Orginal Article

Quantitative estimation and spatial reconstruction of urban and rural construction land in Jiangsu Province, 1820-1985

  • XIE Jinyuan , 1, * ,
  • JIN Xiaobin , 1, 2 ,
  • LIN Yinan 1 ,
  • CHENG Yinong 3 ,
  • YANG Xuhong 1 ,
  • BAI Qing 2 ,
  • ZHOU Yinkang 1, 2
  • 1. School of Geographic and Oceanographic Sciences, Nanjing University, Nanjing 210023, China
  • 2. Natural Resources Research Center of Nanjing University, Nanjing 210023, China
  • 3. School of History and Archives, Yunnan University, Kunming 650091, China

Author: Xie Jinyuan (1991-), MS Candidate, specialized in land resource and management. E-mail:

*Corresponding author: Jin Xiaobin (1974-), specialized in land use and land planning. E-mail:

Received date: 2017-03-04

  Accepted date: 2017-03-30

  Online published: 2017-09-06

Supported by

National Natural Science Foundation of China, No.41340016, No.41671082


Journal of Geographical Sciences, All Rights Reserved


Land cover is the most evident landscape signal to characterize the influence of human activities on terrestrial ecosystems. Since the industrial revolution, the expansion of construction land has profoundly changed the status of land use coverage and changes. This study is proposed to reconstruct the spatial pattern of construction land (urban construction land and rural settlement land) for five historical periods over the past 200 years in Jiangsu Province with 200 m × 200 m grids on the basis of quantitative estimation. Urban construction land is estimated based on data about city walls, four gates along walls, and other socio-economic factors. Rural settlement land is calculated based on the rural population and per capita housing allowance. The spatial pattern of historical construction land is simulated based on the distribution of modern construction land in 1985 with a quantitative-boundary- suitability control method and thorough consideration over connectivity of different land use types. The study concludes that: (1) the amount of construction land in Jiangsu Province is estimated at 963.46 km2 in 1820, 1043.46 km2 in 1911, 1672.40 km2 in 1936, 1980.34 km2 in 1952 and 10,687.20 km2 in 1985; and (2) the spatial distribution of construction land features the great proclivity to water bodies and main roads and the strong polarization of existent residence. The results are verified directly and indirectly by applying the trend verification of construction land changes and patterns, the correlation analysis between rural settlement land and local arable land, and quantitative accuracy test of the reconstructed construction land to actual historical survey maps covering four sample regions in 1936.

Cite this article

XIE Jinyuan , JIN Xiaobin , LIN Yinan , CHENG Yinong , YANG Xuhong , BAI Qing , ZHOU Yinkang . Quantitative estimation and spatial reconstruction of urban and rural construction land in Jiangsu Province, 1820-1985[J]. Journal of Geographical Sciences, 2017 , 27(10) : 1185 -1208 . DOI: 10.1007/s11442-017-1430-4

1 Introduction

Global climate change and human activities act as the dual stressors which drastically change the Earth’s landscapes and increase the fragility of the Earth’s ecosystem. Western industrial revolution, rapid development of science and technology, substantial increase of productivity, population growth, and the expansion of modern cities all lead to the remarkable phenomenon that construction land has become one of the dominant land use types (Bao and Gao, 2016; Saunders, 2012; Yang et al., 2016). Construction land affects the global and regional climate by changing the nature of the underlying surface and intensifying the combustion of fossil fuels (Gu et al., 2011; Yan et al., 2016).
Current researches on the temporal and/or spatial dimensions of construction land changes mainly focus on central cities and employ a comprehensive analysis based on multiple data sources such as historical documents, historical maps, and modern remote sensing images. For example, Liu et al. (2010) collected historical land statistical data and modern remote sensing data to rebuild the urban land statistics from 1700 to 2005 in China while taking population data into consideration; Tian et al. (2014) used remote sensing imagery and historical statistical records as data sources and amended land use data from 1880 to 2010 in India; He et al. (2002) estimated urban built-up areas in 18 provinces in 1820 by using city wall perimeter data recorded in historical records in the Qing Dynasty; Fang et al. (2002), Yin et al. (2013), Ji et al. (2009) and Li et al. (2014) explored the urban changes of Beijing, Shanghai, Guangzhou, and Guiyang respectively in the centennial scale based on historical documents, historical maps and modern remote sensing imagery.
There are two major methods of spatial reconstruction: the top-down “static allocation” and the bottom-up “dynamic evolution”. “Static allocation” is generally accompanied with variables such as historical population density, modern land use pattern, or land suitability. Tian et al. (2014) allocated historical land use types based on the modern population density and historical population data. Liu et al. (2010) framed a historical land use pattern based on historical statistics and modern land use distribution as interpreted from remote sensing imagery. Pan et al. (2015) and Lin et al. (2015) selected the topography, distances from rivers, and other natural characteristics as well as socio-economic factors to structure land suitability models as base maps to allocate historical construction land. “Dynamic evolution” simulates historical land use patterns through successive loop iterations upon grids that incorporate indicators and rules representing human land use behaviors and artificial intelligence algorithms. Schaldach et al. (2011) conducted a land suitability analysis based on factors such as topographic slopes, the road infrastructure, and protected areas by using population change data as the main trigger. Population for each urban and rural grid was then allocated according to population density and estimated land suitability. Whenever the population density for a grid was calculated greater than the set (default) value, the preset land use type was then switched to the other. Ray et al. (2010) progressively rebuilt the construction land in the Muskegen River Basin in Michigan based on land use and transformation characteristics identified from different periods of land use remote sensing data by applying the artificial neural network algorithm. Bai et al. (2007) reconstructed a spatial pattern of land use based on modern topographical maps and historical aerial photographs. Light condition, temperature and water resource were defined as independent variables while the type of historical land use was the dependent variable. The logistic binary regression model was applied to determine the transition probability.
Although these researches have provided some tentative perspectives to reconstruct historical construction land, some aspects in this research domain still remain in an exploratory phase. These aspects include but not limited to the integrity of objects of land reconstruction, identification and quantification of key factors, and refinement of the spatial resolution. According to the typicality of historical background and accessibility of historical data, this study selected five time intersects in the last 200 years during which China was under three different regimes: the year 1820 in the middle of the Qing Dynasty, the year 1911 in the end of the Qing Dynasty and early stages of the Republic of China, the year 1936 in the middle of the Republic of China, the year 1952 in the early stages of the People’s Republic of China, and the year 1985 in the early stages of China’s Reform and Opening up. While construction land (including urban construction land and rural settlement land) is proposed as the study object and the modern geographic coverage of Jiangsu Province is set to be the research scope, this study aims to provide a fresh perspective to rebuild the spatial and temporal pattern for historical construction land with contemporary natural resources and environmental data, historical documents, concerns over urbanization and the human-land relationship. A desired verification method is also provided and discussed. This study attempts to provide methodological references and data support for the larger scale and longer time span reconstruction of construction land or multiple (all) land use types.

2 Characteristics of construction land changes and reconstruction perspectives

2.1 Characteristics of construction land changes

Most contemporary construction land is the outcome of the evolution of settlements. Settlements were initially to meet the basic needs of human habitation and progressively developed into larger plots to accommodate economic development, political evolution, cultural cultivation, and/or life assistance, and/or as a response to defense requirements, and/or religious propaganda (Ma et al., 2016; Zhou et al., 2013). Spatially, settlements generally evolve from “points” to “axes” and then to “surfaces”. Some “points” become regional centers with strong and dense aggregations, which influence and stimulate the development of surrounding areas through the “diffusion effect”. This process generally begins with the traffic vicinity, and forms a central area along traffic lines (“axis”). An overlapping of “axis” and the combination of “axis” and “point” result in a bigger “expansion effect” toward the “surface”, which forms denser and more modern regional traffic networks and urban structures (Hu, 1994).
In this study, the urban construction land includes construction land in cities and towns. Construction land labeling with “city” refers to the land at and above the county level while used for residence, worship, roads, markets, defense, storage, management, and other purposes. Construction land identified as “town” refers to the land below the county level with relatively concentrated economy and population while mainly used for residence, commerce, warehousing, and production.
Historically, urban development was slow in China. Cities and towns were generally small in scale and mostly with explicit boundaries (walls and four gates1(1 Four-gates refer to the handicraft industry and commercial shops that were distributed in the center of a town, and may be considered a city street or street market.)). The handicraft industry, markets, residences, government offices, and temples, were usually distributed in the central urban area. It is common that a certain amount of agricultural land or open space was present between the wall and the central area. With the increase of urban population, the urban land expanded by “filling in” the space within the city walls until the urban space was “saturated”. Subsequently, the urban land would break through the limitation of the city walls.
In the early modern period, the sprout of capitalism intensified the aggregation of various elements in urban space, pushed the expansion of urban scales and the births of new industrial cities and towns, and stimulated the diffusion of urban construction land toward its periphery. Later, urban construction land further expanded to surrounding suburbs associated with the rapid growth of industry and population. In some cases, the cities rapidly expanded to reach the far suburbs. The urban structure tended to develop in a network style and the expansion of urban construction land progressed from the “point-axis” style to a “point-axis- surface” style (Xu et al., 2009).
Rural settlement land refers to the land outside the urban area but not used for agricultural production. It includes farmers’ residential areas, courtyards, and drying areas as well as land for village infrastructure and public facilities (Hu et al., 2007; Jin et al., 2016).
Historically, the development of most rural settlements was slow and many rural settlements’ centers had remained stable for a long time (Jin, 1988; Yang et al., 2016). In the early modern period, the development of agricultural economy was lagging because of little advances in agricultural technologies and the deficiency of policy incentives. The development and the renovation of rural housing were also slow. Most houses were built along the main traffic routes in villages. The rural settlement land typically expanded in a “point-axis” style. With the expansion of the central area, small-scale and scattered settlements nearby were attracted or progressively merged into large settlements. In recent decades (especially after the 1980s), with the improvement of agricultural production conditions, the rapid development of the agricultural economy, and the substantial improvement of living conditions, the building of new houses and renovation of existing houses have become a popular trend. Rural settlement land expands typically in a spreading style, i.e., often occupying surrounding agricultural land directly.

2.2 Perspectives on reconstructing historical construction land

In contrast to modern construction land data, the availability of historical data is affected by the scope, integrity, and accessibility of historical documents. In general, historical construction land data can be obtained by applying any of three methods: (1) reasonable amendments according to the credibility of original data such as historical documents, modern statistics, and other historical resources; (2) estimation with the historical population, structure, per capita land use, and other indirect indictors; (3) calculation with substitute indicators (such as city wall perimeter and shape) related to land use except population data.
This study uses the spatial distribution of construction land in 1985 as the basis to derive the spatial pattern of construction land in four typical historical intersects (1820, 1911, 1936, and 1952) by employing quantitative-boundary-suitability control methods with the connectivity of land use, the different expansion processes and characteristics of urban land and rural settlements during the study period. Figure 1 illustrates the methodology applied in this study.
Figure 1 Technical route of the study
“Quantitative estimation” has been mainly conducted with the examination of historical documents and the acquisition, arrangement, and amendment of related literatures.
(1) Urban land
For the year 1820, the area of a city/town was determined mainly based on data such as the perimeter of city/town walls, the shape of a city/town, the combination of cities/towns, and the four gates surrounding a city/town. With all cities and towns, the amount of urban land can be obtained.
For the other three time intersects of 1911, 1936 and 1952, the amount of urban land was estimated mainly based on per capita land use level with delicate calibration with corresponding socio-economic data and the population size and structure. The amount of urban construction land was then further allocated to cities and towns according to the structure of urban population.
(2) Rural settlement land
In rural areas, per capita settlement land has been inferred from per capita house size and the average house capacity ratio along with the proportion of public construction land. Then the amount of rural settlement land can be calculated based on per capita settlement land and the corresponding rural population for any of our four proposed time intersects.
The spatial grids of historical construction land in Jiangsu Province therefore can be rebuilt with 200 m × 200 m resolution in association with historical records about city walls, construction land suitability evaluation and the connectivity of different land use types.
The results then can be verified in terms of process analysis, correlation analysis and spatial pattern comparison to illustrate the effectiveness of our research outcomes.

3 Research area and data sources

3.1 Research area

The defined research area is Jiangsu Province with the modern administrative boundary. It is located at 116°18′-121°57′E and 30°45′-35°20′N. Its climate is in a transitional zone from semi-tropical to warm temperate. It is of a flat and low terrain with an overall topography of high in the northern and southern parts, low in the central, declining from west to east. Plains are the typical landforms in the area, taking about 68.81% of the total land. It possesses several conjoining rivers and dense water networks. 16.86% of the total land is occupied by water bodies with a coastline of 954 km in length.
Jiangsu Province has been a traditional agricultural production region for centuries due to its mild temperature, great moisture conditions, plain territories, thick soil layers, fertile soil conditions, abundant water bodies, etc. Meanwhile, along with the development of the handicraft industry, urban commodity economy, and the cultural advancement, it has become one of the most developed regions in China since the Ming Dynasty.
Jiangsu Province was officially set up in 1667 (the 6th year of Emperor Kangxi’s reign in Qing Dynasty) and came into its current shape by 1767 (Fu, 2009). Although severe changes occurred to its administrative boundary during our study period, there was little change in terms of administration structure. From the mid-Qing Dynasty (1820) to early years of the Republic (1911), there were seven Fu (cities), four Zhou (cities) and one Ting (city). Specifically, seven units of “Fu” refer to Jiangning, Suzhou, Changzhou, Zhenjiang, Yangzhou, Huai’an and Xuzhou; four units of “Zhou” include Tongzhou, Haizhou, Sizhou and Taicang, and one unit of “Ting” is Haimen. In 1936, an administrative inspection district system was installed. Jiangsu was divided into a number of districts with subordinate counties, i.e. Jiangning, Liyang, Wuxi, Nantong, Jiangdu, Yancheng, Huaiyin, Donghai, Tongshan and Sixian. In 1983, the contemporary administrative system of cities with subordinate counties was put in place, including 11 cities: Nanjing, Wuxi, Xuzhou, Changzhou, Suzhou, Nantong, Lianyungang, Huaiyin, Yancheng, Yangzhou and Zhenjiang. Figure 2 shows historical maps of the administrative structure in Jiangsu Province for these typical historical time intersects.
Figure 2 Changes of administrative division and land use of research area in typical periods

(1. Arable land data derived from Yang et al. (2015); 2. Water areas data drawn from Historical Geographic Information System)

3.2 Data sources

Data involved in this study are of three types: literature and statistical data, land use data, and basic geographical data. Specifically,
(1) Literature and statistical data include data for urban and rural settlements, and population, socio-economic data. These data are primarily used for the quantitative control of the construction land.
3.2.1 Urban construction land data
Perimeters of city/town walls in 1820 are drawn from Jiaqing Rebuilt Chi Unification. The administrative system in Qing Dynasty is from Draft of History of Qing Dynasty (Zhao, 1976) and Geography History Table of Qing Dynasty (Zhao, 1941). Per capita urban construction land around the founding of the People’s Republic is from Urban Historical Geography of Jiangsu (JGI, 1982). Data about four gates along the city walls in the Qing Dynasty is derived from Fan (1990).
3.2.2 Rural settlement data
Per capita rural house size in 1933 is from Bo (1941). The rural house capacity ratio is drawn from Song et al. (2008) and Lin et al. (2015).
3.2.3 Population data
City (Fu) population (household) data in 1820, 1911, 1936 and 1952, are adapted from Cao (2002), Population Geography Information System (, and Fifty Years of Jiangsu Province (EB, 1999). The population structure data is based on Wang (1984) and Fifty Years of Rural Economy of Jiangsu Province (EB, 2000).
3.2.4 Socio-economic data
Associations and chamber of commerce data at the end of the Qing Dynasty and early years of the Republic are drawn from Wang (1984). Rural agricultural company data for the end of the Qing Dynasty and early years of the Republic are from Archives Compilation of History of Republic of China (Third Series Agriculture and Commerce) (SHAC, 1998a). The gross value of urban industrial products in 1930 is from Archives Compilation of History of Republic of China (Fifth Series Finance and Economy) (SHAC, 1998b). Rural cooperatives data in 1930 are selected from archival files from The Second Historical Archive in China and an adaption of Sun (2009). Urban added value data of the secondary and tertiary industries in 1952 are derived from Fifty Years of Jiangsu Province (EB, 1999). Rural industrial labor force data in 1952 is also from Fifty Years of Rural Economy of Jiangsu Province (EB, 2000).
(2) Land use data includes modern land use spatial pattern, historical distribution of arable land and water bodies. It is expected to define spatial boundaries of construction land.
The spatial data of historical arable land data is based on Yang et al. (2015). Data of historical water bodies is drawn from Historical Geographic Information System (, National Provincial Map of Republic of China in 1933, and Atlas of Jiangsu (EG, 1978). Modern (1985) land use data is from the Earth Resources Observation and Science Center (EROS) ( The urban land use efficiency data (i.e., vacancy ratio) in Qing Dynasty is extracted from Atlas of Ancient China (Qing Dynasty) (Cao et al., 1997) and the survey map of Jiangsu Province in the 1930s.
(3) Basic geographical data includes administrative divisions and topography, which were used for the determination of spatial distribution of construction land.
Historical administrative boundaries and scopes are drawn from Historical Geographic Information System. The elevation and slope data are from the International Scientific Data Mirror Site ( Road data is from National Provincial Map of Republic of China in 1933 ( and Transportation History of Jiangsu Province (TCCC, 1995).

4 Research methods and spatial pattern reconstruction

The reconstruction of spatial pattern for historical construction land in four typical historical intersects (1820, 1911, 1936, and 1952) includes two procedures, i.e., quantitative estimation and spatial reconstruction. Quantitative estimation provides basis of quantitative control, along with boundary and suitability control, the spatial pattern of historical construction land in Jiangsu can be reconstructed.

4.1 Quantitative estimation of construction land in typical time intersects

4.1.1 Quantitative estimation of urban construction land
Cities and towns are highly concentrated with population, resources, environment, and socio-economic elements. Urban construction land expands along with the development of any of these elements. Such expansion has the characteristics of temporal sequentiality, periodicity, irreversibility, and spatial particularity. Between 1820 and 1911, the urbanization level (i.e., the proportion of urban population) was maintained at around 10% (Li and Xu, 2008). Most cities were political centers established by the governments and had specially constructed boundary walls. Therefore, it is quite anticipated that the urban population density might fluctuate gradually within city walls and so does the value of per capita urban construction land, the estimation of urban construction land by using data such as city walls and four gates along walls can escape meddles with population changes.
In this study, urban construction land in 1820 is estimated mainly based on the perimeter and shape of city/town walls. Data of four gates along the walls and the land vacancy ratio representing arable land and open space that was present within the walls are also used in case for calibration. Moreover, the estimation of the urban construction land at the Fu (city) level also takes the special form of urban administrative divisions (for example, “Fu” and “County” in the same city, or multiple counties within the same city) (Jin et al., 2016) into consideration. Such estimation has been conducted as in Equation 1:
$\left\{ \begin{align}& {{A}_{i}}=\sum\limits_{j=1}^{s}{A{{U}_{ji}}} \\& A{{U}_{ji}}={{L}_{ji}}\times {{W}_{ji}}\times \left( 1{{\alpha }_{ji}} \right) \\& {{L}_{ji}}={{W}_{ji}}=\frac{{{P}_{ji}}}{4} \\ \end{align} \right.$
$A_i$ is the area of urban construction land in Fu $ i$;
s is the number of cities/towns in Fu $ i$;
$ AU_{ji}$ is the area of construction land in city/town j of Fu $ i$;
$L_{ji}$ and $W_{ji}$ are the length and width of the border of city/town j in Fu $ i$, respectively;
$\alpha_{ji}$ is the vacancy ratio in the border of wall in city/town j of Fu $ i$; and
$P_{ji}$ is the perimeter of the walls of city/town j in Fu $ i$.
The unified form of a city/town is set as rectangular.
(1) In the year 1820
Referred to cities/towns in the south of the Yangtze River that have rich data about four gates (Fan, 1990), the length×width ($L_{i}\times W_{i}$) of a large town is defined at 3.75×106 m2, a medium town at 2.5×106 m2, and a small town at 2.5×105 m2. $\alpha$ for a provincial capital is set at 65% (Cao et al., 1997), that for a normal Fu is 50%, and that for a county/town is 40%. In cases where overlay occurs, Fu takes the priority in area allocation, i.e., the area will be assigned to Fu but not county; and only one county will have the allocation if more counties than one are located within the same city.
The development of society, economy and politics leads to changes in the size and structure of population, and in sequence effects on the demand for land resources, i.e., for residence, production, and business activities. In the early modern period, it is more obvious for the city/town walls to act as the restriction to urban development. Urban construction land progressively broke through the walls and expanded greatly. In this study, the construction land in recent and early modern periods (i.e., in 1911, 1936 and 1952) has been estimated and calculated based on the population size and structure as in equations 2-5.
$\left\{ \begin{align} & {{A}_{i\left( t \right)}}=\sum\limits_{j=c,z}{A{{P}_{ji\left( t \right)}}\times {{P}_{Ti\left( t \right)}}\times {{\beta }_{ji\left( t \right)}}\left( t=1,2,3 \right)} \\ & A{{P}_{ui\left( t \right)}}=A{{P}_{ci\left( t \right)}}=A{{P}_{zi\left( t \right)}} \\ \end{align} \right.$ (2)
$A{{P}_{ui\left( t \right)}}=\left\{ \begin{align} & A{{P}_{ui\left( t-1 \right)}}+\left( A{{P}_{\overline{us}\left( t \right)}}-A{{P}_{ui\left( t-1 \right)}} \right)\times {{N}_{i\left( t \right)}},t=1,2,3 \\ & \frac{{{A}_{i\left( t \right)}}}{{{N}_{\overline{up}i\left( t \right)}}},t=0 \\ \end{align} \right.$ (3)
$A{{P}_{\overline{us}\left( t \right)}}=\frac{1}{n}\sum\limits_{m=1}^{n}{\frac{{{L}_{m}}\times {{W}_{m}}}{{{H}_{m\left( t \right)}}\times {{P}_{Hm\left( t \right)}}}}$ (4)
${{N}_{i\left( t \right)}}={\frac{{{N}_{\overline{uc}i\left( t \right)}}}{{{N}_{\overline{up}i\left( t \right)}}}}/{\left( \frac{1}{k}\sum\limits_{m=1}^{k}{\frac{{{N}_{\overline{uc}m\left( t \right)}}}{{{N}_{\overline{up}m\left( t \right)}}}} \right)}\;$ (5)
$A_{i(t)}$ is the area of urban construction land of Fu $i$ in the period of $t$;
$AP_{ui(t)}$, $AP_{ci(t)}$, and $AP_{zi(t)}$ represent per capita land use of urban, city and town of Fu $i$ in the period of $t$, respectively;
$P_{Ti(t)}$ is the total population size of Fu $i$ in the period of $t$;
$\beta_{ji(t)}$ is the population structure of Fu $i$ in the period of $t$;
$AP_{ui(t-1)}$ is the per capita urban construction land of Fu $i$ in the period of $t-1$;
$A{{P}_{\overline{us}\left( t \right)}}$ is the average value of per capita urban construction land in the period $t$;
Ni(t) is the urban socio-economic modification indicators of Fu $i$ in the period of $t$;
t=0, 1, 2 and 3 represent the year 1820, 1911, 1936 and 1952, respectively;
$H_{m(t)}$ is the number of households within city/town $m$ in the period of $t$;
$P_{Hm(t)}$ is the average household population of city/town $m$ in the period of $t$;
n is the number of cities and towns in Jiangsu in the period of $t$;
${{N}_{\overline{uc}i\left( t \right)}}$ is the urban socio-economic indicators of Fu $i$ in the period of $t$;
${{N}_{\overline{up}i\left( t \right)}}$ is the urban population of Fu $i$ in the period of $t$;
$\frac{1}{k}\sum\limits_{m=1}^{k}{\frac{{{N}_{\overline{uc}m\left( t \right)}}}{{{N}_{\overline{up}m\left( t \right)}}}}$ is the average value of per capita urban socio-economic indicators in the period of t; and
$k$ is the number of Fu in Jiangsu in the period of $t$.
(2) In the year 1911
Population in most counties grows and diffuses into surrounding areas. Per capita urban construction land is estimated based on the land area within the walls and population within cities/towns. Per capita urban construction land in each city/town is then amended with the number of business associations and chamber of commerce in the city.
(3) In the years 1936 and 1952
In the middle period of the Republic of China, factories gathered in cities and towns and the national capital began to emerge. During the Anti-Japanese War (1937-1945), all cities in Jiangsu were invaded more or less. Transportation system collapsed, industrial factories were occupied and the handicraft industry had almost been halted. Urban development in this period was in a stagnant status. From the end of the WWII to 1949, the urban population had increased and the urban development had recovered.
Assuming that the urban development level in 1936 was equivalent to that in 1952, per capita urban construction land can be determined based on data in 1952. Per capita urban gross value of products in 1936 can be used to fine calibrate such base data to reflect variations in economies and urban development across regions in 1936. Similarly, per capita added value of the secondary and tertiary industries in 1952 can be utilized as socio- economic indicators as well to refine the base data to evidence the imbalance across regions in 1952.
4.1.2 Quantitative estimation of rural settlement land
The village is an important component of the settlement system. Rural settlements were initially connected by affinities or clans and formed slowly and stably. The amount and distribution of rural settlement land is affected by the population, family structure, life style, as well as terrain, water resources, and cultivation radius. In general, they were not separated by defined spatial physical barriers (such as city/town walls). The form of social function is generally simple in rural areas. The relationship between the population and arable land affected the evolution of rural settlements. Compared with the urban land, the rural settlement land is expected to be positively correlated with the rural population.
In this study, the rural settlement land area in each period is estimated based on the population size, per capita land area and residential forms as in equations 6-11.
$A{{R}_{i\left( t \right)}}=A{{P}_{ri\left( t \right)}}\times {{P}_{TRi\left( t \right)}}$ (6)
$A{{P}_{ri\left( t \right)}}={A{{P}_{bi\left( {{t}_{0}} \right)}}}/{{{\lambda }_{i\left( t \right)}}}\;$ (7)
$A{{P}_{bi\left( {{t}_{0}} \right)}}={A{{P}_{hi\left( {{t}_{0}} \right)}}}/{R{{C}_{i\left( {{t}_{1}} \right)}}}\;$ (8)
$A{{P}_{hi\left( {{t}_{0}} \right)}}={A{{H}_{hi\left( {{t}_{0}} \right)}}}/{{{P}_{Fi\left( {{t}_{0}} \right)}}}\;$ (9)
${{\lambda }_{i\left( t \right)}}=\left\{ \begin{align} & 1,\text{ }t=0 \\ & {{\lambda }_{h\left( t \right)}}\left( {{\lambda }_{h\left( t \right)}}{{\lambda }_{l\left( t \right)}} \right)\times N{{R}_{i\left( t \right)}},\text{ }t=1,2,3 \\ \end{align} \right.$ (10)
$N{{R}_{i\left( t \right)}}={\frac{{{N}_{\overline{rc}i\left( t \right)}}}{{{N}_{\overline{rp}i\left( t \right)}}}}/{\left( \underset{m=1,2,\cdots ,k}{\mathop{\max }}\,\left\{ \frac{{{N}_{\overline{rc}m\left( t \right)}}}{{{N}_{\overline{rp}m\left( t \right)}}} \right\} \right)}\;$ (11)
$AR_{i(t)}$ is the area of rural settlement land of Fu i in the period of t;
$R_{TRi(t)}$ is the rural population size of Fu i in the period of t;
$AP_{ri(t)}$ is the per capita rural settlement land of Fu i in the period of t;
$AP_{bi(t_0)}$ and $A{{P}_{hi\left( {{t}_{0}} \right)}}$ are the per capita land of rural house base area and house size of Fu i in the period of t0, respectively;
$\lambda_{i(t)}$ is the ratio of house base land in the rural settlement land of Fu $ i$ in the period of \lambda;
$\lambda_{h(t)}$ and $\lambda_{l(t)}$ are the highest and lowest value of $\lambda_{i(t)}$;
t0 and t1 represent the reference years, i.e., 1933 and 1978;
t = 0, 1, 2 and 3 represent 1820, 1911, 1936 and 1952, respectively;
$R{{C}_{i\left( {{t}_{1}} \right)}}$ is the rural house capacity ratio of Fu i in the period of t1;
$A{{H}_{hi\left( {{t}_{0}} \right)}}$ and ${{P}_{Fi\left( {{t}_{0}} \right)}}$ are the rural average house size per household and average household population of Fu i in the period of t0;
$N{{R}_{i\left( t \right)}}$ is the rural socio-economic modification indicators of Fu i in the period of t;
${{N}_{\overline{rc}i\left( t \right)}}$ is the rural socio-economic indicator of Fu i in the period of t;
${{N}_{\overline{rp}i\left( t \right)}}$ is the rural population size of Fu i in the period of t;
$\underset{m=1,2,\cdots ,k}{\mathop{\max }}\,\left\{ \frac{{{N}_{\overline{rc}m\left( t \right)}}}{{{N}_{\overline{rp}m\left( t \right)}}} \right\}$ is the maximum value of per capita rural socio-economic indicators in the period of t; and
k is the number of Fu in Jiangsu in the period of t.
Before the 1980s, the rural socio-economic development was slow, the land use structure was relatively stable, and the housing structure was mostly single-floor. After the Reform and Opening in China, with the improvement of the agricultural economy, the living conditions have been continuously improved (Zheng, 2011; Wu et al., 2008).
(1) In the year 1820
The rural house capacity ratio in 1820 is assumed to be as the same as that in 1978 with a value of 0.245 (Lin et al., 2015). It is then finely adjusted in different agricultural regions by taking into consideration the effects of the number and species of livestock and poultry. Sorted by the total number of livestock and poultry from the maximum to the minimum, the whole province can be grouped into three regions: the rice-tea area (including Jiangning, Zhenjiang, Changzhou, Suzhou and Taicang), Yangtze rice-wheat area (including Tongzhou, Haimen, Yangzhou, Huai’an and Sizhou), and wheat-kaoliang area (including Hai and Xuzhou). The corresponding adjusted rural house capacity ratio value for these three regions is 0.27, 0.245 and 0.22 respectively.
Based on the survey data of 1933 (Bo, 1941), per capita rural house size is set at 16.81 m2, 20.44 m2 and 14.08 m2 in sequence. Therefore, per capita rural house base area can then be calculated as 62.26 m2, 83.43 m2 and 64 m2 respectively for three regions. Combined with the rural population in an individual Fu, the amount of rural settlement land in each Fu can be obtained.
(2) In the year 1911
Per capita house base area remains stable for all three time intersects in this study.
Moreover, it is expected that more and more rural land would be used for transportation, infrastructure, and provision of public facilities along with the rural development. Thus the proportion of public construction land also needs to be taken into consideration. During 1820 and 1911, the government enacted several policies to encourage agricultural production to meet needs due to the war, famine, and population increase. Rural agricultural companies were progressively emerged and the rural area experienced a series of farmland and water conservancy facility construction. In this case, per capita rural agricultural company is also going to be used as an adjusting parameter for the proportion of public construction land at Fu level.
(3) In the year 1936
The rural cooperative movement rapidly developed under the guidance and promotion from the central government of the Republic of China. By 1936, the average rural cooperatives at county level in south and north of the Yangtze River had reached 71 and 56 respectively (Sun, 2009). Setting the village as the organization unit, mutual aid groups subordinating to rural cooperatives were established. It is expected that farmers would be entitled to join mutual aid groups as members. Along with the bao-jia system (an old administrative system organized on the basis of households, each bao consisting of 10 jias, and each jia consisting of 10 households), mutual aid groups vigorously developed local economic output, water conservancy, industries, and comprehensive improvement such as birth control, education, hygiene, and religion (Zhu and Wang, 2008). During 1911 and 1936, there was some increase in the proportion of rural public construction land. Therefore, per capita rural cooperatives are used as the basis for the differential treatment of the proportion of public land.
(4) In the year 1952
People’s Republic of China started a comprehensive land reform in 1950. A number of measures were put into action to boost agricultural production, construct water conservancy infrastructure and improve soil quality. In 1951, preliminary agriculture cooperatives began to be established. In 1952, the output of main agricultural products reached the maximum in history. At the time, among the whole agricultural labor force, besides farming, forestry, animal husbandry and fishery, most concentrated in the agro-industry such as exploiting natural resources (i.e., mining, salt extraction, forest cutting, etc.), processing agricultural byproducts, and repairing industrial products. Therefore, the percentage of the agro-industrial labor force in the rural labor force has been chosen to differentiating the percentage of public land in each individual Fu.
Combined with data in 1985 interpreted from remote sensing images, the estimation of construction land in Jiangsu for all time intersects is shown in Table 1 and Figure 3.
Table 1 Reconstructed historical construction land (km2) at each Fu (city) in Jiangsu Province
Fu Year City Year
1820 1911 1936 1952 1985
Jiangning Fu 61.97 66.22 105.27 Nanjing 107.28 792.68
Zhenjiang Fu 78.26 60.47 74.58 Wuxi 107.17 394.34
Changzhou Fu 149.53 76.06 157.05 Xuzhou 225.63 1832.12
Suzhou Fu 197.32 87.03 111.81 Changzhou 76.01 336.83
Taicang Zhou 19.36 12.72 27.40 Suzhou 145.35 532.60
Tong Zhou 54.56 145.11 219.91 Nantong 301.81 331.75
Haimen Ting 11.41 17.52 37.58 Lianyungang 103.66 996.26
Yangzhou Fu 169.79 192.11 315.29 Huaiyin 135.44 1564.72
Huai’an Fu 85.02 164.55 254.04 Yancheng 227.85 987.07
Sizhou 36.73 19.21 49.03 Yangzhou 163.31 595.68
Haizhou 30.07 80.90 116.67 Zhenjiang 61.27 306.11
Xuzhou Fu 69.45 121.55 203.76 Taizhou 193.35 454.94
Suqian 132.19 1562.12
Total 963.46 1043.46 1672.40 Total 1980.34 10687.20

4.2 Spatial reconstruction of construction land

4.2.1 Spatial allocation principles
In this study, the spatial reconstruction of construction land conforms to a few principles:
(1) Natural resources such as the climate and geomorphology related to the distribution of construction land do not change along with the time line;
(2) Modern construction land is the outcome of the gradual expansion of historical construction land. The historical construction land should not locate beyond the boundary of corresponding modern urban construction land and rural settlement land;
(3) Historical urban administrative divisions are distributed within the scope of modern cities and towns. Extinct cities and towns that once existed during our study period are supplemented based on the literature reviews;
(4) Only one-way evolution is taken into consideration, i.e., urban construction land evolves from town land or settlement land in the earlier period but not vice versa.
4.2.2 Reconstruction methods
The spatial distribution of construction land is affected by both natural and human factors. Indicators of the natural conditions include the elevation, slope, and distance to water areas. Socio-economic indicators mainly include the distance to roads, the distance to cities/towns, and the distance to rural settlements. Meanwhile, the base map of land suitability has been configured by weights which are determined by applying the entropy weight method. Figure 4 shows the detailed technical process.
Figure 4 Spatial reconstruction of historical construction land
4.2.3 Reconstruction results
For the year 1820, the urban construction land and rural settlement land are defined by the number of grids corresponding to the estimated quantity while shapes of city/town walls are used as control borders. The suitability and connectivity of grids are also been considered (Lin et al., 2015).
As for the years 1911, 1936 and 1952, the urban construction land and rural settlement land are assumed to expand only within the scope of corresponding city/town/village at contemporary time. The city land expanded within walls and then beyond the walls according to the suitability and connectivity characteristics. The result of the land suitability evaluation has been deemed as a controlling parameter.
In order to improve the visual effect, the Boolean data in 200 m × 200 m grids has been converted into proportional data in 1 km × 1 km grids in Figure 5. As shown in Figure 5, the total construction land in Jiangsu has generally increased over the study period. However, there are significant differences across regions and periods. The spatial distribution of construction land exhibits its proclivity to water bodies and road networks and the great polarization effect of existent residence. Specifically, the construction land in 1820, 1911, 1936 and 1952 is shown below.
Figure 5 Spatial pattern of historical construction land in Jiangsu Province
(1) In the year 1820
The construction land was distributed mainly in southern Jiangsu, about 53% of it is in Jiangning, Zhenjiang, Changzhou, Suzhou and Taicang.
It is straightforward that water bodies were of a great effect on the spatial pattern of the construction land. Rural settlements in central Jiangsu were mainly located at the junction of the main tributaries and along rivers/lakes to meet human production and living needs.
The densest distribution of rural settlements occurred in Zhenjiang, Changzhou, and the central and northern Suzhou from the south of the Yangtze River to Taihu Lake. The second densest area was found in western and southwestern Yangzhou along the Gaoyou Lake, western Tongzhou along the north of the Yangtze River, northwestern Huai’an along the downstream area of the Yellow River in northern Jiangsu Province.
(2) In the year 1911
The density of construction land in southern Jiangsu decreased rapidly because of the civil war. The total construction land slightly increased in central Jiangsu, and Huai’an, Xuzhou and Haizhou in northern Jiangsu. The construction land distribution density slightly increased in central Huai’an in northern Jiangsu and southwestern Xuzhou.
(3) In the year 1936
The construction land was densely distributed in western Tongzhou, central Yangzhou in central Jiangsu, and southeastern Xuzhou along the Luoma Lake, southern Haizhou along the Guanhe River, central and northern Huai’an, and southwestern Sizhou in northern Jiangsu. At the time, the total construction land rapidly increased with the construction of roads and railroads.
(4) In the year 1952
The construction land continued to increase. In particular, the construction land density rapidly increased in the coastal areas of Nantong.
In contrast, the construction land in central and southern Huai’an slightly decreased probably because of the reduction of water bodies.

5 Verification of results

5.1 Verification of trends and correlation

(1) Urban land
In the mid-Qing Dynasty (i.e., 1820), the percentage of urban population is higher in Jiangning and Suzhou than that in other cities since they are where the Jiangning Administrative Commissioner’s Office and Jiangsu Provincial Administrative Commissioner’s Office locate respectively. The construction land in these two cities takes approximately an half of the total urban land in Jiangsu Province.
Urban development in central and northern Jiangsu is slow. By the end of the Qing Dynasty and early years of the Republic of China (i.e., 1911), urban population in southern Jiangsu had dropped dramatically but slightly increased in central and northern Jiangsu due to the imbalanced occurrences of the war aftermath.
Till 1936, national industry and commerce had been reestablished and developed in cities and towns while urban land had been consistently increased. Urban land takes approximately 16% of the total construction land in Jiangsu in 1936. Urban land in southern Jiangsu is still of the greater percentage than that in central and northern Jiangsu.
In 1952, the national economy was almost completely recovered from years of wars while urban construction and development rapidly increased.
The above detailed outline of urban land changes in our study period anchoring on our selected several time intersects shows that our reconstructed urban land is consistent with historical events and our reconstruction data is compatible with the overall trend.
(2) Rural settlement land
Arable land is essential for rural households while the agricultural cultivation economy is the base for the existence and development of rural settlements. Development of arable land and that of rural settlements are interrelated and mutually promoted. Therefore, the quality of our quantitative estimates can be accessed by a correlation analysis between the amount of arable land and that of the rural settlement land.
The arable land data for the Qing Dynasty is compiled based on Cao et al. (2014), for the Republic is generated from China’s Agricultural Production and Commerce Statistics (Xu, 1983), Statistical Analysis of Chinese Land Use (GSB, 1936), Agriculture Development in China: 1368-1968 (Dwight, 1984), and for after the founding of the People’s Republic is taken from National Land Survey and Chinese Economic Statistics.
The correlation coefficient between the amount of rural settlements and that of arable land at Fu (city) level has been estimated at 0.531 for the year 1820 with p<0.1, 0.720 for the year 1911 with p<0.01, 0.746 for the year 1936 with p<0.01, and 0.769 for the year 1952 with p<0.01. Figure 6 illustrates the correlation analysis results.
Figure 6 Correlation between the amount of arable land and that of rural settlement land at Fu (city) level

5.2 Quantitative verification

To verify the effectiveness of our methodology, our reconstructed dataset is compared to corresponding historical maps in which historical data for Jiangsu Province is accessible. The quantitative verification is conducted with indicators such as the absolute error, relative error, and coverage ratio between the number of towns/villages identified in the reconstruction results to that present in historical maps. The calculations of these indicators are conducted by applying equations 12-14:
$AE=\left( rr-hm \right)$ (12)
$RE={\left( rr-hm \right)}/{hm}\;\times 100%$ (13)
$CR={ca}/{hm}\;\times 100%$ (14)
where AE and RE are respectively absolute error, relative error between the number of reconstruction towns/villages and the number from historical maps;
CR is the coverage ratio of towns/villages in historical maps in a range of 1 km of the rebuilding towns/villages;
rr is the number of towns/villages identified in the rebuilding process;
hm is the number of towns/villages in historical maps; and ca is the number of towns/villages in historical maps in a 1 km buffer of the towns/villages identified in the rebuilding process.
Such verification process has been conducted upon four selected counties/cities:
Danyang County representing hilly land in Nanjing-Zhenjiang-Yangzhou region (Region I);
Changshu County representing watershed plain land around Taihu Lake (Region II);
Nantong City representing alluvial plain of the Yangtze River (Region III); and Suqian County representing the Yellow River Flood Plain (Region IV).
Military surveying map in the 1930s ( has been used as reference in this verification. 1 km has been set as the buffer zone for error detection, i.e., the deviation within the distance of 1 km is deemed to be compatible or ignorable. Table 2 andFigure 7 show the verification results.
Table 2 Accuracy of the reconstruction results in sample areas in 1936
Sample Location Region represented AE RE CR
Danyang Southern Jiangsu Region I: hilly land areas in Nanjing-Zhenjiang- Yangzhou -139 -30.82% 76.72%
Changshu Southern Jiangsu Region II: watershed plain land areas of Taihu Lake -40 -4.45% 62.03%
Nantong Central Jiangsu Region III: alluvial plain of the Yangtze River -31 -4.45% 56.67%
Suqian Northern Jiangsu Region IV: the Yellow River Flood Plain 176 32.47% 74.72%
Average -34 -1.31% 65.80%
Figure 7 Accuracy of reconstruction results in 1936 referring to historical maps
In our four selected sample areas, there are a total of 2554 reconstructed towns/villages and 2588 corresponding towns/villages in the historical maps. The total relative error is -1.31%. In particular, the relative error in Region II and Region III is at -4.45%. The relative error in Region I and Region IV is -30.82% and 32.47% respectively.
As to the accuracy test, valid reconstructed towns/villages within 1 km buffer zone of their counterparts in the historical map reached 65.8% of the total. Specifically, such accuracy reached 76.72% in Region I, 74.72% in Region IV, 62.03% in Region II and 56.67% in Region III.

6 Discussion and conclusions

This paper aims to provide a new perspective on reconstructing the spatial pattern for historical construction land, especially urban construction land and rural settlement land, by focusing on characteristics of historical land use changes while incorporating data from various sources including historical documents, statistics, land use, and basic geographic data.
Urban construction land is proposed to be estimated based on indicators such as city walls and four gates along city walls while in a few cases need fine calibration with population and per capita construction land data. The rural settlement land is calculated by multiplying rural population data and per capita housing allowance.
Thus, the spatial pattern of historical construction land in Jiangsu can be reconstructed in 200 m × 200 m grids for our four study periods in the last 200 years with quantitative control, border control, and land suitability control and the connectivity among land use types.
Our study outcome indicates that great differences present across regions in 1820. For reference, for Jiangsu Province at the similar time being, He et al. (2002) estimates that the total urban land in Jiangsu Province was 185.77 km2 (in which Jiangning was 60 km2) by using the perimeter of walls recorded in Jiaqing Rebuilt Chi Unification; Pan et al. (2015) finds out that the total urban land was 115.54 km2 based on data derived from He et al. (2002) and calibrated with the percentage of urban population; Lin et al. (2015) concludes that the total urban land was 222.9 km2 and Jiangning alone took 146.35 km2.
This study classified urban construction land into city land and town land based on a refined analysis of historical documents. Some survey maps at that time show that a large amount of arable land and open space presented within the walls of cities in historical periods, e.g., the land vacancy ratio in Jinling was 65%. Thus, although city walls can be considered a physical boundary between urban and rural landscapes, it might not be appropriate to treat city walls as a boundary between urban construction land and rural settlement land in terms of the type and extent of land use. Therefore, this study uses the land vacancy ratio to refine the total urban construction land calculated by the perimeters of city walls to come up with the value of total urban construction land in 1820 at 58.18 km2 (of which 21 km2 is in Jiangning). This result is compatible with the dynamic evolution theory about the development of urban construction land. It has been widely observed that in history urban construction land expands with the increase of population by “filling in” the space within city walls then expands further beyond city walls.
The variation in the estimated amount of rural settlement land across studies is mainly resulted from the choice of one substitute indicator, i.e., per capita rural land use.
Pan et al. (2015) estimates that the total rural settlement land in 1820 was 4724.68 km2 based on per capita rural land use level in 1985. Lin et al. (2015) shows that the total rural settlement land was 783.93 km2 in history by using per capita rural housing allowance value and the house capacity ratio in 1978. This study finds out that the total rural settlement land was 853.23 km2 in 1820 with the house capacity ratio in 1978 and the value of per capita rural housing allowance in the 1930s. These historical values are preferred in this study for the closeness of their time to 1820.
Generally, academic contributions of this study mainly include three aspects:
(1) Expanding the list of reconstruction objects and the time span of reconstruction
Most contemporary studies focus on either urban construction land or the composite of urban land and rural settlements. This study further divides urban land into cities and towns, which supplements the previous untouched/neglected construction land at the county level. Moreover, this study replenishes three more historical periods covering recent and early modern times to classic studies with only one historical period by using proxy coefficients, which enables the continuity of reconstructing the spatio-temporal pattern of historical land use.
(2) Including more variables and strengthening the data calibration
Besides classic variables such as historical population and per capita land use, this study integrates more variables such as data about four gates along the city walls and the rural house capacity ratio into consideration based on extensive literature review. Regarding the variation among variables in terms of suitability across regions, coefficients such as the number of per capita urban business associations and chambers of commerce and that of per capita rural cooperative entities are also employed in this study to calibrate the reconstructed historical construction land to further approach its historical true values.
(3) Reinforcing the verification process
To make up the deficiency of contemporary studies in the research outcome verification process, this study combines three direct and indirect methods to strengthen the assessment effectiveness over the rationality of our reconstructed historical land use pattern. These three methods include the consistency analysis of the process of construction land changes with the tendency of corresponding regional land use pattern, the correlation analysis between the distribution of rural settlement land and the arable land, and the quantitative analysis of the spatial distribution for historical construction land at typical time intersects.
Historical events are of the nature non-firsthand, irreversibility and lack of spatial pattern information. Meanwhile, the spatial pattern of construction land is influenced by various factors, especially those social factors, which imply great uncertainty. This study aims to reconstruct the historical construction land based on a few assumptions about natural resources, socio-economic conditions, and land use principles. Therefore, each individual typical time intersect only stands for the average status during that historical period and the outcome of such reconstruction is only a feasible scenario based on rational reasoning. However, our proposed comprehensive reconstruction approach is of the great reference value to reconstructing historical construction land and even multiple (or all) land use types with greater time span and/or greater geographic coverage.

The authors have declared that no competing interests exist.

Bai S Y, Zhang S W, Zhang Y Z, 2007. Digital rebuilding of LUCC spatial-temporal distribution of the last 100 years: Taking Dorbod Mongolian Autonomous County in Daqing City as an example.Acta Geographica Sinica, 62(4): 427-436. (in Chinese)lt;p>The Yangtze River Delta is one of the economically developed coastal areas. From the late 1970s, its urbanization process has been quickened greatly, which resulted in the number increase and the spatial expansion of urban areas. The Landsat MSS, TM/ETM satellite images, which were respectively acquired in 5 periods of 1979, 1990, 1995, 2000 and 2005, were used to extract urban land information and analyze urban growth data with the help of remote sensing and GIS softwares. We analyzed the spatio-temporal characteristics including urban growth speed, growth intensity, fractal dimension and urban growth pattern. Additionally, dynamics of urban expansion in the Yangtze River Delta were also analyzed. The results are drawn as follows: (1) From 1979 to 2005, the growth speed of urbanization area was accelerating obviously. The quantities of increasing area of urbanized land were 37.66 km<sup>2</sup> 112.43 km<sup>2</sup> 274.86 km<sup>2</sup> and 421.73 km<sup>2</sup> in the past four periods (1979-1990, 1990-1995, 1995-2000 and 2000-2005), respectively. Meanwhlie, the growth intensities of urbanized land enhanced gradually. From 1979 to 1990, the growth intensity was only 0.03, then reaching 0.10, 0.24 and 0.37 in the following three periods. (2) The spatial structure of urbanization area in the Yangtze River Delta was fractal. The fractal dimension and stability coefficient of urbanized land structure fluctuated to a certain extent. From 1979 to 2000, the fractal dimension of urbanized land structure decreased yearly. The shape of urbanized land tended to be regular. After 2000, the area increase of urbanized land on a large scale led to more complicated shape of urbanized land. The stability coefficient also had similar characteristics to that of fractal dimension. So the change of urbanized land in spatial structure was relating to the growth process of urbanized land. (3) The growth process of urban agglomeration in the Yangtze River Delta was from one pole and two belts to five poles and five belts. From 1979 to 1990, Shanghai was the only first-grade growth pole of urbanized land and Shanghai-Nanjing railway and Shanghai-Hangzhou railway were the two first-grade growth belts of urbanized land in the Yangtze River Delta. At the latest period (from 2000 to 2005), the first-grade growth poles included 5 cities, i.e., Shanghai, Nanjing, Hangzhou, Suzhou and Ningbo. Besides Shanghai-Nanjing railway and Shanghai-Hangzhou railway, Shanghai-Jingjiang railway, Hangzhou-Ningbo railway and the highway linking Nanjing to Gaochun also became growth belts of urbanized land in the Yangtze River Delta in that period.</p>


Bao J L, Gao S, 2016. Traditional coastal management practices and land use changes during the 16-20th centuries, Jiangsu Province, China.Ocean & Coastal Management, 124: 10-21.Within the framework of global change research, the issues related to coastal management and adaptation represent an important focus. During the 16 19th centuries, there were complex coastal evolution patterns, diverse human activities and rich historical documents for the Jiangsu coast, eastern China. Hence, this region represents a unique place in the world for the study of the traditional coastal management practice. In the present contribution, we use the historical archives/documents and adopt multidisciplinary methods, to reveal the characteristics and changing patterns of the traditional management practice for the Jiangsu coast. We have found that the Jiangsu coastal management practice was dominated by the central government, including policies and legal norms for the coastal areas, e.g., tax policy. Although there were differences compared with the present-day coastal management, it was already more advanced than the management systems merely based on custom, taboos and cultural norms. To some extent, Jiangsu coastal management practice is typical for the pre-industrial period, which was an initial stage of modern coastal management. The governments of the Ming and Qing Dynasties paid more attention to sea salt industry than reclamation/agriculture in the land use policies, and set a strict administrative management system to implement coastal management for land resources, labors and coastal engineering. It had a profound impact on the Jiangsu coastal land development and socio-economic change. In the 16 19th centuries, with rapid accretion on the Jiangsu coast and the intense control of the central government, salt production was the main coastal exploitation activity. After the 19th century, the land use was converted to agricultural development. However, along with the coastal natural change, population growth and destruction of war, the demand of multifarious land uses increases rapidly. As a result, the traditional management pattern, which emphasized monopoly profits but lack of negotiation and coordination between different stakeholders, became gradually unsustainable and collapsed eventually; subsequently, agricultural development was the dominant socio-economic activity of the Jiangsu coastal region. Thus, the highly centralized management system in the pre-industrial periods for the China coasts was evolved gradually into a decentralization system, which was consistent with the trend of the development of modern coastal management.


Bo K, 1941. China Land Use. Department of Economics: University of Nanking Agricultural College. (in Chinese)

Cao S J, 2002. Urban population of Jiangsu Province in the Qing Dynasty. Journal of Hangzhou Normal University (Social Sciences), (4): 50-56. (in Chinese)

Cao W R, Zhen X H, Huang S Z, 1997. Atlas of Ancient China (Qing Dynasty). Beijing: Cultural Relics Press. (in Chinese)

Cao X, Jin X B, Wang J Set al., 2014. Reconstruction and change analysis of cropland data of China in recent 300 years.Acta Geographica Sinica, 69(7): 896-906. (in Chinese)Historical land-use and land-cover changes caused by human activities during the last three centuries have been regarded as one of the five key frame issues in the LUCC project. China, with a history of 5000 years, has had its population boom ever since the early Qing Dynasty (around AD1700), and unprecedented development of national agricultural reclamation had started, left China as one of areas with rapid land-use and land-cover changes. Currently, there are two global historical land use datasets, generally referred as the &lsquo;RF datasets&rsquo; and &lsquo;HYDE database&rsquo; but at the zonal level, these global datasets are widely doubted with coarse resolution and inevitable errors. Academics have tried to reconstruct China's historical land-use and land-cover both quantitatively and spatially, but there are remarkable differences in their results, thus bringing troubles to relevant researches. Since the quantity forms the backbone of cropland restructuring, this paper grounded itself on China's historical records and related research achievements, and reconstructed China's provincial cropland data at the modern boundaries from 1661 to 1985, using a variety of methods based on resources and population, such as factor revision, man-land relationship test, and reclamation trend examination, etc. Our results differ less from HYDE, CHCD and Zhang with an average difference rate of less than 15%. But at the provincial level, our results are closer to CHCD, with 22% of provinces' average difference rate being over 30% . But significant diversities were found in a few provinces and further researches are needed. Then we analyzed China's cropland growth process and regional change characteristics. The results show that ever since the population boom in the Qing Dynasty, China's cropland trebled from 42.4&times;10<sup>6</sup> ha in the early Qing Dynasty to 136.9&times;10<sup>6</sup> ha in 1985. In terms of the growth rate, the process of China's cropland rise can be identified into five periods. Significant differences existed among the provincial cropland change. At the beginning of the Qing dynasty, China's farming activities mainly existed in the Yangtze River Plain, the Huang-Huai-Hai Plain, Guanzhong Basin and Yinchuan Plain. Thereafter, reclamation activities expanded to outer agriculture areas. Since the founding of the People's Republic of China, Northeast China and Northwest China have been major sources of additional cropland. National policy, disasters, wars, and economic development, are main factors affecting cropland changes.


Dwight P H, 1984. Agriculture Development in China: 1368-1968. Edinburgh: Edinburgh University Press.

Editorial Group (EG) of Atlas of Jiangsu, 1978. Atlas of Jiangsu. Nanjing: Jiangsu Map Publishing House. (in Chinese)

Editorial Board (EB) of Fifty Years of Jiangsu Province, 1999. Fifty Years of Jiangsu Province. Beijing: China Statistics Press. (in Chinese)

Editorial Board (EB) of Fifty Years of Rural Economy of Jiangsu Province, 2000. Fifty Years of Rural Economy of Jiangsu Province. Beijing: China Statistics Press. (in Chinese)

Fan S Z, 1990. Jiangnan Towns in the Ming and Qing Dynasty. Shanghai: Fudan University Press. (in Chinese)

Fang X Q, Zhang W B, Zhang L Set al., 2002. The urban expansion and the evolution of urban fringe in Beijing in the 20th century.City Planning Review, 26(4): 56-60. (in Chinese)This paper analyses the urban expansion of the Beijing City based on the TM images and historical maps. The growth of Beijing has gone through three phases in the last century. It expanded slowly in the first half of the 20th century, fast between 1950s and 1980s, even faster since the middle of 1980s. Meanwhile, the structure of the urban area has changed from that of the early 20th century.

Fu L X, 2009. A new study on the establishment of Jiangsu Province in the Qing Dynasty. Studies in Qing History, 73(2): 23-31. (in Chinese)When Jiangsu Province was established in the Qing Dynasty, and when Jiangnan Province was divided into Jiangsu and Anhui Provinces, are the most important questions in the study of the local history of Jiangsu and Anhui Provinces and have always been topics of concern within academic circles. Based on textual analysis on the related historical materials from around 1667, this article argues that the process of establishing Jiangsu Province began in 1661 ( the 18th year of the Shunzhi reign) when “Left” and “Right” Provincial Administrative Commissioners were divided and ended in 1667 ( the 6th year of Kangxi reign) when the newly established Provincial Governors were named. The Qing court and local highranking provincial officials both played a role in this process. The related issues regarding the establishment of Jiangsu Province can be only clarified based on a comprehensive and deep study of the related historical data.

Government Statistic Bureau (GSB), 1936. Statistical Analysis of Chinese Land Use. Nanjing: Zhongzheng Publishing House. (in Chinese)

Gu C L,Hu L Q, Zhang X Met al., 2011. Climate change and urbanization in the Yangtze River Delta. Habitat International, 35: 544-552.The Yangtze River Delta (YRD), one of China’s most developed, dynamic, densely populated and concentrated industrial area, is growing into an influential world-class metropolitan area and playing an important role in China’s economic and social development. The formation and the urbanization process of YRD are inseparable from climate change. This paper explores such interrelationship from two perspectives. On one hand, using historic data, we summarized the urbanization process in the YRD, and concluded that climate change has been shaping the Delta and its socioeconomic development. On the other hand, the urbanization process of the Delta has shaped its geography and built environment, which, however, are not adaptable to future climate change. Potential disruptive effects include large flooded land area, flood disasters, production and energy inefficiency, and other environmental threats. It is imperative to adopt policies and programs to mitigate and adapt to climate change in the fast urbanization process.


He F N, Ge Q S, Zhang J Y, 2002. Reckoning the areas of urban land use and their comparison in the Qing Dynasty in China.Acta Geographica Sinica, 57(6): 709-716. (in Chinese)lt;p>Based on the numerable and standardized data collected from historical documents, we reckoned the areas of urban land use for 18 provinces in the Qing Dynasty, analyzed the changing situation, regional differentiation, and contrasted them with those of the contemporary age. The results are shown in the following: (1) The method, by which the areas of urban land use in the Qing Dynasty are reckoned in light with at the administrative division levels, the number of towns and city wall perimeters, was provided with rationality to a certain degree. (2) The results showed that the area of urban land use was 1,987.44 km<sup>2</sup> in 18 provinces of the Qing Dynasty, merely 0.05% of the total land area in the region. Among which the scopes of Zhili and Jiangsu provinces were the greatest, being 316.34 km2 and 185.77 km<sup>2</sup> respectively 0.097% and 0.188% of the jurisdiction area, and Guangxi and Guizhou provinces were the smallest, 27.92 km<sup>2</sup> and 32.78 km<sup>2</sup> respectively (only 0.012% and 0.033% of the jurisdiction area). (3) There were obvious spatial differences in urban land use in the region. For the scale of urban land use, the northern provinces were greater than the southern provinces, the eastern provinces were greater than the western provinces, and southwest provinces were the smallest in all the provinces. The provinces of Zhili, Shaanxi, Sichuan, Jiangsu and Zhejiang were obviously greater than the other provinces.</p>


Hu X H, Yang G Q, Zhang Xet al., 2007. The change of land use for rural residency and the driving forces: A case study in Xiantao City, Hubei Province. Resources Science, 29(3): 191-197. (in Chinese)The research on land use/cover and its driving force is one of the main research subjects worldwide.The land use for rural residency is an important type of land use because of its quantity and scale,which is more than 6 times of the scale in city and town.Moreover,most of the rural settlements surround by the high quality cultivated land,so the sprawl of rural settlements is more threatening to the cultivated land than the extension of city and town.One of the outstanding problems of land use is that the scale of rural settlements has not been decreased with the urbanization.In fact,it is increasing quickly.So the study on the changes of land use rural residency and their driving forces is important for protecting cultivated land and accelerating the new rural construction and rural development theoretically and realistically.First of all,the driving forces of land use change for rural residency were explained in the paper.Natural factors and socio-economic factors,and the socio-economic factors would be paid more attention because the effects of natural factors have little influence on rural residential land within a relatively short time.Then,the Xiantao City in Hubei province was taken as an example to study the changes of land use for rural settlements and the driving forces based on the land conversion data and socio-economic statistic data from 1996 to 2004.The results showed that the quantity of rural residential land has decreased 172.59 hm~2 for the past 8 years in Xiantao City,and the land use intensity was low.Per capital land area for rural settlement was 171.76m~2,exceeding the average criterion of 21.76m~2 in China.By the factors analysis,the main driving forces of land use for rural settlements were analyzed,including the economic development and urbanization,the development of science and technology and the policy factors.The regression method was set up based on the main three driving forces of the rural residential land use changes.At the end,some valuable conclusions and recommendations were put forward based on the analysis aforementioned.

Hu Z Z, 1994. Settlement Geography. Taiwan: San Min Book. (in Chinese)

Ji Y, Sun W, Li Get al., 2009. Space-time features analysis of land use/cover change in Guangzhou urban built-up area from 1907 to 1968. Journal of South China Normal University (Natural Science Edition), 123(1): 121-126. (in Chinese)After pre-treatments with the maps from multiple sources by using geographic information system,the authors apply a new method to extract the urban boundary information and reconstructs the land using map of Guangzhou in 1907,1928 and 1968.Combining the materials in other literatures,it analyzes and summarizes the space-time features of land use/cover change in Guangzhou urban built-up area from 1907 to 1968.The results show that during 1907 to 1968,the main land used types in Guangzhou urban were the residential area,the industrial area,the water area,and the traffic land.The area change rate of land used type are fluctuate.During 1907 to 1928,the area of traffic land increased most.The area of the residential area,the industrial area and unutilized land increased dramatically from 1928 to 1968.During the 60 years,with the increasing of the built-up area,the river's length and the water area increased constantly,but the average width of the trunk stream of the Pearl River were gradually decreased.

Jiaqing Rebuilt Chi Unification, 1986. Beijing: Zhonghua Book Company. (in Chinese)

Jiangsu Geography Institute of Geography Department in Nanjing Normal University (JGI), 1982. Urban Historical Geography of Jiangsu. Nanjing: Jiangsu Science and Technology Press. (in Chinese)

Jin Q M, 1988. Rural Settlement Geography. Beijing: Science Press. (in Chinese)

Jin X B, Pan Q, Yang X Het al., 2016. Reconstructing the historical spatial land use pattern for Jiangsu Province in mid-Qing Dynasty.Journal of Geographical Sciences, 26(12): 1689-1706.This study is proposed to reconstruct a high-resolution spatial distribution of historical land use pattern with all land use types to overcome low-accuracy and/or the monotonic land use type in current historical land use reconstruction studies. The year of 1820 is set as the temporal section and the administrative area of Jiangsu Province is the study area. Land use types being reconstructed include farmland, residential land(including both urban land and rural residential land), water body, and other land(including forest land, grassland, and unused land). Data sources mainly refer to historical documents, historical geographic research outcomes, contemporary statistics, and natural environmental data. With great considerations over regional natural resources and social and economic conditions, a few theoretical assumptions have been proposed to facilitate the adjustment on prefecture farmland, urban land, and rural residential land. Upholding the idea that the contemporary land use pattern has been inherently in sequence with the historical land use pattern as well as the land use pattern shall be consistent to its accessibility, this study reconstructs the land use pattern in Jiangsu Province in 1820 with 100 m*100 m grids based on accessibility analysis and comprehensive evaluation. The outcome has been tested as valid by regionalization and correlation analysis. The resulted spatial distribution shows that back in 1820 in Jiangsu Province:(1) farmland, urban land, rural residential land, water body, and other land take about 48.49%, 4.46%, 0.16%, 15.03%, and 31.86% of the total land area respectively;(2) the land use pattern features high proportion of land in farming while low-proportion land in non-farming uses while population, topography, and the density of water body lead to great spatial variations; and(3) the reconstruction methodology has been tested as reasonable based on significant positive correlations between 1820 data and 1985 for both farmland and rural residential land at the prefecture level.


Li B B, Xu F, 2008. The research on urbanization rate and staging in modern China. Journal of East China Normal University (Philosophy and Social Sciences), 197(3): 34-41. (in Chinese)

Li S, Luo X Q, 2014. Study on spatiotemporal expansion feature of urban based on multisource data using remote sensing technique in Guiyang city in the latest 700 years.Journal of Natural Resources, 29(10): 1734-1745. (in Chinese)At first, a multitemporal urban land use information is extracted from the Landsat images of Guiyang in 1947, 1973, 1990, 2001, 2006 and 2010 using multiple source data including historical literature, thematic maps and Landsat images, in which the spatiotemporal feature of urban expansion is analyzed after 1279. These characteristics involve expansion trend, direction, expansion intensity, scale and spatial differentiation and so on. The results show that: 1) before1947, the main expanding directions were the south(Huaxi) and the west(Jinyang), and the main type was sprawling outwards, the velocity of urban expansion was slow. 2) The urban expansion of Guiyang filled in a small basin region from 1947 to 1973 and started an expansion out along the valley topography. The urban expansion showed a diffusion trend outward in the main urban region after 1990. 3) The expansion of urban fringe of Guiyang was arrived to the south(Huaxi), the east(Longdongbao), the northeast(Xintianzhai) and the west(Jinyang). Urban expansion of Guiyang after 2006 showed a significant Y-shaped pattern. At last, current rapid expansion of Guiyang is not suitable to the local sustainable development, Guiyang city is suggested stopping such rapid spatial expansion.


Lin Y N, Jin X B, Yang X Het al., 2015. Data set establishment and spatial reconstruction of built-up area in the Mid Qing Dynasty: Method and case study.Geographical Research, 34(12): 2329-2342. (in Chinese)Human-induced land use/cover change (LUCC) has significant effects on the climatic and ecological processes at both global and regional scales, especially in the last 300 years. It is an important driving force of global environmental change and has been one of the hot topics of international concerns. Since the modernization started after industrial revolution, the expansion of built-up areas (construction land) profoundly changed the status of LUCC and made the landscape of urban areas become a notable feature of a certain region. Limited by the available data, reconstruction of historical built-up areas confronts with some difficulties. Supported by the data of historical records, modern statistical and natural resources, the method of reconstructing historical construction land (urban land and rural residential land) of the typical time section in the mid-Qing Dynasty was proposed in this paper. We reconstructed the scale and spatial pattern of urban land by using data including the perimeter of ancient city wall, the shape of the city and the points of city location. Meanwhile, the characteristics of livability and continuity of land, rural population, per capita living space requirements, and the points of potential rural settlements were used as the foundation for rural residential land reconstruction. Based on the theoretical analysis, this study took Jiangsu Province as the study case and chose the year of 1820 as the research year for the availability of historical data. A spatial distribution of construction land with a resolution of 100 m 100 m was finally established. The results showed that: (1) Under the circumstances of lacking effective data, it is a viable attempt to reconstruct the spatial pattern of historical construction land by exploring the most potential of historical documents and making proper deduction based on the modern construction land patterns. (2) As for relative factors calculation, the differences in features of combination of towns, the urbanization rate, farming radius and residential patterns should be considered. At the same time, different historical situations of each area should be taken into account in order to determine the scale and spatial pattern of land rationally. (3) The total area of construction land of Jiangsu Province in 1820 was 1006.84 km, in which urban land was 222.90 kmand rural residential land was 783.93 km. Compared with related studies, the main difference came from research assumptions and computation basis of study units.

Liu M L, Tian H Q, 2010. China’s land cover and land use change from 1700 to 2005: Estimations from high-resolution satellite data and historical archives. Global Biogeochemical Cycles, 24(3): 285-286.1] One of the major limitations in assessing the impacts of human activities on global biogeochemical cycles and climate is a shortage of reliable data on historical land cover and land use change (LCLUC). China had extreme discrepancies in estimating contemporary and historical patterns of LCLUC over the last 3 centuries because of its geographical complexity, long history of land use, and limited national surveys. This study aims to characterize the spatial and temporal patterns of China's LCLUC during 1700-2005 by reconstructing historical gridded data sets from high-resolution satellite data and long-term historical survey data. During this 300 year period, the major characteristics of LCLUC in China have been shrinking forest (decreased by 22%) and expanding cropland (increased by 42%) and urban areas (including urban and rural settlements, factories, quarries, mining, and other built-up land). New cropland areas have come almost equally from both forested and nonforested land. This study also revealed that substantial conversion between forest and woodland can be attributed to forest harvest, forest regeneration, and land degradation. During 1980-2005, LCLUC was characterized by shrinking cropland, expanding urban and forest areas, and large decadal variations on a national level. LCLUC in China showed significant spatial variations during different time periods, which were caused by spatial heterogeneity in vegetation, soils, and climate and regional imbalance in economy development. During 1700-2005, forests shrunk rapidly while croplands expanded in the northeast and southwest of China. During 1980-2005, we found a serious loss of cropland and urban sprawl in the eastern plain, north, and southeast regions of China and a large increase in forested area in the southeast and southwest regions. The reconstructed LCLUC data sets from this study could be used to assess the impacts of land use change on biogeochemical cycles, the water cycle, and the regional climate in China. To further eliminate uncertainties in this data set and make reliable projections of LCLUC for the future, we need to improve our understanding of the drivers of LCLUC and work toward developing an advanced, spatially explicit land use model.


Ma R F, Wang T F, Zhang Wet al., 2016. Overview and progress of Chinese geographical human settlement research. Journal of Geographical Sciences, 26(8): 1159-1175.Increasing Chinese urbanization and industrialization has prompted greater attention to the study of human settlement and the human-land relationship in the fields of geography, architecture, and urban planning. We used bibliometric methods and statistical software to review 180 articles on human settlement in 16 Chinese geographical journals. We found that Chinese geographical human settlement research is characterized by the following: (1) Most research focuses on human settlement extension, valuation indicators, models for urban and rural settlements, theoretical exploration and the planning practices of single-factor, human settlement and complex, geographical livability in macro-scale, urban settlement differentiation and ideal patterns in medium scale, the comprehensive evaluation of settlement environment, and the planning of community units in micro-scale, community settlements; socio-cultural investigation and warnings about advancing human settlement. (2) No progress has been made in synthesizing and integrating method systems. PSR models and DPSIR models are used for targeting mechanisms, while the standard settlement evaluation system was composed of physical & economic indicators by questionnaire surveys. On the other hand, spatial clustering based on GIS has been a frequent focus in recent years. Pioneering research on human settlement and theoretical systems within the context of China urbanization and industrialization will provide guidance on the sustainability of Chinese cities and regions. The following five aspects require greater attention: (1) Natural suitability research on human settlement, and a survey of human settlement demands to reflect the range of different demands concerning ecologically suitable settlements in urban environments, the corresponding valuation indicators, systems, and evolution, and the impact of the residents socio-economic attributes. (2) Spatial-temporal evaluation and sustainability research on urban and rural human settlement at various scales, focusing on evolution and spatial differentiation at various scales such as city clusters and comparisons between cities, within the cities and communities. (3) Development of theory and technology for human settlement evolution research, including detection technology and methods, data mining measures, and forecasting and emulation of regional and urban human settlement evolution processes, mechanisms and patterns. (4) Research on the control of human settlement that focuses on optimization, patterns, and policies for effective management and development. (5) Estimating the human settlement system service value and establishing suitable human settlement systems, including social, economic, cultural and ecological service values.


Pan Q, Jin X B, Zhou Y K, 2015. Gridding reconstruction of land use pattern in Jiangsu Province in the mid-Qing Dynasty.Acta Geographica Sinica, 70(9): 1449-1462. (in Chinese)Reconstructing the spatial data of historical LUCC could fill the gap between modern remote sensing interpretation results and long- term LUCC change analysis, and promote the study of spatiotemporal dynamics of land use and its impacts on climate and ecology. The current studies has obtained certain achievements and revealed the spatial pattern of historical LUCC to some extent. However, the results were normally in the form of single land use type and low spatial resolution. Thus, this study targeted to propose an integrated method of land use reconstruction with relatively high spatial resolution. We set 1820 as the time section and took the administrative boundary of contemporary Jiangsu Province as the study area. Supported by historical documentary records, historical geography research results,modern statistical data and natural environmental data, we divided the land use types into cropland, settlement(including urban land and rural residential land), water body and other lands(including forest, grassland and unused land). Considering the characteristics of regional natural resources and socioeconomic conditions, theoretical hypotheses were proposed to obtain the areas of cropland, urban land and rural residential land at prefectural level. Based on the modern land use pattern, from the perspective of man-land relationship, the land use map of Jiangsu Province in 1820 with a spatial grid of 100 m 100 m was established by administrative seat proximity analysis, integrated suitability indices analysis and so on. Then, statistics of different geographical divisions and comparative analysis of data in sub- regions were used respectively to analyze and validate the results indirectly. The results showed that:(1) In 1820,the area of cropland, urban land, rural residential land, water body and other lands of Jiangsu Province accounted for 48.49%, 4.46%, 0.16%, 15.03% and 31.86% of the total study area respectively.(2) The land use pattern in Jiangsu Province has general features of higher cultivation rate and lower construction rate. Meanwhile, affected by the discrepant population distribution, topography and drainage density, land use in different geographical divisions varied significantly.(3) The proportion of the area of residential land and cropland at county level occupying the correspondent city level in 1820 and 1985 showed a significant positive linear correlation. Therefore, the results of this study had certain reasonability and could provide methodological references for related historical LUCC reconstruction.


Ray D K, Pijanowskia B C, 2010. A backcast land use change model to generate past land use maps: Application and validation at the Muskegon River watershed of Michigan, USA. Journal of Land Use Science, 5(1): 1-29.We developed a GIS and neural network-based land use/land cover change model for backcasting land use change and applied it to the Muskegon River watershed, a typical upper midwestern watershed in the USA. We developed 12 variants of the model, based on different structural assumptions, to simulate urban, forest, agriculture, and shrubland transitions. We compared the model variants against 12,598 land use interpreted locations from 235 aerial photographs acquired from the study region between the late 1930s through to the early 1970s. The model variants produced around 41 70% accuracy (integrating both omission and commission errors) in simulating the spatial locations of the dominant land use category, forests and agriculture, and lower accuracy for the shrub and urban land use categories. We describe the assumptions made in developing the model and discuss the implications of the assumptions to model goodness-of-fit analysis and to forecasting land use. The Windows executable version of the model and data sets are available for download from


Saunders D, 2012. Arrival City. Shanghai: Shanghai Translation Publishing House.

Schaldach R, Alcamo J, Koch Jet al., 2011. An integrated approach to modelling land-use change on continental and global scales.Environmental Modelling & Software, 26(8): 1041-1051.Land-use and land-cover change are important drivers of global environmental change, affecting the state of biodiversity, the global carbon cycle, and other aspects of the earth system. In this article we describe the development of the land-use model LandSHIFT, which aims to simulate land-use and land-cover change on the continental and global scale. The model is based on a “land-use systems” approach, which describes the interplay between anthropogenic and environmental system components as drivers of land-use change. LandSHIFT’s modular structure facilitates the integration of different components that cover key parts of land-use systems. The model prototype combines a module for the simulation of land-use change dynamics with a module for calculating crop yields and net primary productivity of grassland. LandSHIFT is driven by country-level model inputs including time-series of socio-economic variables as well as agricultural production data. This information is regionalized to land-use grid maps with a cell size of 5 arc-minutes. Here, the model clearly differentiates between the land-use activities settlement, crop cultivation and grazing. By using standardized input–output formats, LandSHIFT can be combined with other models for conducting complex simulation studies.


Song W, Chen B M, Yang Het al., 2008. Analysis on the status quo of residence base resources in rural areas of China.Chinese Journal of Agricultural Resources and Regional Planning, 29(3): 1-5. (in Chinese)This paper analyzes in detail the total amount of residence bases in rural areas of China,the average residence base areas per household and the development,change and status quo of residence base cubage rate.From 1996 to 2005,the total amount of residence bases in rural areas of China and residence base cubage rate all presented an increasing tendency,while the average residence base areas per household show a decreasing tendency.In 2005,the total amount of residence bases in rural areas of China was 9.1161 million hm2,the average residence base areas per household were 361.43m2.The average residence base cubage rate in rural areas was 0.268.At present,the average residence base areas per household exceeds the normal standard with very low cubage rate.In order to implement the intensive utilization of rural residence bases,we should strengthen the residence base system of one household one residence base,strictly examine and approve the residence base,complete the moving policy for residence base and vigorously promote the changeover from flat house to storied building.

Sun Z X, 2009. The research on Jiangsu rural cooperatives in the Republican Period (1927-1937) [D]. Nanjing: Nanjing Agricultural University. (in Chinese)

The Second Historical Archives of China (SHAC), 1998a. Archives Compilation of History of Republic of China (Third Series Agriculture and Commerce). Nanjing: Jiangsu Ancient Books Press. (in Chinese)

The Second Historical Archives of China (SHAC), 1998b. Archives Compilation of History of Republic of China (Fifth Series Finance and Economy). Nanjing: Jiangsu Ancient Books Press. (in Chinese)

Tian H Q, Kamaljit B, Tao Bet al., 2014. History of land use in India during 1880-2010: Large-scale land transformations reconstructed from satellite data and historical archives. Global & Planetary Change, 121(10): 78-88.In India, human population has increased six-fold from 200 million to 1200 million that coupled with economic growth has resulted in significant land use and land cover (LULC) changes during 1880–2010. However, large discrepancies in the existing LULC datasets have hindered our efforts to better understand interactions among human activities, climate systems, and ecosystem in India. In this study, we incorporated high-resolution remote sensing datasets from Resourcesat-1 and historical archives at district (N02=02590) and state (N02=0230) levels to generate LULC datasets at 502arc minute resolution during 1880–2010 in India. Results have shown that a significant loss of forests (from 8902million02ha to 6302million02ha) has occurred during the study period. Interestingly, the deforestation rate was relatively greater under the British rule (1880–1950s) and early decades after independence, and then decreased after the 1980s due to government policies to protect the forests. In contrast to forests, cropland area has increased from 9202million02ha to 140.102million02ha during 1880–2010. Greater cropland expansion has occurred during the 1950–1980s that coincided with the period of farm mechanization, electrification, and introduction of high yielding crop varieties as a result of government policies to achieve self-sufficiency in food production. The rate of urbanization was slower during 1880–1940 but significantly increased after the 1950s probably due to rapid increase in population and economic growth in India. Our study provides the most reliable estimations of historical LULC at regional scale in India. This is the first attempt to incorporate newly developed high-resolution remote sensing datasets and inventory archives to reconstruct the time series of LULC records for such a long period in India. The spatial and temporal information on LULC derived from this study could be used by ecosystem, hydrological, and climate modeling as well as by policy makers for assessing the impacts of LULC on regional climate, water resources, and biogeochemical cycles in terrestrial ecosystems.


Traffic Chronicles Compilation Committee (TCCC), 1995. Transportation History of Jiangsu Province. Beijing: China Communications Press. (in Chinese)

Wang S H, 1984. Regional Studies of China’s Modernization: Jiangsu Province. Taipei: Institute of Modern History. (in Chinese)

Wu W H, Niu S W, Guo X Det al., 2008. The empirical analysis of the village pattern evolution in the central part of Huang-Huai-Hai Plain.Geographical Research, 27(5): 1017-1026. (in Chinese)The study on the village pattern evolution is important and significative to constitute the policy of land utilization in the new period and to build new socialist countryside.The Huang-Huai-Hai Plain,located in the central part of East China,is the important farm belt,and consists of multitudinous villages with a high population density.In this region,the marketization level and urbanization level are high because of close to the coastal developed areas,and the interior development problems of the villages are prominent such as a mass of farmlands being occupied,and too many houses being abandoned.So it is one of the representative regions to research problems of village development.Progress in the study related with countrysides,settlements,farmers,and villages have been reviewed.Studies of the village pattern evolution on macroscopic scale were much more than those on microscopic scale,so microscopic studies should be strengthened to better understand village development.The Wulou village in the central part of the Huang-Huai-Hai Plain was chosen as a typical case.The face-to-face interview and the global position system(GPS)measurement method were used in the study.The main results are as follows:Firstly,the Wulou village went through a series of evolution pattern including the absolute tardiness development before the establishment of People's Republic of China in 1949,the relative tardiness development before the implementation of the economic reforms in 1978,the fast expansion in the 1980s,the steady scale in the 1990s and the recessionary development in the early 21st century.In the meantime,the roads and ponds in the village have also been changed correspondingly.Secondly,the main factors of the village pattern evolution included economic improvement,the change of social structure,urbanization and the effect of national policy.However,population variation was still the decisive factor.Thirdly,the change of village scale was not obvious before the implementation of the economic reforms in 1978,the expansion of village scale was prominent from the 1978 economic reform to the 1990s,and the empty and abandoned houses have increased largely since about 2000.Fourthly,the previous village expansion attributed to four aspects as follows:No family planning after 1949 resulted in population increasing enormously;the economic development enabled people to improve their habitations;the status of nuclear families(namely a family consists of few people which only include husband,wife and their children in general)was enhanced step by step;and farmers were not conscious of the importance of protecting plantations.Finally,the causes leading to abandoned houses lately are that the urbanization and market economy are becoming or will become the direct drive of the village hollowing at present;the profit of planting grains is so low that it has become the important impetus of the village hollowing;and the fast increase of population has been successfully controlled by the policy of family planning in China since the 1980s.Furthermore,according to the status quo of the village development in the study area and the social development demands,the corresponding countermeasures have been brought forward.


Xu D, 1983. China’s Agricultural Production and Commerce Statistics in Recent History. Shanghai: Shanghai People’s Publishing House. (in Chinese)

Xu X Q, Zhou Y X, Ning Y M, 2009. Urban Geography. Beijing: Higher Education Press. (in Chinese)

Yan Z W, Wang J, Xia J Jet al., 2016. Review of recent studies of the climatic effects of urbanization in China.Advances in Climate Change Research, 7: 154-168.

Yang X H, Jin X B, Du X Det al., 2016. Multi-agent model-based historical cropland spatial pattern reconstruction for 1661-1952, Shandong Province, China.Global and Planetary Change, 143: 175-188.To advance the research of global land use/cover change (LUCC), biodiversity, global carbon cycle, and other aspects of the earth system, it is essential to reconstruct changes in historical cropland cover with long time series and high-resolution grid. Currently, it is a general approach which is based on the view of combining the overall control of cropland area, selecting grid of high land suitability, and ‘top-down’ decision-making behaviors to reconstruct the historical cropland. Considering various factors that influenced cropland distribution, including behavioral agent's selection by itself and the limitation of nature and human factors, a spatiotemporal dynamical reconstruction model of historical cropland based on the multi-agent systems has been developed from the perspective of ‘bottom-up’, which combine macroscopic and microscopic decision-making behaviors of agents to simulate the government and farmer autonomously implementing the selection behaviors of farming area. Taking Shandong Province as the study area, this model was used to imitate its cropland spatiotemporal pattern with 102km grid-resolution from 1661 combining the contemporary pattern and reconstructed amount of historical cropland as a maximum potential scope and control variable of reconstruction model, respectively, furthermore, followed the accuracy valuation and comparative analysis. The reconstructed results show that: 1) It is properly suitable for Multi-Agent to simulate and reconstruct the spatial distribution of historical cropland; 2) compared with historical map datasets (1930s) from the view of point to point, the correctly classified producer accuracy, user accuracy and overall accuracy of reconstructed result totally up to 59.09%, 80.62% and 62.31%, respectively, and shows our reconstruction map achieved a better agreement with the historical maps; 3) from the view of grid-level or county-level, our reconstruction approach can effectively keep away from the grid with mountain-hilly and easily flooding probability, it showed a good similarity and higher consistency when compared with the research result based on archives and historical records in overall pattern and tendency of cultivation, and difference error decrease gradually, moreover, the cultivation ratios in county-level is more close to the historical situation.


Yang X H, Jin X B, Guo B Bet al., 2015. Research on reconstructing spatial distribution of historical cropland over 300 years in traditional cultivated regions of China.Global & Planetary Change, 128: 90-102.Constructing a spatially explicit time series of historical cultivated land is of upmost importance for climatic and ecological studies that make use of Land Use and Cover Change (LUCC) data. Some scholars have made efforts to simulate and reconstruct the quantitative information on historical land use at the global or regional level based on “top–down” decision-making behaviors to match overall cropland area to land parcels using land arability and universal parameters. Considering the concentrated distribution of cultivated land and various factors influencing cropland distribution, including environmental and human factors, this study developed a “bottom–up” model of historical cropland based on constrained Cellular Automaton (CA). Our model takes a historical cropland area as an external variable and the cropland distribution in 1980 as the maximum potential scope of historical cropland. We selected elevation, slope, water availability, average annual precipitation, and distance to the nearest rural settlement as the main influencing factors of land use suitability. Then, an available labor force index is used as a proxy for the amount of cropland to inspect and calibrate these spatial patterns. This paper applies the model to a traditional cultivated region in China and reconstructs its spatial distribution of cropland during 6 periods. The results are shown as follows: (1) a constrained CA is well suited for simulating and reconstructing the spatial distribution of cropland in China's traditional cultivated region. (2) Taking the different factors affecting spatial pattern of cropland into consideration, the partitioning of the research area effectively reflected the spatial differences in cropland evolution rules and rates. (3) Compared with “HYDE datasets”, this research has formed higher-resolution Boolean spatial distribution datasets of historical cropland with a more definitive concept of spatial pattern in terms of fractional format. We conclude that our reconstruction is closer to the actual change pattern of the traditional cultivated region in China.


Yin C Y, Shi Y S, Wang H F, 2013. Process and characteristics of boundary expansion of built-up area of Shanghai City since the late Qing dynasty.Progress in Geography, 32(12): 1793-1803. (in Chinese)Urban form is defined as a figure presentation of spatial pattern of city's external shape, and, together with urban pattern and urban morphology, constitutes the three subjects at different levels in the research field of urban morphology. Current studies on this subject ignored the identification of the growth process from sprouting, growth, maturity, to ageing, and its impacting factors such as historical, political and social structure. Urban form is the lowest level in the urban morphology field. However, it has a very close relationship with urban growth process and the influence factors, it is an important knowledge basis for further understanding urban pattern and urban morphology, and it is also beneficial to grasping the expansion process of the boundary of urban built-up areas and making rational measures to control the trend of urban growth. Furthermore, coastal cities that sprouted in the estuarine always have unique growth characteristics different from others cites, because of their dependence on waterway transportation in early times. For example, early urban expansion nearby a port may be related to estuaries, rivers i.e., and the correlation would weaken in the process of social and economic development. How to quantitatively describe this evolution process and influence factors? This is a question that still needs to be further discussed by using new methods and defining new spatially related indices. Therefore, based on multi-type historical maps and multi-temporal remote sensing images, this paper first employs geographic information system(GIS) software ArcGIS Desktop 10.01 and remote sensing(RS) images processing platform ERDAS Imagine 2011 to extract the information of built up area boundary of Shanghai from 1842 to present, a 170-year time period, and secondly, calculates the fractal dimension(FD) and related circumscribing circle(RCC) of built-up area boundary by Fragstats 4.1 software to measure the complexity of urban boundary expansion, use ArcGIS Desktop 10.01for statistical analysis of the fan-shaped areas of built-up area expansion in different directions such as East, West, South, North, to identify the directional characteristics, and define an index urban hydrophilic property(UHP) to express the attraction of Huangpu River and Wusong River to built-up area boundary extension. All of these metrics finally are used for analyzing the historical evolution process and characteristics of the built-up area boundary of Shanghai City. In addition, with the reference to the historical literature, this paper discusses the impact of social transformation on the urban form evolution of Shanghai. The results show that:(1) FD and RCC of urban boundary showed a rise pattern of "M";(2) urban boundary expansion showed a "single-direction" or "full-direction" mode during different historical periods;(3) the attractive effect of Huangpu River and Wusong River on built-up area boundary expansion has weakened due to the development of public transport system on the land;(4) complexity and directions of urban built-up area boundary showed significant stage characteristics along with the historical evolution process, which is mainly driven by different levels of social and economic development in the historical periods. The conclusion of the study is of reference value to Shanghai government in planning a more sustainable urban form in the future, and also makes a significance contribution to the extension of the theories and methods of urban morphology study.


Zhao E X, 1976. Draft of History of Qing Dynasty. Taipei: Zhonghua Book Company. (in Chinese)

Zhao Q C, 1941. Geography History Table of Qing Dynasty. Beijing: Zhonghua Book Company. (in Chinese)

Zheng B, 2011. Researches on rural residential areas and variation characters of housing construction[D]. Wuhan: China University of Geosciences. (in Chinese)

Zhou G H, He Y H, Tang C Let al., 2013. Dynamic mechanism and present situation of rural settlement evolution in China.Journal of Geographical Sciences, 23(3): 513-524.This paper provides a detailed analysis of the factors influencing the evolution of rural settlements, including natural environmental constraints, infrastructure, regional cultural inheritance and integration, urbanization and rural industrial transformation, land use reformation and innovation, rural household behavior conversion, macro-control policy factors, and so on. Based on differences between the ways and degree of effect on rural settlement evolution, these factors are classified into basic factors, new-type factors and mutation factors. The drive of basic factors mainly focuses on the traditional inheritance of rural settlements, the new-type factors mainly affect rural settlement transition, and the mutation factors may bring about sudden changes. All these factors constitute a hree-wheel driving mechanism for the evolution of rural settlements, and shape three typical driver paths: slow smooth path under the basic factors, new path to rapid development under the new-type factors, and the sudden change path under the mutation factors. The paper also investigates the overall situation of rural settlement evolution in the aspects of settlement system, settlement scale, settlement morphology, settlement function, settlement culture, settlement environment, etc. The general process of rural settlement evolution is divided into four stages: initial, transitional, developmental, and mature stages.


Zhu K J, Wang S M, 2008. The research on rural construction movement of Jiangsu during the period of Republic of China.Agricultural History of China, 27(4): 85-92. (in Chinese).The campaign over rural construction in Jiangsu province formed an important part of the ru- ral construction during the period of Republic of China. This thesis states basic facts of the campaign over rural construction in Jiangsu province during the period of Republic of China. Meanwhile, there are some represen- tative rural construction zones, such as experimental units in Wuxi, rural improvement experimental units in Xugongqiao, rural church experimental zones at Chunhua town, rural improvement experimental zones at Wu- jin district and self- government Jiangning experimental county. Through an overall view of the characteristics of rural construction in Jiangsu province, Jiangsu rural construction could be summed up by four construction models, that is, public education, vocational education, church construction and political reform. At that time, the campaign brought about great influence on Jiangsu rural areas. It strengthened the cooperation between peasants and promoted rural democratic construction. The economy was well developed and farmers' income was greatly increased. What's more, the standard of education on peasants was sharply increased, which, to some degree, improved the backward rural education. Above all, it fairly changed the lifestyle of peasants and advanced the modern civilization in countryside.