printer version of this document
Each administrative unit (of county level or higher status) is given a new entry in the source notes table. The content of the source notes table includes descriptions and dates of changes in placename, administrative area, and administrative seat. Original quotations from source texts are provided, along with commentaries and citations. The source notes table serves as a chronology and bibliographic reference for the administrative units included in CHGIS datasets.
To begin the examination of spatial changes over time, a base map is prepared for the selected target area (usually a single county or prefecture). The base map includes three contemporary layers circa 1990 (ArcChina county boundaries, ArcChina towns and administrative unit points, ArcChina rivers) and one newly developed historical layer (CHGIS 1911 county boundaries). The 1911 county boundaries are used as the primary reference for reconstructing the historical boundaries. The Compilation of the 1911 County Boundaries is explained in a separate section, below.
The base maps are printed out on A3 size paper. The following image shows an unedited base map.
Draft Version of Historical Maps
Researchers use cross-hatch patterns and different colored pens to compile space-time-composite polygons directly on the base map. In some cases a section of the historical boundary line is judged to coincide with the 1911 or the 1990 boundary line, in other cases a new line is drawn directly on the base map to indicate the valid boundary for a particular period of time. The color coded patterns are dated and annotated alongside the sketch, directly on the base map.
An example of the annotated version of the historical boundary changes for the same area as in the preceding image is shown below:
In this example you can see the historical changes for Yong’an Xian (which occupies roughly the same area as the contemporary Yong’an Shi). The county was established in 1452 with an area indicated by diagonal red lines. In the year 1471 the jurisdictional area shrank by losing a section on the northern edge, shown in red and gray cross-hatched lines. In 1536 the jurisdictional area shrank once again, this time by losing a section on the eastern edge, shown in red and blue cross-hatched lines. Finally in 1567, the jurisdiction lost two more sections on the southern edge, shown in red and black lines.
Each county or prefecture area is traced backwards in time using this method.
Historical Point Locations
Historical towns and administrative seats are determined using a similar method as the historical areas, sometimes on the same base map sheet as the areas, sometimes on a separate sheet. For historical points, the placenames and their valid dates are written directly onto the base map. For each location, such as an administrative seat for a prefecture, the base map is examined to see if the correct location is now occupied by a contemporary point feature in the ArcChina layer (circa 1990). If the historical point is located exactly at the same position as the ArcChina feature, the ArcChina point is marked with a colored pen and labeled with the historical name and dates of validity. If a matching position in ArcChina is not available, the historical location is estimated based on the relative context of the ArcChina points, boundaries, rivers and the 1911 boundaries. When establishing the location of historical points that cannot be “snapped” directly to existing ArcChina features, many historical sources including maps, gazetteers, and secondary sources are consulted. Having established the best estimate for the location of a particular historical point, the point is added with a colored pen and labeled with the historical point’s name and dates. The resulting base map contains a number of colored points with labels indicating the placename and the begin and end dates for which that name and location were valid. An example is shown in the following image.
Compilation of the 1911 County Boundaries
One of the GIS layers used for the base maps contains 1911 county boundary lines. The method of compiling the 1911 county boundaries is both complex and illustrative of how the historical areas of jurisdiction are determined. The basic idea involves the addition of points to the CHGIS datasets, based on their existence in a particular county level jurisdiction in 1911. By judging which historical jurisdiction those points lie within, an approximate boundary line including those points can be drawn.
The original maps produced in 1911 were not drawn to exact scale or known projection, but were created using a surveying and calculation technique which varied from region to region. Because of the scale and projection anomalies it is not possible to use the “rubber-sheeting” method to digitize the maps into GIS. Rubber-sheeting is the process of stretching and skewing a scanned digital image of a map so that a sampling of particular locations on the map match their correct positions in a known projection.
Rather than scanning and rubber-sheeting the 1911 maps, the original maps, (such as the example shown in the following image), are consulted alongside the GIS workstation.
For sections of the 1911 boundary which are determined to be the same as 1990 boundary lines, those line sections are digitized by copying the existing arc segment to a new GIS layer. Similarly some of the historical boundary sections can be matched to sections of rivers, and these river arc segments are copied to the new layer. For parts of the border which do not correspond directly to the 1990 boundaries or the river lines, the points shown as lying within the jurisdiction of the 1911 county are carefully noted and compared with contemporary points. Many of these points can be snapped to their contemporary 1990 positions. Once the points known to have been part of the 1911 county have been digitized into the GIS layers, the remaining boundary sections encompassing those points can be drawn.
The finished 1911 county boundary line will therefore include information derived in at least three ways: