Climate change in China during the past 2000 years: An overview

Ge Quansheng , Zheng Jingyun

Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China Email: [email protected] Outline

1 Introduction

2 Historical Documents as Proxy

3 Reconstructions and Analyses

4 Summary and Prospects 1. Introduction: E Asia2K

Climate System Socio-economic System •typical East Asian •dense population and rapid monsoon climate economic development • significant seasonal and • be susceptible to global inter-annual and inter- warming and extreme decadal variability climate events Climate change study in the past 2ka in East Asian is both beneficial and advantageous.

• various types of natural proxy • Plenty of historical documents Fig. Active regional working groups under as proxy the past 2ka theme (PAGES 2009) 2. Historical Documents as Proxy

Type Period Amount

 Chinese classical 1,531 kinds, 137 BC~1470 AD documents 32,251 volumes More than 8,000 1471~1911 (The Ming  Local gazettes books (部), 110, and Qing Dynasty) 000 volumes  Memos to the About 120,000 1736~1911 emperor pieces  Archives of the 1912~1949 20,000 volumes Republic of China More than 200  Private diaries 1550~ books (部) Chinese classical documents

AD 833, North China plain: Extreme drought event was occurred, crops were shriveling, no yields, people were in hungry….

Fig. Example for Ancient Chinese writings Local gazettes

The 28th year of the Daoguang reign (1848 AD), the 6th (lunar) month, strong wind and heavy rain, the Yangtze River overflowed; the 7th month, strong wind Fig. Gazettes of Yangzhou Prefecture and thunder storm, field published in 1874 AD and houses submerged. Memos to the emperor, Yu-Xue-Fen-Cun

• Yu (雨, rainfall)—the infiltration depth into soil after each rainfall; Fig. Distribution of YXFC •Xue (雪, snowfall)—the snowfall and the amount of every reign depth after each snowfall; •Fen (分, Chinese length unit )— 0.32 cm approximately; •Cun (寸, Chinese length unit, =10 Fens) —3.2 cm approximately. Memos to the emperor, Yu-Xue-Fen-Cun

On the Dec. 26 (11th day of the comment by eleventh month in lunar calendar), Emperor Zhengding Fu: Gaocheng county 3 Qianlong: “read cuns; Jinzhou county 2 cuns; Wuji it” county 2 cuns; Pingshan county 2 cuns. Fig. The first and last page (right to left) of the original 6-page memo reported by Gao Bin, Governor General of Zhili province 3. Reconstructions and analyses

3.1 Winter Half-year Temperature Variation in the past 2000a

3.2 Precipitation or Dry-wet Reconstructions

• Variation • The past 1500a • Extreme events • Spatial patterns • The past 300a

3.3 Inter-comparison between independent reconstruction and uncertainty assessment on regional temperature reconstructions 3.1 Winter Half-year Temperature Variation in the past 2000a —— Methodology • 1 AD~1500 AD

Relationship between Modern Spatial phenological phenomena Instrumental representativeness of and temperature change data each station

The past and present phenology’s Temperature Seasonal Winter half-year latitude and date difference temperature of temperature in difference a single site East China

10a or 30a winter half- Phenological records in year temperature in East Chinese historical documents China in the past 3.1 Winter Half-year Temperature Variation in the past 2000a —— Methodology • 1501 AD~1740 AD Proxy indicator or temperature 10a winter half-year anomaly series in some sub- temperature in East China regions by other authors Spatial • 1741 AD~1950 AD representativeness

Winter snow-days extracted Winter temperature from Yu-Xue-Fen-Cun in Hefei

• 1951 AD~1999 AD The mean instrumental temperature data of twenty stations 3.1 Winter Half-year Temperature Variation in the past 2000a —— Results

Fig. Winter half-year temperature anomaly in East China with 30a resolution in the past 2ka (below), and 10a resolution during 961~1110 and 1501~1999 (above). Curve is three-year running mean. 3.1 Winter Half-year Temperature Variation in the past 2000a —— Results

 0~500s: with a cooling trend of -0.17℃/100a; around the 490s reached the lowest temperature level (about 1℃ colder than 1951~1980)  510s~1310s: with a warming trend of 0.04℃/100a; the 30a mean temperature of two warm peaks were generally 0.3~0.6℃higher than the present, while a maximum warming of 0.9℃ occurred in 1230s~1250s  1320s~1870s: with a rapid cooling rate of 0.10℃/100a; the 30a mean temperatures of four cold troughs were generally 0.6~0.9℃ colder with a maximum cooling of 1.1℃ during 1650s~1670s  1880s~1990s: rapidly warming to the 20th century; temperature increased dramatically for the most recent two decades whose mean temperature was 0.5℃ than 1951~1980 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Basic dataset

Annual dry-wet grades derived from Chinese historical documents at 48 stations • Spatial coverage: East China (three sub-regions including North China, Jiang-Huai and Jiang-Nan) • Temporal coverage: 137 BC~2000 AD; but before 500 AD existed a lot of discontinuous records 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Dry-wet variation

Fig. Precipitation change and its regional difference in East China during 500~2000 • Upper panel: Regional D-W index series (Blue dash: 0.01HZ FFT low-pass filter) • Lower panel: standard deviation (in 100a running) of the D-W index series starting from 500-599 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Dry-wet variation • Strong Multi-decadal to centennial changes and regional differences

Region Whole East China North China Plain Jiang-Huai Jiang-Nan Dry-wet Wet Dry Wet Dry Wet Dry Wet Dry period -510s 520s-570s -510s 520s-580s -520s 530s-560s -530s 540s-570s 580s-630s 640s-700s 590s-620s 630s-700s 570s-640s 650s-690s 580s-850s 860s-920s 710s-750s 760s-900s 710s-750s 760s-900s 700s-760s 770s-820s 930s-970s 980s-1010s 910s-1000s 1010s-1240s 910s-990s 1000s-1280s 830s-990s 1000s-1230s 1020s-1060s 1070s-1130s Duration 1250s-1430s 1440s-1520s 1290s-1430s 1440s-1520s 1240s-1320s 1330s-1380s 1140s-1160s 1170s-1220s 1530s-1620s 1630s-1640s 1530s-1560s 1570s-1540s 1390s-1460s 1470s-1530s 1230s-1420s 1430s-1540s 1650s-1910s 1920s- 1650s-1910s 1920s- 1540s-1880s 1890s-1950s 1550s-1630s 1640s-1690s 1960s- 1700s- Tab. Dry-wet phases in East China and its 3 sub-regions in the past 1500a 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Dry-wet variation • The range of precipitation variability observed in the 20th century in East China can not fully represent that for the past 1500a. • The decadal pattern of warming with drought in North China from 1980s is likely to be the pattern of the natural climate variability. 570s-770s: 14 of 21 decades, dry Warm 930s-1310s: 27 of 39 decades, dry epoch: Totally, 41 of 60 (68%) decades, show dry 500s-560s: 4 of 7 decades, dry Cold 1320s-1890s: 21 of 58 decades, dry epoch: Totally, 25 of 65 (38%) decades, show dry 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Extreme events With precipitation anomaly of at least 1.645 times the standard deviation distancing the mean of the series

Fig. Severe persistent drought (red bar) and flood (blue bar) events. Pink bar: three most severe droughts in the past 300a. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 1500a  Extreme events  There occurred 16 severe, persistent droughts and 18 severe, persistent floods over East China during 501~2000. And the most severe sustained drought occurred in 1634~1644.

 Droughts were more frequent during the 12th ~14th century while floods were more frequent since the middle of the 17th century.

 The severity and duration of the most severe flood in the 20th century was comparable to that in the past 1500a. However, the severity and duration of the most severe drought in the 20th century can't compare with that in the history. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a

② The Semi-arid ① The Middle-Lower Reaches of Region: including 3 Yellow River: including 17 sites representative sites and dividing into 4 sub-regions.

③ The Middle-Lower Reaches of Yangtze River: including 5 representative sites

④ The East China: including North China and the Middle-Lower Reaches of Yangtze River 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ① The Middle-Lower Reaches of Yellow River Yu-Xue-Fen-Cun

Snowfall depth Infiltration depth

Snowfall and depth Principle of ＋Physical model equation for each site water balance of infiltration

Precipitation of Precipitation of each snowfall each rainfall

Monthly ~ seasonal ~ annual precipitation at each site or sub-region or the whole region 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ② Semi-arid Region Yu-Xue-Fen-Cun

Infiltration depth Rainfall Date Soil infiltration experiment Precipitation of each rainfall

Criteria: at least 6 rainy days in a 25-day period; less than 10 days continuous rainless days

The length, initial/final date of rainy season at each site 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ③ The Middle-Lower Reaches of Yangtze River

Yu-Xue-Fen-Cun Rainfall date and rain-days

Criteria of Meiyu period: Identified by the • from May to August, there are at least six instrumental data rainy days during any continuous 10 days; during 1951-2000 • Meiyu period is starting (ending) over the whole region only if those at 4 of the 5 sites are starting (ending). Conversion equation The length, starting/ending data of Meiyu period over MLRYR Rainfall of Meiyu period over MLRYR 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a

(a) Hebei region 800 Hebei region region 200 Spring 600

100 400 (b) Jinnan region Jinnan region 800 0 Summer 900 700 600 500 300 400

100 Weihe region(c) Weihe region Autumn 800 200 600 100 400 0 Winter (d) Shandong region 1200 100 Shandong region 50 1000 0 1200 800 Annual 1000 600

800 400 (e) Whole region 800 600 Whole region

400 600 200 400 1750 1800 1850 1900 1950 2000 1750 1800 1850 1900 1950 2000 yr Fig. Annual and seasonal precipitation at Jinan (Left). Annual precipitation over the Middle-Lower Reaches of Yellow River and its 3 sub-regions (Right). 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a

Fig. Rainy season length, initial/final date at each site in the semi-arid region (left); wavelet analysis of rainy season length at each site (right). 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a

Fig. Starting date (a), ending date (b), length (c) and rainfall of Meiyu (d) during 1736~2000 (left). Wavelet analysis of rainfall in Meiyu period (above) 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ① Over the middle-lower reaches of Yellow River: The abrupt change of precipitation from high to low occurred around 1915.  Precipitation during 1791~1805, 1816~1830 and 1886~1895 was markedly higher than the mean, while that during 1916~1945 and 1981~2000 was markedly lower.  The precipitation variation had 2~4a, quasi-22a and 70~80a cycles.

② In the semi-arid region, the length of rainy season:  showed evident inter-annual and inter-decadal changes.  was statistically significantly longer during 1736~1840 than that of 1955~2000. This finding suggested that the intensity of the East Asian Summer Monsoon became weaker from the mid-19th century.  had 16~20a, 30~40a and 50~60a cycles. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ③ Over the middle-lower reaches of Yangtze River:  The starting and ending dates and the length of Meiyu had significant inter-annual and inter-decadal variations.  The length of Meiyu had 7~8a, 20~30a, 40a and centennial oscillation.  The length of Meiyu and the strength of East Asian Summer Monsoon had a very good correlation. • During 1736~1770, 1821~1870 and 1921~1970, the EASM was stronger, corresponding to the decreased length of Meiyu. • During 1771~1820, 1871~1920 and 1971~2000, the EASM was weaker, corresponding to the increased length of Meiyu. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ④ The East China North China Precipitation • Precipitation over the MLRYR (①) • Drought/flood grade (the mean of 22 sites) D-F grade

Spatial Patterns

The Middle-Low Reaches of Yangtze River • Meiyu (③) Fig. The 0.1Hz low pass filter series for precipitation variations. Blue solid line: • Drought/flood grade (the MLRYR; red dashed line: NC; and the mean of 27 sites) small picture is power spectrum analysis. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ④ The East China

Fig. Cross-wavelet spectrum analysis between precipitation (above) and D-F grade (below) over NC and MLRYR. The red circle: reverse phase; the blue: consistent phase. 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ④ The East China

Fig. The 4 typical spatial patterns for the summer monsoon rainbands from D-F grades in East China from 1736~1911 3.2 Precipitation or Dry-wet Reconstructions —— The Past 300a ④ The East China

 The location of the summer monsoon rainbands in East China had inter-decadal oscillations, like 20-30a and quasi-10a, as well as inter-annual oscillations, like 5-7a, 2-4a.

 The spatial pattern in the south-north regions was different on various time scales (e.g. reversal pattern at 60-80a, consistent at quasi-30a, and, at 20a, a reversal phase existing before 1870 and a consistent phase after).

 Although a reverse pattern had dominated the past 50a, this had not been the case for all periods over the past 300 years. 3.3 Inter-comparison between independent reconstruction and uncertainty assessment on regional temperature reconstructions

14 series at 10a resolution more than 500 years and reconstructed from historical documents, and series reconstructed from tree-ring, stalagmite, as well as multi-proxy were selected for comparison

Fig. Different types of temperature reconstructions used in the uncertainty assessment of temperature in the past 2ka. The whole China is divided into 5 sub-regions. Methodology for Inter-comparison and Uncertainty Assessment

Temperature Correlation Analysis Standardization Reconstructions Cluster analysis

The Envelope Assessment Method Methodology of Series Comparison

Uncertainty of the series

Methodology of Uncertainty Assessment Results from Inter-comparison

Fig. Correlation coefficients and the clustering spectrum system of the 14 reconstructions in East China from historical documents Six reconstructed temperature series from historical documents and natural evidence (left) and the wavelet of these series (right). (a) Temperatures of winter-half- year in the East of China (Ge et al., 2003); (b) Annual temperatures in the East of China (Wang and Gong, 2000a); (c) Average temperatures in China (Yang et al., 2002); (d) Temperatures from July to August in Beijing, China (Tan et al., 2003); (e) Width indexes of tree ring in Qilian Mt., Qinghai, China (Liu et al., 2004); (f) Average temperature in Qinghai-Tibet plateau, China (Yang et al., 2003). Results from Inter-comparison

•The coherence of climatic series in different region actually depends on regional climate similarity, and regional climatic difference should be responsible for the discrepancy between reconstructed climate series.

•Comparison on wavelet suggests that 30-year temporal resolution might be the most reasonable for temperature change study by using Chinese documentary data.

•There are approximately simultaneous temperature fluctuations between most of series derived from historical documents and the ones from natural evidence in China, especially under a lower temporal resolution. Cold/warm stages of 50–100 years can be well revealed by both ‘documentary’ and ‘natural’ series. Results from Uncertainty Assessment

Northeast

Fig. Different temperature series (above). The uncertainty range of reconstructions (below), the bold-black curve is the regional temperature coherent series, the part with less series number used is marked with the dashed, and the bold-gray is 10a smoothed instrumental data. Results from Uncertainty Assessment

Tibet Central East

Fig. The same as the last slide Results from Uncertainty Assessment

Fig. Comparison of composite reconstruction temperature in 5 Southeast sub-regions (0.01Hz filter), the shaded areas are the coldest periods over the past 500a. Fig. The same as the last slide Results from Uncertainty Assessment

 Large uncertainty existed for the period before the 16th

century, while some temperature variations had high level of

consistency in the past 500a.

 The warming during the 10–14th centuries in some regions

might be comparable in magnitude to the warming of the last

few decades of the 20th century which was unprecedented

within the past 500 years. 4 Summary and Prospects

In the recent years:  By using historical phenological cold/warm records, we reconstructed winter half-year temperature in the past 2ka in East China, and examined its amplitude, change rate and periodicity.  Utilizing various kinds of Chinese historical documents and developing quantitative conversion methods, we reconstructed precipitation or Dry-wet changes in different regions (including North China, the middle-lower reaches of Yangtze River, South China and the whole East China) on different timescales (including the past 1500a and 300a). Further, the precipitation variations and spatial patterns in the history were analyzed.  We compared different temperature reconstructions from various proxies and quantified the uncertainty existing in the series of different regions. 4 Summary and Prospects

In the future:  Improve the quantitative methods extracting climate information from historical documents in order to reduce the uncertainty of reconstructions;

 Develop long-term (millennial long) and high-resolution (annual~decadal) temperature or precipitation series, especially emphasize the regions short of reconstructions and explore new historical archives potentially for climate reconstruction;

 Study the characteristics and spatial patterns of climate change in China more comprehensively, and the spatial synchronism between the variations in China and the Northern Hemisphere; 4 Summary and Prospects

In the future:  Understand the relationship between climate change and some driving forces such as solar activity, volcanic eruption, the concentration of green house gas and several atmospheric circulation modes;

 Analyze the impacts of warm/cold epochs or severe climate change events in the history on socio-economical system and the corresponding adaptation of human, and compare the spatial differences. Reference: 1. Ge Q S, Zheng J Y, Fang X Q, et al. Winter half-year temperature reconstruction for the middle and lower reaches of the Yellow River and Yangtze River, China, during the past 2000 years. The Holocene, 2003, 13(6): 933~940. 2. Zheng J Y, Wang W C, Ge Q S, et al. Precipitation variability and extreme events in eastern China during the past 1500 years. Terrestrial Atmospheric and Oceanic Sciences, 2006, 17(3): 579~592. 3. Zheng J Y, Hao Z X, Ge Q S. Variation of precipitation for the last 300 years over the middle and lower reaches of the Yellow River. Science in China, Series D-Earth Sciences, 2005, 48(12): 2182~2193. 4. Ge Q S, Hao Z X, Tian Y Y, et al. The rainy season in the Northwestern part of the East Asian Summer Monsoon in the 18th and 19th centuries. Quaternary International, 2010, doi:10.1016/j.quaint.2010.02.025. 5. Ge Q S, Guo X F, Zheng J Y, et al. Meiyu in the middle and lower reaches of the Yangtze River since 1736. Chinese Science Bulletin, 2007, 52(1): 1~8. 6. Hao Z X, Zheng J Y and Ge Q S. Variations in the Summer Monsoon Rainbands across eastern China over the past 300 years. Advances in Atmospheric Sciences, 2009, 26(4): 614~620. 7. Ge Q S, Zheng J Y, Tian Y Y, et al. Coherence of climatic reconstruction from historical documents in China by different studies. International Journal of Climatology, 2007, 28(8): 1007~1024. 8. Ge Q S, Zheng J Y, Hao Z X, et al. Temperature variation through 2000 years in China: An uncertainty analysis of reconstruction and regional difference. Geophysical Research Letters, 2010, 37, L03703, doi: 10.1029/2009GL041281. Thank you!