1. Introduction
Urban land expansion is the inevitable result of urbanization and is also the most obvious spatial characteristic of urbanization. With the process of urbanization, urban economic activities develop vigorously, and more people are attracted and flock to cities, thus promoting the rapid growth of the urban population [
1,
2,
3]. Economic development and population growth constantly increase the demand for land resources, which inevitably leads to urban land expansion [
4,
5]. As a developing country, China is in a period of rapid urbanization. With the urban population growth and economic development, the scale of urban land will increase accordingly. Urban land expansion provides a material basis for urban development. Still, there are some problems in the process of urbanization, such as the disorderly expansion of urban land, uncontrolled spatial expansion, and low efficiency of land use [
6,
7,
8], which lead to increasing tension between the people and the land and restrict regional sustainable development [
9]. Based on the basic national conditions of more people and less land, China has put forward the land management policy of economical and intensive land use, which improves the efficiency of urban land use and prevents the disordered expansion of urban land to protect cultivated land resources and solve the contradiction between “eating” and “construction”. Thus, it is very important to effectively control urban land expansion and formulate sustainable land use policies to analyze the characteristics of urban land expansion and predict the trend of urban land expansion scientifically. Recent research on urban land expansion mainly focused on characteristics and patterns of urban land expansion, driving factors, and prediction of urban land use scale [
10,
11,
12,
13,
14,
15].
Researchers have extensively discussed the spatial-temporal characteristics and driving forces of urban land expansion from different scales, which are the basis of scientific regulation of urban land use scale. Based on remote sensing data, some GIS analysis methods are used to study the spatial-temporal expansion characteristics of urban lands, such as the expansion index method, fractal dimension method [
16], and center-of-gravity model [
17]. It has been shown that urban land expansion has obvious spatial-temporal differences [
18]. The driving factors of urban land expansion are analyzed with mathematical models, such as the spatial probit model [
4], logistic regression analysis [
19], vector error correction model [
20], regression model [
21], and spatial-temporal weighted regression model [
22]. Marshall et al. [
23] pointed out that urban land expansion is closely related to population size in most regions of the world. Poghosyan [
11], Morshed [
13], and Alsharif et al. [
24] showed that population and urbanization are the main influencing factors of urban land expansion. Alice et al. [
25] further showed that population increase, income increase, and continuous improvement of infrastructure are the driving factors for urban land expansion. In addition, some studies showed that the driving forces of urban land expansion have spatial heterogeneity, and the constraining effects of natural conditions on urban land expansion have gradually weakened over time [
4]. In China, it is generally believed that social and economic development is the main driving force of urban land expansion [
26,
27,
28].
Some of the researchers have tried to use models to predict urban expansion, such as land change modeler [
10], variable weights logistic cellular automata model [
29], FLUS model [
30,
31], CA model [
32], logistic regression [
33], artificial neural networks [
34], and so on. These researches mainly predict the scale of urban land at a specific time point in the future or study the reasonable scale of urban land in a certain period from a certain factor affecting urban land expansion. Based on the characteristics of land use change in the process of urbanization, some researchers have used a variety of models to study the limit scale of urban land expansion, such as the model of stepwise regression mode and per capita land area [
35], classification and regressive method [
36], logistic model and autoregressive distributed lag model [
37]. There are few research achievements in predicting the expansion limit of urban land from the perspectives of time and space.
Shandong province is a province with a large population and economy in China, and its urbanization and national economy are at the forefront of the country. Still, Shandong province is also China’s main grain-producing area, and the contradiction between humans and land cannot be ignored. Controlling the rational expansion of urban land is not only the requirement of land resource management but also the practical requirement of cultivated land protection. Therefore, taking Shandong province as the study area, based on the impact of economic development and urbanization on urban land expansion, this paper constructs the marginal effect model and logistic model to predict the expansion limits of urban land from two perspectives of space and time, that is the scale limit and time limit. Compared with predicting the scale of urban land in a certain period, predicting the expansion limit of urban land can better reflect the future trend of urban land expansion and provide references for macro-control of urban land expansion. Predicting the expansion limit of urban land from the perspectives of time and space enriches urban land expansion research methods. Furthermore, the research results can provide reference for formulating land use policies conducive to promoting sustainable land use and long-term sustainable social and economic development in Shandong province.
3. Results
3.1. Correlation Analysis between Urban Land Expansion and Socio-Economic Development in Shandong Province
The condition of Pearson correlation analysis is that the data follow a normal distribution. Before the Pearson correlation analysis, the Shapiro–Wilk test was applied to test the normality of the output value of secondary and tertiary industries, urban population, and the urban land area. The test results show that the p-values are 0.270, 0.586, and 0.587, respectively, greater than 0.05, and the variable data follow normal distribution at the 95% confidence level (
Table 1).
The results show that the Pearson correlation coefficients between urban land area and secondary and tertiary industry output value and urban population in Shandong province are 0.97 and 0.95, respectively, with high correlation and p < α = 0.01. This indicates that economic development and urban population growth are the main influencing factors of urban land expansion in Shandong province.
3.2. Characteristics of Urban Land Expansion in Shandong Province
From 2001 to 2018, the urbanization rate of Shandong province increased from 39.20% to 61.18%. In order to ensure the needs of various activities of urban residents, the total scale of urban land in Shandong province has continued to expand with the continuous improvement of urbanization. Urban land increased from 2197.97 km2 to 6692.15 km2, an increase of 4494.18 km2 with an average annual increase of 264.36 km2 from 2001 to 2018. It can be seen that urban land expansion is the inevitable result of urbanization.
In Shandong province, urban land has also been expanding with the increase in the output value of secondary and tertiary industries. Still, the expansion range and expansion rate show a fluctuating downward trend with obvious phased characteristics (
Table 2).
Before 2006, the output value of secondary and tertiary industries in Shandong province was less than 2 trillion yuan, with relatively low economic development, but the economic growth rate was fast. From 2001 to 2006, the output value of secondary and tertiary industries increased by an average annual growth of 18.66%, and the average annual expansion rate of urban land was as high as 11.94%, with an average annual increase of 332.33 km2. During this period, economic development has a great demand for urban land.
With the continuous improvement of the economy, the economic growth rate gradually decreased. From 2007 to 2013, the output value of secondary and tertiary industries in Shandong province increased from 2197.18 billion yuan to 3877.56 billion yuan, with the average annual growth rate dropping to 10.64%. At the same time, the expansion area and expansion rate of urban land also decreased. The average annual expansion area of urban land was 252.91 km2, with an average annual expansion rate of 5.55%.
In 2014, the output value of secondary and tertiary industries in Shandong Province exceeded 4 trillion yuan, reaching a higher economic level. After that, the economic growth rate further slowed down, and the expansion rate of urban land also decreased. From 2014 to 2018, the average annual economic growth rate was 5.33%, and urban land’s average annual expansion rate decreased to 3.52%, with an average annual increase of 212.38 km2. Obviously, with the continuous improvement of the economic level and the gradual slowdown of the economic growth rate, the expansion area and expansion rate of urban land are declining. This indicates that the demand for newly-added urban land in economic development continues to decrease, and the role of urban land expansion in promoting economic growth is decreasing. From the perspective of urban land use efficiency, the output value per area of urban land in Shandong province continues to increase, from 371 million yuan/km2 in 2001 to 751 million yuan/km2 in 2018. The economic output per area of urban land is increasing. This indicates that with technological progress and industrial restructuring, urban land use efficiency keeps improving, which is also one of the reasons for the decrease in the demand for newly-added urban land, reflecting the gradual reduction of the dependence of economic development on newly-added urban land. Therefore, it can be predicted that the scale of urban land will not continue to expand. When the economy develops after a certain level, both the slowdown of economic growth and the improvement of land use efficiency will reduce the demand for urban land, and the scale of urban land will gradually stabilize, which conforms to the law of diminishing marginal effects.
3.3. Limit Prediction of Urban Land Expansion Based on Economic Development
3.3.1. The Limit of Time
In order to reduce the impact of heteroscedasticity problems on model estimation, the heteroscedasticity test should be conducted in linear regression. In this paper, the White heteroscedasticity test method was used, and the test results are shown in
Table 3. The results showed that
p < 0.05, indicating that the null hypothesis was rejected and there was heteroscedasticity.
In this paper, the commonly used logarithmic transformation method of raw data was used to eliminate the influence of heteroscedasticity. The logarithms of urban land area and the output value of secondary and tertiary industries in Shandong province from 2001 to 2018 were taken and expressed as ln
C and ln
D, respectively. The regression analysis of ln
C and ln
D is carried out by Eviews software, and the regression results are shown in
Table 4.
The regression results showed that the correlation between ln
C and ln
D reached 0.995, indicating that the fitting degree between them was good and significantly correlated. The constant is 7.155, and the regression coefficient is 0.577. Accordingly, the expression of their relation is obtained.
Take the derivative of Equation (8) to calculate the marginal urban land:
Take the derivative of Equation (9) to calculate the change rate of marginal urban land:
The marginal urban land and its change rate in Shandong province are shown in
Figure 2.
Figure 2 shows that the marginal urban land in Shandong province showed an obvious downward trend from 2001 to 2018, and the change rate of marginal urban land was negative and increased towards zero. It shows that in the process of urban development, technological progress and industrial restructuring have effectively improved urban land use efficiency, thereby weakening the demand for new urban land in economic development. The marginal urban land will decrease gradually, and its change rate will approach zero eventually. When the change rate of marginal urban land is zero, marginal urban land will also approach zero, and urban land will reach the limit of expansion.
The results of the regression analysis of ln(|
M′|) and
t are shown in
Table 5.
The relation between ln(
M′|) and
t is obtained from the results of regression analysis:
Let |M′| = 0.00001, which can be regarded as the change rate of marginal urban land is infinitely close to 0, and obtain t = 2032.56. That is, the expansion of urban land in Shandong province will reach its limit in 2033 (rounded).
3.3.2. Limit Scale
Put the urban land area and the output value of secondary and tertiary industries in each year into Equation (6) to obtain the logistic expression (Equation (12)) with the highest fitting degree after multiple fittings. Relevant parameter values are shown in
Table 6. According to the values of R
2 and F, it can be judged that the fitting effect is good with statistical significance. Based on this, the limit scale of urban land expansion in Shandong province is estimated to be 7982 km
2.
3.4. Limit Prediction of Urban Land Expansion Based on Urban Population Growth
3.4.1. The Limit of Time
Taking the urbanization rate of Shandong province from 2001 to 2018 as the dependent variable and the year as the independent variable (set 2001 as 1), the logistic model was selected for regression, and the K value was set as 100. The results are shown in
Table 7. The fitting goodness of regression is 0.9907, and the sig value is less than 0.01, which indicates that it has a good fitting effect. The logistic expression of the urbanization rate (Equation (13)) is obtained by substituting the values of each parameter into the formula. When
y = 80,
t = 36.39, i.e., the urbanization rate of Shandong province will reach 80% in 2037, and the scale of urban land will reach the limit of expansion.
3.4.2. Limit Scale
In order to improve the prediction accuracy, the resident population data from 1978 to 2018 were selected to fit the logistic curve of the total population change in Shandong province (
Figure 3).
The fitting parameters are shown in
Table 8, and the logistic model of the total population growth in Shandong province was constructed (Equation (14)).
R
2 = 0.9904 and Sig. < 0.01 indicate that the curve-fitting effect is ideal (
Table 8). The average error between the fitting value and the actual value is only 1.29%, which shows that the prediction result of the logistic model is accurate and reliable and can reflect the population change trend of Shandong province. According to this model, it is predicted that the peak growth of the total population in Shandong province is 113.36 million, so the urban population can be calculated when the urbanization rate is 80%. In combination with the Code for Classification of Urban Land Use and Planning Standards of Development Land (GB50137-2011) [
49], and the requirements of economic and intensive land use, which is calculated by 100 square meters per person, the limit scale of urban land expansion in Shandong province based on urban population growth is 9068.8 km
2 by Equation (7).
4. Discussion
4.1. Comparison of the Two Predictions
Comparing the two predictions, it can be found that the limit time of urban land expansion in Shandong province is relatively close under these two perspectives. The limit time of urban land expansion from the perspective of economic development is slightly earlier than that from the perspective of urban population growth. The limit scale of urban land expansion from the perspective of urban population growth is larger than that from the perspective of economic development. The main reason is that according to the relationship between urban land expansion and economic development, industrial structure optimization and technological progress improve urban land use efficiency, and economic growth gradually reduces the demand for newly-added urban land along with social and economic development. When the demand for newly-added urban land tends to zero, it is considered that urban land has reached the expansion limit. At this time, although economic development has got rid of the dependence on urban land expansion, the urban population is still increasing along with the urbanization process. Moreover, with the improvement of socioeconomic level, the requirements of urban residents for production and living conditions and the urban environment will also increase. The increased demand for various types of land use such as residential, transportation, infrastructure and public service facilities, and ecological environment construction will also lead to urban land expansion. Thus, from the predicted results, the limit scale of urban land expansion from the perspective of urban population growth is larger than that of economic development.
Based on the comprehensive analysis of the two prediction results, the urban land in Shandong province will reach the expansion limit in 2037 with a limit scale of 9068.8 km2. Compared with the urban land area of 6521.61 km2 in 2018, the potential of urban land expansion in Shandong province is 2547.19 km2 in the future, and the duration of expansion is 19 years.
4.2. Urban Land Expansion Trends and Urban Development Policy Requirements
In this study, the results show that with urbanization and economic development, the dependence of economic development on urban land expansion has gradually decreased in Shandong province. According to the prediction results, after 2033, the role of urban land expansion in promoting economic development in Shandong province will be significantly reduced, and more newly-added urban land will be used to improve the urban residents’ production and living conditions and the urban ecological environment. This trend is in line with the requirements of China’s strict land use management system and high-quality urban development policies. Under the guidance of the policies of intensive land use and the transformation of old growth drivers into new ones, in the future, the urban industrial structure of Shandong province will be continuously optimized, and land use efficiency will continue to improve. This will lead to the demand for economic development for new urban land gradually decreasing. In addition, in the context of the current high-quality urban development, improving people’s livelihood and strengthening urban ecological construction in the future will become the focus of urban development. Thus, the demand for infrastructure land, public facilities land, ecological land, residential land, and traffic land will further increase, leading to urban land expansion.
4.3. Research Methods
Socioeconomic development is the main influencing factor of urban land expansion. Many studies have predicted the scale of urban land in a certain period based on the influencing factors of urban land expansion. The research on the prediction of urban land limit scale is mainly carried out from the perspective of the demand for construction land in the process of urbanization, and there is limited research on the expansion limit of urban land from the perspective of the economic effect of urban land use. This study introduces the marginal effect principle in economics and constructs the marginal effect model of urban land expansion from the perspective of economic development demand for urban land to predict the expansion limit of urban land. This model helps improve the research method of urban land expansion. In addition, compared with the prediction of urban land in a certain period, based on urbanization and economic growth, predicting the expansion limit of urban land from the perspectives of time and space can better reflect the trend of urban land expansion. Moreover, from the perspectives of time and space, the prediction of the expansion limit of urban land also enriches the research content of urban land expansion to a certain extent.
4.4. Deficiencies and Future Research
This paper only predicts the expansion limits of urban land from the perspectives of economic development and population growth. However, urban land use change is a dynamic and complex process, and urban land expansion is affected by a variety of factors, such as natural factors, policies, and city functional orientation. In the future, it is necessary to consider the comprehensive impact of various factors on urban land expansion and further improve the prediction model. Moreover, due to the regional differences in population distribution and socioeconomic development, the characteristics of urban land expansion are different in different regions and cities of different levels. Therefore, empirical studies on the expansion limits of urban land according to the development characteristics of different cities can improve the efficiency of urban land expansion regulation, which is also the direction of further research.