Article | Open | Published:

Mysterious abrupt carbon-14 increase in coral contributed by a comet

Scientific Reports volume 4, Article number: 3728 (2014) | Download Citation


A large and sudden increase in radiocarbon (14C) around AD 773 are documented in coral skeletons from the South China Sea. The 14C increased by ~ 15‰ during winter, and remain elevated for more than 4 months, then increased and dropped down within two months, forming a spike of 45‰ high in late spring, followed by two smaller spikes. The 14C anomalies coincide with an historic comet collision with the Earth's atmosphere on 17 January AD 773. Comas are known to have percent-levels of nitrogen by weight, and are exposed to cosmic radiation in space. Hence they may be expected to contain highly elevated 14C/12C ratios, as compared to the Earth's atmosphere. The significant input of 14C by comets may have contributed to the fluctuation of 14C in the atmosphere throughout the Earth's history, which should be considered carefully to better constrain the cosmic ray fluctuation.


Carbon-14 (14C) is a cosmogenic isotope of C formed on Earth primarily through radiation of atmospheric nitrogen by the reaction:14N(n,p)14C (refs. 1–4). Its abundance in the atmosphere varies with time5, which is generally attributed to variations in the earth's magnetic field, solar activity and changes in the carbon cycle6. A large and sudden increase in 14C of ~12‰ was reported from a tree ring study in Japan to have occurred between AD 774 and AD 775 (hereafter M12)7. Their modeling showed that the atmospheric level of 14C must have jumped over the course of no longer than a year, corresponding to an increase 10 times larger than the average production from Galactic cosmic rays and 20 times larger than that expected over 2 × 11 yr solar cycles. The measured values were shown to be too large for a solar flare or local supernova. Given that no detectable increase in 14C corresponding to supernovas SN 1006 and SN 1054 were observed7,8, it is argued that much higher energies would be required for the M12 event, if it is related to a supernova7. Alternative explanations for this mysterious 14C elevation include a highly energetic radiation burst, e.g., proton storms from giant solar flares9,10, a giant cometary impact upon the Sun11, or floods of γ-rays from supernova explosions12. Such high levels of radiation however, might also cause mass extinctions13, which are absent following the M12 event. Moreover, it has been argued, based on historical records, that no superflares have occurred in the Sun during the last two millennia14.

A simulated carbon cycle model10 suggested that the strength of the M12 event was significantly overestimated by the previous study7. One key issue is the duration of the 14C input. Based on modeling, it has been proposed that a tree ring record of the event could be explained by a spike in 14C production that lasted less than 1 year7. However, owing in part to the annual resolution of the 14C data, they could not assess the duration in more detail7. Porites coral with an annual growth rate ≥ 10 mm/yr has now provided a high temporal-resolution 14C record15.

One 1.2-m fossil Porites coral, XDH, was drilled from the Xiaodonghai Reef (18°12.46′N, 109°29.93′E) from the northern South China Sea in 1997. We analysed 14C contents for half-annual-resolution subsamples at depths of 1.04–42.65 cm and ~2-year biweekly-resolution subsamples at depths of 12.25–17.19 cm (Fig. 1, Table S2 and S3).

Figure 1: Measured radiocarbon content in coral and trees7.
Figure 1

The concentration of 14C is expressed as Δ14C. For trees, Δ14C is the deviation (in‰) of the14C/12C ratio of a sample with respect to modern carbon (standard sample) after correcting for the age and isotopic fractionation7; For coral, Δ14C is the direct deviation (in‰) of the14C/12C ratio of a sample with respect to modern carbon after isotopic fractionation correction. (a) Half-annual and (b) Biweekly resolution record of Δ14C in the Porites coral (open blue circles with error bars) from the South China Sea (SCS) from this study. (c) Comparison of coral δ18O (solid green circles) [plotted to identify the seasonal cycle of sea surface temperature (SST), with the maximum seasonal coral δ18O value corresponding to February, the coldest month at our sample site]. (d) Annual to biennial resolution record of Δ14C in two cedar trees (open light blue diamonds with error bars) from Japan7. The vertical pink bars indicate the mysterious 14C increase event (M12). Japanese tree data is plotted on their original time scales while the 230Th age of our coral data is shifted, within quoted errors, 3.5 years young to correlate with the event. As indicted by our high-resolution coral record, this event happened in the winter, which is consistent with a big Comet event (Dai7). The thick red lines in (a) and (b) indicate the average values before and after M12. The gray triangle in (a) indicates position of 230Th-dated layer. Note that the δ18O in (c) is plotted with a reverse axis.


The 14C increased by ~ 15‰ in the winter of AD 773 and remained roughly constant for ~ 4 months, and then jumped up by another ~45‰ within four weeks and then dropped down in late spring, forming a spike of 45‰ high. This is followed by two smaller spikes of > 20‰ over the next 6 months until fall, and then maintained ~15‰ higher than normal values over the following several months (Fig. 1b). We obtained a 230Th date of AD 783 ± 14 (table S1) at a depth of 2.15 cm, which is 7 annual growth bands above the layer containing the onset of 14C anomalies at a depth of 16.11 cm and corresponding to an age of AD 776 ± 14. When the previously published tree ring spectrum7 was examined, the 14C content had actually started to climb in AD 773 (Fig. 1d). There are no other 14C increases until 200 yrs later16. Considering dating errors, the major 14C increases we observed are also likely to have occurred in AD 773 (Fig. 1a).


The coral 14C spectrum shown in Fig. 1 is difficult to be explained using normal production pathways from Galactic cosmic rays. The abrupt 14C increase by ~45‰ within two weeks (Fig. 1b) requires a radiation intensity 100 times stronger than the previous estimation for M12. Since the residence time of carbon dioxide in the atmosphere is 5–15 years17,18, 14C spikes in coral suggest highly uneven distribution. It is well established that a comet collided with the Earth's atmosphere from constellation Orion (or Shen in traditional Chinese astronomy) on 17 January AD 773, the 7th year of Emperor Dai Zong of the Tang Dynasty. The phenomenon (hereafter Dai7) lasted less than one day and had an accompanying coma that stretched across the whole sky19,20. “Dust rain” in the daytime before the “comet” implies that a considerable amount of cometary material was added to the atmosphere assuming these two events are associated. Celestial observations were especially significant to the emperors of ancient China, especially in the Tang Dynasty, and these were carefully recorded. This event was recorded in several different official archives in China19,20, included by royal celestial officers in Chang'an (now Xi'an), the capital city of the Tang dynasty (34°16′N, 108°54′E).

It is quite possible that Dai7 resulted in the M12 global abrupt 14C increases recorded in tree rings and corals. Comas are known to have percent levels of nitrogen by weight (in the forms of NH3, NH2, NH, etc)21,22, and are heavily exposed, as compared to nitrogen within the earth's atmosphere because of lacking a magnetic field protection23. Considering that meteorite usually has 14C and10Be about two orders of magnitude higher than those of rocks from the Earth's surface24,25,26, it is reasonable to propose that coma and comet may be expected to have 14C/12C ratios several orders of magnitude higher than that of the Earth's atmosphere23. Generally, 14C occurs in very low concentrations in the Earth's atmosphere, i.e., no more than one part per trillion of the total carbon content of the atmosphere27. The total amount of preindustrial 14C in the atmosphere was ~150 metric tonnes. Assuming an average 14C/12C ratio of 1 × 10−6 in the Dai7 comet, ~150 million metric tonnes of C from the Dai7 event would double the 14C content of the Earth's atmosphere. Assuming a C abundance of 10% in the comet, a total of ~30–150 million metric tonnes of materials would then be required to explain the 14C anomalies. This is only about 1–3% of the estimated total mass-loss of Haley's Comet in 1910 (ref. 28). With the considerable uncertainties surrounding the dispersal of cometary material throughout the atmosphere and shallow oceans, such a process seems commensurate with the observed 14C increases (Fig. 1).

The coma 14C would have been dispersed into the Earth atmosphere heterogeneously (Fig. 2). Because the coma is far better exposed to cosmic radiations than the nucleus, it should have a much higher 14C/12C ratio. A considerable proportion of the coma with its higher 14C/12C content is probably scattered and absorbed into the outer atmosphere. The bulk of the cometary material with14C/12C values that are much lower than that of the coma, but still considerably higher than the Earth's atmosphere, may be expected to descend into the troposphere and become incorporated into corals and trees. Four months later, the high14C/12C material captured in the outer atmosphere (stratosphere) mixes downward into the troposphere, a process facilitated by summer storms, and is absorbed by corals, resulting in their high and fluctuating 14C spikes in coral (Fig. 1b). After another six months, the enriched 14C material becomes well mixed and imparts elevated 14C levels to the whole atmosphere (Fig. 1b).

Figure 2: A cartoon illustrating our proposed mechanism causing a 14C spike--the collision of the Dai7 “Comet” with high 14C and 10Be contents with the Earth's atmosphere.
Figure 2

As it descends, 14C and10Be is released until the comet burns out. This spike of cosmogenic 14C is first added to the atmosphere with its originally very low 14C, and the additional carbon is then incorporated into coral from the South China Sea and Japanese trees. The original record of the Dai7 “Comet” event (in Chinese with translation) is also shown in the lower left corner of the cartoon. Photos are provided by Yi Liu.

Consistent with the 14C increase, there was a 30% increase in the decadal10Be flux record in Dome Fuji from AD 755 to 785 (refs. 7,16,29), which has been attributed to a burst of high energy γ-rays12. We were not able to obtain 10Be data in this study. Nevertheless,10Be is another cosmogenic isotope formed through spallation of nitrogen12,14N(n,p + α)10Be, or oxygen, which often co-varies with 14C. The increase in 10Be, can also be interpreted by the Dai7 event. The comet with abundant oxygen and nitrogen, could likewise produce high amounts of 10Be under exposure to cosmic radiation.

As an alternative, short radiation bursts, e.g., the merger of two magnetized neutron stars, can produce a spinning black hole and launch a relativistic energy jet as observed in short γ-ray bursts30 that might also explain the brief input of 14C and10Be (ref. 12). This could conceivably produce an interaction between the short γ-ray burst and the magnetic field of the Earth which might appear to be a comet. However, the γ-ray burst is fast and interacts with the entire magnetic field of the earth in seconds; therefore it is not easily explained as having “entered from the constellation of Shen (Orion)”19,20. It is also difficult to explain the ‘dust rain” beforehand, unless the dust rain was only a coincidence.

It has long been recognized that 14C and10Be in the Earth's atmosphere varied dramatically throughout the history of the Earth5,16,31, which has previously been solely attributed to cosmic radiations1,2,3,4,13. The coincidence of Dai7 and the 14C,10Be spikes in tree rings and coral suggests that comets might also contributed significant amount of 14C to the Earth's atmosphere episodically.


Coral core

A 1.2 - m long core of fossil Porites coral XDH was drilled from Xiaodonghai Reef in the northern South China Sea in 1997. Slabs of 7 mm in thickness, were sectioned, washed with ultrapure water, and dried for X-ray images. X-ray diffraction analysis shows our coral samples are 100% aragonite and scanning electron microscopy image indicates the absence of secondary aragonite around the coral part having the 14C spike. The subsamples were crushed and homogenized one by one in an agate mortar.


Sample XDH-2 at depth of 2.15 cm was dated by 230Th techniques32 in the High-Precision Mass Spectrometry and Environment Change Laboratory (HISPEC), at National Taiwan University, on a multi-collector inductively coupled plasma mass spectrometer (MC-ICP-MS) (Table S1).

Carbon-14 sample preparation was carried out in the State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry. About 8–9 mg coral sample power was weighed and put in a special reaction quartz tube reacted with purified H3PO4 for more than 24 hours at room temperature after being kept continuously in a 1.0 × 10−3 torr vacuum system for at least 4 hours. CO2 from the reaction tube is purified and then transferred to a tube and graphitized33. The graphite samples were analyzed in the AMS laboratory at Peking University34, the standards used during the analysis are NIST OXI and OXII, the analytic precision for our samples are better than 3‰ and 5‰ for half-annual and biweekly samples, respectively.

δ18O measurements from the same biweekly subsamples were carried out using MAT-252 mass spectrometry equipped with Kiel II micro carbonate automatic sample input device at the Institute of Earth Environment, Chinese Academy of Sciences. The results are expressed in the delta (δ) notation relative to the Vienna Pee-Dee Belemnite (V-PDB) standard. The analytical error of the laboratory standard is approximately ± 0.2‰ for δ18O (ref. 35).


  1. 1.

    , , , & Radiocarbon production by the gamma-ray component of supernova explosions. Radiocarbon 37, 599–604 (1995).

  2. 2.

    & Radiocarbon calibration and application to geophysics, solar physics, and astrophysics. Radiocarbon 42, 137–150 (2000).

  3. 3.

    , , , & Solar proton events in cosmogenic isotope data. Geophys. Res. Lett. 33,10.1029/2006gl026059, doi:10.1029/2006gl026059 (2006).

  4. 4.

    Core-collapse supernovae and the Younger Dryas/terminal Rancholabrean extinctions. Icarus 215, 101–106, 10.1016/j.icarus.2011.06.043 (2011).

  5. 5.

    et al. INTCAL98 radiocarbon age calibration, 24,000-0 cal BP. Radiocarbon 40, 1041–1083 (1998).

  6. 6.

    Causes of Temporal 14C Variations, in Encyclopedia of Quaternary Science. Scott A. (ed) 2931–2940 (Elsevier, Oxford, England 2007).

  7. 7.

    , , & A signature of cosmic-ray increase in AD 774–775 from tree rings in Japan. Nature 486, 240–242, 10.1038/nature11123 (2012).

  8. 8.

    et al. in Proc. 29th Cosmic Ray Conf. Vol. 2, Acharya B. S. (ed) 357–360 (Tata Institute of Fundamental Research, Mumbai, 2005).

  9. 9.

    , , & Terrestrial effects of possible astrophysical sources of an AD 774–775 increase in C-14 production. Geophys. Res. Lett. 40, 1237–1240, 10.1002/grl.50222 (2013).

  10. 10.

    et al. The AD775 cosmic event revisited: the Sun is to blame. Astron. & Astrophys. 552, 10.1051/0004-6361/201321080, doi:10.1051/0004-6361/201321080 (2013).

  11. 11.

    & Comet encounters and carbon 14. Astrophys. J. Lett. 761, 10.1088/2041-8205/1761/1082/l1027, doi:10.1088/2041-8205/761/2/l27 (2012).

  12. 12.

    & A Galactic short gamma-ray burst as cause for the C-14 peak in AD 774/5. Month, Not, Royal Astron. Soc. 430, 32–36, 10.1093/mnras/sts378 (2013).

  13. 13.

    Evidence for a solar flare cause of the Pleistocene mass extinction. Radiocarbon 53, 303–323 (2011).

  14. 14.

    , & Superflares on ordinary solar-type stars. Astrophys. J. 529, 1026–1030, 10.1086/308325 (2000).

  15. 15.

    et al. Coral records of central tropical Pacific radiocarbon variability during the last millennium. Paleoceanography 25, Doi 10.1029/2009pa001788, doi:Artn Pa4212, Doi 10.1029/2009pa001788 (2010).

  16. 16.

    , & Another rapid event in the carbon-14 content of tree rings. Nat. Comm. 4, 10.1038/ncomms2783 (2013).

  17. 17.

    IPCC. In Climate Change 2007: The Physical Science Basis. Working Group 1 Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Solomon S., Qin D., Manning M., Marquis M., Averyt K., Tignor M. M. B., Miller Jr H. L. & Chen Z. (eds) (Cambridge University Press, Cambridge 2007).

  18. 18.

    On the Atmospheric Residence Time of Anthropogenically Sourced Carbon Dioxide. Energ Fuel 25, 5503–5513, Doi 10.1021/Ef200914u (2011).

  19. 19.

    , , & (eds) Jiutangshu Vol. 36, 1327 (Royal History, Hou Jin Dynasty, Kaifeng, China 945).

  20. 20.

    & (eds) Compilation of Astronomic Records of Ancient China. 408 (Science and Technology Press of Jiangsu Province, 1988).

  21. 21.

    & NH2 and NH spatial intensity distribution in the coma of Halley's comet. Adv. Space Res. 40, 155–159, DOI 10.1016/j.asr.2007.04.024 (2007).

  22. 22.

    , & Composition of the volatile material in Halley's coma from in situ measurements. Space Sci Rev 90, 3–18, Doi 10.1023/A, 1005256607402 (1999).

  23. 23.

    & Cosmogenic nuclide enhancement via deposition from long-period comets as a test of the Younger Dryas impact hypothesis. Earth Planet Sc Lett 377–388, 55–61 (2013).

  24. 24.

    Terrestrial ages of meteorites. In Meteorites and the early solar system II. Lauretta D., & McSween Jr H. Y. (eds), 889–905 (The University of Arizona Press, Tucson, AZ 2006).

  25. 25.

    , , & Measurements of in situ produced C-14 in terrestrial rocks. Nuclear Instru & Methods in Physics Res Section B-Beam Interact Materials Atoms 223, 253–258, DOI 10.1016/j.nimb.2004.04.051 (2004).

  26. 26.

    et al. A multi-radionuclide approach for in situ produced terrestrial cosmogenic nuclides: Be-10, Al-26, Cl-36 and Ca-41 from carbonate rocks. Nuclear Instru & Methods in Physics Res Section B-Beam Interact Materials Atoms 268, 1179–1184, DOI 10.1016/j.nimb.2009.10.128 (2010).

  27. 27.

    ANALYTICAL CHEMISTRY Ultrasensitive radiocarbon detection. Nature 482, 312–313 (2012).

  28. 28.

    The Size, Mass, Mass-Loss and Age of Halleys-Comet. Month Not Royal Astronom Soc 213, 103–109 (1985).

  29. 29.

    et al. Ice core record of Be-10 over the past millennium from Dome Fuji, Antarctica: A new proxy record of past solar activity and a powerful tool for stratigraphic dating. Quater Geochron 3, 253–261, 10.1016/j.quageo.2008.01.003 (2008).

  30. 30.

    et al. The Missing Link: Merging Neutron Stars Naturally Produce Jet-Like Structures and Can Power Short Gamma-Ray Bursts. Astrophys J Lett 732, Doi 10.1088/2041-8205/732/1/L6 (2011).

  31. 31.

    et al. Intcal09 and marine09 radiocarbon age calibration curves, 0–50,000 year cal BP. Radiocarbon 51, 1111–1150 (2009).

  32. 32.

    et al. High-precision and high-resolution carbonate Th-230 dating by MC-ICP-MS with SEM protocols. Geochim Cosmochim Ac 99, 71–86, DOI 10.1016/j.gca.2012.09.018 (2012).

  33. 33.

    et al. Modifying a sealed tube zinc reduction method for preparation of AMS graphite targets: Reducing background and attaining high precision. Nuclear Instru & Methods in Physics Res Section B-Beam Interact Materials Atoms 259, 320–329, DOI 10.1016/j.nimb.2007.01.175 (2007).

  34. 34.

    et al. A new compact AMS system at Peking University. Nuclear Instru & Methods in Physics Res Section B-Beam Interact Materials Atoms 259, 23–26, DOI 10.1016/j.nimb.2007.01.314 (2007).

  35. 35.

    et al. Monsoon precipitation variation recorded by tree-ring delta O-18 in arid Northwest China since AD 1878. Chem Geol 252, 56–61, DOI 10.1016/j.chemgeo.2008.01.024 (2008).

Download references


This work was supported by Natural Science Foundation of China (No. 41090374), National Key Basic Research Program of China (No. 2013CB956102), Natural Science Foundation of China (No. 41121002 and 41003002) and State Key Laboratory of Isotope Geochemistry grants (SKLIG-KF-12-01, SKLIG-KF-12-02 and SKLIG-JY-12-01). National Science Council, and National Taiwan University grants (101-2116-M-002-009, 102-2116-M-002-016, and 101R7625). Thanks to Drs Chung-Che Wu and Ping Ding for assistant in sample analyses. This is contribution No. IS-1794 from GIGCAS.

Author information

Author notes

    • Yi Liu
    •  & Zhao-feng Zhang

    These authors contributed equally to this work.


  1. CAS Key Laboratory of Crust-Mantle Materials and Environments, School of Earth and Space Sciences, University of Science and Technology of China, Hefei 230026, China

    • Yi Liu
    •  & Zi-cheng Peng
  2. State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, The Chinese Academy of Sciences, Guangzhou 510640, China

    • Zhao-feng Zhang
    • , Ming-xing Ling
    •  & Cheng-de Shen
  3. High-Precision Mass Spectrometry and Environment Change Laboratory (HISPEC), Department of Geosciences, National Taiwan University, Taipei 10617, Taiwan

    • Chuan-Chou Shen
  4. State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, The Chinese Academy of Sciences, Xi'an 710075, China

    • Wei-guo Liu
  5. Institute for the History of Natural Sciences, The Chinese Academy of Sciences, Beijing 100190, China

    • Xiao-chun Sun
  6. State Key Laboratory of Nuclear Physics and Technology & Institute of Heavy Ion Physics, School of Physics, Peking University, Beijing 100871, China

    • Ke-xin Liu
  7. CAS Key Laboratory of Mineralogy and Metallogeny, Guangzhou Institute of Geochemistry, The Chinese Academy of Sciences, Guangzhou 510640, China

    • Weidong Sun


  1. Search for Yi Liu in:

  2. Search for Zhao-feng Zhang in:

  3. Search for Zi-cheng Peng in:

  4. Search for Ming-xing Ling in:

  5. Search for Chuan-Chou Shen in:

  6. Search for Wei-guo Liu in:

  7. Search for Xiao-chun Sun in:

  8. Search for Cheng-de Shen in:

  9. Search for Ke-xin Liu in:

  10. Search for Weidong Sun in:


Y.L., W.D.S., Z.F.Z. and Z.C.P. designed and initiated the research. C.D.S. and K.X.L. analysed 14C. Y.L., C.C.S. and W.G.L. analysed 230Th age and O isotopes. Y.L., W.D.S., Z.F.Z. and M.X.L. plotted all the figures. X.C.S. provided information on the AD 773 Comet. W.D.S., Y.L., Z.F.Z. and M.X.L. prepared the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to Weidong Sun.

Supplementary information

Word documents

  1. 1.

    Supplementary Information

    Mysterious abrupt carbon-14 increase in coral contributed by a comet

About this article

Publication history






By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.