Long-lasting Extreme Magnetic Storm Activities in 1770 Found in Historical Documents

Hisashi Hayakawa; Kiyomi Iwahashi; Yusuke Ebihara; Harufumi Tamazawa; Kazunari Shibata; Delores J. Knipp; Akito D. Kawamura; Kentaro Hattori; Kumiko Mase; Ichiro Nakanishi; Hiroaki Isobe

doi:10.3847/2041-8213/aa9661

1. Introduction

Solar eruptions can severely impact the geospace environment and human activities (Schwenn 2006; Cannon et al. 2013; Knipp et al. 2016; Oughton et al. 2016). The so-called Carrington space weather event in 1859 is considered to be the most extreme in the history of telescopic observations (Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Cliver & Dietrich 2013; Hayakawa et al. 2016a). Modern civilization heavily relies on satellites and large-scale power grids. If such events were to strike the Earth now, the consequences could be catastrophic (Schwenn 2006; Cannon et al. 2013; Knipp et al. 2016; Oughton et al. 2016). Understanding the occurrence frequency and upper intensity limit of solar flares and resulting magnetic storms is therefore essential, although the short history of modern scientific observations and the rarity of such events make this difficult (Riley & Love 2017). Recent studies suggest that the Sun may be capable of producing much stronger space weather events than modern society has experienced; for example, numerous "superflares" have been found in solar-like stars that are orders of magnitude more energetic than the strongest solar flares ever recorded (Maehara et al. 2012). Independently, sharp spikes in the cosmogenic isotope have been found in tree rings, which suggest an extraordinarily large cosmic-ray flux, possibly originating from intense solar flares (Miyake et al. 2012; Mekhaldi et al. 2015).

The Carrington flare, observed simultaneously by Carrington (1859) and Hodgson (1859), has been considered a benchmark event in the study of extreme space weather (Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Cliver & Dietrich 2013; Hayakawa et al. 2016a). This flare produced a great magnetic storm and auroral activity. The aurora was witnessed in very low-latitude regions such as Chile, Hawaii, the Caribbean Coast, and Japan, resulting in numerous records of diverse kinds. The aurora was witnessed as low as about 20° ∼ 23° in dipole magnetic latitude (MLAT; Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Cliver & Dietrich 2013; Hayakawa et al. 2016a).

A large magnetic storm is often a consequence of a solar coronal mass ejection (CME; Tsurutani et al. 2003), although there is not a one-to-one correspondence because the southward component of an interplanetary magnetic field is not always embedded in the sheath fields or interplanetary CME clouds (Cliver & Dietrich 2013). Measuring the strength of the magnetic storms is challenging when geomagnetic field data are limited, or unavailable (Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Cliver & Dietrich 2013). Based on observations, the latitudinal extent of the auroral oval is correlated with worldwide geomagnetic disturbances (Schultz 1997; Yokoyama et al. 1998); thus, the equatorward extent of the auroral oval estimated from historical records has been used as a proxy for measuring magnetic storm intensity.

2. Methods and Source Documents

2.1. Auroral Records in Historical Documents

Aurora-like records in East Asian historical documents are frequently described as luminous phenomena during the night, namely vapor, light, cloud, and so on (Osaki 1994; Watanabe 2007; Vaquero & Vázquez 2009; Zhuang et al. 2009; Kawamura et al. 2016; Tamazawa et al. 2017). For Chinese documents, we examined official histories from the central government and many local treatises in various regions of China, which involve aurora-like records in their chapters of omens. We found 22 records of relevant auroral observations. From Korea, we consulted governmental diaries, such as the Ilseongnok and Seungcheonwon Ilgi, written in the palace of the Joseon dynasty with detailed descriptions of weather and astronomical observations. However, we found no relevant records in Korea, likely due to bad weather as previously concluded by Willis et al. (1996). According to the Seungcheonwon Ilgi (v.73, pp. 180–184), it was cloudy on September 16, cloudy in the morning and rainy in the evening on September 17, rainy on September 18, no weather record on September 19, and cloudy on September 20. From Japan, we consulted contemporary diaries written by people in various social classes, finding 88 relevant records. Most diaries were written by observers themselves, making their descriptions more reliable. All references for original source documents are listed in the Appendix A.2.

2.2. Contemporary Sunspot Drawings

As for contemporary sunspot drawings, we consulted a series of the sunspot drawings by Johann Caspar Staudacher covering the period from 1749 February 15 to 1796 January 31, which have been preserved, and were digitized by Arlt (2008). Here, we find great sunspot groups (active regions) from 1770 September 12 to 22. The area of sunspot groups on September 16 is measured up to 6000 millionths of hemispheres of the Sun. Using Equation (1) in Shibata et al. (2013), the upper limit of the flare energy from this sunspot group can be estimated at about 10³⁴ erg.

2.3. Computing the Magnetic Latitude of Observational Sites

The extent of aurora in terms of the lowest MLAT is a proxy for the strength of the magnetic storm (Schultz 1997; Yokoyama et al. 1998). To estimate the strength of the magnetic storm, we computed contemporary MLAT of observational sites, defined by the angle between the observational site and the geomagnetic equator. The geomagnetic equator is the great circle of the Earth whose plane is perpendicular to the axis of Earth's dipole field. We calculated the location of the contemporary magnetic pole (the axis of Earth's dipole field) with the spherical harmonic coefficients provided for the geomagnetic field model GUFM1 covering the recent four centuries (Jackson et al. 2000). We used the MLAT derived based on the dipole component of the geomagnetic field, which is called dipole MLAT, unless otherwise mentioned.

3. Results

Witnessing the 1859 aurora, a Japanese chronicler noticed that the event was similar to the event on 1770 September 17 (Hayakawa et al. 2016a). The 1770 aurorae, which occurred just one year after the solar cycle 2 maximum in 1769 (Clette et al. 2014), were very prominent in Japan (Willis et al. 1996; Nakazawa et al. 2004). Figure 1 and Appendix Figures 5(a) and (b) show contemporary drawings of the aurora in a vivid red color covering a wide area of the sky. The 1770 aurorae were also observed in the southern hemisphere by Joseph Banks and Sydney Parkinson on board HMS Endeavour as specialist members of Captain Cook's crew, and hence known to be the earliest record of simultaneous auroral observations in both hemispheres (Banks 1962; Parkinson 1773; Willis et al. 1996). Since the East Asian observations have not been compiled to study the overall scale of the event, we survey these contemporary historical documents to collect 111 relevant historical records from the 1770 events and compare them with the Carrington event in terms of the auroral visibility.

**Figure 1.** J091762 = MS Special 7–59, National Diet Library, ff. 6b–7a (at Nagoya): corresponding to record J091762 in Table 1 in the Appendix . (Courtesy: the National Diet Library).

Download figure:
Standard image High-resolution image

Figure 2 shows the locations of auroral observations during 1770 September 16–18 together with the MLAT contours calculated using the magnetic field model GUFM1 (Jackson et al. 2000). The most equatorward observation was at Dòngtínghú 18 fdg 8 MLAT (C091707: N28°51', E112°37') on September 17 and the second-most was near Timor Island, −20 fdg 6 MLAT, by specialist members of Captain Cook's crew on September 16 (CJC0916: S10°27', E112°49'). Thus, the latitudinal extents of 1770 auroral events were at least comparable with those of the Carrington event.

fdg — **Figure 2.** Locations of the aurora observations during 1770 September 16–18. According to the historical magnetic field model GUFM1, the magnetic north pole was at N 798, E 3034 in 1770. The date, color, term, direction, duration, observational site, geographical coordinate, geomagnetic latitude, and bibliography of each record are summarized in the tables in the Appendix .

Download figure:
Standard image High-resolution image

Of particular interest is the record by Captain Cook (CJC0916; Willis et al. 1996) at −20 fdg 6 MLAT with the angular height of the aurora "reaching in height about twenty degrees above the horizon." Assuming that the upper part of the highly visible red aurora reached 300 km altitude, the equatorward edge of the aurora oval can be estimated to 27° MLAT at 300 km altitude. The magnetic footprint on the ground is located at 29° MLAT. (We omitted the sign because of north–south symmetry, and used the dipole magnetic field.) It is expected that at 27°–29° MLATs, the reddish aurora spread over the sky including the zenith. This is supported by the record in China at 27 fdg 1 MLAT (C091605), stating that "red light crossed the heaven" on the same day.

Figure 3 shows the distribution of MLATs of auroral observations as a function of time. On September 17, the aurora was observed at many points ranging from 18 fdg 8 to 31 fdg 6 MLAT. It is important to note that the aurorae were witnessed at MLAT < 30° almost continuously during September 10–19, except for September 12.

Figure 4 shows the sunspot drawings from 1770 September 14–18 by Staudacher (Arlt 2008), which depict an extremely large and complex sunspot group. This large sunspot was even observed by naked eye in Japan (J091713) around 1770 September 17. From the drawing on September 16, for example, the corrected sunspot area is measured up to 6000 millionths of the Sun's visible hemisphere, more than twice the size of the sunspot group during the Carrington event in 1859 (Cliver & Dietrich 2013; Hayakawa et al. 2016a) and comparable to the largest known sunspot group in 1947 April (Newton 1955; Koyama 1985; Knipp et al. 2017). Nevertheless, we admit that the accuracy or level of detail of Staudacher's drawings still requires special attention, as noted by Arlt (2008).

4. Discussions

The low-latitude aurorae (MLAT < 30°) were almost continuously observed for nine nights. The duration is exceptionally long. In many magnetic storms, the duration is only one or two nights (Shiokawa et al. 2005) because the magnetic storm usually lasts for one to two days. In the Carrington event, the low-latitude aurorae (MLAT < 30°) were observed on 1859 August 28–29 and September 1–2 (Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Cliver & Dietrich 2013). The long duration of the auroral observations in 1770 indicates long-lasting magnetic storm activities resulting from continuous solar activity, such as multiple, consecutive CMEs possibly originating from the same sunspot group. It is known that active and large sunspot groups produce flares repeatedly (Kimball 1960; Tsurutani et al. 2003; Green & Boardsen 2006; Tsurutani et al. 2008; Cliver & Dietrich 2013; Hayakawa et al. 2016a). When huge CMEs are consecutively ejected, the first one tends to be decelerated by drag force. The subsequent CMEs, however, should experience less drag, so that they tend to maintain their speed. Such consecutive solar eruptions occurred in 2012 July when a series of earlier eruptions removed the pre-existing solar wind, and the latter eruptions moved through the density depletion region. Complex CME–CME interactions also produce extremely enhanced the magnetic field. The 2012 July ejecta fortunately did not hit the Earth. If they had hit the Earth, they could have caused severe magnetic storms comparable to that of the Carrington event (Liu et al. 2014). Consecutive solar eruptions might have caused the long-lasting magnetic storm activity that occurred in 1770 September. The low-latitude aurorae in 1770 September are most likely caused by equatorward displacement of the auroral oval in association with severe enhancement of the magnetospheric convection. The magnetospheric convection might be powered by subsequent CMEs. On 1770 September 17, one of the consecutive solar eruptions, or a product arising from complex CME–CME interactions might further power the magnetospheric convection, so as to result in a fantastic aurora at extremely low latitude, as low as ∼18 fdg 8 MLAT. This auroral display was extreme in terms of its brightness as well as examined in Ebihara et al. (2017).

Sunspot location is an important factor for the geo-effectiveness of solar eruptions (Freed & Russell 2014). From Staudacher's drawings, the large sunspot group became recognizable near the east limb on September 12. On the other hand, as shown in Figure 3, the low-latitude aurorae were observed even before September 12. These early aurorae were likely not associated with the large sunspot group. However, the sunspot drawing on September 12 shows moderate size sunspots near the disk center and in the western hemisphere. We speculate that eruptions from these sunspot groups, or perhaps from filaments associated with these emerging spots, produced the aurorae observed before September 12. From the drawings, one can see that the large sunspot group was located eastward from the disk center on September 15–17, which is a reasonable location for CMEs to hit the Earth and produce magnetic storms.

The large sunspot group and the series of low-latitude aurorae may have been a part of the longer enhanced activity. Low-latitude aurorae with an approximate 27-day return period were also observed on 1769 October 24–26 and on 1770 January 18 in the Iberian Peninsula (Aragonès & Ordas 2010), and in 1770 October 15–16 in East Asia as well (Tables 1 and 2). Active regions have been traced for up to 10 months during solar minimum and about 4 months during solar maximum (Schrijver & Harvey 1994). Furthermore, active regions near the equator have long correlation times with active complexes and active longitudes having even greater longevity (Pelt et al. 2010). The events of 1770 September may have been extreme manifestations of an active complex that existed from 1769 October to 1770 October.

5. Conclusion

In summary, we identified a series of low-latitude auroral records from East Asian historical documents in 1770 September and October, likely suggesting great magnetic storms. Investigating the sunspot drawings by Staudacher (Arlt 2008), we found a huge sunspot group up to 6000 millionths of hemispheres of the Sun in mid-September, twice as large as that related to the Carrington event of 1859. The lowest magnetic latitude of the auroral observations in the 1770 aurorae was 18 fdg 8 MLAT, at least comparable or perhaps lower than the Carrington event. In addition, these aurorae were almost continuously observed in the low-latitude regions (<30°MLAT) for nine nights, while the Carrington event aurorae were intermittently seen on August 28–29 and September 1–2. This huge sunspot group and resultant magnetic storms provide context for the longevity of enhanced solar activity that lasted at least one year. Therefore, we conclude that in comparison to the Carrington event, the scale of the magnetic storm is comparable as inferred by magnetic latitude of auroral visibility, but the duration of the storm activity was much longer than usual.

In 2012 July, a possible Carrington-class CME fortunately missed the Earth (Baker et al. 2013; Liu et al. 2014). Historical evidence shows that extreme Carrington-class storms do occur repeatedly, as already recognized by the Japanese chronicler who noticed the similarity of the 1770 and 1859 aurorae (Hayakawa et al. 2016a). The timing of the 1770 event is also consistent with the apparent 60–100 year period of extreme solar flares known from their resultant great magnetic storms in 1859, 1921, and 1989 (Cliver & Dietrich 2013), or the return period of the 90 ± 60 year Carrington-like flare estimated by an empirical relationship between the variation of radiation and the magnetic storm and the assumption of X45 class (Curto et al. 2016). Historical documents allow us to trace solar activity back for millennia (Stephenson et al. 2004; Vaquero & Vázquez 2009; Hayakawa et al. 2016b), and possibly open new doors to determine the occurrence probability, longevity, and intensity of great magnetic storms caused by extreme solar flares.

We acknowledge the Supporting Program the "UCHUGAKU" project and RISH (2016, 2017) and SPIRITS (2017) of the Kyoto University, the CPIS of SOKENDAI, and Grants-in-Aid from the MEXT of Japan (JP15H05816, JP15H03732, JP16H03955, JP15H05815, and JP17J06954). D.J.K. is partially supported by AFOSR grant FA9550-17-1-0258. We thank R. Arlt, Tohoku University Library, National Diet Library of Japan, Shizuoka Municipal Library, and Hosa Library for permissions to reproduce Staudacher's sunspot drawings, auroral drawings, and relevant folios. We also thank Y. Watanabe, M. Abe, and H. Miyahara for valuable advices, T. Iyemori, K. Ichimoto, E.W. Cliver, and B.T. Tsurutani for their preliminary review of our manuscript, and an anonymous referee for helpful recommendations.

H.H. analyzed historical documents, surveyed contemporary sunspot drawings, and prepared this manuscript. K.I., K.H., K.M., and I.N. analyzed Japanese historical documents. Y.E. analyzed these historical documents in relation to aurora science. H.T. analyzed sunspot drawings in relation to solar physics. K.S. offered advice on discussions of solar physics. D.J.K. analyzed the longevity of sunspot active regions, A.D.K. examined nitrate signals in relation to coronal mass ejections and solar energetic particles. H.I. supervised the study and contributed to the scientific discussion.

Appendix:

A.1. Tables of Contemporary Observations

In these tables, we summarize the auroral records during September and October in 1770 that we collected from contemporary East Asian historical documents. We show their ID, observational date (year, month, and date), color, duration (start and end), geographical latitude and longitude of observational sites (G. lat. and G. long.), and contemporary magnetic latitude. As the contemporary East Asians calculate time dividing a day into 12 double-hours, in this table, we showed their central value for convenience. Note that some of these historical documents have already been summarized in previous studies (e.g., Willis et al. 1996; Nakazawa et al. 2004; Ebihara et al. 2017) and previous catalogs (e.g., Osaki 1994; Watanabe 2007; Zhuang et al. 2009; Kawamura et al. 2016; Tamazawa et al. 2017), while we show the results of surveys on the original historical documents. Table 1 lists contemporary observations in China and Table 2 lists contemporary observations in Japan. Table 3 lists contemporary observations by Captain Cook. Here we define observational dates from 06:00 to 30:00 (06:00 on the following day) and describe the observational time beyond the midnight continuously as local time + 24 in the same day (e.g., 26:00 on September 17 for 02:00 on September 18), in order to categorize the starting time and ending time of auroral displays as a contiguous record, as many auroral observations lasted beyond midnight. The abbreviations of the colors are as follows: R (red), W (white), P (purple), B (blue), Y (yellow), and Gl (gold).

Table 1. Contemporary Observations in China

ID	Year	Month	Date	Color	Start	End	G. lat.	G. long.	MLAT
C091101	1770	Sep	11	R	20:00	24:00	N36°50'	E112°51'	26.8
C091501	1770	Sep	15	R			N36°07'	E112°52'	26.1
C091601	1770	Sep	16	five-color			N39°53'	E118°52'	29.7
C091602	1770	Sep	16	five-color			N39°45'	E118°45'	29.6
C091603	1770	Sep	16	five-color			N40°02'	E117°25'	29.9
C091604	1770	Sep	16	five-color			N39°43'	E119°10'	29.5
C091605	1770	Sep	16	R			N37°16'	E118°53'	27.1
C091701	1770	Sep	17	R		24:00	N36°11'	E116°46'	26.0
C091702	1770	Sep	17	R, W		26:00	N35°30'	E117°38'	25.4
C091703	1770	Sep	17	W		24:00	N36°11'	E116°46'	26.0
C091704	1770	Sep	17	R, W	20:00	24:00	N36°11'	E116°47'	26.0
C091705	1770	Sep	17	R			N36°32'	E113°03'	26.5
C091706	1770	Sep	17	R, W		26:00	N35°30'	E117°38'	25.4
C091707	1770	Sep	17	R, W	20:00	sunrise	N28°51'	E112°37'	18.8
C091708	1770	Sep	17	R, W	20:00	26:00	N37°11'	E119°56'	27.0
C091709	1770	Sep	17	R			N32°36'	E114°23'	22.5
C091801	1770	Sep	18	R			N37°10'	E112°29'	27.1
C091802	1770	Sep	18	R			N37°32'	E116°32'	27.4
C091803	1770	Sep	18	R			N37°10'	E112°29'	27.1
C100301	1770	Oct	3	R, W			N33°52'	E115°47'	23.8
C101501	1770	Oct	15	jadish			N28°54'	E111°31'	18.9
C101502	1770	Oct	15	jadish			N26°28'	E111°36'	16.5

Download table as: ASCII Typeset image

Table 2. Contemporary Observations in Japan

ID	Year	Month	Date	Color	Start	End	G. lat.	G. long.	MLAT
J090301	1770	Sep	3	R	18:00	sunrise	N40°48'	E140°27'	31.0
J090302	1770	Sep	3	R	18:00	sunrise	N40°37'	E140°28'	30.8
J090303	1770	Sep	3		18:00	sunrise	N40°37'	E140°28'	30.8
J090304	1770	Sep	3	R	18:00	sunrise	N38°35'	E140°57'	28.8
J091001	1770	Sep	10	R			N34°31'	E135°47'	24.5
J091301	1770	Sep	13	R	18:00	sunrise	N40°37'	E140°28'	30.8
J091302	1770	Sep	13	R	22:00	26:00	N34°34'	E133°14'	24.5
J091303	1770	Sep	13				N38°26'	E141°18'	28.7
J091401	1770	Sep	14				N38°26'	E141°18'	28.7
J091601	1770	Sep	16	R	24:00	sunrise	N34°46'	E138°00'	24.9
J091602	1770	Sep	16	R, W			N35°06'	E138°52'	25.2
J091701	1770	Sep	17	R-W	18:00	sunrise	N35°01'	E135°46'	25.0
J091702	1770	Sep	17	R-BW	22:00	sunrise	N35°42'	E139°45'	25.9
J091703	1770	Sep	17	R		sunrise	N35°01'	E135°46'	25.0
J091704	1770	Sep	17	R			N35°34'	E138°34'	25.7
J091705	1770	Sep	17	R	16:00		N35°01'	E135°46'	25.0
J091706	1770	Sep	17	R			N35°30'	E135°46'	25.5
J091707	1770	Sep	17	R			N35°01'	E135°46'	25.0
J091708	1770	Sep	17	R-W	20:00	26:00	N35°01'	E135°46'	25.0
J091709	1770	Sep	17	R			N35°42'	E139°45'	25.9
J091710	1770	Sep	17				N35°30'	E134°14'	25.5
J091711	1770	Sep	17	R	20:00	24:00	N37°21'	E137°04'	27.4
J091712	1770	Sep	17	R-W			N41°26'	E140°07'	31.6
J091713	1770	Sep	17	R-W	20:00	sunrise	N35°11'	E136°54'	25.2
J091714	1770	Sep	17	R	18:00		N35°00'	E135°46'	25.0
J091715	1770	Sep	17	R	sunset	sunrise	N35°42'	E139°45'	25.9
J091716	1770	Sep	17	R-W	20:00	24:00	N34°41'	E135°32'	24.7
J091717	1770	Sep	17	R-W		sunrise	N35°01'	E135°46'	25.0
J091718	1770	Sep	17	R-W	evening	26:00	N34°31'	E135°51'	24.5
J091719	1770	Sep	17	R-W			N35°34'	E137°27'	25.6
J091720	1770	Sep	17	R	22:00		N34°49'	E135°26'	24.8
J091721	1770	Sep	17	R	20:00		N34°49'	E135°39'	24.8
J091722	1770	Sep	17	R-W	20:00		N33°44'	E135°23'	23.7
J091723	1770	Sep	17	R			N34°41'	E135°32'	24.7
J091724	1770	Sep	17				N34°46'	E136°08'	24.8
J091725	1770	Sep	17	five-color			N34°45'	E135°32'	24.8
J091726	1770	Sep	17	R			N37°23'	E138°50'	27.5
J091727	1770	Sep	17	R-P	18:00	30:00	N34°41'	E135°32'	24.7
J091728	1770	Sep	17	R	sunset		N35°05'	E138°55'	25.2
J091729	1770	Sep	17	R-Gl	22:00	26:00	N34°52'	E137°49'	25.0
J091730	1770	Sep	17	R			N37°55'	E140°06'	28.1
J091731	1770	Sep	17	R			N33°41'	E130°47'	23.6
J091732	1770	Sep	17	R	20:00		N34°33'	E135°48'	24.6
J091733	1770	Sep	17	R			N33°44'	E135°59'	23.8
J091734	1770	Sep	17	R			N34°36'	E135°45'	24.6
J091735	1770	Sep	17	R	18:00	sunrise	N34°36'	E135°48'	24.6
J091736	1770	Sep	17	R-W	sunset		N34°31'	E135°51'	24.5
J091737	1770	Sep	17	R	18:00	sunrise	N35°23'	E136°56'	25.4
J091738	1770	Sep	17	R			N35°26'	E137°08'	25.5
J091739	1770	Sep	17	R			N38°03'	E139°24'	28.2
J091740	1770	Sep	17	R	22:00		N34°58'	E138°56'	25.1
J091741	1770	Sep	17	R	24:00		N34°46'	E137°24'	24.8
J091742	1770	Sep	17	R			N34°46'	E137°23'	24.8
J091743	1770	Sep	17	R-W	20:00	sunrise	N34°46'	E137°23'	24.8
J091744	1770	Sep	17	R			N38°39'	E141°11'	28.9
J091745	1770	Sep	17	R	22:00		N38°26'	E140°19'	28.6
J091746	1770	Sep	17	R		sunrise	N36°34'	E136°40'	26.6
J091747	1770	Sep	17	bright	19:00	30:00	N37°30'	E140°27'	27.7
J091748	1770	Sep	17	P-R-W-Y			N38°25'	E141°18'	28.7
J091749	1770	Sep	17	R			N31°43'	E131°04'	21.6
J091750	1770	Sep	17	R	24:00		N39°26'	E140°46'	29.6
J091751	1770	Sep	17	R-W	18:00		N41°26'	E140°07'	31.6
J091752	1770	Sep	17	R-W			N34°48'	E134°59'	24.8
J091753	1770	Sep	17	R		sunrise	N37°50'	E139°13'	28.0
J091754	1770	Sep	17	R-W	sunset	24:00	N35°11'	E136°54'	25.2
J091755	1770	Sep	17	R-W	morning?	sunrise	N34°52'	E134°33'	24.8
J091756	1770	Sep	17	R	22:00		N34°59'	E134°58'	25.0
J091757	1770	Sep	17	R			N35°19'	E137°56'	25.4
J091758	1770	Sep	17	R-W			N36°52'	E140°25'	27.1
J091759	1770	Sep	17	R-W			N36°52'	E140°25'	27.1
J091760	1770	Sep	17	R, five-color	18:00	sunrise	N39°27'	E140°22'	29.6
J091761	1770	Sep	17	R-W	22:00	26:00	N34°35'	E136°32'	24.6
J091762	1770	Sep	17	R-W	evening	sunrise	N35°11'	E136°54'	25.2
J091763	1770	Sep	17	R	evening	sunrise	N39°42'	E141°09'	29.9
J091764	1770	Sep	17	R	21:00		N35°02'	E135°46'	25.1
J091765	1770	Sep	17	R-W	18:00	28:00	N34°58'	E135°46'	25.0
J091766	1770	Sep	17	R	sunset		N34°17'	E135°34'	24.3
J091767	1770	Sep	17	R-W			N35°01'	E135°46'	25.0
J091768	1770	Sep	17	R	20:00	26:00	N40°37'	E140°28'	30.8
J091769	1770	Sep	17	R	26:00		N35°42'	E139°45'	25.9
J091770	1770	Sep	17	R	18:00	sunrise	N40°37'	E140°28'	30.8
J091801	1770	Sep	18	R-W			N35°21'	E137°23'	25.4
J091802	1770	Sep	18	five-color			N34°45'	E135°32'	24.8
J091803	1770	Sep	18	R			N35°01'	E135°46'	25.0
J092201	1770	Sep	22	R	21:00	26:00	N35°10'	E138°41'	25.3
J092501	1770	Sep	25	R	22:00		N35°01'	E135°46'	25.0
J100201	1770	Oct	2	R	22:00		N35°39'	E137°26'	25.7
J101601	1770	Oct	16	R-W-P	21:00		N38°41'	E141°11'	28.9

Download table as: ASCIITypeset images: 1 2

Table 3. Contemporary Observations by Captain Cook

ID	Year	Month	Date	Color	Start	End	G. lat.	G. long.	MLAT
CJC0916	1770	Sep	16	R	22:00	24:00	S10°27'	E112°49'	−20.6

Download table as: ASCII Typeset image

A.2. References of Historical Documents

In order to find out records of naked-eye sunspot and low-latitude auroras in 1770 September–October, we examined the contemporary historical documents. We show their references below. In case they are published as books or critical editions, their format is given as follows: "Record ID number = Romanized title of source document, page/folio number: Reference of source document in original language." Korean historical documents are without record ID numbers because they do not include auroral candidates as discussed in the main text. The references of source documents are given as follows: "Abbreviation: Author name, Document name, year of publication." In case they are included in some publication entitled with another name, their format is given as follows: "Abbreviation: Their title; Author name, Document name, year of publication." In case they are unpublished and preserved as manuscripts, their format is given as follows: "Abbreviation: Document name, [Owner; shelf mark]." In order to keep traceability of historical documents for readers of this paper, we show names of authors and documents in their original languages, while showing abbreviations in English.

China

C091101 = Wŭxiāngxiànzhì, II, 75a: 白鶴, 史傳遠. 『武郷縣志』, 1790/91.
C091501 = Zhǎngzǐxiànzhì, XIV, 17a: 紀在譜, 黃立世. 『長子縣志』, 1778/79.
C091601 = Yŏngpíngfŭzhì, III, 19b: 李奉翰, 王金英. 『永平府志』, 1774/75.
C091602 = Luánzhōuzhì, I, 48a: 呉士鴻, 孫學恒. 『灤州志』, 1810/11.
C091603 = Jìzhōuzhì, II, 52b: 沈鋭, 章過. 『薊州志』, 1831/32.
C091604 = Chānglíxiànzhì, I, 6a: 何崧泰，馬恂. 『昌黎縣志』, 1866/67.
C091605 = Shòuguāngxiànzhì, IX, 19a: 劉翰周. 『壽光縣志』, 1799/1800.
C091701 = Qngshǐgǎo, Distasters II, 1572: 趙爾巽. 『清史稿』, 1976.
C091702 = Qngshǐgǎo, Disasters II, 1572: 趙爾巽. 『清史稿』, 1976.
C091703 = Qngshǐgǎo, Disasters IV, 1611: 趙爾巽. 『清史稿』, 1976.
C091704 = Féichéngxiànzhì, XVI, 3b: 曾冠英. 『肥城縣志』, 1815/16.
C091705 = Xiāngyuánxiànzhì, VIII, 6b: 李廷芳, 徐玨. 『襄垣縣志』, 1782/83.
C091706 = Chángshānxiànzhì, IV, 12a: 倪企望, 鍾廷瑛. 『長山縣志』, 1801/02.
C091707 = Dòngtínghúzhì, VII, 38a: 綦世基, 沈廷瑛, 夏大觀, 陶澍, 萬年淳. 『洞庭湖志』, 1825/26.
C091708 = Yēxiàn quánzhì, III, 44a: 魏起鵬. 『掖縣全志』, 1893/94.
C091709 = Zhèngyángxiànzhì, III, 38b: 陳全三, 潘守謙. 『正陽縣志』, 1936/37.
C091801 = Qngshǐgǎo, Disasters II, 1572: 趙爾巽. 『清史稿』, 1976.
C091802 = Qngshǐgǎo, Disasters II, 1572: 趙爾巽. 『清史稿』, 1976.
C091803 = Róngchéngxiànzhì, I, 23a: 李天騭, 岳賡廷. 『榮成縣志』, 1840/41.
C100301 = Bózhōuzhì, XL, 8a: 任壽世, 劉開. 『亳州志』, 1825/26.
C101501 = Yuánjiāngxiànzhì, XXII, 46b: 唐古特, 駱孔僎. 『沅江縣志』, 1810/11.
C101502 = Yŏngzhōufŭzhì, XVII, 114b: 呂恩湛，宗績辰纂. 『永州府志』, 1828/29.

Korea

Ilseongnok: Anonymous. 『日省録』, 1991.
Seungcheonwon Ilgi: Anonymous. 『承政院日記』, 1961.

Japan

J090301 = Tosho Nendaiki, 354–355: 当所年代記; 青森県文化財保護協会編『みちのく双書第22集平山日記』, 1967.
J090302 = Eiroku Nikki, Meiwa 07/07/14¹² : 永禄日記 [弘前市立図書館, KK-215-エイ]
J090303 = Honban Meijitsuroku, Meiwa 07/07/14: 本藩明実録, 92; 青森県文化財保護協会『みちのく双書第45集本藩明實録・本藩事実集（中）』, 2003.
J090304 = Narumi Nikki, 33: 鳴海日記; 佐藤素子，菅野照雄，谷地森隆『鳴海日記』, 2005.
J091001 = Sogamura Horiuchi Chogen Oboegaki, 162: 曽我村堀内長玄覚書; 平井良明『大和国庶民記録堀内長玄覚書』, 1993.
J091301 = Tsugaru Henran Nikki VII, Meiwa 07/07/24: 津軽編覧日記七 [弘前市立図書館, K-215-ツガ-7]
J091302 = Koshike Kyuki Kafu Eidairoku, 561–562: 古志家旧記家譜永代録; 府中市編『府中市史』史料編Ⅲ, 1988.
J091303 = Nendaiki, 695: 年代記; 石巻市史編纂委員会編. 『石巻の歴史』 IX, 3, 1990.
J091401 = Nendaiki, 695: 年代記; 石巻市史編纂委員会編. 『石巻の歴史』 IX, 3, 1990.
J091601 = Yokosuka Nemoro Rekidai Meikan, 34–35: 横須賀根元歴代明鑑; 原田和『遠江資料集』, 1960.
J091602 = Fujiyama Houei Funka-no-Zu: 5, 富士山宝永噴火の図 [静岡県立図書館歴史文化情報センター, 03036-近世写真集08-K314]
J091701 = Zokushigusho, 689: 続史愚抄; 黒板勝美，国史大系編纂会編『新訂増補国史大系第15巻続史愚抄』, 1966.
J091702 = Zoku Konendai Ryakuki, 13: 続皇年代略記; 国書刊行会『続々群書類従』II, 1, 1969.
J091703 = Honcho Tenmonshi, Meiwa 07/07/28: 本朝天文志 [国会図書館古典籍資料室, 205–158]
J091704 = Yashi, 14a 野史, VIII [国会図書館古典籍資料室, 133–99イ]
J091705 = Taihei Nenpyo, 73b-74a: 泰平年表 [早稲田大学, 文庫01 01–773]
J091706 = Atomigusa, Meiwa 07/07/28: 後見草 [国立公文書館, 216-0008]
J091707 = Zoku Nihon Oudai Ichiran, Meiwa 07/07/28: 続日本王代一覧, IX [国文学研究資料館, ヤ3-49-9,10a]
J091708 = Hannichi Kanwa, XII, Meiwa 07/07/28: 半日閑話 [国会図書館古典籍資料室, 195-54]
J091709 = Bukou Nenpyo V, Meiwa 07/07/28: 武江年表 [国会図書館古典籍資料室, 213.6-Sa222b]
J091710 = Inpu Nenpyo, Meiwa 07/07/28: 因府年表 [国会図書館古典籍資料室, わ217-4]
J091711 = Shinnenji Kihaku-cho-sho, 1399: 真念寺鬼簿帳抄: 柳田村史編纂委員会編『柳田村史』1975.
J091712 = Hokkaidoshi, Meiwa 07/07/28: 北海道志 [東京大学史料編纂所, 1041.10-9]
J091713 = Hennen Tairyaku, III, 80b-81a: 編年大略 [名古屋市蓬左文庫，32-4-34]
J091714 = Oriorigusa, III, Meiwa 07/07/28: 折々草 [東京大学総合図書館, A00-4197]
J091715 = Zokudankai, XXXIX, Meiwa 07/07/28: 続談海 [国立公文書館, 150-0094]
J091716 = Setsuyo Kikan, XXXIII, 177: 摂陽奇観; 船越政一郎（編）『浪速叢書』IV, 1927.
J091717 = Seikai, 26b-27a, 30b-31b: 星解 [東北大学図書館, 8-21347-1]
J091718 = Nennen Chinjiki, 457-458: 年々珍事記; 桜井市史編纂委員会『桜井市史』史料編II, 1981.
J091719 = Tanose-mura Shoji Tomegaki-cho, 525: 田瀬村諸事留書帳; 福岡町史編纂委員会『福岡町史』II, 1983.
J091720 = Inatsukake Nikki, 185: 稲束家日記; 池田市史編纂委員会編『池田市史』史料編IV, 1980.
J091721 = Kenbunroku, 538: 見聞録; 枚方市史編纂委員会編『枚方市史』IX, 1974.
J091722 = Mandaiki, 496: 万代記; 田辺市教育委員会編『紀州田辺万代記』IV, 1992.
J091723 = Asunarou, 83: あすならふ; 大阪市史編纂所編『大阪市史史料』XXIV, 1988.
J091724 = Uenomachi Kyuki Mokuroku, 178: 上野町旧記目録; 上野市古文献刊行委員会編『上野町旧記目録』, 1990.
J091725 = Eitai Kakocho, 623–624: 永代過去帳; 吹田市編纂委員会編『吹田市史』VI, 1974.
J091726 = Senpukuji Shoji Kenbun Zakki, 19: 専福寺諸事見聞雑記; 長岡市史編集委員会近世史部会編『市史双書NO. 24　専福寺諸事見聞雑記』, 1993.
J091727 = Kagomimi-shu, 405–406: 籠耳集; 原田伴彦他編『日本都市生活史料集成』I, 1977.
J091728 = Ohira Nendaiki, 68: 大平年代記; 沼津市史編集委員会編集『大平村古記録』, 2000.
J091729 = Uchiyama Matatsu-Ou-den, 303: 内山真龍翁伝; 天竜市役所編『天竜市史』VI, 1979.
J091730 = Sanju Nenpyo, 138: 三重年表; 山形県編『山形縣史』III, 1960.
J091731 = Banzai Reki, 16: 万歳暦; 方城町史編纂委員会『方城町史資料』, 1972.
J091732 = Yorozu Mezurashikikoto Oboecho, 921: 万珍敷事覚帳; 田原本町史編纂委員会編『田原本町史』I, 1988.
J091733 = Kumano Nendaiki, III, 3a: 熊野年代記 [東京大学史料編纂所, 2012-233]
J091734 = Kokon Kikkyo Kokoroe Shirushicho, 601: 古今吉凶並心得記帳; 安堵町史編纂委員会編『安堵町史』II, 1991.
J091735 = Aramaki-mura Miyaza Nakama Nendaiki, 374: 荒蒔村宮座中間年代記; 天理市史編纂委員会編『改訂天理市史』I, 1977.
J091736 = Nennen Chinjiki, 458: 年々珍事記; 桜井市史編纂委員会『桜井市史』II, 1981.
J091737 = Meiwa Daikanbatsuki, 209: 明和大旱魃記; 犬山市教育委員会・犬山市市史編纂委員会編『犬山市史』V, 1990.
J091738 = Mitake Wakamiya Hachiman Jinja Sairei-cho, 871: 御嵩若宮八幡神社祭礼帳; 御嵩町史編纂室編『御嵩町史』, 1987.
J091739 = Kawauchi-shi Kiroku, 877: 河内氏記録; 中条町史編纂委員会編『中条町史』III, 1985.
J091740 = Seken Kikigaki Oboe, 3: せ間聞書覚; 沼津市史編集委員会編『原宿植松家日記・見聞雑記』1995.
J091741 = Ryunenji Tomeki, 185: 龍拈寺留記; 近藤恒次編『三河文献集成』I, 1963.
J091742 = Beika Nenpyo, 728: 米価年表; 近藤恒次編『三河文献集成』II, 1965.
J091743 = Beikaki, 708: 米価記; 近藤恒次編『三河文献集成』II, 1965.
J091744 = Odake-san Kofukuji Juji Oboegaki, 299: 大嶽山興福寺住持覚書; 南方町史編纂委員会『南方町史』, 1975.
J091745 = Omachi Nenbutsu Kocho, 149: 大町念仏講帳; 河北町史編纂委員会編『大町念仏講帳−河北町誌編纂史料−』, 1991.
J091746 = Seirinki Taiunko Nenpu, 653–654: 政隣記，秦雲公御年譜; 侯爵前田家編輯部『加賀藩史料』VIII, 1935.
J091747 = Nendaiki, 622: 年代記; 白沢村史編纂委員会編『白沢村史』1991.
J091748 = Nendaiki, 694: 年代記; 石巻市史編纂委員会編『石巻の歴史』IX, 1990.
J091749 = Nendai Jitsuroku, 32: 年代実録; 安山松巌著『年代実録』1974.
J091750 = Sakugara Hikae, 110: 作柄控; 高橋梵仙編『近世社会経済史料集成　第４巻　飢渇もの』I, 1977.
J091751 = Kyukishoroku, 431: 旧記抄録; 松前町史編纂室編『松前町史』I, 1974.
J091752 = Ruinen Oboegaki Shuyo, 143: 累年覚書集要; 三木郷土史の会編『累年覚書集要−明石藩三木郡小川組大庄屋安福家七代の記録−』1994.
J091753 = Echigo Yashi, 123: 越後野志; 源川公章『越後野志』I, 1974.
J091754 = Shozan Chomonshu, 59–60: 想山著聞集; 谷川健一他編『日本庶民生活史料集成』XVI, 1970.
J091755 = Yuran-do Nenpu-sho, 15–16: 幽蘭堂年譜抄; 竹下喜久男『清文堂史料叢書第72刊　播州龍野藩儒家日記−幽蘭堂年譜』I, 1995.
J091756 = Ozawa Nikki, 19: 小沢日記; 脇坂俊夫編『村役人日記−天領と三草藩領−』1986.
J091757 = Kumagaya-ke Denki, 376: 熊谷家伝記; 市村咸人校訂『熊谷家伝記』IV, 1934.
J091758 = Yorodzu Oboe-cho, 65: 萬覚帳; 矢祭町史編さん委員会編『矢祭町史研究』II, 1977.
J091759 = Ware Isho Mikiki Shirishi Oboe-no-sho, 231: 我一生見聞知覚書; 矢祭町史編纂委員会編『矢祭町史研究』II, 1977.
J091760 = Nikki, 967: 日記; 秋田県編『秋田県史』I, 1963.
J091761 = Nikki, IV, 318: 日記; 大久保正編『本居宣長全集』XVI, 1974.
J091762 = Enkoan Zuikan Zue, 6b-7a: 高力種信「猿猴庵随観図絵」[国会図書館古典籍資料室，特7-59]
J091763 = Morioka-han Karo-seki Nikki Zassho, 278: 盛岡藩家老席日記雑書; 盛岡市教育委員会『盛岡藩家老席日記雑書』XXVII, 2011.
J091764 = Hinami, Meiwa 07/07/28: 日次明和七庚寅年 [明治大学博物館下賀茂神社文書：鴨脚家文書, 目録第5号昭和28年, 冊の部, 家86]
J091765 = Nikki, 86a-b: 日記　明和七年 [東丸神社東羽倉家文書 B-2-164]
J091766 = Koin Hinami, 44a: 庚寅日次 [国文学研究資料館中橋家文書, 44G-00027]
J091767 = Hakkai Onki, Meiwa 07/07/28: 広橋兼胤「八槐御記」 [国立公文書館1781-35 古40-666]
J091768 = Hirosaki-han Onkuni Nikki, Meiwa 07/07/28: 弘前藩御国日記明和七年 [弘前市立図書館）, TK215-1-2052]
J091769 = Hirosaki-han Edo Nikki, Meiwa 07/07/28: 弘前藩江戸日記 [弘前市立図書館, TK215-2-858]
J091770 = Funai Jijitsu-en, XV, Meiwa 07/07/28: 封内事実苑 [弘前市立図書館, GK-215-9]
J091801 = Kamaya Shoya Nendaiki, 661: 釜屋庄屋年代記; 山岡町史編纂委員会編『山岡町史』II, 1978.
J091802 = Eitai Kakocho, 623–624: 永代過去帳; 吹田市編纂委員会編『吹田市史』VI, 1974.
J091803 = Zokushi Gusho, 689: 続史愚抄; 黒板勝美・国史大系編修会編『新訂増補国史大系第15巻　続史愚抄』, 1966.
J092201 = Imaizumi-mura Takarakagami, 567–569: 今泉邑宝鑑; 富士市教育委員会編『富士郡今泉邑宝鑑往古抜差』富士市教育委員会，1989.
J092501 = Zokushigusho, 689: 続史愚抄; 黒板勝美, 国史大系編修会編『新訂増補国史大系第15巻　続史愚抄』, 1966.
J100201 = Nendaiki, 1246: 年代記; 付知町編『付知町史』通史編史料編，1974.
J101601 = Genna Gannen Yori Toshi no Kikkyo Tome Cho, 86: 元和元年より歳之吉凶留帳; 迫町町史編纂委員会『迫町史資料』I, 1974.

Southern Hemisphere

CJC0916 = Joseph Banks, The Endeavour Journal of Joseph Banks 1768–1771 (ed. Beaglehole, J. C.), Angus & Robertson Ltd., Sydney, 1962, v.2, p.149.; Sydney Parkinson, 1773. A Journal of a Voyage to the South Seas in his Majesty's Ship, The Endeavour. Faithfully transcribed from the Papers of the late Sydney Parkinson, Draughtsman to Joseph Banks, Esquire, on his late Expedition, with Dr. Solander, round the World. Embellished with Views and Designs, delineated by the Author, and engraved by capital Artists. London, Printed for Stanfield Parkinson, the Editor, and sold by Messrs Richardson and Urquhart, at the Royal Exchange (and others). Facsimile edition, 1972, p.161.

A.3. Descriptions of Historical Records with Contemporary Auroral Drawings

Figure 1 in the main text shows MS Special 7–59 from the National Diet Library, ff. 6b-7a (at Nagoya). It corresponds to the record J091762 in Table 2. The aurora is drawn as the vertical red brightening in the northern sky. People's response to the aurora is also drawn in the bottom part of the figure. Their original text and translations are given below.

Original text:

七月廿八日夕かた北の空うす赤く遠方の火事かと沙汰するうちに次第次第に色こくなり夜に入て明ルき事月夜の如し戌ノ刻此より赤氣甚しく中に竿の如き白筋幾すじも顕れ半天に覆広がりて西東ニ広く白氣数多し地一面に真赤になりて諸人おどろきさわぎ所の生祇にて神楽をあげ或は念仏をとなへて生たる心地なしこれハ世がめつしるか火の雨でもふりハせぬかと屋根に水をかけるも有高き所に登りて見れハ赤氣のうちに物の煮ゆる音聞ゆと夜明にハ東西へわかつ様にて消たり

Translation:

On 1770 September 17, the sky was slightly red in the evening. Discussing if it is a fire, its color gets denser and denser to get as bright as a night with a full moon. From 20:00, many bands of white stripes like rods appeared within intensive red vapor to spread to cover half of the sky with numerous white vapors from east to west. The Earth got totally illuminated very red to make people surprised. Some dedicated divine dances and some prayed to Buddha, as they felt so scared to feel their life in danger. Some even suspected if the world was about to perish. Some spread water on the roofs of their houses suspecting rain of fire was coming. Some claimed that they even heard a sound as if something was boiled within red vapor, when seeing the red vapor, climbing up somewhere high. The red vapor disappeared in the dawn as if getting divided into east and west.

Figure 5 shows Japanese drawings of the aurora on 1770 September 16 and 17: J091602 and J091717. Figure 5(a) is a Japanese drawing of the aurora on 1770 September 16: J091602 (MS Shizuoka Municipal Library, 03036-Early Modern Photographs 08-K314, f. 5). Seven white stripes are drawn in a background of faint red vapor with Mt. Fuji.

Original text:

明和七年戊寅七月廿七日夜九ツ時富士あし高之間より赤ク成りし如斯中ニ白キすし十四五本ツヽ出ル右之赤ミ東西へ啓ク

Translation:

On 1770 September 16, around 24:00, the sky got reddened between Mt. Fuji and Mt. Ashitaka and there appeared 14–15 bands of white stripe. The redness on the right side spread toward the east and west.

Figure 5(b) is J091717 = MS 21347-1, Tohoku University Library, ff. 26b-27a, ff.30b-31b (at Kyoto). It corresponds to the record J091717 in Table 2. This drawing has the other two variants, one of which is shown with a translation in Ebihara et al. (2017). Their original text and translations are given below.

Original text:

明和七年庚寅七月廿八日夜紅氣弥天子刻正見圖

七月廿八日夜北方隔山左右一面空中赤色也見人正是大火也於山北雖在大布施八升等之村々又久多庄五村人家不多恐是若州一国大火而模於空見此矣倩按若州一国無一度於可焼様但若州城下者小濱也民多雖然一度可為焼亡哉必可有次第遅速殊小濱者隔路十有七里也何於京都不可見是正土中之水気登於空應受日光模而成赤色空中一面如然譬如遠村之出火雖不見火本模於雲遙見日輪者隠地球而雖不見光者模水気輝於空于見之如大火日已在北順夜更可東囘謂竟去矣到丑刻見之東北間尚盛也赤色中有同色之筋譬如日没之前浮雲覆空隠日日光従雲間洩光則顯光筋水気之中厚薄有不同依之従薄所日光指登而見如柱立曾非怪事至於暁尚囘於東至曙赤光滅是全雨気之所為也雖然晴天相続不発雨併到今月度々之夕立者正其所為歟宝永年中亦享保年間在之曾不及論善悪吉凶而已

Translation:

The drawing on how the red vapor was seen filling the heaven around 24:00 on 1770 September 17.

On 1770 September 17, at night, it got totally reddened within the sky beyond the northern mountains. Those who saw thought it was a great conflagration. Although we had villages such as Wofuse or Yamasu, or five villages in Kuta-no-sho, there are not many houses. This is probably because it was Obama that was in fire. We saw the sky, thinking that probably the whole land of Wakasa was in fire. However, thinking seriously, it is impossible for the whole land of Wakasa to be burnt at once. The capital city of Wakasa is Obama with a considerable population. Even if it had gotten totally burnt, we would definitely have needed to consider the difference of timing when the fire came. Although Obama is 17 ri (=68 km) away, why can we see not a fire at Obama from Kyoto? This may be because water vapor in the earth ascended to the air and got reddened by the sunlight. It is also possible that fires in distant villages seem like clouds even though the fires themselves are not seen. Even if the solar disk is hidden by the earth, the light becomes water vapor to seem like fire. People had left, discussing that the Sun was already in the north and heading eastward to the midnight. Around 02:00, red vapor was seen at northeast evidently. Within red vapors, we saw stripes in the same color and it seemed as if the Sun shed some light through floating clouds. The luminous stripe was very evident. It was unevenly thick or faint within the water vapor. It seemed like a fiery pillar of ascending sunlight at the parts where water vapor was weak. This is an unexampled wonder. At dawn, red vapor floated eastward. At dawn, the red light had vanished. This is totally because of rain vapor. This can be related with the fact that it finally squalled sometime this month after several months without rain. We had similar phenomena in the years of Houei and Kyoho as well. It is impossible to discuss whether it is good or bad.

A.4. A Japanese Record of Sunspots Visible with the Naked Eye

A Japanese report (J091713) includes not only records for auroras but also sunspots around 1770 September 17. This record tells us to know that the great sunspots captured by Staudacher (Figure 4) were so big that even contemporaries saw it with the naked eye in Japan (see, Figure 6)

Original text:

一七月廿八日暮合頃ゟ北方火事之如く夜ニ入戌之刻時分赤氣

甚敷中ニ竿の如き白氣幾筋も顕れ半天ニ覆ひ次第ニ廣かり東西ニ赤氣広く白氣数多夜半ゟ薄く成り夜明ニ消江戸ニ而も北ニ見る由也

一頃日中日中ニ黒き点有之也

Translation:

I. On 1770 September 17, from around dusk, it was like a conflagration in the north. Around 20:00, several strokes of white vapors appear in intensive red vapor. They covered half of heaven and gradually spread with red vapor expanded east and west and numerous white vapors. From around midnight, they became faint and disappeared in the dawn. It is said that they were also seen northward in Edo.

I. Nowadays, there were black spots in the sun.

**Figure 6.** Record of naked-eye sunspot observation in Nagoya (J091713). Original text and translation are included in the text (courtesy: the Hosa Library).

Download figure:
Standard image High-resolution image

Long-lasting Extreme Magnetic Storm Activities in 1770 Found in Historical Documents

Article metrics

Permissions

Author e-mails

Author affiliations

Author notes

ORCID iDs

Dates

Abstract

1. Introduction

2. Methods and Source Documents