the corresponding dates in the MSL records are ~58.7±0.3 and ~57.4±0.2 ka, and ~57.4±0.2 and ~55.8±0.4 ka, respectively, indicating the duration of D/O 16 was 1.6 ka (Wang et al., Reference Wang, Cheng, Edwards, An, Wu, Shen and Dorale2001). Second, the δ18O values of JFYK7 and MSL gradually become more negative at the onset of D/O 16, and the amplitudes of the δ18O fluctuations were approximately 1.4‰ and 0.4‰, respectively. In contrast, after the peak of D/O 16, the δ18O values gradually become less negative and reach maximum values in JFYK7 and MSL within 1.5 ka and 0.6 ka, respectively. Third, during D/O 14, a notable bimodal pattern is recorded in JFYK7, but the MSL record shows a single symmetrical pattern (Fig. 5b and c). The strongest summer monsoon during the peak of D/O 14 recorded in JFYK7 occurred at ~52.3±0.1 ka compared with ~53.2±0.4 ka in the MSL record. The δ18O values of stalagmite MSL gradually become more negative at the start of D/O 14, reach the minimum value at ~53.2±0.4 ka, and then gradually become less negative (Fig. 5d). This symmetrical pattern during D/O 14 in the MSL δ18O records is not fully reflected in the JFYK7 record (Fig. 5c).
Compared with the MSL record, the JFYK7 record provides more details about the evolution of the ASM during this period because of the higher δ18O resolution (~24 yr) and the better chronology because of more 230Th dates and higher dating precision (Fig. 5c and d). The series of D/O events in JFYK7 is more similar to the NGRIP record (Fig. 5a, c, and d). We propose that the original identification of D/O 17 in MSL at ~58 ka should be shifted to ~59 ka and that the original D/O events 16 and 17 should belong to D/O 16 (Fig. 5d). In addition, the original identification of D/O 14 in MSL at ~53 ka should be shifted to ~54.3 ka (Fig. 5d). The MSL record will then be more consistent with the JFYK7 and NGRIP records.
The consistency between JFYK7 and MSL may partly be attributed to the change of solar insolation in the Northern Hemisphere, which influenced the evolution of the ASM. On the other hand, the discrepancies in the patterns and phases of δ18O records from JFYK7 and MSL should be attributed to the regional differences within the summer monsoon regimes, such as the individual characteristics of the caves, vegetation feedbacks, and others that could dampen the abrupt signals, and/or uncertainties in the age models and resolution of different records (Zhao et al., Reference Zhao, Wang, Edwards, Cheng and Liu2010).
Comparison with NGRIP
Five δ18O excursions in the JFYK7 records correspond to centennial- to millennial-scale climatic oscillations (D/O events 13–17) that have been documented in the Greenland ice cores (Fig. 5). The δ18O values of Greenland ice cores are mainly influenced by the local air temperature (Dansgaard, Reference Dansgaard1954; Dahl-Jensen and Johnsen, Reference Dahl-Jensen and Johnsen1986; Jouzel et al., Reference Jouzel, Alley, Cuffey, Dansgaard, Grootes, Hoffmann and Johnsen1997; Thomas et al., Reference Thomas, Wolff, Mulvaney, Johnsen, Steffensen and Arrowsmith2009). This suggests coupling between the ASM and the temperature changes over the North Atlantic. In general, the ASM recorded in JFYK7 was intensified/weakened when the temperatures recorded in the NGRIP cores increased/decreased (Fig. 5a and c). The westerlies have been proposed as the linkage between the ASM and climate changes in the high latitudes of the Northern Hemisphere (Porter and An, Reference Porter and An1995; Cosford et al., Reference Cosford, Qing, Yuan, Zhang, Holmden, Patterson and Cheng2008). This mechanism was confirmed by a simulation that coupled the North Atlantic meridional overturning circulation and the East Asian monsoon (Sun et al., Reference Sun, Clemens, Morrill, Lin, Wang and An2012).
Despite the overall similarity between JFYK7 and NGRIP, there are several differences. The temperature changes recorded in the NGRIP record show abrupt increases at the onsets of D/O 16 and 14 (Fig. 5a) (North Greenland Ice Core Project members, 2004), in contrast to the relatively gradual increase recorded by JFYK7 (Fig. 5c). In addition, the temperature changes indicate rapid cooling at the ends of D/O events 17 and 15 (Fig. 5a), but this pattern was not prominent in the variation of ASM recorded by JFYK7 (Fig. 5c). Other discrepancy between JFYK7 and NGRIP is the absence of D/O 15.1 in JFYK7 record, similar to the record from northern China (Fig. 5b) (Duan et al., Reference Duan, Cheng, Tan and Edwards2016).
The abrupt temperature changes recorded in the NGRIP record at the onsets of D/O 16 and 14 and the onsets and ends of D/O 17 and 15 are tightly related to the climate changes around the North Atlantic, whereas the Asian monsoon system is mainly driven by the thermal contrast between the Indian Ocean, the Pacific Ocean, and the Asian continent (Webster et al., Reference Webster, Magaña, Palmer, Shukla, Tomas, Yanai and Yasunari1998; Wang and Lin, Reference Wang and Lin2002). Yangkou Cave is located in the mid-low latitudes of the Asian monsoon region, which is influenced by the Indian summer monsoon and the East Asian summer monsoon (Fig. 1a) (Ding et al., Reference Ding, Li and Liu2004; Li et al., Reference Li, Yuan, Li, Yang, Wang, Wang, Li, Qin, Zhang and Lin2007). Previous studies indicate a rapid atmospheric teleconnection between the North Atlantic and the ASM (Porter and An, Reference Porter and An1995; Sun et al., Reference Sun, Clemens, Morrill, Lin, Wang and An2012; Duan et al., Reference Duan, Cheng, Tan and Edwards2016); other factors must have modulated the evolution of the ASM and resulted in the discrepancies between China and Greenland. The tropical Indian Ocean, the Pacific Ocean, even temperature changes in the Southern Hemisphere have been deemed to modulate the evolution of the ASM (Cai et al., Reference Cai, An, Cheng, Edwards, Kelly, Liu, Wang and Shen2006; Li et al., Reference Li, Shen, Huang, Jiang, Yang, Mii, Lee and Lo2014; Zhou et al., Reference Zhou, Zhao, Feng, Chen, Mi, Shen and He2014, 2016; Han et al., Reference Han, Li, Cheng, Edwards, Shen, Li and Huang2016), but the credible and convincing mechanism remains unclear.
CONCLUSIONS
A new, high-precision stalagmite δ18O record from southwest China, JFYK7, revealed the evolution of the ASM between 49.1 and 59.5 ka and reproduced the D/O 13–17 events that have been documented in the Greenland ice cores. The new record shows more details about the changes of the ASM and revises the previous identification of D/O 14 and 17 from 53 to 54.3 ka and 58 to 59 ka, respectively. The different pattern of the stalagmite records from northern China (Duan et al., Reference Duan, Cheng, Tan and Edwards2016) and southern China (JFYK7)—for example, with respect to the onset of D/O 14 and 16—suggests a different response of the ASM to climate change in the high northern latitudes. Further research is needed to clarify the influence on the ASM of the climate changes in tropical oceans and the Southern Hemisphere.
ACKNOWLEDGMENTS
Great thanks to the editors Matthew Lachniet and Derek Booth and the anonymous reviewers for their constructive suggestions and comments to improve the manuscript. A professional English translator, Dr. Xiang Xinyi, Southwest University, China, is also greatly appreciated for her effort in polishing the manuscript. This research was supported by National Natural Science Foundation of China (Nos. 41172165, 41302138, and 41440020), “the Fundamental Research Funds for the Central Universities” (Nos. XDJK2013A012, XDJK2014C010, and XDJK2017A010), “Karst Dynamics Laboratory, MLR and GZAR” (No. KDL201301), and the open project of the “State Key Laboratory of Loess and Quaternary Geology” (No. SKLLQG1310) to T.-Y. Li and J.-Y. Li. This study was also supported by a Taiwan Ministry Of Science and Technology (MOST) grant (103-2119-M-002-022) to C.-C. Shen.
