Test Calibration of the Paleoclimatic Proxy Data with Chinese Historical Records

Abstract

The calibration of paleoclimate proxies is one of the key problems in the study of paleoclimate at present. Historical documentary records of climate are suitable for calibration on dating and the climatic implication of the proxy data in a climatological sense. A test calibration on correcting the Delingha tree ring precipitation series using Chinese historical documentary records shows that among the 44 extreme dry cases in 1401–1950 AD, 42 cases (or 95.5%) are believable. Thus the long series of Delingha rings-denoted precipitation is highly reliable. Another test to validate the monsoon intensity proxy data based on the Zhanjiang Huguangyan sediments using historical records indicates that the years of Lake Maar Ti content series-designated winter monsoon intensities are entirely opposite to historical documents-depicted years of harsh winters in 800–900 AD. As a result, serious doubt is raised about the climatic implication of this paleo-monsoon proxy series.

Keywords

proxy data ; calibration ; paleoclimate ; historical documentary records of climate

1. Introduction

Any proxy dataset obtained from natural sources (e.g., lake sediments, stalagmite, and tree rings) has to be compared with data obtained from other sources (including instrumental and historical documentary records), in order to confirm the correctness of interpreted climate implication involved in these proxy data, the precision of measured or induced data, and the accuracy in dating. This is an indispensable step in determining the usefulness of such scientific records. The author believes that the calibration of paleoclimate proxies is one of the key problems in the study of paleoclimate today. In recent years, rapid progress has been made in studies of paleoclimate, the approaches to obtain paleoclimate proxies have been expanded, data acquired from field samples and laboratory measurements have been drastically increased, and many high-resolution proxies have been published one after another. But the inconsistency and even contradiction among them has become a disturbing issue in ongoing studies of paleoclimate. Historical documentary records of climate characterized by accurate and reliable recording of time (to year, even to month or day), and by the certainty of locations and understandable implication of climatic elements (temperature, rainfall, etc.) make it possible to compare them to the proxy data for calibration on dating and the implication of the proxy data in a climatological sense.

Early encouraging attempts include a comparison of the curve of δ¹⁸ O ratios of ice core of Dunde in the Qilianshan Mountains published in 1991 [ Yao et al., 1991 ] with the 500-year winter temperature series of Shanghai derived from historical records and reported by Zhang [1980] (Fig.1 ), which showed their conformity on a 10-year scale [ Zhang , 1995 ]. In addition, the comparison of the curve of dust depositing events in Chinese history [Zhang , 1983] published in 1982 with the later reported content of fine particles in ice cores in Western China [ Yao et al., 1995 ] and with that in the Greenland Site-J ice core [ Tegen and Rind, 2000 ](figure is omitted), indicated consistency between historical and natural records. These results provided us with confidence in the proxy data.

Figure 1. Comparison of δ18 O series of Dunde ice cord [ Yao et al., 1991 ] and winter temperature series of Shanghai inferred from the Chinese ancient documents [ Zhang, 1980 ] for the last 500 years redrawn from [ Zhang, 1995 ]

2. Calibration of dating with Chinese historical records

Proxy data of rainfall derived from tree rings have a temporal resolution of a year, that can be employed to make comparison with historical record inferred dry/wet climate regimes. The inconsistency between them either provides clues for further investigation of the historical records or corrects the dating of tree ring series. More importantly, of course, the consistency in their intercomparison will provide convincing evidence for the accuracy of these scientific data.

Test calibration was performed on the Delingha tree ring-denoted precipitation series in 1001–2000 AD using Chinese historical documentary records. This yearly rainfall series was constructed by use of interring width in well-replicated samples from 7 sites including Delingha and Wulan of Qinghai province and the reconstructed series is equivalent to the annual precipitation in a general sense [ Shao et al., 2004 ]. The series (1001–2000 AD) has an annual mean precipitation of 132 mm, and a standard deviation σ=34 mm. If the annual precipitation anomaly at 1.7σ (= 75 mm) is set for the extreme dry year, then such years of R < 75 mm in modern times are 1957 (62 mm), 1979 (74 mm), and 1995 (74 mm). Since the historical records have better continuation after 1400 AD, the validation is made only of the ring-denoted precipitation thereafter. The extremely dry years with R < 75 mm in 1401–2000 AD on the curve of Delingha are as follows: 1436, 1444, 1450, 1451, 1455, 1468, 1478, 1480, 1481, 1484, 1492, 1495, 1496, 1524, 1528, 1539, 1588, 1602, 1633, 1649, 1657, 1665, 1672, 1676, 1687, 1688, 1689, 1693, 1710, 1713, 1715, 1727, 1749, 1770, 1792, 1796, 1816, 1824, 1831, 1843, 1861, 1877, 1895, 1918, 1957, 1979, and 1995, 47 years in all.

The 3 driest years in 1951–2000 were borne out by modern meteorological observations and no need is done to validate the reliability. The 44 extreme dry years in 1401–1950 AD will be validated with historical records. According to the precipitation patterns of “drought covering a large area and flood occurs only in a lined zone” in China, the present writer contends that the extreme scarcity of rainfall in those years did not occur on a local basis but over a considerably large region, for which a survey was undertaken of a larger region under droughts covering the whole of Qinghai and its neighboring provinces of Gansu and Ningxia. And the used historical records were taken from “A Compendium of Chinese Meteorological Records of the Last 3,000 Years” [Zhang, 2004 ], which was quoted from 48 ancient Chinese books subjected to systematic collection and rectification. The validation of the 44 rings-shown lowest precipitation years against the historical drought records in 1401–1950 AD, with the results given in Figure 2 .

Figure 2. An examination of extremely low precipitation of Delingha proxy series from tree rings [ Shao et al., 2004 ] using Chinese historical documentary records [ Zhang, 2004 ] in 1401–2000 AD, where the extreme dry years supported by historical documents are denoted by brown dots, one-year bias between the two are shown by brown open circles and the years of missing historical descriptions are denoted by pink dots

Thirty of the 44 driest years shown in the precipitation proxy of Delingha tree rings during 1401–2000 AD can be validated by the historical records, of which 28 events are confirmed by the historical description of droughts (brown dots) and the remaining 2 droughts in 1481 and 1665 AD lagged by one year in comparison to historical evidence (brown open circles), that is, the rings-denoted lowest rainfall events occurred in the following years. These two cases need to be investigated further and may provide a clue for both tree ring dating and historical record correction. Furthermore, 12 driest years shown in the Delingha ring proxy, which are 1436, 1444, 1478, 1480, 1492, 1524, 1649, 1672, 1676, 1693, 1710, and 1843 AD (denoted by pink dots), had no corresponding historical records in the province of Qinghai and its neighboring provinces. These extreme dry years indicated by Delingha tree rings can be utilized to supplement the deficit of historical documents. This has great significance for the vast western portion of the country where historical documents were lacking.

The comparison and validation show that the long series of Delingha rings-denoted precipitation is a highly reliable proxy with high resolution. Among the 44 extreme dry cases validated 42 cases (or 95.5%) are believable, and 63.6% of which are definitively verified by the historical documents, 31.8% of which can be used to supplement the lack of documentary records, and only 2 cases (or 4.5%) are doubtful.

3. Determining the usability of climatological implications of the proxy series using historical records

Proxy climate data obtained from various natural samples such as ice cores, tree rings, coral, and lake sediments are the substituted values of climate element (e.g., temperature, precipitation, etc.) as an interpretation of laboratory measurements of the samples based on the fundamental principles in physics and biological sciences. Due to great difference in the variety of samples, the sampling environment, and various values of measurement, whether the measured value is really representative of a climate element remains to be validated. This is directly related to the validity of the conclusions on climate change based on these laboratory measurements. The past documentary records often clearly described the synoptic phenomena and climatic events such as warm and cold temperatures, precipitation and others. As a result, the descriptions in ancient documents are of unique value in validating the climate implication of proxy dataset.

A recent attempt to use historical records was made to validate the monsoon intensity proxy data based on the Zhanjiang Huguangyan sediments, and raised doubt about the implication of this paleomonsoon proxy [ Zhang and Lu, 2007 ]. The paleomonsoon intensity sequence is denoted by the content of Titanium (Ti) measured in the Huguangyan cores. Researchers claimed that the Ti in the cores of the Lake Marr came from North China and was brought into South China by strong winter monsoon and thus changes in the Ti content could represent the variability of the intensity of East Asian winter monsoon [ Yancheva et al., 2007 ].

Harsh winters were recorded in a full and accurate manner in Chinese historical documents, with which the years of severe winters and their frequent period over the monsoon region of eastern China can be identified [ Zhang and Lu, 2007 ]. Obviously, harsh winter means strong winter monsoon, and the years of strong winter monsoon were years with severe winters, and the period with intense winter monsoons were times of frequent harsh winters. Figure 3 presents the comparison of historical documents-depicted years of harsh winters to the years of Lake Maar Ti content series-designated winter monsoon intensities in 800–900 AD. The following significant differences are apparent in the figure: (1) without exception all years of severe winters correspond to lower-values of Ti, that is, these years correspond to the years of weak winter monsoon (below the red dashed horizontal line in Figure 3 ) rather than those of intense monsoons; (2) the maximum Ti values appear in the neighborhood of 870 AD, indicative of the strongest winter monsoon and thus coldest winter, but this period is not for the years of severe winter and even in 845–875 AD (spanning 30 years) there was only one year of harsh winter but in relation to an extremely low value of Ti (which implied the year was of feeble winter monsoon and hence a somewhat warmer winter); (3) the two intervals of lower-valued Ti relate to the years of frequent severe winters in the historical records, with 8 harsh winters over 810–844 AD as the coldest period of Tang Dynasty and 4 cold winters in 880–900 AD as written in the documentary records (see the blue color-denoted bar of Figure 3 ). It is, however, seen that in the 2 cold periods Ti magnitudes show the weak monsoon indicative of somewhat warmer winters, a situation that is entirely opposite to historical records. It seems that the phase of monsoon variation indicated by Ti content is ahead or behind of historical records for 40 years.

Figure 3. Comparison between harsh winters recorded in history (blue vertical bar) and Ti content inferred winter monsoon intensity for 800–900 AD (red line, the bold red line is from a 9-point running average redrawn from [ Zhang and Lu, 2007 ]

Two explanations can be given to the contradiction in Figure 3 . One possibility is that an about 40 years bias of dating exists in the measurement of Ti samples resulting in the divergence in the phase of historical records and Ti content for harsh winters. But due to the advanced measuring techniques used in the laboratory and a variety of methods employed it is almost impossible to produce such a big error in dating. In particular, the article [ Yancheva et al., 2007 ] pointed out that the sequence they constructed had the time resolution of “a year” so that it was used to address the occurrences of historical events and dynasty replacement in Chinese history. The other explanation is that there is error in interpretation of the climatic implication of Ti content. Specifically, the change in Ti content does not represent the variation of winter monsoon intensity and Ti may not come entirely from the north. It is worth noting that such doubts are also raised by geochemists, who argue that Ti of the Huguangyan cores originates locally and its variation is related to the hydrological conditions of Lake Maar and is not controlled by Asian winter monsoon [ Zhou et al., 2007  ; 2009 ]. It follows that the scientific implication of paleo-climate proxy series should be determined and validated from all perspectives and with a multidisciplinary approach.

Acknowledgements

This work was supported in part by National Science Foundation of China (41075055).

References

1. Shao et al., 2004 X. Shao, L. Huang, H. Liu, et al.; Variation in rainfall shown in 1000-yr tree rings of Delingha of Qinghai; Science in China Series D (in Chinese), 34 (2) (2004), pp. 145–153
2. Tegen and Rind, 2000 I. Tegen, D. Rind; Influence of the latitudinal temperature gradient on soil dust concentration deposition in Greenland; J. Geoph. Res., 105 (D6) (2000), pp. 7199–7212
3. Yancheva et al., 2007 G. Yancheva, N.R. Nowaczyk, J. Mingram, et al.; Influence of the intertropical convergence zone on the East-Asian monsoon; Nature, 445 (2007), pp. 74–77
4. Yao et al., 1991 T. Yao, Z. Xie, X. Wu, et al.; Climatic change since little ice age recorded by Dunde ice cap; Science in China Series B, 34 (6) (1991), pp. 760–767
5. Yao et al., 1995 T. Yao, K. Jiao, Z. Yang, et al.; Climate change shown in the Guliya ice cores since Little Ice Age; Science in China Series D (in Chinese), 25 (10) (1995), pp. 1108–1114
6. Zhang, 1980 D. Zhang; Winter temperature changes during the last 500 years in South China; Chinese Science Bulletin, 25 (6) (1980), pp. 497–500
7. Zhang, 1982 D. Zhang; Analysis of dust rain for the historical times in China; Chinese Science Bulletin, 28 (3) (1982), pp. 361–366
8. Zhang, 1995 D. Zhang; High resolution records available from Chinese historical documents; Quaternary Research (in Chinese), 15 (1) (1995), pp. 75–80
9. Zhang, 2004 D. Zhang; A Compendium of Chinese Meteorological Records of the Last 3,000 Years (in Chinese), Jiangsu Education Press (2004), p. 3666
10. Zhang and Lu, 2007 D. Zhang, L. Lu; Anti-correlation of summer and winter monsoons; Nature, 450 (E7-E8) (2007) 06338 http://dx.doi.org/10.1038/nature 06338
11. Zhou et al., 2007 H. Zhou, H. Guan, B. Chi; Record of the winter monsoon strength; Nature, 450 (E10-E11) (2007) http://dx.doi.org/10.1038/nature06408
12. Zhou et al., 2009 H. Zhou, B. Wang, H. Guan, et al.; Constraints from strontium and neodymium isotopic ratios and trace elements on the sources of the sediments in Lake Huguang Maar; Quaternary Research, 72 (2009), pp. 289–300