Historical Geographic Data of the Silk Road Zone II Versions EN1 Vol 3 (3) 2018
A GIS dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties
: 2018 - 05 - 18
: 2018 - 05 - 30
: 2018 - 09 - 26
2500 15 0
Abstract & Keywords
Abstract: Urban construction is a major form of human land use activities. It records the history of urban system evolution and reflects the changes of a city's location, size, and form. Historical data of urban built-up area along the Silk Road provide statistical support for studying the evolution process of the cities involved, and for restoring other urban elements at a longer time span. This paper looks at the Chinese cities along the Silk Road during the Ming and Qing dynasties. Here, built-up area refers to the area of a city fenced off by the city walls, and the data were derived from historical documents and remote-sensing images. Acknowledging the "lifecycle" of historical toponyms, we combined historical geography methods and GIS to build the dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties (totaling 393 records). It provides detailed and systematic data for in-depth study of the Silk Road and relevant historical GIS restoration.
Keywords: the Silk Road; urban built-up area; the Belt and Road Initiative; GIS; city walls
Dataset Profile
English titleA GIS dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties
Data corresponding authorCheng Yinong (342791@qq.com)
Data authorsXue Qiaofeng, Cheng Yinong, Jin Xiaobin
Time range1368 – 1911
Geographical scopeProvinces/regions along the ancient Silk Road in China, including Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang provinces.
Data volume146 KB
Data formatESRI Shapefile
Data service system<http://www.sciencedb.cn/dataSet/handle/645>
Source of fundingMajor projects of the National Social Science Fund of China (14ZDB031)
Dataset compositionThe dataset is composed of a shapefile. Among it, *.shp file stores the geometry of geographic features; *.shx file stores the index of graphic features and attribute information; *.dbf file stores the dBase table for element information attributes; *.prj file stores spatial references; *.sbn and *.sbx are spatial index files for geometry; *.shp and *.xml files store metadata in XML format.
1.   Introduction
A large number of cities were distributed along the ancient Silk Roads. As hubs for economic, political and cultural exchange between East and Central Asia, these cities maintained the prosperity and stability of their local societies. This research studied the changes in the urban built-up area along the Silk Road in Shaanxi, Gansu, Qinghai, Ningxia and Xinjiang provinces in China during in the Ming and Qing dynasties (1368 – 1911 AD). The study was funded by the National Social Science Fund of China "Silk Road Historical Geographic Information System Construction", which aimed to build a historical geographic information system for the Silk Road around the Belt and Road Initiative. The project planed to establish a diachronic and open database of the corresponding area from the Western Han dynasty to the founding of the People's Republic of China (202 BC – 1949 AD), in the hope of providing historical geographical basic data for comprehensive researches of the Silk Road.1–2
Urban built-up area records the history of a city and reflects changes in its location, size and form. Historical extents of the urban built-up area provide data support for studying the evolution of the cities along the Silk Road. Previous studies on the restoration of urban built-up area in the Ming and Qing dynasties usually take city wall enclosure as the indicator. The reasons were as follows. First, most of the Chinese cities in the Ming and Qing dynasties built city walls, which served as a major delimiter of urban and rural areas in terms of function, landscape and psychology; second, existing historical documents have informative, reliable records on the construction and renovation of the city walls, which are conducive to built-up area restoration. However, it should be noted that the area enclosed by city walls was not completely equivalent to the actual urban built-up area. This is because: there was often non-construction land within the enclosure, such as farmland and wasteland, and also urban blocks outside the city walls. However, as it was difficult to restore these two types of land use accurately, city wall was still the basis for the restoration of the urban built-up area during this period. Previous studies have used multi-source data (such as historical texts, maps and remote sensing images) to restore the land use of a small number of cities,3–5 or mathematical models to restore the land use of multiple cities,6–7 which have achieved satisfactory results. In comparison, the latter was more efficient, while the former generated more accurate outcomes. As there are abundant materials of Chinese cities in the Ming and Qing dynasties which enable more elaborate reconstructs, this study collected and integrated multi-source data, including historical documents, ancient maps, remote sensing data, local histories and research results, and established the GIS dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties. This dataset contained surface vector data indicating the range of the built-up area of these cities, and attribute data such as the time spans and names of specific periods in which the built-up area remained unchanged. In order to facilitate data reuse and standardization, the time span was recorded as the "lifecycle" of a historical place name in use, a recommendation proposed by the Chinese Historical Geographic Information System (CHGIS).8 This dataset provided a basis for longer-term historical studies of the Silk Road cities and restoration of other urban elements of the cities.
2.   Data collection and processing
2.1   Data collection
In this study, "city" is defined as an administrative settlement under the effective jurisdiction of the central governments of Ming and Qing dynasties, or a military settlement with administrative jurisdiction. It includes Fu, Zhou, Xian, Wei, and Suo in the Ming dynasty, and Fu, Zhou, Ting, Xian and the Manchus military cities in the Qing dynasty, excluding towns, military forts, folk forts and chieftains’ cities. Urban built-up area refers to the area enclosed by city walls, and the change of city walls ranges means the change of the corresponding urban built-up area. The study focuses on the time range of the Ming and Qing dynasties (1368 – 1911). For cities whose city walls were built before 1368, the status of the city walls in 1368 was restored; for cities whose city walls were first built after 1368, 1368 was designated as the year when the walls were built, due to the lack of relevant data; for a few cities without walls, we tried to restore the distribution of relevant buildings according to the multi-source data collected, or temporarily gave them up. The main technical process for dataset generation is shown in Figure 1.

Fig.1   Technical flowchart of the dataset generation
2.2   Data sources
There were two types of data sources: one used for the reconstruction of the built-up area of historical cities, including historical geography documents (ancient literature, ancient maps, local chronicles), modern surveying and mapping data (remote sensing data, topographic maps), and historical geography research results; the other about the change of administrative divisions and place names, which was used to determine their "lifecycle" and to bridge this project with the "Silk Road Historical Geographic Information System".
(1) Data for the reconstruction of the built-up area of historical cities. They were from ancient and modern local chronicles of respective cities along the Silk Road, including A Collection of Historical and Modern Documents: Local Chronicles, Renovation Records Under the Reign of Jiaqing in the Qing Dynasty, survey maps in the 1930s, remote sensing data in the 1970s,9 modern remote sensing data,10–11 and archaeological and historical geography researches.
(2) Data for the administrative divisions and place names: Comprehensive Table of Administrative Evolution in Ming Dynasty,12Comprehensive Table of Administrative Evolution in Qing Dynasty , 13 and The History of Chinese Local Administrative System. 14–15
2.2   Data preprocessing
We integrated the above data to determine the location of the city walls, and then used Google Earth as the base map for drawing the urban built-up area according to the fence of the city walls, based on which the surface vector data were generated and saved in kml format. Then we used the ArcGIS 10.3 Arctoolbox tool to convert the kml format data into a shapefile, and the GCS_WGS_1984 coordinate system was adopted as the geo-referencing system. Finally, according to the construction and renovation time of the city wall, and the change of the place name, a "lifecycle" was determined for each place to designate the time period in which its state was stable. Through the above steps, a total of 393 data entries for the urban built-up area along the Silk Road during the Ming and Qing dynasties were restored, each with a defined location, boundary and "lifecycle" (Figure 2).

Fig.2   Schematic diagram illustrating the reconstruction process of the built-up area of a typical city (Qianzhou, Shaanxi Province)
The biggest difficulty in reconstruction was to manually determine the remains of the city walls and the referencing objects indicating the location of the demolished walls. Since most of these walls were not preserved intact today or completely destroyed, historical geography was used to restore their historical location. Many cities along the Silk Road in the Ming and Qing dynasties that were not abandoned today but were incorporated into modern cities and become part of them more or less preserved traces of ancient urban forms, as they developed from their historical sites. According to the integrity of the city walls, they were divided into the following four types:
(1) City walls that were well preserved, such as Fucheng of Xi’an and Qianhusuocheng of Xingwuying. This type require minimum efforts, as a historical built-up area could be restored by simply marking the existing walls on Google Earth. This type accounts for 27.23% of the total data.
(2) City walls with few remains, but well-preserved referencing objects that indicated the location of the walls. These referencing objects were often relics of the city walls, mainly moats and special urban streets. In ancient times, the moats were usually excavated along the exterior of the city walls. After the walls were demolished, remains of the moats were retained, so the walls could be located by referring to the moat remains. In addition, for some cities, ring roads were built on the ruins of the demolished walls. If a ring road was found to be around the old city, the road could generally be considered as the location of the city walls, such as Gongning City and Dihua City in the modern Urumqi city. After these indicators were identified, they were validated against existing literature or ancient maps, and the urban built-up area could be delimited and drawn. This type accounts for 39.44% of the total data.
(3) City walls and relics that were obscure. In this case, it was necessary to search existing literature, ancient maps and modern maps for cultural relics or historic sites related to an ancient city (such as Confucious Temple, Chenghuang Temple, Bell or Drum Tower, other temples, memorial gateway, former residence of celebrities in the old cities), as well as old place names of the ancient city (such as South Gate, East Gate, Southwest City Wall Corner, etc.). We marked these locations on Google Earth, and then cross-referenced the relics locations and place names with existing literature and ancient maps, and combined the records of the city wall perimeter to determine the city wall enclosures. This method proved to be accurate in the case of rich data, and deviations were small even when less data were available. This type accounts for 24.43% of the total data.
(4) City walls in lack of materials. In this case, according to the cities location points of the CHGIS V6 TIME SERIES layer,16 we drew a square vector preface of 1 km×1 km, and mark "null" in the "references" field, so as to make it convenient for users to eliminate this data entry when needed, or for us to supplement it in the future. This type accounts for 8.90% of the total data.
Among the above four types, the reconstruction accuracy of type 1 and 2 was high, and the accuracy of type 3 was controllable. The first three types accounted for more than 90% of the total data, and the reconstruction results were generally reliable.
3.   Sample description
In the attribute table of this (shapefile) dataset, the field “name” represented the official name of the city. Some cities were composed of several small cities. For example, Xi'an was composed of Xi'an Fucheng and the East, West, South and North Guan cities. In addition, there were duplicates in the official names of the cities, so the field “other_name” would help further to identify specific cities. The field “begin_yr” indicated the start date of the lifecycle, while the field “end_yr” indicated the end, and the duration was designated as the lifetime of the built-up area of the city. The field “references” briefly recorded corresponding data sources and text description, which allowed data provenance. Cities marked by 1 km × 1 km squares indicated a lack of data, annotated by “data null” under this field. The field “landmark” recorded key landmarks such as city walls, moats, ring roads, cultural relics and place names, which could help rebuild the walls of these cities. The field “accuracy” recorded the accuracy of each restoration result, where numbers 1, 2, 3, or 4 was assigned to each, corresponding to the four types mentioned above, of which 1 had the highest accuracy and 4 the lowest. The dataset enables data query whereby a place name can be entered to generate temporal changes of the built-up area of a specified city in 1368 – 1911. The “time period” search allows the display of the extents of the urban built-up area in any year from 1368 – 1911, or to examine relevant distribution of their contemporaneous periods. Figure 3 shows a sample of urban built-up area in 10 cities including Xi’an (Fu), Yulin Wei (Fu), Lanzhou (Fu), Hancheng Xian, Ganzhou Wei (Fu), Xi’ning Wei (Fu), Xingwuying Suo, Ningxia Wei (Fu), Huiyuan City and Urumqi (Gongning City and Dihua City) in 1380, 1490, 1600, 1710, 1820, and 1900. This example showcases an example of data query results generated by this dataset.

Fig.3   Schematic diagram showcasing the data sample
4.   Quality control and assessment
Despite a lack of surveying and mapping data for the Ming and Qing dynasties, historical information on the city wall construction during this period was rich. Therefore, city walls were used as a substitute indicator for delimiting the urban built-up area in the Ming and Qing dynasties, providing a good data basis for this study. A few cities with unclear information on wall constructions, which were mainly in present Xinjiang, had been marked in the dataset. The basic data of this dataset were derived from publicly published ancient literature or modern academic works, which had undergone a rigorous quality control process. The use of comprehensive multi-source data restoration of the city walls is a mature and accurate method which has been used by the historical community for a long time. In the data production process, spatial registration and spatial positioning were performed based on modern remote sensing data, which ensured data integrity, reliability and standardization. The concept of “lifecycle” not only facilitated easy bridging with other parts of the project, but also ensured data accuracy, uniqueness and reliability on time scale.
5.   Usage notes
This dataset is a shapefile, which could be read by mainstream GIS software. It can also be read and written by mainstream programming languages such as Python, thus enabling batch processing. This dataset records annual changes in the built-up areas of most cities in the five Chinese provinces along the Silk Road from 1368 – 1911. It has a query function, which allows the display of changes of the built-up area of a certain city, and/or the overall situation of the urban built-up area in any years or time periods. This dataset provides a data foundation for studying the historical development, urban system changes, and land use changes of cities along the Silk Road. Users can select different points of interest for specific analysis.
Zhang P. Proposition of construction of historical geographical information system about the Silk Road and its value and significance. Journal of Shaanxi Normal University (Philosophy and Social Sciences Edition) 45(2016): 5-11.
Zhang P. Historical geographic information platform for the Silk Road: design, concept and application. Journal of Yunnan University(Social Sciences Edition) 16(2017): 50-54.
Li X & Wu H. Application of color infrared aerial photography to study urban historical geography: taking the relationship between cultural landscape evolution and river course change in the three riverside cities of Jiujiang, Wuhu and Anqing as an example. Journal of Peking University (Historical Geography Special Issue) 1992: 37-41.
Zhong C. Lone-term morphological changes of the old Shanghai town: a meso-scale study in town-plan analysis. Journal of Chinese Historical Geography 3(2015): 56-70.
Cheng Y. The method of restoring external forms of ancient Chinese cities: taking Urumqi as an example, in New Ideas in Geography, Archaeology and History, 322-335, ed. Center for research on ancient Chinese history Peking University. Beijing: Zhonghua Book Company, 2012.
He F, Ge Q & Zheng J. Reckoning the areas of urban land use and their comparison in the Qing Dynasty in China. Acta Geographica Sinica 57(2002): 709-716.
Lin Y, Jin X, Yang X et al. Dataset establishment and spatial reconstruction of urban and rural construction land of Jiangsu Province in the past 200 years. Acta Geographica Sinica 72(2017): 488-506.
Center for Historical Geography, Fudan University. CHGIS v4.0. Available at: <http://yugong.fudan.edu.cn/views/chgis_index.php?list=Y&tpid=700>.
United States Geological Survey. Available at: <https://earthexplorer.usgs.gov/>.
United States Geological Survey. Available at: <http://glovis.usgs.gov/>.
National Earth System Science Data Sharing Infrastructure. Available at: <http://www.geodata.cn/>.
Niu P. Comprehensive Table of Administrative Evolution in Ming Dynasty. Beijing: Sinomaps Press, 1997.
Niu P. Comprehensive Table of Administrative Evolution in Qing Dynasty. Beijing: Sinomaps Press, 1990.
Guo H & Jin R. The General History of Chinese Administrative Divisions (Ming Dynasty). Shanghai: Fudan University Press, 2007.
Fu L & Lin J. The General History of Chinese Administrative Divisions (Qing Dynasty). Shanghai: Fudan University Press, 2013.
Fairbank Center for Chinese Studies of Harvard University and the Center for Historical Geographical Studies at Fudan University. CHGIS, Version: 6. Available at: <https://sites.fas.harvard.edu/~chgis/>.
Data citation
1. Xue Q, Cheng Y & Jin X. A GIS dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties. Science Data Bank. DOI: 10.11922/sciencedb.645
Article and author information
How to cite this article
Xue Q, Cheng Y & Jin X. A GIS dataset of urban built-up area along the Silk Road in the Ming and Qing dynasties. China Scientific Data 3(2018). DOI: 10.11922/csdata.2018.0023.zh
Xue Qiaofeng
data collection and dataset generation.
doctoral student; research area: historical urban geography and ancient maps.
Cheng Yinong
data collection and technical support.
PhD, Professor; research area: historical urban geography and ancient maps.
Jin Xiaobin
technical support.
PhD, Professor; research area: land resource management.
National Social Science Fund of China (14ZDB031)
Publication records
Published: Sept. 26, 2018 ( VersionsEN1
Released: June 22, 2018 ( VersionsZH2
Published: Sept. 26, 2018 ( VersionsZH3