Brief Introduction to China’s 1:1,000,000 Soil Database and Soil Reference System

Shi Xuezheng Yu Dongsheng and Pan Xianzhang
(Institute of Soil Science, The Chinese Academy of Sciences, Nanjing, China)

The Institute of Soil Science, Chinese Academy of Sciences, has successfully produced a 1:1,000,000 soil database of China, which is composed of 3 parts, 1.) soil spatial data – the 1:1,000,000 digital soil map of the country, 2.) soil attribute data and 3.) Chinese soil reference system. The soil spatial data are the most detailed countrywide digital soil map available and the Chinese soil reference system has just been completed as a result of years of arduous research. The database is applicable to scientific research and education and is also an indispensable fundamental data in conducting national or regional modernized agricultural production, rational exploitation of land, water and forest resources, environment protection and ecological restoration. It will certainly facilitate international exchange and cooperation.

1. The second national soil survey and compilation of the 1:1,000,000 soil map of China
Beginning in 1979 and ending in 1994, the second national soil survey lasted 16 years. With over 84,000 leading members and soil scientists as well as technical workers taking part in the activity, the survey covered a total of 2444 counties, 312 state-run farms (ranches or forest farms) and 44 forestry zones. The survey began with a detailed investigation at the county and township level guided by the country’s uniform soil survey technical regulations and the soil classification system. Soil samples were collected and detailed large-scaled soil maps were plotted. Then they were scaled step by step upwards to prefectural, city, provincial and eventually the national level, forming a general map.

In South China, soil maps of agricultural regions at the base level were drawn generally on the scale of 1:10,000 (some at the scale of 1:5,000 or 1:20,000). In North China, agricultural maps were produced at scales of 1:10,000 or 1:25,000, while maps of forest or pasture regions were generated at the scale of 1:50,000. Maps of grasslands and deserts in the Xizang Plateau and Xinjiang were created at a scale of 1:200,000. The soil survey was conducted by well-trained, specialized teams. In advance of the field work, preliminary maps were generated from aerial and satellite soil image interpretation. The preliminary maps helped insure a consistent and accurate basis for the ground survey.

Compilation of the 1:1,000,000 soil map of the country was initiated in 1986. As a first step, 7 typical regions were selected for detailed map plotting. In the process of plotting the maps, comparisons were made between different mapping units, and an assessment was made of the effect of different mapping interpretations. Based on the analysis and study of the sample maps and a number of small-scaled soil maps gathered from inside and outside the country, “Criteria for plotting China’s 1:1,000,000 soil map” were formulated in 1990 and the compilation of the 1:1,000,000 soil map of the People's Republic of China was formally started in 1992.

Compilation used provincial-leveled soil maps prepared by provinces, metropolises, and autonomous regions, topographic maps of the same scale and satellite images as its fundamental base. In addition, extensive reference to soil records, geological maps, forest distribution maps, landuse maps and old soil survey data such as North China Plain soil map, etc. of various provinces were made. The soil maps prepared for different provinces differ in scale. For instance, the soil maps for the western part of China are dominantly of the scale of the 1:1,000,000 and for the eastern part 1:500,000. The soil maps of Taiwan, Hong Kong and Macao were directly formed through interpretation of satellite images in reference to the soil distribution pattern and satellite image interpretation marks of Guangdong Province. In the 1:1,000,000 soil map of the People's Republic of China, there are altogether 909 soil mapping units and 4 non-soil formations. The minimum mapping units for forest and pasture areas are 25 mm2 in map units, 25 square kilometers in real world units. Agricultural areas and regions of great productivity or academic interest were mapped with minimum units of 16 and 4 mm2, respectively. These translate to 16 and 4 square kilometers in real world coordinates. This is the first set of 1:1,000,000 soil maps that have been prepared based on the solid work conducted in the past decades and a rich resource of supporting ancillary data.

2. 1:1,000,000 digital soil map of the country
On the basis of the “1:1,000,000 soil map of the People's Republic of China” published by the National Soil Survey Office in 1995, the 1:1,000,000 digital soil map of the country was compiled and completed by a research group (including Prof. Shi Xuezheng, Asso. Prof. Yu Dongsheng and Pan Xianzhang) through digitizing and edge matching. The 1:1,000,000 digital soil map of the country is composed of 64 standard sized maps. The digital soil map truly reflects the features of the original soil map and has inherited the mapping units used for the original map, which are mostly soil families. There are a total of 12 soil orders, 61 great groups, 235 subgroups and 909 families. Listed in Table 1 and Table 2, respectively, are the projection type and parameters and basic elements of the layers.

Table 1 Geodetic coordinate system used in the digital soil map

Projection Lambert conic Conformal  
Spheroid KRASOVSKY    
  Units Meters  
Parameters:      
1st standard parallel 25 0 0.000
2nd standard parallel 47 0 0.000
Central meridian 105 0 0.000
Latitude of projection’s origin 0 0 0.000
False easting (m)   0  
False northing (m)   0  

The digital soil map has three types of layers, which represent the polygon, line and point shape of ground objects, see Table 2.

Table 2 Basic elements of digital soil map

Layers of the map Content
Polygon shape of ground objects Soil types; water surface (rivers, lakes, reservoirs)
Line shape of ground objects Monoline rivers; roads (highway, railway); administration boundary lines; the Great Wall.
Point shape of ground objects Cities; vertical control points

The soil map digitization process included multiple steps. First, the polyester film maps used for publication of the soil atlas and the published colored paper maps formally were scanned separately. Polyester film maps are less deformable than the paper maps, and so were used in the actual digitizing of soil units. The paper maps were used as a reference for rectification of the polyester maps. ARCVIEW software was used to digitize the maps. The digital files were converted into ARC/INFO format for editing non-point information, including deleting dangles and erroneous labels and eventually converting the ARCVIEW files into polygon layers. The edited polygon format files were then converted into the ARCVIEW form for comparison with the rectified graphs of the colored paper maps to detect errors in the line rendering and mistakes in soil attribute coding. The edited polygon layers were checked again under ARC/INFO to ensure the accuracy of boundary and attribute information.

3. Soil attribute data
The soil attribute data in the China’s 1:1,000,000 soil database are taken from the “Soil Series of China”, which consists of 6 volumes and covers nearly 3000 soil series. The data of each soil series can be divided into two parts, the description of soil properties and the soil analytical data. In the former, attribution and distribution, major characters, typical soil profile and productivity of each series are depicted. The section of “attribution and distribution” addresses which soil order, great group, subgroup and the soil family should belong to and where it should be distributed; the section of “major characters” deals with what kind of parent materials it is derived from and general information about the series. General information includes structure of its profile, thickness of its soil layer, etc. The section titled “typical soil profile” discusses exact location, altitude and parent material of the profile, related meteorological data such as annual mean temperature, etc., and natural vegetation or crops at the site it is located. In addition, it also provides detailed field description of every layers in the profile, for instance, color, texture and structure of the soil and distribution of plant root systems, etc.. The analytical data varies from soil type to soil type, but can be generally sorted into physical properties such as soil particle size composition, soil texture, etc., chemical properties such as pH, organic matter, CEC, exchangeable base, exchangeable H, exchangeable Al, etc., and soil nutrients such as total N, total P, total K, available P, available K, etc.

4. Integration of the Chinese soil database
Soils differ sharply in spatial distribution according to the joint effect of landscape, parent material, climate, vegetation and other various factors. It is hard to get satisfactory results by extrapolating data obtained through investigation, test and analysis of the soil at the profile site to an area in a small-scaled soil map using spatial interpolation methods. How to apply the soil typing method in making use of the attribute data of the soil profiles to characterize accurately spatial distribution of soil properties proved to be a complex problem. For that purpose, well-experienced soil taxonomists, veterans of the second national soil survey, were invited to match the attribute data of every soil profile with spatial units of the soil map by the soil typing method. Soil typing involved linking the attribute data of every soil profile to soil map units with the same soil title. For the sake of easy operation and accurate linkage, this work was performed on a province-by-province basis, with county (city) as control regions. The “Soil Series of China”, the fruit of the second national soil survey and the 1:1,000,000 soil map of the country are basically consistent in dividing soil great groups and subgroups. Consequently, the matching at the soil great group and subgroup levels did not present any substantial obstacles. It was quite the other way round with the matching at the soil family level, because the names of soil types were not uniform and differed sharply between soil profiles and spatial units on the soil map. Therefore, taxonomists of the project team had to refer to soil and soil series records kept in their respective provinces, cities and regions, to define accurate links between attribute data of soil profiles to corresponding soil map units. These judgments were made primarily on the similarity of soil parent material or type, consistent or adjacent in location and distribution of the soil profiles. Thus a soil database merging both the attribute data of the soil profiles and soil spatial data of the soil map was completed.

5. Chinese soil reference system
Soil classification is the basis for integrating soil data as well as for exchanging and extrapolating findings between soil science and other related disciplines. As of yet, however, the world has not adopted a uniformly accepted soil classification system. Among various systems in the world, Soil Taxonomy (ST) of the USDA and the WRB (World Reference Base for Soil Resources) formulated jointly by FAO and some other international organizations are the most influential and have become the mainstay soil classification systems. As such they are often used as a basis and medium for international academic exchange. That is to say, during the process, whenever a soil is mentioned, the academic terms of the two classification systems have to be used. The overwhelming majority of the soil data that China has accumulated up to now, no matter whether they are soil spatial data – soil maps or soil profile attribute data, use terms of the “Soil Genesis Classification System”, such as red soil, paddy soil, brown earth, cinnamon soil, etc.

The genesis based system debuted when China began her study of soil science in the 1930s and grew stronger and stronger under the influence of the soil geographic genesis school after the foundation of the new China. By 1978 it was fairly comprehensive and was used as an important foundation for the second national soil survey, which further promoted its development. This system, however, is completely different from the “Chinese Soil Taxonomy” now still under study, the USDA Soil Taxonomy and WRB, which makes it very difficult for Chinese soil scientists to participate in international academic exchange and for soil scientists of other countries to understand the soils of China.

To overcome this obstacle, we are endeavoring to set up a “Chinese Soil Reference System”, which is based on data and information gathered during field investigations of approximately 3000 soil profiles of different series all over the country and test results of the soil samples collected from the profiles (published in Soil Series of China in 6 volumes). Soil scientists well-experienced in soil classification are organized to sort out respective attributions of each soil in the four systems by referring to diagnostic horizons and diagnostic characteristics of the four systems, thus setting up a reference system between them. Once it is completed, you, no matter whether a Chinese or a foreigner, a soil scientist or a non-soil scientist, can easily find out the names of a soil in other three systems as long as you know the name of the soil in the Chinese Soil Genesis Classification System and type it into the “Chinese Soil Reference System”. Such a system will surely promote academic exchange between Chinese and foreign soil scientists.

6. Various soil-classification-system-based versions of China's 1:1,000,000 soil database
Whether in international exchange or in higher education, and in China or in other countries of the world, soil scientists and scientists in other related fields are all eager to learn the distribution patterns and regional characteristics of the soils of China through the use of USDA Soil Taxonomy and WRB. The adoption of the two systems by China facilitates academic exchange, international cooperation and higher education. For that purpose, we are going to set up different versions of the “China's 1:1,000,000 soil database” based on different soil classification systems, i.e. USDA Soil Taxonomy, WRB and Chinese Soil Taxonomy. The process of the setting up is to link the soil reference information of the above-mentioned nearly 3,000 soil profiles to the 1:1,000,000 digital soil map of the country, thus forming a version of “China’s 1:1,000,000 digital soil map” using USDA Soil Taxonomy, a version using WRB and a version using Chinese Soil Taxonomy. These different versions of digital soil maps can also be linked to the soil profile attribute data, thus eventually forming four different versions of China's 1:1,000,000 soil database, i.e. “Chinese Soil Genesis Classification System”-based version of China's 1:1,000,000 soil database, “Chinese Soil Taxonomy”-based version of China's 1:1,000,000 soil database, USDA “Soil Taxonomy”-based version of China's 1:1,000,000 soil database and WRB-based version of China's 1:1,000,000 soil database, which will be very convenient tools for boosting international academic exchange and cooperation and higher education in countries all over the world.

Chinese Soil Database