Lake Suwa (area: 12.8 km2, average water depth: 4.7 m) is a small lake located in central Japan, which is known for its mysterious phenomenon called Omiwatari in the winter season. When the lake freezes, shrinkage and expansion of the ice due to diurnal temperature variations causes an ice-cracking phenomenon Omiwatari, which was said to resemble a bridge crossing the lake. The ancient village people might have believed it to be the track of a god visiting a goddess on the opposite shore.
The dates of complete freezing and Omiwatari have been recorded by the Suwa shrine for the last 574 years since 1444. During a cold winter, Omiwatari would already have occurred by mid-December, whereas in a warm winter, it would be delayed until the end of February, or no Omiwatari would have occurred at all. In recent years, the occurrence frequency of Omiwatari phenomenon has been decreasing rapidly, probably due to global warming. In the most recent winter (2017-2018), we had very cold weather, and Omiwatari phenomenon occurred after the lake had almost completely frozen over. An Omiwatari viewing ceremony was held by “Yatsurugi” Shinto shrine priests on 5th February, and good harvests and a prosperous economy in this year were predicted by referring to the past documentary data for Omiwatari crossing routes.
(Photograph taken by Kiyoshi MIYASAKA on January 31, 2018; Explanation: Takehiko MIKAMI)
Past meteorological records are important for improving our understanding of past, present, and future climates. Imaging and digitization of historical paper-based instrumental meteorological records must be carried out before these records are lost to decay. This kind of activity called “data rescue” is now taking place at many institutions around the world. A data rescue project is underway to preserve Japanese instrumental meteorological records from the 19th century. These data were collected by foreign residents and visitors, Japanese scientists influenced by Dutch science, and by Japanese merchants. Recently, meteorological measurements taken at Mito from 1852 to 1868, and at Yokohama in 1872 and 1873 have been found. Based on instrumental records collected through this data rescue project, a warmer climate in the 1840s and 1850s around the South-eastern Kanto Region has been identified. Large year-to-year variations of winter temperatures have also been detected.
Historical documentary sources provide important data for the reconstruction of typhoons before the collection of systematic instrumental records by official government meteorological agencies. In this paper, we describe five sources that we use to provide details about typhoons affecting the four main islands of Japan in 1877. The sources described are: 1) English-language newspapers published in Japan; 2) the Historical Weather Database for Japan; 3) Japanese lighthouse records; 4) logbooks from British and American ships; and 5) data tables from the Imperial Meteorological Observatory in Tokyo. We then show the data and reports from these sources that provide evidence for four typhoons (June 11, July 26-27, August 25-27 and October 11) that affected Japan in 1877. All five sources give information about these typhoons indicating that the historical sources can provide a detailed record of storms affecting Japan including information about location and storm track, wind speed, barometric pressure, and damage. Such historical documentary records of typhoons can be used to extend our knowledge of typhoon frequency, intensity and behavior into periods that previously lacked information.
Tropical cyclone (TC) Talas caused heavy rainfall and landslides in the Kii Peninsula of Japan in 2011, and TC Haiyan caused storm surges in Samar and Leyte islands of the Philippines in 2013. There are records of TC tracks similar to TC Haiyan and TC Talas in the past which show similar damages. In this study, we focus on these two TCs and compare their tracks with TCs having similar tracks observed during the past 117 and 122 years in the Philippines and Japan, respectively. Two TCs in 1897 and 1912 made landfall in the same area as TC Haiyan, which caused storm surges. TC Haiyan made landfall in the Visayas area of the Philippines. About 15 TCs land in the Visayas area every 10 years. However, these three TCs (including Haiyan) were among the strongest TCs landing in the Visayas area during the past 117 years. Intense TCs landed in the Visayas area posed a risk of storm surges to Samar and Leyte Islands. A TC in 1889 brought heavy rainfall and landslides to the same area as TC Talas did. About 3.7 TCs make landfall in the Kochi prefecture in Japan every 10 years. TC Talas and the TC in 1889 were among the slowest TCs that made landfall in the Kochi prefecture. Although more intensified TCs landed in the same area, they tended to propagate faster and produced less rainfall in the Kii Peninsula. Slow movements of TCs make landfall in the Kochi prefecture pose a risk of heavy rainfall and landslides in the Kii Peninsula. Imaging and digitization of historical paper-based instrumental meteorological records including TC tracks are referred to as “data rescue”. Longer TC records allow us to study TC cases with similar damage even when the frequencies of TC disaster are low.
During the second half of the 1850s, Louis Furet (1816-1900), a French priest sent by the Paris Foreign Missions Society, undertook the first extended scientific meteorological observations on Okinawa, the main island of the kingdom of Ryukyu, today's Okinawa Prefecture. Using instruments borrowed from the French Navy Depot, he collected data five times a day according to the then recently standardized protocols. These data were all addressed to Charles Sainte-Claire Deville, the founder, in 1852, of the French Meteorological Society. In November 2015, Gaston Demarée, a historian of meteorology, and the author, a historian of Okinawa, had the good fortune to unearth the totality of Furet's meteorological material, which consists of observation sheets and letters, in the archives of Météo-France, the French national meteorological service and a distant heir to the French Meteorological Society. The present article is not directly concerned with data analysis proper (see the article by Demarée, Mailier et al.). It nevertheless closely examines the variables recorded in the different tables—daily, monthly, yearly—used by Furet and the additional information contained in the margins. It also aims at presenting the course of Furet's life and scientific education, which singled him out among fellow missionaries, and the social context of his observations. His research interests on Okinawa were actually far from being limited to meteorology, even to the point of earning him criticism from his superior for an excessive commitment to science. They ranged in scope from the natural sciences to the description of the inhabitants' lives. During his six-year stay on Okinawa, Furet corresponded with several important scientific institutions and learned societies. The ten or so articles he published in their journals deal with the geology, fossils—for the study of which he is also recognized as a pioneer—and natural resources of Okinawa, as well as with its culture, history and language. The present article also provides a full list of the typhoons or tropical storms, along with elements of description, that have been recorded by the missionaries who resided on Okinawa between May 1844, when the first missionary set foot on the island, and October 1862, when Furet's departure for Japan put an end to the missionary presence for decades.
Father Louis Théodore Furet (1816-1900) was a missionary of the Société des Missions étrangères de Paris (Paris Foreign Missions Society) who was posted in the Far East from 1853 to 1869. The discovery of his manuscript of meteorological observations undertaken at Naha, Okinawa, Ryukyu Islands, Japan, opens up new perspectives on the historical climatology of 19th century Japan. Furet arrived at Naha (spelled Nafa in the 19th century literature), the main port of Okinawa, on 26 February 1855. From December 1856 to September 1858, he carried out meteorological observations five times a day: at 6 and 10 am, 1, 4 and 10 pm. The hydrological engineer Alexandre Delamarche (1815-1884) calibrated the meteorological instruments entrusted to Father Furet by the French Dépôt de la Marine. The observations were carried out following the meteorological procedures in use in France in the 1850s. The atmospheric pressure data are given by the barometer readings, the barometer-temperature readings and the computed temperature-corrected atmospheric pressure. The pressure data are controlled and corrected, where necessary, using the Delcros and Haeghens formula which was in use in the 1850s. The historical atmospheric pressure observations are compared to the present-day long-term averages at Naha. During his observation period a typhoon was witnessed by Father Louis Furet on 18 May 1857. Other low atmospheric pressure observations probably correspond to extra-tropical storms. In such an event, the Dutch ship van Bosse was wrecked near the island Tarama but the captain, his wife and the whole crew survived. They were helped by the inhabitants of the island and were later transferred to Naha, Okinawa, where they met the three French missionaries and finally got a passage from the Dutch trading post at Decima to Batavia (present-day Jakarta).
Climate variations in East Asia during the Little Ice Age are reconstructed using wind records together with chronicles of weather disasters in Japan and China, as well as weather observations from Okinawa in the 1850s. The results of analyses are as follows: (1) On the basis of numerous records that remain for the southern coast of Japan and the south-eastern coast of China, it appears that many rainstorm disasters occurred during the summer period in both Japan and China. They usually took place one or two months later in Japan than in China; (2) Observations in Naha show that, when air pressure drops considerably, wind speed is very high and its direction rotates. Following unusual air pressure falls in Naha, strong winds caused disasters in China or in Japan. It is thought that these events are linked to movements of typhoons; (3) Increases in windstorm damage occurred during different periods in Japan and China. This may be related to changes in atmospheric circulation. Cool summer periods are observed around 1705, 1740, 1765, 1785, 1830, and 1845. The year 1855 marks a turning point between periods with prevailing cool states and periods with prevailing warm states.
This study analyzed precipitation data during the 1860s in Yokohama, Japan, using observations made by an American Christian missionary, J.C. Hepburn. Hepburn's meteorological data included monthly maximum, mean, and minimum temperatures, monthly precipitation, and the number of rain days in each month. The climate during the late 1860s was characterized by large inter-annual precipitation variability during summer. We detected unusually heavy rainfall during the summer of 1868, whereas the summer of 1867 was characterized by extremely low precipitation. These trends are in accordance with precipitation patterns reconstructed from historical weather documents. Furthermore, we investigated the spatial distribution of extreme climate events in 1867 and 1868 throughout Japan. The results suggest that the unusual dryness in 1867 and wetness in 1868 were related to synoptic-scale climate in Japan, rather than local-scale patterns in Yokohama. The results from this study could be used to understand the nature of summer climate variations in Japan in the 1860s.
Many historical documents in Japan include daily weather records. To reconstruct historical climates from daily weather records, Ichino et al. (2001) constructed a method for estimating global solar radiation based on daily weather conditions. Presented is an attempt to reconstruct monthly mean solar radiation during the first part of the 19th century based on weather records found in 11 historical documents, with special attention to records in the year 1836, when a famine occurred. Global solar radiation is an important factor in the energy balance of the Earth, and is also fundamental to the hydrological cycle and agricultural productivity. This implies that reconstructing global solar radiation involves investigations into climatic physics and historical human societies. The estimated monthly means of global solar radiation in 1836 are compared to average estimations for 30 years (1821-1850), including the 1830s, at 11 points in Japan where data are generally available. According to our estimation, the values of solar radiation in July and August 1836 were smaller than their provisional normal values, which is the average for 30 years (1821-1850), at all points except those in Tohoku and southern Kyushu, although these differences were not extreme. On the other hand, fluctuations during the 1830s were larger than other decades before and afterwards. Moreover, the estimations in spring (February, March, and April) and autumn (September, October, and November) were no smaller than the average estimations. These results show an anomalously low solar radiation in July and August, which might have been a key climatic condition affecting society in the famine year.
Inpu-Nenpyo is compared to weather diaries. Inpu-Nenpyo, an official historical document from Tottori, northwest coast of Japan, was compiled from 1630 to 1841, and consists of four books. The density of mentions, for example, of “heavy snow” and “drought”, in the document is higher in the second half of the period than in the first. By comparison, many mentions of “heavy snow” are made in the 1810s, and many mentions of “little snow” are made from the late 1720s to the 1750s and in the late 1830s. On the other hand, “drought” is mentioned often in July from the late 1800s to the early 1810s. Comparing the documents and the diaries, it is found that years of “heavy snow” in the Inpu-Nenpyo are consistent with years with many days of snowfall in the diaries and that “drought” correlates with many sunny days; the former are thought to be cold and the latter hot in some cases. However, it is difficult to clarify their relations and to draw conclusions. One reason may be a lack of other historical documents and diaries. Therefore, local public bodies, which often keep them in museums and archives should be encouraged to collate them and make them more accessible to the public.
A unique database of weather diaries kept by the Hirosaki Clan Agency in Hirosaki city from the late 17th century to the late 19th century is introduced. The database is published in the form of a CD-ROM, which includes information on daily weather and natural disasters, such as floods and avalanches. The frequency of meteorological disasters in 40-year periods is summarized in Table 1.