Soil TypeDatabase of China —the National Soil Dataset based on the Second National Soil Survey

No comments yet

Submit questions or advice:



You are not logged in, please[Login]or[ Register]!

Soil TypeDatabase of China —the National Soil Dataset based on the Second National Soil Survey

Total number of views and downloads

View in HTML Paper download
2426 17

Soil TypeDatabase of China —the National Soil Dataset based on the Second National Soil Survey

The author's papers

Sorry, failed to retrieve the author's related papers.

            Data source: Chinese Science Citation Database(CSCD)

Soil Type Database of China: A nationwide soil dataset based on the Second National Soil Survey

Shi Jianping1*, Song Ge1

1. Institute of Soil Science, Chinese Academy of Sciences, Nanjing 210008, P. R. China

* Email: jpshi@issas.ac.cn

Abstract: Derived from the Second National Soil Survey, Soil Type Database of China is the most comprehensive soil data resource on the national scale. It embodies the distribution, area, characters, landuse and productive performance of most soil types throughout the country in the 1980s. The database contains 2,473 typical profiles, 8,751 soil genetic horizons and the corresponding physico-chemical properties. The project was initiated by the Institute of Soil Science, Chinese Academy of Sciences (CAS) in 2000. Up to now, an entity-relationship (E-R) data model has been built which contains soil type-location relationships and stratified classifications based on Genetic Soil Classification of China (GSCC, GB/T 17296), and the data products have been developed to cover most of the soil types in China. The database can be widely used in land degradation assessments, environmental impact studies, and soil carbon reserves studies. It may also be used as basic data to guide agricultural production.

Keywords: soil; Soil Type Database; Second National Soil Survey; data integration

Database Profile

Chinese title

中国土种数据库——基于第二次土壤普查的全国性土壤数据集

English title

Soil Type Database of China: A nationwide soil dataset based on the Second National Soil Survey

Corresponding author

Shi Jianping (jpshi@issas.ac.cn)

Data author(s)

Shi Jianping, Song Ge

Database composition

The database is derived from the Second National Soil Survey. It embodies the distribution, area, characters, landuse and productive performance of most soil types throughout the country in the 1980s. The database contains 2,473 typical profiles, 8,751 soil genetic horizons and the corresponding physico-chemical properties. It is made up of nine tables: basic information table, typical profile landscape table, typical profile physico-chemical properties table, statistical profile physico-chemical properties table, province name table, county/city name table, location-soil local type relationship table, soil group table, and soil subgroup table.

Time range

1978 – 1984

Geographical scope

China (excluding Taiwan)

Data format

Excel

Data volume

7 MB

Data service system

 

Source(s) of funding

The “Science Data Bank” Project of the 11th and 12th Five-Year Informatization Construction Special Programs of CAS (XXH12504-1-02), and Field Frontier Program of the Institute of Soil Science, CAS (ISSASIP1626)

1. Introduction

Being the material base of agriculture, soil is one of the most strategic natural resources for a country. In China, there are diverse soil types due to complex natural conditions and profound anthropic factors. In soil science, natural conditions influencing soil genesis can be summarized as five soil-forming factors: topography, climate, parent material, vegetation and soil age. In different areas, these factors manifest different ingredients and features, interacting with different intensities to form a wide variety of soils. Soil local type is the collection of a set of soil pedons with similar characteristics. Soil pedons with distinctive features and clear-cut boundaries are selected to represent the soil local types which these soil pedons belong to. These representative soil pedons provide references for estimation of soil characters and discrimination of soil local types.

Soil Type Database of China comes from Soil Species of China1, the final outcome of the Second National Soil Survey. Soil Species of China was compiled through arranging, screening and summarizing all soil type documents at province, city and district levels, providing the most comprehensive national soil data information until now. Covering 2,473 typical profiles of soil local types, the book records the affiliation in soil classification system, distribution, area, major characters, typical profiles, productive performances and physico-chemical properties of most soil types throughout China in the 1980s.

Soil data has both temporal and spatial attributes. Temporally, soil properties change under the influence of natural conditions and human activities. Spatially, soil types vary in both horizontal and vertical directions. Different layers of a pedon even manifest different physico-chemical properties. Like plant taxonomy, the quantitative and standardized classification of soil types facilitates the interpretation, simulation and comparison of data of different soils. The Second National Soil Survey spanned from 1978 to 1984. As it progressed differently in different regions, the temporal change of soil survey data cannot be considered. The soil spatial distribution and classification should be considered simultaneously in the structural design for Soil Type Database of China.

The database takes soil local type as the unit to extract such information as affiliation, distribution, topography, landuse, major characters, as well as physico-chemical properties of typical and statistical profiles. The physico-chemical properties of typical profiles involve 28 parameters in multiple genetic horizons including soil nutrients (soil organic matter, total nitrogen, total phosphorus, total potassium, alkali-hydrolyzable nitrogen, available phosphorus, readily available potassium and pH), soil physical properties (particle-size composition, texture, bulk density and porosity) and soil chemical properties (exchangeable cations, cation exchange capacity (CEC), effective cation exchange capacity (ECEC) and calcium carbonate). Referring to the structural models of similar soil databases in other countries2 – 3, the horizontal distribution relationships between locations and soil types were constructed. Spatially, 2,473 soil local types were distributed among 1,642 counties in 30 provinces excluding Taiwan. The vertical distribution relationships between soil local type and genetic horizons were constructed. Vertically, 8,751 genetic horizons data were collected from 2,473 typical profiles, with 3.5 horizons per profile on average. According to GSCC, the stratified classification relationships among soil groups, subgroups and soil local types were constructed. By means of the entity-relationship (E-R) data model, users can conduct queries or retrievals by soil classifications and/or locations. The database can be used for land degradation assessments, environmental impact studies, and soil carbon storage researches, or used as basic data to guide agricultural production.

2. Data collection and processing

2.1 Soil local type classification

Soil local type is the basic category of GSCC. It is a cluster of multiple soil pedons which come from the same or similar landscape, experience similar hydrothermal conditions, possess relatively consistent morphological characteristics in soil profiles, and have approximate physico-chemical and biological properties. To categorize soil local types should comply with soil genesis theory and soil attributes.

(i) Soil local type should be identified by the relatively stable attributes of a soil pedon. For a soil local type, the soil attributes are similar and the attribute values are in the same magnitude order. Among different soil local types, soil attribute values differ in magnitude order.

(ii) A soil local type is a tessera, which shares similar micro landscapes and hydrothermal conditions, as well as identical parent material, vegetation and landuse.

(iii) For a certain soil local type, the arrangement order and thickness of soil genetic horizons (or other soil layers) in soil profile are similar.

(iv) Soils of the same soil local type show the same soil morphological characters and equivalent soil development level.

(v) Soils of the same soil local type have similar productive performance and potential.

2.2 Content standardization

Soil Species of China is one of the serial outcomes of the Second National Soil Survey. It was compiled through arranging, screening and summarizing the soil type documents at province, city, and district levels. The book contains six volumes introducing the affiliation in GSCC, distribution, area, major characters, typical profiles, productive performances, and physico-chemical property data of major soil types in six regions of China. The standardized contents are as follows:

(i) Soil local type nomenclature. Nomenclature methods for soil local type were different among different provinces. Some provinces used popular or vernacular names, while others adopted continuous nomenclature. Soil Species of China uses uninominal nomenclature. To avoid multiple names for the same soil local type or the same name for different soil local types, a geographical name would be added before the name of soil local type. Taking Qishan-Chengnitu (祁山橙泥土) as an example, Qishan (祁山) is a county name. In some regions of China, paddy soil was named with a suffix “田” (field), which has been retained in this database, such as Chaoshanitian (潮砂泥田), Niroutian (泥肉田), Baifentian (白粉田), etc. 

(ii) Soil type affiliation. Shallow mountain soils without B horizon development are identified as lithosol or skeletal soil. Shallow mountain soils with poor B horizon development are still identified as their original soil groups, but denominated as weakly developed soil-group-name by adding a character “性土” (weakly developed soil) in their soil-subgroup-names. Soil cultivated on reticulated mottling horizon in south China still belongs to a soil group of red earth, although its B horizon has been eroded by water. However, the cultivated red clay in northwest and northeast China is identified as a subgroup of red primitive soil. Soils with A-C profile pattern developed on limestone area of south China are categorized into the soil group of limestone soil, while those developed in north China are categorized into automorphic soil. For paddy soil, subgroups are discriminated according to the ratio of crystalline iron oxides to amorphous iron oxides in soils.

(iii) Symbol of soil genetic horizon. Considering different symbols used by different soil type records, each soil genetic horizon was given a uniform symbol in consultation with relevant documents. In Soil Type Database of China, the symbol of soil genetic horizon has been unified as shown in Table 1.

Table 1  Symbols of soil genetic horizons

(iv) Major characters. Major characters of a soil type include soil genetic characters and soil nutrient conditions. Soil genetic characters involve the description of profile pattern, for instance, A-B-C for brown earth, A-Ah-B-C for black loessial soil, and Aa-Ap-P-C for percogenic paddy soil. Nutrient condition refers to the statistical description of topsoil nutrient contents, including organic matter, nitrogen, phosphorus, potassium and microelements. 

(v) Typical profile description. A typical profile embodies the essential characters of its soil local type. Description of the habitat condition of a typical profile includes its sampling site, topography, altitude, parent material or derived soil, vegetation, landuse and meteorological indicators. The database standardizes morphological descriptions of typical profiles, such as calibrating soil color by Munsell soil chart. Soil texture is amended according to the International Society of Soil Science texture classification and the laboratory analysis results from the Second National Soil Survey.  

(vi) Almost all soil local types have both typical and statistical profile data on physico-chemical properties, which are compiled into two tables separately. A few soil local types only have a physico-chemical properties table of typical profile, which is retained because it reflects the essentials of the soil local type.

2.3 Analysis methods

According to the Technical Essentials for the Second National Soil Survey (Draft) issued by the Ministry of Agriculture4 – 5, soil nutrient analysis adopts the following methods as listed in Table 2.

Table 2  Routine analysis methods for soil nutrients

3. Database structure and example

3.1 Database structure                 

Referring to the structure of domestic and foreign soil databases6 – 9, a relational database is constructed, which involves spatial distribution and stratified classification relations, and provides queries based on location or GSCC.

The database is made up of five parts:

(i) SOIL_TYPE, a basic information table for soil local type. The table consists of soil local type ID, soil local type name, general description, parent material, profile pattern, major characters and production performances. In addition, it records affiliated soil group ID and sub-group ID for linking to soil classification information. 

(ii) PROFILE_LANDSCAPE, a landscape table for typical profiles. A typical profile corresponds to a soil local type. PROFILE_LANDSCAPE table links to SOIL_TYPE table by soil local type ID. PROFILE_LANDSCAPE table contains sampling site of a typical profile, topography, elevation, parent material, mean annual temperature, annual precipitation, natural vegetation, landuse, and so forth.

(iii) VALUM 1, a physico-chemical properties table for typical profiles; and VALUM_sta, a physico-chemical properties table for statistical profiles. Vertically, the profile of a soil local type consists of several soil genetic horizons with different depths and distinct physico-chemical properties. Therefore, there is a one-to-many vertical distribution relationship between a soil local type and its soil genetic horizons. Both VALUM 1 table and VALUM_sta table embody total soil nutrients, available soil nutrients, physical properties, exchange properties, and so on. The two tables link to SOIL_TYPE table by soil local type ID.

(iv) LOCATION_NAME, a table for province names; sublocation, a table for county/city names which contains soil local types; and LOCATION_USAGE, a location relation table for soil local types. The three tables describe the spatial distribution of soil types. Because a province has many counties or cities, LOCATION_NAME table and sublocation table are associated by a one-to-many relationship. A county or city may have various soil local types, and a soil local type can distribute in many counties or cities. Therefore, there is a many-to-many relationship between county/city and soil local types, which is expressed by the LOCATION_USAGE table containing soil local type ID, soil local type name, county/city code, county/city ID, county/city name, and so forth.

(v) SOIL_GROUP and SOIL_SUBGROUP, two tables involve classification relations in GSCC. SOIL_GROUP table includes soil group name in Chinese, soil group name in English, soil order name in Chinese, soil order name in English, and soil group description. SOIL_SUBGROUP table contains subgroup names in Chinese and English. A soil group has many subgroups, so soil groups and subgroups constitute a one-to-many relationship. Likewise, a subgroup has many soil local types, so subgroups and soil local types are also in a one-to-many relationship. SOIL_GROUP table and SOIL_SUBGROUP table link to SOIL_TYPE table by soil group ID and subgroup ID respectively, which embodies the hierarchical relationships in soil classification system. Queries can be sequentially performed via the standardized soil group, subgroup and local type names.

Data snapshots for soil type information are generated to facilitate rapid queries via location and/or soil classification.

3.2 Example usage

Due to the large volume of this database, location and soil classification queries were taken as examples to illustrate the database characteristics, and inform users to trace, locate and understand the datasets. Below, we generate typical samples of Nichitu (soil local type) under latosolic red earth (soil group) in Zhangzhou city, Fujian province.

There are five soil local types in Zhangzhou (county/city ID 1): Nichitu (泥赤土), Chiniantu (赤黏土), Huangchitu (黄赤土), Xiatan-Chiniantu (霞潭赤黏土) and Qingdihuinitian (青底灰泥田) with soil local type ID 10003, 10005, 10007, 10009, and 10503, respectively. In LOCATION_USAGE table, the distribution relations are formed by linking county/city ID with soil local type ID (Figure 1 & Table 3). By the distribution relationship, it is found that Nichitu (泥赤土) distributes at Zhangzhou (county/city ID 1), Quanzhou (county/city ID 2) and Hua’an (county/city ID 23).

Figure 1  Example for location-soil local type relationship

Table 3  Location query results by Zhangzhou city, Fujian province

If we select Nichitu (泥赤土) under the subgroup of typical latosolic red earth, detailed information of this soil local type can be displayed. If we link it to  the soil local type ID of Nichitu (泥赤土), landscape information (Table 4) and physico-chemical properties (Table 5) of the typical profile of Nichitu (泥赤土) can be acquired.

Table 4  Landscape information of the typical profile of Nichitu (泥赤土)

Typical profile of a soil local type contains several genetic horizons. Each genetic horizon has different depths and physico-chemical properties. There is a one-to-many relationship between a soil local type and its soil genetic horizons. Physico-chemical properties of all soil genetic horizons in the typical profile of Nichitu (泥赤土) is listed in Table 5.

Table 5  Major physico-chemical properties of Nichitu’s (泥赤土) typical profile

In GSCC, a soil group contains many subgroups, and a subgroup contains many soil local types. Figure 2 shows how soil group, subgroup and soil local type correlate to each other on selected samples. In this way, queries can be performed via soil group, subgroup and soil local type in sequence. For example, if we select latosolic red earth, typical latosolic red earth and Nichitu (泥赤土) in order from soil group, subgroup and soil local type in the soil classification query, some basic information of Nichitu (泥赤土) can be extracted, as shown in Table 6.

Figure 2  Example for soil group-soil local type relationship

Table 6  Query results according to Genetic Soil Classification of China (GSCC)

Tables 4 and 5 show landscape information and physico-chemical properties of the typical profile of Nichitu (泥赤土) respectively. The physico-chemical property table for statistical profiles records the statistical values of multiple soil profile samples, which contains the same data fields to the physico-chemical property table for typical profile.

4. Data quality control

4.1 Data quality assurance

The database was completed by the Institute of Soil Science, CAS in 2010. Data integrity and consistency were validated through random manual inspection. County/city names were updated from 2011 to 2013. The data fields, units and values were reviewed according to the original documents, and calibrated in accordance with the legal units of measurement of China in 2014. Besides, a set of database design files was archived, including data dictionaries and structural diagrams.

Annually, the database contents are checked and errors corrected in response to users’ feedback. All maintenance records have been archived in log files. 

4.2 Measures of data quality control

Measures have been formulated to address the following aspects.

(i) Legal units of measurement conversion

In the physico-chemical properties tables of typical and statistical profiles, the units of soil organic matter, total nitrogen, total phosphorus and total potassium were converted from % to g/kg. The original fields of raw data were retained, and new data fields were added with the legal units of measurement.

All the units were converted into legal measurement units: soil particle-size composition was unified by the International Society of Soil Science texture classification and in unit of percentage (%); the unit of soil bulk density was g/cm3; and both CEC and exchangeable cations were unified in unit of cmol/kg (+).

(ii) Expression of sampling depth

In raw data, the depth of genetic horizon is expressed by relative thickness (e.g., 20 cm). It is difficult for users to distinguish absolute depth and depth scope. After all data was imported and validated, two data fields were added to designate the top and bottom depths (e.g., 0 cm and 20 cm) of the genetic horizon referring to ISO 28258. 

(iii) Standardization of soil classification

To standardize soil classification within existing data resources is an important measure of data integration in soil sciences. In accordance with Classification and Codes for Chinese Soil (GB/T 17296-2009), the names of soil group and subgroup in the database were standardized. When a subgroup’s name is exactly the same with the name of the soil group to which it affiliates and this soil group includes other subgroups in Classification and Codes for Chinese Soil, Chinese characters “典型” (typical) are added before the original name of this subgroup. For example, “typical red earth” is used to designate the subgroup of “red earth” in the soil group of “red earth”. Totally, 41 subgroup-name changes have been made because of the duplication between the names of subgroup and soil group. In the database, a new field “national standard subgroup name” was added into the SOIL_SUBGROUP table, and the original subgroup name field was retained in order to match the original data. However, soil local type’s affiliation in the soil classification system succeeded to the original document.

In metadata, all soil type names were listed, and soil references between GSCC and Chinese Soil Taxonomy (CST) were given.

(iv) Update of administrative unit

Since the 1980s, the scope and name of some administrative divisions have greatly changed. For example, Chuansha county of Shanghai was renamed as Pudong New District; Chongqing municipality was established in 1997, meaning that its counties and districts which the soil survey involved historically belonged to Sichuan province. Thirty-eight county/city names were updated, but their original names were retained in another data field. Eight counties’ affiliation was also changed. To facilitate the query by location, the administrative division codes of province, city and county (GB/T 2260-2007) were added, and the approximate latitude and longitude of city/county were also given.

(v) Agreement on data type

In the database, data types are given mostly in text, number and note. The precision of numeric data is specified based on the original documents. For the fields with substantial raw data missing, the data type is set as text because that null value in digital fields is likely to become zero during data transmission.

5. Usage notes

Users should previously know the data’s distribution, scale and application scope. The database is based on Soil Species of China, whose raw data does not cover all the counties and cities of China. Typical profiles of 2,473 soil local type are spread among 1,642 counties/cities/districts or other geographical areas (such as Taihang Mountains, Greater Khingan Mountains).

Figure 3 shows the distribution of soil local types among 30 provinces. Liaoning possesses the largest number of soil local types (159), whereas Beijing has the lowest (11). The highest density of typical profiles is in Shanghai (36 per 10,000 km2), in contrast to the lowest in Xinjiang (0.6 per 10,000 km2), and the national average is 2.6 typical profiles per 10,000 km2.

Figure 3  Distribution of soil local types among 30 provinces of China

Table 7 shows the distribution of soil nutrients data of typical profiles among 30 provinces. In sum, the database provides 28 kinds of soil nutrients and their physico-chemical properties data from 8,751 genetic horizons of 2,473 typical profiles.

Table 7  Distribution of soil nutrients data of typical profiles among 30 provinces of China

The distribution of soil physico-chemical property data among 30 provinces is also statistically analyzed. Soil particle-size composition is graded according to the International Society of Soil Science texture classification. Results show that Liaoning has the largest particle-size composition data volume and Beijing the smallest (Figure 4). Soil bulk density was less commonly surveyed among these 30 provinces, and even not determined in 11 provinces of northwest and southwest China (Figure 5). The total CEC data volume was 5,826 in China, to which Liaoning contributed the largest (518), and Guangdong the lowest (7) (Figure 6).

Figure 4  Distribution of soil particle-size composition data among 30 provinces of China

Figure 5  Distribution of soil bulk density data among 30 provinces of China

Figure 6  Distribution of soil cation exchange capacity (CEC) data among 30 provinces of China

Acknowledgments

The study was sponsored by the “Science Data Bank” Project of the 11th and 12th Five-Year Informatization Construction Special Programs of CAS (XXH12504-1-02), and Field Frontier Program of the Institute of Soil Science, CAS (ISSASIP1626). Thanks also go to Zhou Huizhen, Sun Bo, Pan Xianzhang and Xu Jiangbing from the Institute of Soil Science, CAS, who participated in the database construction.

Authors and contributions

Shi Jianping, BS, Senior Engineer; research area: soil information management. Contribution: database design, data model establishment, data import, quality control, and database maintenance.

Song Ge, PhD, Engineer; research area: soil nitrogen cycle and data quality control of long-term soil monitoring. Contribution: data quality control and database update.

References

1. National Soil Survey Office. Soil Species of China (Vol. 1 6). Beijing: China Agriculture Press, 1996.

2. Soil Data Subcommittee, Federal Geographic Data Committee. Soil Geographic Data Standard (FGDC-STD-006), 1997. Available at: [Accessed November 3, 2015].

3. United States Department of Agriculture, Natural Resources Conservation Service. SSURGO Data Packaging and Use, 2012. Available at: [Accessed April 15, 2015].

4. Landuse Bureau of the Ministry of Agriculture of P. R. China. Technical Essentials for the Second National Soil Survey (Draft), in Soil Survey, 1 – 9, ed. Soil Fertilizer Workstation of Hunan Province. Changsha: Hunan Science & Technology Press, 1979.

5. Department of Soil Chemistry, Southwest Agricultural College. Analysis Method in Soil Survey (for the course of Soil Survey), 1981.

6. Shi XZ, Yu DS, Warner ED et al. Soil database of 1:1 000 000 digital soil survey and reference system of the Chinese genetic soil classification system. Soil Survey Horizons 2004 (45): 129 – 136.

7. Nachtergaele F, Velthuizen H & Verelst L. Harmonized World Soil Database (version 1.2), 2009. Available at: [Accessed December 1, 2015].

8. GlobalSoilMap.net Science Committee. Specifications Version 2 GlobalSoilMap.net products, 2011. Available at: [Accessed April 15, 2016].

9. International Organization for Standardization. ISO 28258:2013 Soil quality – Digital exchange of soil-related data. Available at: [Accessed October 1, 2013].

10. Yan X & Cai Z. Number of soil profiles needed to give a reliable overall estimate of soil organic carbon storage using profile carbon density data. Soil Science and Plant Nutrition 54 (2008): 819 – 825.

11. Zhang G, Wu Y & Zhao Y. Physical suitability evaluation of reserve resources of cultivated land in China based on SOTER. Transactions of the Chinese Society of Agricultural Engineering 26 (2010): 1 – 8.

12. Qin ZC & Huang Y. Quantification of soil organic carbon sequestration potential in cropland: a model approach. Science China: Life Sciences 53 (2010): 868 – 884.

Data citation

Shi J & Song G. Soil Type Database of China: A nationwide soil dataset based on the Second National Soil Survey. Science Data Bank. DOI: 10.11922/sciencedb.170.43

 

How to cite this article: Shi J & Song G. Soil Type Database of China: A nationwide soil dataset based on the Second National Soil Survey. China Scientific Data 2 (2016), DOI: 10.11922/csdata.170.2015.0033

Download