NATURE Vol 450/15 November 2007 BRIEF COMMUNICATIONS ARISING Anti-correlation of summer/winter monsoons? Arising from:G.Yancheva et al.Nature 445,74-77(2007) On the basis of the anti-correlation of their palaeoclimatic proxy for climate phases that bracketed a dry spell during AD 700-900,rather the strength of the East Asian winter monsoon from Lake Huguang than a "general shift towards drier climate";wet conditions are Maar,China,with stalagmite records of the strength of the summer recorded in AD 711-770 and 811-1050,with a dry phase in AD 771- monsoon,Yancheva et al.'claim that the strengths of the summer 810(ref.3;Fig.1),so the last thirty years of the Tang dynasty were and winter monsoons are anti-correlated on a decadal timescale. relatively wet,not dry. They argue that the summer rainfall deficit during AD 700-900 that The historical records show that the climate entered a cold phase they infer from their evidence of a stronger winter monsoon,in from the middle of the eighth century,when the number of cold conjunction with a Tanros battle,led to the collapse of the Tang events between AD 756 and 907 were more than double the number dynasty (AD 618-907).Using historical climate records,we show here in AD 618-755 (refs 4,5;Fig.1),implying that the winter monsoon that most cold winters during AD 700-900 were associated with rela- was strong during the late Tang dynasty.Furthermore,out of 22 cases tively wet summers,indicating that the strengths of the winter and of severe cold(with anomalously frozen and heavy snow)during AD summer monsoons were not negatively correlated during this period. 700-900,only two occurred in years with a summer rainfall deficit?, Yancheva et al.'deduce an inverse correlation of winter and sum- suggesting that 90%of the harsh winter events were associated with mer monsoons by comparing centennial-to millennial-scale proxy plentiful summer rainfall.This implies that cold winters and wet records of winter monsoon strength from Lake Huguang Maar with summers were characteristic of the climatic conditions of the late records of summer monsoon strength from Chinese caves.Their Tang period,suggesting that the strengths of the winter and summer inference that summer rainfall was low during AD 700-900 contra monsoons were not anti-correlated during this interval. dicts historical climate records for the same period.Chinese histor- Yancheva et al'interpret their record of titanium (Ti)counts, ical records can provide records of climatic events on annual to daily which has near-annual resolution,as a measure of winter monsoon timescales2.A wetness index for the past 2,000yr based on 36,750 strength'.However,the intervals of increased winter monsoon historical climate records shows that China experienced two wet strength indicated by this record are not always coincident with the intervals indicated by the historical records of severe winter fre- quency (Fig.2).During AD 800-900,the highest Ti counts occur 900 850800 .750 700 650 600 around the AD 860s,which is within a period when severe winters were infrequent;according to the historical records,only one 2 Year AD 900890.880870860850840830820810800 M 8 140 10 0.8 900 850 800 750 700 650 600 Year AD 200018001600140012001000800600400200 1.8 兽1.6 170 14 1.2 1.0 4L 10501060107010801090110011101120113011401150 Age (cal.yr BP) 0.6 200018001600140012001000800600400200 Figure 2Comparison between severe cold winters recorded in Chinese Year AD historical documents2 and the Ti counts from Lake Huguang Maar.Severe cold winters recorded are shown by blue arrows.The Ti counts from Lake Figure 1|Climate fluctuations during the past 2,000 yr based on Chinese Huguang Maar are shown by the thin red curve(the thick red curve has been historical documentary records.a,The moisture index series at a 5-yr smoothed with a 9-point running average,after Yancheva et al.from data resolution for East China during AD 1-1900;wet and dry intervals are archived at the World Data Center for Paleoclimatology,Boulder,Colorado, indicated by the blue and yellow shading,respectively(redrawn from ref.3) USA)during AD 800-900.The high Ti counts correspond to the lowest b,Comparison of the number of severe cold winters per decade(dark blue frequency of severe cold winters(AD 850-870;yellow shading),while the low bars)and the moisture index series for the Tang dynasty (AD 618-907). Ti counts correspond to increased frequency of cold winters(AD 810-840, During the late Tang dynasty(AD 810-907),the increase in the frequency of 880-900;blue shading).The winter monsoon strength inferred from Ti severe cold winters is associated with plentiful summer rainfall (pale blue counts contradicts that based on the occurrence of cold winters in ancient shading). Chinese records. E7 2007 Nature Publishing Group
Anti-correlation of summer/winter monsoons? Arising from: G. Yancheva et al. Nature 445, 74–77 (2007) On the basis of the anti-correlation of their palaeoclimatic proxy for the strength of the East Asian winter monsoon from Lake Huguang Maar, China, with stalagmite records of the strength of the summer monsoon, Yancheva et al.1 claim that the strengths of the summer and winter monsoons are anti-correlated on a decadal timescale. They argue that the summer rainfall deficit during AD 700–900 that they infer from their evidence of a stronger winter monsoon, in conjunction with a Tanros battle, led to the collapse of the Tang dynasty (AD 618–907). Using historical climate records, we show here that most cold winters during AD 700–900 were associated with relatively wet summers, indicating that the strengths of the winter and summer monsoons were not negatively correlated during this period. Yancheva et al.1 deduce an inverse correlation of winter and summer monsoons by comparing centennial- to millennial-scale proxy records of winter monsoon strength from Lake Huguang Maar with records of summer monsoon strength from Chinese caves. Their inference that summer rainfall was low during AD 700–900 contradicts historical climate records for the same period. Chinese historical records can provide records of climatic events on annual to daily timescales2 . A wetness index for the past 2,000 yr based on 36,750 historical climate records shows that China experienced two wet climate phases that bracketed a dry spell during AD 700–900, rather than a ‘‘general shift towards drier climate’’1 ; wet conditions are recorded in AD 711–770 and 811–1050, with a dry phase in AD 771– 810 (ref. 3; Fig. 1), so the last thirty years of the Tang dynasty were relatively wet, not dry. The historical records show that the climate entered a cold phase from the middle of the eighth century, when the number of cold events between AD 756 and 907 were more than double the number in AD 618–755 (refs 4, 5; Fig. 1), implying that the winter monsoon was strong during the late Tang dynasty. Furthermore, out of 22 cases of severe cold (with anomalously frozen and heavy snow) during AD 700–900, only two occurred in years with a summer rainfall deficit2 , suggesting that 90% of the harsh winter events were associated with plentiful summer rainfall. This implies that cold winters and wet summers were characteristic of the climatic conditions of the late Tang period, suggesting that the strengths of the winter and summer monsoons were not anti-correlated during this interval. Yancheva et al.1 interpret their record of titanium (Ti) counts, which has near-annual resolution, as a measure of winter monsoon strength1 . However, the intervals of increased winter monsoon strength indicated by this record are not always coincident with the intervals indicated by the historical records of severe winter frequency (Fig. 2). During AD 800–900, the highest Ti counts occur around the AD 860s, which is within a period when severe winters were infrequent; according to the historical records, only one Number of cold winters per 10 yr 850 800 750 700 650 600 900 850 800 750 700 650 a 0 1 2 3 Year AD Year AD 600 1.0 1.2 1.4 Moisture index Moisture index 0.8 1.6 1.0 1.2 1.4 0.6 0.8 1.6 1.8 b 900 2000 1800 1600 1400 1200 1000 800 600 400 200 2000 1800 1600 1400 1200 1000 800 600 400 200 0 0 Figure 1 | Climate fluctuations during the past 2,000 yr based on Chinese historical documentary records. a, The moisture index series at a 5-yr resolution for East China during AD 1–1900; wet and dry intervals are indicated by the blue and yellow shading, respectively (redrawn from ref. 3). b, Comparison of the number of severe cold winters per decade (dark blue bars) and the moisture index series for the Tang dynasty (AD 618–907). During the late Tang dynasty (AD 810-907), the increase in the frequency of severe cold winters is associated with plentiful summer rainfall (pale blue shading). 1060 1070 1080 1090 1100 1110 1120 1130 1140 890 880 870 860 850 840 830 820 810 Cold winter 160 150 1050 1150 Age (cal. yr BP) 900 800 Year AD Ti (counts s–1) 170 140 Figure 2 | Comparison between severe cold winters recorded in Chinese historical documents2 and the Ti counts from Lake Huguang Maar. Severe cold winters recorded are shown by blue arrows. The Ti counts from Lake Huguang Maar are shown by the thin red curve (the thick red curve has been smoothed with a 9-point running average, after Yancheva et al.1 from data archived at the World Data Center for Paleoclimatology, Boulder, Colorado, USA) during AD 800–900.The high Ti counts correspond to the lowest frequency of severe cold winters (AD 850–870; yellow shading), while the low Ti counts correspond to increased frequency of cold winters (AD 810–840, 880–900; blue shading). The winter monsoon strength inferred from Ti counts contradicts that based on the occurrence of cold winters in ancient Chinese records. NATURE| Vol 450| 15 November 2007 BRIEF COMMUNICATIONS ARISING E7 ©2007 NaturePublishingGroup
BRIEF COMMUNICATIONS ARISING NATURE Vol 45015 November 2007 (AD 865)occurred between AD 850 and 880.In contrast,AD 810-840 dynasty began with the An-Shi riot in AD 755,owing to the resulting witnessed seven very severe winters,including the sea ice event of AD disorders,which exhausted resources".The view of Yancheva et al.on 822 (ref.2),which is often considered to be the coldest interval the historical significance of the border conflict in AD 751 is therefore during the Tang dynasty.According to the interpretation of in doubt. Yancheva et al.',the lower Ti counts recorded in Lake Huguang De'er Zhang'&Longhua Lu2 Maar'during this interval indicate that the winter monsoon was National Climate Centre,China Meteorological Administration,Beijing weak during this interval (Fig.2),in conflict with the historical evid- 100081,China. ence.Their proxy-based assessment of intervals of increased winter e-mail:derzhang@cma.gov.cn monsoon strength during this period should therefore be called into 2Chinese Academy of Meteorological Sciences,Beijing 100081,China. question. Received 17 February;accepted 24 August 2007. Furthermore,the intervals of increased winter monsoon strength in AD 810-840 and AD 880-900,indicated by an increase in the fre 1.Yancheva,G.et al Influence of the intertropical convergence zone on the East Asian quency of cold winters recorded in the historical records,were both monsoon.Nature 445,74-77 (2007). 2.Zhang,D.E.A Compendium of Chinese Meteorological Records of the Past 3,000 Years associated with plentiful summer rainfall,whereas the interval of [in Chinese with English summary]327-371(Jiangsu Education Press,Nanjing, reduced winter monsoon strength between about AD 850 and 870 2004). was associated with a summer rainfall deficit (Fig.1).This casts 3.Zheng,S.Z.,Zhang.F.C.Gong,G.F.in Collection of Papers Regarding Climate Change further doubt on the proposal of Yancheva et al.that the strengths and Extra Long-Term Prediction [in Chinese]29-32 (compiled by China Meteorological Administration,Science Press,Beijing,1977) of the winter and summer monsoons were anti-correlated at a multi- 4.Man,Z.M.Climate in the Tang dynasty of China:Discussion of its evidence [in decadal scale during AD 600-900. Chinese with English abstract].Quat.Sci.22,20-30 (1998). Yancheva et al.say that the Tang dynasty began to ebb in the eighth 5.Zhang,D.E.Environmental change and agricultural development kept in historical century,after a defeat by the Arab army in AD 751,but that border documentary records for Northwest China.Adv.Clim.Change Res.2(suppl.1),28-34 (2006). battle was not a major event;it is better known for spreading ancient 6.Fan,W.L General History of China [in Chinese]VoL.Ill,113-427;376-377 (People's Chinese paper-making techniques to the Arabic countries by prison- Publishing House,Beijing,1978). ers.Instead,most Chinese historians agree that the decline of Tang doi:10.1038/nature06338 Yancheva et al.reply Replying to:De'er Zhang Longhua Lu Nature 450,doi:10.1038/nature06338(2006) Zhang and Lu'argue that Chinese historical climate records contra- intervals,as indicated by our lake records and by the historical winter dict certain of our interpretations?based on Lake Huguang Maar temperature record.Its correspondence with the wet/dry record of sediment records.Interpreting these records as an indicator for win- Zhang and Lu is uncertain;the apparent differences between the two ter monsoon winds and Chinese cave records as an indicator for wet/dry historical reconstructions may derive from geographic vari- summer monsoon rainfall,we observed?an inverse relationship ability or error in the reconstructions.The previously observed between winter and summer monsoons on a millennial timescale longer-term anti-correlation between winter and summer monsoons over the past 16,000yr.In sediments deposited during the period seems to apply to a significant portion of the(multi)decadal-scale of Classical Chinese dynastic history,we found evidence for a tem- variability investigated here. poral coincidence between winter monsoon strengthening and the We acknowledge that much of the last 20yr of the late Tang terminations of important dynasties2.Extrapolating the inverse dynasty was characterized by wetter conditions'.Nevertheless,the monsoon relationship to these multidecadal timescales,we suggested end of the Tang dynasty at AD 907,which occurs at a sharp increase that reduced summer rainfall contributed to dynastic terminations, in Ti content at Lake Huguang Maar',is also marked by a sharp including that of the Tang (AD 618-907).Zhang and Lu'challenge the drying in the unsmoothed historical reconstruction of Zheng et al. validity of the summer/winter monsoon relationship on the grounds (Fig.1,thin blue lines),as well as in the record presented by that historical records indicate that relatively cold winters tended Zhang and Lu'.Thus,the historical data suggest that the Tang dyn- to be associated with relatively wet summers over the period they asty collapsed during a period of both cold winters and low and considered.They argue that the Tang dynasty decline was associated abruptly dropping rainfall(Fig.1).This may be critical:the rate of with cold winters,in agreement with our findings2,but they find no climate change could be as important as mean climate from the point evidence for rainfall changes having contributed to the decline of the of view of social and political impact,as rapid change might discon- Tang. nect social policies from environmental realities.We acknowledge We compare our Lake Huguang Maar record during AD 550-1550 the clarification of Zhang and Lu that the An-Shi rebellion was more with two historical reconstructions (Fig.1):a hydrologic reconstruc- important than the AD 751 Tanros battle in the weakening ofthe Tang tion from the Jiang-Nan area in southeast China(25-31N),the empire. most proximal record of which we are aware,and a lower-resolution, Both temperature and rainfall seem to have had an important winter,half-year,temperature reconstruction for eastern China3.To influence on Classical Chinese dynastic changes,including the clarify similarities between the Lake Huguang Maar sediment record? decline of the Tang dynasty145.Although the reconstructions are and the historical climate records,we have allowed for stratigraphic somewhat ambiguous on the point,the inverse correlation between adjustments in the former by no more than 40 yr (Fig.1).The higher- winter and summer monsoons observed on millennial timescales? resolution historical reconstruction of wet/dry conditions in south- does describe some of the decadal and multi-decadal climate vari- east China corresponds well with weak/strong winter monsoon ability in Chinese dynastic history. E8 2007 Nature Publishing Group
(AD 865) occurred between AD 850 and 880. In contrast, AD 810–840 witnessed seven very severe winters, including the sea ice event of AD 822 (ref. 2), which is often considered to be the coldest interval during the Tang dynasty. According to the interpretation of Yancheva et al.1 , the lower Ti counts recorded in Lake Huguang Maar1 during this interval indicate that the winter monsoon was weak during this interval (Fig. 2), in conflict with the historical evidence. Their proxy-based assessment of intervals of increased winter monsoon strength during this period should therefore be called into question. Furthermore, the intervals of increased winter monsoon strength in AD 810–840 and AD 880–900, indicated by an increase in the frequency of cold winters recorded in the historical records, were both associated with plentiful summer rainfall, whereas the interval of reduced winter monsoon strength between about AD 850 and 870 was associated with a summer rainfall deficit (Fig. 1). This casts further doubt on the proposal of Yancheva et al. that the strengths of the winter and summer monsoons were anti-correlated at a multidecadal scale during AD 600–900. Yancheva et al.say that the Tang dynasty began to ebb in the eighth century, after a defeat by the Arab army in AD 751, but that border battle was not a major event; it is better known for spreading ancient Chinese paper-making techniques to the Arabic countries by prisoners6 . Instead, most Chinese historians agree that the decline of Tang dynasty began with the An-Shi riot in AD 755, owing to the resulting disorders, which exhausted resources6 . The view of Yancheva et al. on the historical significance of the border conflict in AD 751 is therefore in doubt. De’er Zhang1 & Longhua Lu2 1 National Climate Centre, China Meteorological Administration, Beijing 100081, China. e-mail: derzhang@cma.gov.cn 2 Chinese Academy of Meteorological Sciences, Beijing 100081, China. Received 17 February; accepted 24 August 2007. 1. Yancheva, G. et al. Influence of the intertropical convergence zone on the East Asian monsoon. Nature 445, 74–77 (2007). 2. Zhang, D. E. A Compendium of Chinese Meteorological Records of the Past 3,000 Years [in Chinese with English summary] 327–371 (Jiangsu Education Press, Nanjing, 2004). 3. Zheng, S. Z., Zhang, F. C. & Gong, G. F. in Collection of Papers Regarding Climate Change and Extra Long-Term Prediction [in Chinese] 29–32 (compiled by China Meteorological Administration, Science Press, Beijing, 1977). 4. Man, Z. M. Climate in the Tang dynasty of China: Discussion of its evidence [in Chinese with English abstract]. Quat. Sci. 22, 20–30 (1998). 5. Zhang, D. E. Environmental change and agricultural development kept in historical documentary records for Northwest China. Adv. Clim. Change Res. 2 (suppl. 1), 28–34 (2006). 6. Fan, W. L. General History of China [in Chinese] Vol. III, 113–427; 376–377 (People’s Publishing House, Beijing, 1978). doi:10.1038/nature06338 Yancheva et al. reply Replying to: De’er Zhang & Longhua Lu Nature 450, doi:10.1038/nature06338 (2006) Zhang and Lu1 argue that Chinese historical climate records contradict certain of our interpretations2 based on Lake Huguang Maar sediment records. Interpreting these records as an indicator for winter monsoon winds and Chinese cave records as an indicator for summer monsoon rainfall3 , we observed2 an inverse relationship between winter and summer monsoons on a millennial timescale over the past 16,000 yr. In sediments deposited during the period of Classical Chinese dynastic history, we found evidence for a temporal coincidence between winter monsoon strengthening and the terminations of important dynasties2 . Extrapolating the inverse monsoon relationship to these multidecadal timescales, we suggested that reduced summer rainfall contributed to dynastic terminations, including that of the Tang (AD 618–907). Zhang and Lu1 challenge the validity of the summer/winter monsoon relationship on the grounds that historical records indicate that relatively cold winters tended to be associated with relatively wet summers over the period they considered. They argue that the Tang dynasty decline was associated with cold winters, in agreement with our findings2 , but they find no evidence for rainfall changes having contributed to the decline of the Tang. We compare our Lake Huguang Maar record during AD 550–1550 with two historical reconstructions (Fig. 1): a hydrologic reconstruction from the Jiang-Nan area in southeast China (25–31u N)4 , the most proximal record of which we are aware, and a lower-resolution, winter, half-year, temperature reconstruction for eastern China5 . To clarify similarities between the Lake Huguang Maar sediment record2 and the historical climate records4,5, we have allowed for stratigraphic adjustments in the former by no more than 40 yr (Fig. 1). The higherresolution historical reconstruction of wet/dry conditions in southeast China4 corresponds well with weak/strong winter monsoon intervals, as indicated by our lake records and by the historical winter temperature record. Its correspondence with the wet/dry record of Zhang and Lu is uncertain; the apparent differences between the two wet/dry historical reconstructions may derive from geographic variability or error in the reconstructions. The previously observed longer-term anti-correlation between winter and summer monsoons seems to apply to a significant portion of the (multi)decadal-scale variability investigated here. We acknowledge that much of the last 20 yr of the late Tang dynasty was characterized by wetter conditions1 . Nevertheless, the end of the Tang dynasty at AD 907, which occurs at a sharp increase in Ti content at Lake Huguang Maar2 , is also marked by a sharp drying in the unsmoothed historical reconstruction of Zheng et al.4 (Fig. 1, thin blue lines), as well as in the record presented by Zhang and Lu1 . Thus, the historical data suggest that the Tang dynasty collapsed during a period of both cold winters and low and abruptly dropping rainfall (Fig. 1). This may be critical: the rate of climate change could be as important as mean climate from the point of view of social and political impact, as rapid change might disconnect social policies from environmental realities. We acknowledge the clarification of Zhang and Lu that the An-Shi rebellion was more important than the AD 751 Tanros battle in the weakening of the Tang empire. Both temperature and rainfall seem to have had an important influence on Classical Chinese dynastic changes, including the decline of the Tang dynasty1,4,5. Although the reconstructions are somewhat ambiguous on the point, the inverse correlation between winter and summer monsoons observed on millennial timescales2 does describe some of the decadal and multi-decadal climate variability in Chinese dynastic history. BRIEF COMMUNICATIONS ARISING NATURE| Vol 450|15 November 2007 E8 ©2007 NaturePublishingGroup
NATUREVol 45015 November 2007 BRIEF COMMUNICATIONS ARISING b120 Gergana Yancheva',Norbert R.Nowaczyk',Jens Mingram', Tang dynasty Peter Dulski,Georg Schettler,Jorg F.W.Negendank,Jiaqi Liu?, 140 Wet Daniel M.Sigman,Larry C.Peterson&Gerald H.Haug GeoForschungsZentrum Potsdam,Telegrafenberg,D-14473 Potsdam, 160 Germany. 2 Institute of Geology and Geophysics,Chinese Academy of Sciences,PO 80 -wet Box 9825,Beijing 100029,China. Department of Geosciences,Princeton University,Princeton,New Jersey 08544,USA. Anhahn755-76y Rosenstiel School of Marine and Atmospheric Science,University of 950 900 850800 750 700 650 600 Miami,Miami,Florida 33149,USA. Year AD PGeological Institute,Department of Earth Sciences,ETH Zurich, CH-8092 Zurich,Switzerland. e-mail:gerald.haug@erdw.ethz.ch 100 1.Zhang,D.E.Lu,L.H.Anti-correlation of summer/winter monsoons?Nature 450, dot10.1038/nature(06338(2007). 120 Temperature 1 2.Yancheva,G.et al.Influence of the intertropical convergence zone on the East Asian 140 monsoon.Nature 455,74-77 (2007). 3.Dykoski,C.A.A high-resolution,absolute-dated Holocene and deglacial 60 Asian monsoon record from Dongge Cave,China.Earth Planet.Sci.Lett.233,71-86 (2005). 180 4.Zheng,J.et al.Precipitation variability and extreme events in eastern China during the Normalized past 1500 years.Terr.Atmos Ocean.Sci 17,579-592(2006). 5.Ge,Q.et al.Winter half-year temperature reconstruction for the middle and lower reaches of the Yellow River and the Yangtze River,China,during the past 2000 years Holocene13,933-940(2003). +20 40 doi:10.1038/nature06339 1400 1200 1000 800 600 Year AD Figure 1|Comparison of Chinese historical climate records45 with the Lake Huguang Maar sediment record2.Comparisons are shown for the time intervals AD 550-1500(a)and AD 600-950(b).In a,the red line shows the reconstruction of winter half-year temperatures in eastern China(error bars show uncertainties associated with historical data);a and b show the Lake Huguang Maar titanium record(brown:original data and 15-point running mean),and the reconstruction of hydrologic conditions from the Jiang-Nan area in southeast China'(blue:original data and 15-point running mean).The latter are expressed as a normalized dry-wet index;data are connected by a thin blue line only where data points are available from consecutive years.High Ti values (plotted downward)are interpreted to reflect stronger winter monsoon winds".We have allowed for stratigraphic adjustments in the Huguang Maar sediment record'that are consistent within error with the published age model.None of these were more than 40 yr,and most were much less,including during the interval of the Tang dynasty(black curve at bottom of a shows the adjusted timescale minus the "C-based timescale as described in Yancheva et al.',such that the vertical scale matches that of the x-axis). E9 2007 Nature Publishing Group
Gergana Yancheva1 , Norbert R. Nowaczyk1 , Jens Mingram1 , Peter Dulski1 , Georg Schettler1 , Jo¨rg F. W. Negendank1 , Jiaqi Liu2 , Daniel M. Sigman3 , Larry C. Peterson4 & Gerald H. Haug5 1 GeoForschungsZentrum Potsdam, Telegrafenberg, D-14473 Potsdam, Germany. 2 Institute of Geology and Geophysics, Chinese Academy of Sciences, PO Box 9825, Beijing 100029, China. 3 Department of Geosciences, Princeton University, Princeton, New Jersey 08544, USA. 4 Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida 33149, USA. 5 Geological Institute, Department of Earth Sciences, ETH Zu¨rich, CH-8092 Zu¨rich, Switzerland. e-mail: gerald.haug@erdw.ethz.ch 1. Zhang, D. E. & Lu, L. H. Anti-correlation of summer/winter monsoons? Nature 450, doi:10.1038/nature06338 (2007). 2. Yancheva, G. et al. Influence of the intertropical convergence zone on the East Asian monsoon. Nature 455, 74–77 (2007). 3. Dykoski, C. A. A high-resolution, absolute-dated Holocene and deglacial Asian monsoon record from Dongge Cave, China. Earth Planet. Sci. Lett. 233, 71–86 (2005). 4. Zheng, J. et al. Precipitation variability and extreme events in eastern China during the past 1500 years. Terr. Atmos. Ocean. Sci 17, 579–592 (2006). 5. Ge, Q. et al. Winter half-year temperature reconstruction for the middle and lower reaches of the Yellow River and the Yangtze River, China, during the past 2000 years. Holocene 13, 933–940 (2003). doi:10.1038/nature06339 Tang dynasty Tang dynasty Ti (counts s–1) Ti (counts s–1) 120 140 160 2 0 –2 100 1400 1200 1000 Year AD 800 600 Huang Chao End of Tang (907) rebellion (875-884) An Shi rebellion (755-763) 1 0 –1 Wet Dry Cold Warm Wet Dry a b ∆ age (yr) 950 900 800 750 650 600 Year AD 850 700 180 120 140 160 180 +20 –40 0 2 0 –2 Normalized dry-wet index Normalized dry-wet index anomaly (ºC) Temperature Wet Dry Wet Dry Figure 1 | Comparison of Chinese historical climate records4,5 with the Lake Huguang Maar sediment record2 . Comparisons are shown for the time intervals AD 550–1500 (a) and AD 600–950 (b). In a, the red line shows the reconstruction of winter half-year temperatures in eastern China5 (error bars show uncertainties associated with historical data); a and b show the Lake Huguang Maar titanium record2 (brown: original data and 15-point running mean), and the reconstruction of hydrologic conditions from the Jiang-Nan area in southeast China4 (blue: original data and 15-point running mean). The latter are expressed as a normalized dry-wet index; data are connected by a thin blue line only where data points are available from consecutive years. High Ti values (plotted downward) are interpreted to reflect stronger winter monsoon winds2 . We have allowed for stratigraphic adjustments in the Huguang Maar sediment record2 that are consistent within error with the published age model. None of these were more than 40 yr, and most were much less, including during the interval of the Tang dynasty (black curve at bottom of a shows the adjusted timescale minus the 14C-based timescale as described in Yancheva et al.2 , such that the vertical scale matches that of the x-axis). NATURE| Vol 450|15 November 2007 BRIEF COMMUNICATIONS ARISING E9 ©2007 NaturePublishingGroup