The accomplishments of the survey projects of Inoh Tadataka in early 19th century Japan should not be evaluated merely from the visible results of his maps of the country. The latest surveying technologies and instruments, as well as his knowledge of the astronomical almanac, had a wide range of influences upon the surveying skills and astronomical knowledge of local surveyors and scholars. Inoh's Sokuryo-nikki (Survey diary) and records of local counterparts preserved in throughout Japan are reviewed and connections are evaluated. The records have been unearthed in recent years by historians editing regional histories and local history researchers. These investigations are important aspects of recent studies of Inoh's projects and supplement basic research of Otani (1917) and Hoyanagi (1974). During his journeys to survey Japan over seventeen years, Inoh kept a daily journal. It records some 12,000 people who attended or guided Inoh's team. However, his journal lacks details of connections among them. Local records contain extensive practical information concerning the project. Generally, officials of local lords or village officials accompanied the team of surveyors. They would learn on the job. According to their records and letters, some made and improved upon Inoh's surveying instruments. Others wanted to become students of Inoh and later attended private classes in Edo. Still others discussed calendrical calculations, trigonometric functions, and logarithm. Subsequently, some returned to their home regions and took charge of local surveys. As a result, we can recognize the wide range of influences the surveying project of Inoh Tadataka had.
The Santo Hoiki (Record of Mountain and Island Bearings) is a record of survey results obtained by Inoh Tadataka and his group, of which 67 volumes are existent today. These volumes were handed down through his family and are currently in the possession of The Inoh Tadataka Museum. The Santo Hoiki is among the 2,345 documents of Inoh Tadataka Related Materials, which have been designated a national treasure. The Santo Hoiki is classified into two types, namely Type A, which consists of 65 volumes, and Type B, which consists of two volumes. Both Type A and Type B record the bearings of mountains and islands. However, the former is a record of bearings of mountains and islands by observation spot, while the latter is a record of bearings by mountain and island. Type A Santo Hoiki contains data from a part of the third survey and all of the fourth through ninth survey records. More specifically, the third survey Santo Hoiki is compiled in one volume, the fourth survey in four volumes, the fifth survey in 14 volumes, the sixth survey in five volumes, the seventh survey in 17 volumes, the eighth survey in 20 volumes, and the ninth survey in four volumes. Each volume contains records of bearings in the order of observation date; therefore, it can be assumed that the records were transcribed directly from field record notes in order to determine the locations of mountains and islands and to correct survey errors. Type B Santo Hoiki was probably compiled from Type A volumes to draw bearing lines on Inoh's map. Research comparing bearing lines in the No. 3 Santo Hoiki to the names of mountains and islands recorded on Inoh's Medium Scale Map (1:216,000) after the fifth survey reveals that the No. 3 Santo Hoiki was compiled in order to draw bearing lines on Inoh's Medium-scale Map.
Inoh Maps were the first scientific maps of Japan to be prepared based on survey results of Inoh Tadataka (1745-1818)'s team. Each map is revolutionary in the history of Japanese Cartography. Inoh Maps have been used in nautical charting in two ways: in the preparation of British Admiralty charts and in the preparation of Japanese charts at the early stage of what was the then the Hydrographic Department of Japan. Hoyanagi (1974) summarized the scientific achievements of Inoh Tadataka, including his contributions to nautical charting, and these have been quoted in many papers. The Tokyo Geographical Society conducted the 2017 United Kingdom Geo-Tour as one of programs marking the 200th anniversary of his death. Based on research carried out in the United Kingdom in 2016 and 2017, old and new findings are summarized and unresolved issues are presented. The main findings are as follows:
(1) Inoh Small scale Maps held in the United Kingdom have been permanently transferred to The National Archives.
(2) The Royal Navy evaluated Inoh Maps and used them in preparing eight British Admiralty charts of the area around Japan.
(3) Graticule lines of British Admiralty charts are based on the prime meridian of Greenwich; therefore, the location of Japan was established on the world maps.
(4) British Admiralty charts were compiled from Inoh maps with corrections for longitude and transferred from a graticule similar to Sanson–Flamsteed's projection to the graticule of the Mercator projection.
(5) Hydrographic Department of Japan was founded in 1871. Te first Chief Hydrographer Admiral Yanagi Narayoshi ordered the reproduction of approximately 300 Inoh Large scale Maps held by what was then the Ministry of Interior and utilized them for hydrographic survey, the preparation of nautical charts, and sailing directions.
(6) Exact reproduction history of Inoh Small scale Maps held in the United Kingdom from original Inoh Small scale Maps is not known.
Geographical features in the early 19th century and transformation of landforms during the past 200 years in Japan are studied through an overlay analysis of the current GSI map issued by the Geospatial Information Authority of Japan and Inoh's map issued in 1821. Analyzed are: (i) survey activities of Inoh's team and terrain conditions, shapes of lakes and islands, and distribution of place names at the end of the Edo era using 214 sheets of Inoh's large-scale map (1:36,000), and (ii) retreat and advance of coastlines, and route changes of roads and rivers over the 200 years. Land development is investigated through reclamation and landfills from natural and socio-economic aspects by applying GIS datasets to a digitalized version of Inoh's map. The usefulness of geospatial analysis is demonstrated by employing GIS techniques to understand quantitatively changing land conditions in Japan.
The map of Japan drafted by Inoh Tadataka is considered to be the earliest produced from a scientific survey. The descriptions of longitude and latitude are based on astronomical observations, and Inoh's nationwide survey has been considered to be the origin of modern surveys in Japan. However, he did not succeed in determining longitude. Although he had knowledge of a spherical earth, the results of his survey were projected on a plane, not on a spherical surface. The parallels of latitude drawn on his maps are based on an accurate astronomical survey he carried out, but the meridians on his maps are absolutely inconsistent. His survey method also combined traverse and intersection surveys without control points. Therefore, his nationwide survey cannot be considered to be representative of a survey carried out in the modern period. There are many open traverse lines on his maps. These lines generally extend to temples and shrines, although they are not effective for improving the accuracy of the survey. Because temples and shrines might have been important public facilities at that time, the Tokugawa shogunate government probably requested information concerning their locations. He carried out a nationwide survey ten times, but he could not survey the northern half of Ezo island (Hokkaido). It is said that Mamiya Rinzo, who studied survey technology under Inoh Tadataka, surveyed Ezo island and submitted his survey data to Inoh Tadataka, therefore, Inoh's map of Ezo island might be entirely based on Mamiya's data. Further studies are necessary because Mamiya's survey has not been clarified.
Land-survey expeditions across the Japanese archipelago during the period from 1800 to 1817 and scientific map-making conducted by Inoh Tadataka were monumental achievements in the Japanese history of cartography. How Inoh's astronomical knowledge and skills are reflected in his survey instrumentation and map fabrication of the Japan islands is reviewed. The contents of this paper are as follows: Section II makes an overview of Inoh's astronomical background nurtured both in his early life and at the Asakusa shogunal astronomical office; Sections III and IV discuss the influence of imported Chinese and Dutch books and European astronomical instruments on the survey apparatus used in the above expeditions; Section V describes his concrete star observation methods for determining local latitude, which were drawn by a local painter in the Kure district of Hiroshima in 1806; Section VI introduces survey work and map-making of the Sanuki fief in Shikoku by Kume Michikata, a surveyor and contemporary of Inoh, in contrast to those adopted by Inoh; and, Section VII comments on Inoh's cartographic policy and techniques in completing his famous Dainihon Enkai Yochi Zenzu (The Great Coastal Map of Japan).
In order to measure the longitude difference between two distant sites, Inoh Tadataka and his technical team observed three lunar eclipses when surveying the Japanese Islands at the beginning of the 19th century, and left some observation records. The local times when the partial eclipses began and ended were measured with a pendulum clock. By introducing the equation of time to the calendar date of the observations, the time difference can be obtained between the local time observed by Inoh and the lunar eclipse timetable of NASA shown in Japanese Standard Time at 135E degrees of longitude. Consequently, the accuracy of the longitude survey performed by Inoh Tadataka is evaluated in comparison to the precise map of the Geographical Survey Institute.
In 1880, Inoh Tadataka surveyed the meridian arc length corresponding to one degree of latitude difference. A record of calculations is contained in a handwritten manuscript titled “Sokuchidosetsu” or “On Meridian Arc Length Corresponding to One Degree.” The uncertainty of latitude for a short arc is dominant and is estimated to be about 0.42 minutes. This uncertainty is in accord with the rounding latitude at half a minute in many documents that refer to Inoh's work on coordinates. Some points were used for astronomical observations of latitude in two successive years. A comparison of those results suggests some bias in the results and a standard deviation around the mean of about 1.3 minutes. This is obviously much more than the uncertainty estimated from the scatter of one degree meridian arc length against the N–S component of a great circle connecting the ends of a route. Data sets of astronomical latitude observations surveyed independently help clarify the uncertainty associated with Inoh's latitude observations obtained from field surveys. The scatter of distance and latitude difference data around the linear trend suggests that the uncertainty of the N–S distance is about 7.3/10,000.
In 1801, Inoh Tadataka observed the heights of stars at the meridian passage to learn the latitudes of points, and also surveyed surface distances among the points surveyed astronomically. He obtained the distance of one degree as 28.2 Ri under the Japanese system and regarded this value as a definition for surveying and making maps. Therefore, 2538 Ri corresponds to 10000 km in the metric system. Subsequently, he surveyed most of the coastline of Japan and its major roads, and made basic 1:36000-scale maps using this definition. The accuracy of latitudes at astronomical observation points on the 1/216000-scale map is estimated by comparing those on an actual digital map of Japan. The standard deviation of the difference in latitudes between the two maps is within 0.5 minute (0.926 km).
The map projections used in Japan at the time of Inoh's map are studied, and the graticule and map projection of his map are examined on the basis of research to date. As many researchers have already pointed out, a contradiction exists between the map projection of Inoh's map and the graticule drawn on it. Ohtani (1917) concludes that the graticule on Inoh's map was drawn using the Sanson–Flamsteed projection and Hoyanagi (1974) concurs that this was certainly the projection of Inoh's map. On the other hand, Unno (1985b) disagrees that the Sanson–Flamsteed projection was used, and instead identifies a trapezoidal projection. Ohtani (1917) also introduces the map-making technology of Inoh Tadataka in detail. Using his technology, survey results are expanded on a plane without modification, and no conversion from a spherical surface to a plane is performed. As a result of verifying the graticule on Inoh's map and his map projection, it is highly probably that the graticule on Inoh's map was drawn with a trapezoidal projection, and it is proved that Inoh's map corresponds well to an equidistant secant cylindrical projection. However, the standard parallels of these equidistant secant cylindrical projection maps vary on every map.