Land Cover Changes and Urban Expansion in Chongqing,China: A Study Based on Remote Sensing Images

Abstract

China has experienced an unprecedented rate of urbanization in recent decades. As a city with strong political and economic influences in the southwest of China, Chongqing is a typical example of rapidurban development in this period of time. To study the land cover changes and urban expansion of Chongqing, Landsat images from 1999 to 2018 were selected, processed, and quantitatively analysed The results showed that the built-up area of the city had increased tremendously during these years, yet vegetation still accounted for the vast majority of the city’s land area. Restricted by the local topography including mountains and hills and infrastructure constructions, the urbanization process that occurred in central Chongqing actually showed a dominant expansion direction, an obvious spatial clustering tendency, and significant spatio-temporal differences among various regions.

Keywords

Land cover change urban expansion remote sensing Chongqing China

Introduction

China is one of the biggest developing countries in the world and has experienced rapid urbanization in recent decades (Ren et al., 2018). The ‘Opening up’ policy in 1987 and the establishment of the ‘Socialist Market Economy’ in 1992 greatly promoted the growth of industries and commerce (Hao, 2008) and gave a strong impetus for its urbanization process (Zhang, 2008). China’s city development has entered into the accelerated phase particularly after 1997, and the rate of urbanization increased rapidly from 31.00% to 65.50% during 1997–2018.

Chongqing, the most populous city of China, is located in the southwest part of the country, covering an area of 82,400 square kilometres and administering 38 districts and counties. Due to its special geographic location, Chongqing is widely known as the ‘mountain city’ and the ‘city on rivers’. While hills and mountains account for 76% of Chongqing administrative area; the Yangtze River actually traverses the whole territory of the city with a flow length of 691 kilometres.

Along with Beijing, Shanghai and Tianjin, Chongqing is a provincial-level municipality under the direct administration of the central government. By merging the neighbouring prefectures, including Fuling, Wanxian and Qianjiang, its administrative area expanded exponentially in 1990s hereafter (Cao et al., 2015). Since 2007, when it was first administered as one of the four municipalities, Chongqing was tasked with the resettlement of migrants from the Three Gorges Reservoir and has subsequently become a pioneer of urban development in the western region of China (Li et al., 2011). In 2010, it was identified as one of five national cities and the only one located in western China. In the same year, the Liangjiang New Area of Chongqing was established as the third state-level reform and opening-up area and the only one in the hinterland. Among all of the Chinese mega cities, Chongqing has been expected to become an advanced manufacturing and service industry base.

A lot of literature has been devoted to the study of urbanization in China (Bai et al., 2012; Wang et al., 2015). From first-tier cities (You, 2016; Xu et al., 2019) to the cities of prefecture and county levels (Liu et al., 2015; Ye & Wu, 2014), the variations in population (Wang et al., 2017), land (Wang et al., 2018) and spatial form (Lin et al., 2015) have been widely discussed from different perspectives. While some studies have opted to use quantified indexes to evaluate the dynamics of urban expansion and modes of urban sprawl (Gong et al., 2018), others have done so by means of parameter analysis, through which the characteristics and determinants of urban growth were explored (Zhang & Wang, 2018). In addition to the above, there have also been studies that are focused on the coupling relationships between development plans and practical effects of urbanization in Chinese cities (Chen et al., 2019).

Among all of the approaches, remote sensing data analysis is the most popularly adopted in empirical studies of urbanization processes (Kamusoko, 2017). To understand the spatio-temporal changes of cities at the macro level, various metrics such as area patch density, fractal dimension and shape index can be explored from remote sensing images, which act as quantitative descriptions of the urban landscape (Xu & Min, 2013). With reference to the landscape metrics, the intensity and manner of urban sprawl is able to be well interpreted (Zeng et al., 2015), giving indications for the identification and explanation of urbanization hot spots (Lopez et al., 2017). Beyond that, based on the results of remote sensing image classification (Rozenstein & Karnieli, 2011), the land use changes (Dadashpoor et al., 2019) and spatial evolution patterns of cities (Taubenböck et al., 2017) can also be recognized, thus providing essential information for human settlement and environment impact modelling (Melchiorri et al., 2018; Mohan & Kandya, 2015).

Based on remote sensing images, the analysis and discussions in this article have been carried out with the aim of precisely figuring out the land cover changes and urban expansion of Chongqing under the background of fast urbanization. The article is organized in five sections. Following the introduction section, the study area and satellite image datasets are briefly introduced in the second section. The third section elaborates the land cover change of Chongqing during the last 20 years. The general direction, spatial-temporal trends and patterns of urban sprawl of Chongqing are analysed in the fourth section. The fifth section concludes the article.

Study Area

Among the 38 districts and counties of Chongqing, Yuzhong, Dadukou, Jiangbei, Nanan, Shapingba, Jiulongpo, Beibei, Yubei and Banan districts comprise the main city, where the land cover has changed most significantly and in which most of the urbanization over the past 20 years has occurred. For the purposes of representation, the nine central districts, with a total area of 5,472.68 square kilometres, are chosen as the region of study in this article (Figure 1).

Figure 1.

Source: The authors.

Generally, the main city of Chongqing is featured with polycentric groupings. Urban population and functions have been gradually organized in each of the central districts (Table 1). They are relatively independent but also connected to each other, having developed in a coordinated manner. The city of Chongqing grew out of the districts of Banan and Yubei which, preceding this, initially developed out of the counties of Ba and Jiangbei. The Beibei district was once the relocation area of the provisional capital of Chongqing during the period of the Republic of China, where the administrative body and the Presidential Palace of the Republican China Government was located. Up until now, however, the Yuzhong district has remained the core area and the seat of the municipal government, while Shapingba holds its place as the cultural and educational centre of Chongqing.

Table 1.

Basic Information of the Central Districts of Chongqing

District	Area (km²)	Residents (in million)	GDP (RMB)		Urbanization Rate
District	Area (km²)	Residents (in million)	Total (in billion)	Per Capita (in thousands)	Urbanization Rate
Yuzhong	23.71	0.67	112.22	166.94	100.00
Dadukou	103.00	0.35	19.65	56.47	97.32
Jiangbei	220.77	0.88	87.94	99.76	95.70
Nanan	265.00	0.87	79.16	90.58	94.97
Shapingba	396.00	1.14	86.02	75.20	94.58
Jiulongpo	432.00	1.22	113.04	82.52	92.20
Beibei	755.00	0.81	53.00	65.75	80.83
Yubei	1,452.00	1.62	144.72	89.20	80.24
Banan	1,825.00	1.08	71.66	66.28	79.08

Source: Chongqing Statistical Yearbook 2017.

Note: 1 RMB = $0.1512.##

Methodology

From the open datasets of Landsat Thematic Mapper (TM), we selected and extracted remote sensing images of Chongqing between 1999 and 2018 with a 5-year interval. To minimize potential deviations caused by seasonal solar altitude and vegetation areal changes among different seasons, remote images of the same periods of time in each year were selected as far as possible. Taking into account the accessibility and quality of data, the remote sensing images used for the case study of Chongqing, as presented in Table 2, mainly include the Landsat 4-5 TM images of the city for 1999, 2004 and 2009 and the Landsat 8 OLI_TIRS images of the city for 2014 and 2018 (see Figure 2).

Table 2.

Remote Sensing Datasets of Chongqing

Year	Stripe No.	Row No.	Date Imaged	Data Source
1999	127	39	2000/5/20	Landsat 4-5 TM sat. data product
	127	40	2000/5/20	Landsat 4-5 TM sat. data product
	128	39	2000/8/31	Landsat 4-5 TM sat. data product
	128	40	2000/12/21	Landsat 4-5 TM sat. data product
2004	127	39	2003/3/10	Landsat 4-5 TM sat. data product
	127	40	2003/3/10	Landsat 4-5 TM sat. data product
	128	39	2005/4/7	Landsat 4-5 TM sat. data product
	128	40	2005/4/7	Landsat 4-5 TM sat. data product
2009	127	39	2010/9/21	Landsat 4-5 TM sat. data product
	127	40	2010/8/4	Landsat 4-5 TM sat. data product
	128	39	2010/5/23	Landsat 4-5 TM sat. data product
	128	40	2010/8/11	Landsat 4-5 TM sat. data product
2014	127	39	2014/7/30	Landsat 8 OLI_TIRS sat. data
	127	40	2013/12/2	Landsat 8 OLI_TIRS sat. data
	128	39	2014/7/21	Landsat 8 OLI_TIRS sat. data
	128	40	2014/8/6	Landsat 8 OLI_TIRS sat. data
2018	127	39	2018/1/14	Landsat 8 OLI_TIRS sat. data
	127	40	2017/4/1	Landsat 8 OLI_TIRS sat. data
	128	39	2017/5/26	Landsat 8 OLI_TIRS sat. data
	128	40	2017/5/26	Landsat 8 OLI_TIRS sat. data

Source: The authors.

Figure 2.

Source: The authors.

According to the spectral characteristic curves of various land use types in the main city of Chongqing, the difference between the 3rd, 4th, and 5th bands was actually the largest among the six bands of TM remote sensing images. Thus, to effectively detect the land cover changes and figure out the urban expansion manner, the bands 3, 4 and 5 were selected as the components of false colour images for empirical studies, assigning band 5 as the red colour, band 4 as the green and band 3 the blue.

Based on the remote sensing images, spatial analysis was conducted from multiple perspectives with the aim of identifying the direction and scale of urban expansion, spatial structural changes taking place in the city and studying the evolution pattern of land covers (see Figure 3).

Figure 3.

Source: The authors.

First, to ensure the quality and precision of information extraction, the remote sensing images were pre-processed. Specifically, the image data was geometrically corrected, radiometrically calibrated, band combined, stitched and cropped (Zhu, 2017) using the Environment for Visualizing Images (ENVI).

Second, with the pre-processed image data, the land cover of central Chongqing was recognized through a supervised classification method. According to the widely accepted Chinese land classification system recommended by the national standard, the ‘current land use classification’ (GBT 2010–2017), six land cover types, namely, cultivated land, woodland, grassland, built-up area, water body and bare land were considered in the procedure of image sample training and classification.

Third, judging from classified remote sensing images across different years, the historical land cover changes and variations in built-up areas in the main city of Chongqing were detected and quantitatively analysed. Then, by means of scale and intensity comparison of urban expansion in different directions, the outlines and spatial characteristics of urban sprawl during these periods of time were examined.

Finally, the Moran’s I and Getis’ G indexes were calculated for quantitative analysis on the relationships between each local part of the main city and its neighbouring areas in the process of land cover change. Based on the two parameters calculated, Local Indicator of Spatial Association (LISA) plots and heat maps were further generated, with which the spatial clustering features and evolution patterns of urbanization were recognized.

Results and Discussions

Land Cover Change and Urban Expansion of Chongqing

Great changes have taken place in Chongqing over the last two decades. Driven by industrial development, population mobility and policies, the city has been internally reorganized and is externally expanding. However, different from the urbanization process of plain cities, the unique historical social conditions and natural ecological environment of the mountain city is what drove its special features and the pattern of its spatial evolution.

Land Cover Change Detection

As the Jeffries-Matusita distances (Bruzzone et al., 1995) among cultivated land, woodland and grassland in TM remote sensing images is less than 1.0 (see Table 3), the three classes of land were not analysed separately. The land covers of the city were just classified as four types, the built-up area, water body, bare land and vegetation. Based on this principle, the remote sensing datasets in 1999, 2004, 2009, 2014 and 2018 were treated with the maximum likelihood approach, resulting in the maps as shown in Figure 4.

Table 3.

Jeffries-Matusita Distance Matrix of Different Land Cover Classes

Land Cover Classes	Cultivated Land	Woodland	Grassland	Built-up Area	Water Body	Bare Land
Cultivated land	0.00	0.56	0.44	1.50	1.77	1.38
Woodland	0.56	0.00	0.68	1.61	1.79	1.42
Grassland	0.44	0.68	0.00	1.58	1.76	1.37
Built-up area	1.50	1.61	1.58	0.00	1.77	1.18
Water body	1.77	1.79	1.76	1.77	0.00	1.69
Bare land	1.38	1.42	1.37	1.18	1.69	0.00

Source: The authors.

Figure 4.

Source: The authors.

Comparing the results with the manual interpretation samples of the remote sensing images, the accuracy of land cover classification was evaluated. According to the assessments (see Table 4), the overall accuracies of land cover classification of the 5 years were all above 80%, with the kappa coefficients no less than 0.70. This is especially so for the built-up area, in which the producer’s and user’s accuracies were all above 90%.

Table 4.

Accuracy Assessments of Land Use Land Cover Classification

Year	Class	Producer. Accuracy (%)	User Accuracy (%)	Overall Accuracy (%)	Kappa Coefficient
1999	Built-up area	90.33	92.59	80.07	0.70
	Water body	100.00	91.23
	Vegetation	75.68	80.72
	Bare land	93.04	90.47
2004	Built-up area	96.52	98.90	81.61	0.72
	Water body	80.71	73.44
	Vegetation	80.26	87.58
	Bare land	95.34	91.86
2009	Built-up area	92.59	92.34	81.52	0.71
	Water body	90.80	92.66
	Vegetation	85.56	78.47
	Bare land	92.19	80.99
2014	Built-up area	98.21	95.43	82.27	0.73
	Water body	89.56	87.87
	Vegetation	83.82	90.15
	Bare land	84.63	90.27
2018	Built-up area	95.95	97.49	83.90	0.76
	Water body	80.37	88.42
	Vegetation	88.62	80.86
	Bare land	95.90	91.88

Source: The authors.

As shown in Figure 5, within the main city of Chongqing, mountains and green vegetation represent the largest contributors, with their proportion declining a little in 2004 to about 68.1% but continuing to rise thereafter. From 1999 to 2004, bare land contributed to approximately 23% of the total land area, attributable to the stagnant urbanization of the city in those years. Many engineering projects were stopped during this time, leading to an increase in bare land and heavy dust pollution. Following this, as the ‘farmland to forests’ initiative was carried out and people became more aware of the issues of ecological sustainability, the proportion of the green vegetation has registered a year-by-year increase in contrast to that of bare land, which dropped to around 5% in 2018. By contrast, the overall area of water body has remained stable with minor increases, contributing to approximately 3.0% of the total area.

Figure 5.

Source: The authors.

Urban Expansion Analysis

The built-up area in Chongqing has increased continuously from 150.28 km² in 1999 to 830.41 km² in 2018 (see Figure 6). The constant high-speed expansion of construction land in the main city from 1999 to 2018 has influenced the area of bare land to a great deal, especially in the period between 2009 and 2014.

Figure 6.

Source: The authors.

Judging from the spatial distribution of built-up area in Figure 7, it is easy to see that steep changes have taken place in central Chongqing over the past two decades. The built-up area gradually expanded outward from the Yuzhong district, wherein the two rivers joined and eventually gave shape to the main city of Chongqing, which today encompasses nine urban districts. Separated by longitudinally extending mountains, the central district today is divisible into three parts: the left, the middle and the right region.

Figure 7.

Source: The authors.

To analyse the extent of urban expansion in different directions, the entire built-up area of central Chongqing was divided into 16 sectors and numbered anti-clockwise from N, NEE, NE, NEE, E, SEE, SE, SSE, S, SSW, SW, SWW, W, NWW, NW to NNW. With the Liberation Monument of Yuzhong district as the centre and 64.45 km as the radius, the entire built-up area was covered in a circular region. Table 5 presents the proportions of construction land area in central Chongqing in different directions from 1999 to 2018. Statistical analysis was carried out on each of the 16 sectors and the result presented in a radar chart (see Figure 8).

Table 5.

Proportion (%) of Construction Land Area by Direction During 1999–2018

	1999	2004	2009	2014	2018
N	0.44	1.43	3.90	6.77	9.67
NNE	1.02	2.40	6.35	9.21	11.43
NE	0.30	0.58	1.18	2.16	4.28
NEE	0.27	0.48	2.62	6.89	10.12
E	0.22	0.28	0.66	1.91	2.94
SEE	0.05	0.08	0.65	1.45	2.08
SE	0.08	0.22	0.54	0.91	1.29
SSE	0.08	0.20	0.57	1.12	1.70
S	0.22	0.36	0.85	1.95	3.21
SSW	1.23	1.79	2.70	3.58	4.60
SW	2.46	3.36	5.75	8.28	10.86
SWW	3.03	4.33	6.46	8.92	11.11
W	3.34	4.11	6.44	9.47	11.90
NWW	2.62	3.99	7.83	11.15	13.98
NW	1.34	2.54	4.40	6.26	7.96
NNW	1.11	2.24	5.38	8.52	10.88

Source: The authors.

Figure 8.

Source: The authors.

Comparing the expansion scale of built-up area in different directions and during different periods of time, the spatial characteristics and temporal variations of urban sprawl of central Chongqing were analysed.

The urban sprawl generally displayed a development trend that followed in the north-western direction. This was attributable to the natural environment, economic context, government policy framework and transportation development of the city at the time. In fact, the expansion of the built-up area in the city took place in the north of the Yangtze River throughout all periods of time. Centring around the Yuzhong district, the main city of Chongqing has mainly developed along the north-western direction, which is flatter. The parallel ridge-valley geomorphology of the main city restricted further expansion of the old urban area in earlier years, resulting in limited increases in the built-up area before 2001. After 2001, although the impact of natural conditions on the urban expansion was not reduced, it could not reverse the trend of urban spatial structure change from compact to loose. Besides, subject to mountain barriers, the built-up area in the main city of Chongqing also showed a ‘southward first, northward next’ pattern of sprawl along the valleys.

Before 1997, Chongqing used to be part of the Sichuan province and its economic development was restricted. After 1997, the city has reported a continuously rising growth in GDP and remarkable increases in the proportion of built-up area. At the end of 2016, the built-up area in the main city almost equalled that in the jurisdiction of Chengdu City and the total GDP exceeded two trillion yuan for the first time in 2018.

Before 2001, the construction of transportation infrastructure in the main city of Chongqing used to be weak and the urban groupings were poorly interconnected. After 2001, rapid economic development has backed up the structuring of a transportation network. The construction of Metro Line 1, the Chengdu-Suining high-speed railway and the outer-ring expressway boosted the formation of some new groupings or industrial zones, giving rise to a high spatial concentration of regional-scale built-up area expansion along the transportation lines.

Also, from 1997 to 2009, when remediation and re-planning was required for the city, the urbanization process was quite slow, resulting in limited urban expansion. Since 2009 onwards, the local government strengthened urban construction, particularly the infrastructures, transport systems and local accessibility which led to explosive urbanization in the city.

Discussions on Spatial Clustering Features

The spatial autocorrelation of property values indicates the similarity of nearby objects in the geographic space (Getis, 2008). It can be measured with both global and local indexes. While the former points to the existence of a spatial clustering phenomenon in the studied region, the latter is able to figure out the exact locations and features of spatial clustering areas. Thus, to depict the spatial structural changes and historical developments of the city, we divided the study area into gridded regions with a side length of 1.5 km and calculated the parameters of global and local Moran’s I.

Analysis with Global Moran’s I Index

As presented in Table 6, the global Moran’s I index value of the built-up area in central Chongqing was greater than 0.7 for all the years from 1999 to 2018, which suggests that the spatial distribution of urban land is significantly correlated. Furthermore, the resulted Z score values from the random distribution test were also far higher than the critical value of 1.96 at the confidence of 0.05, demonstrating an obvious convergent clustering tendency in the study area.

Table 6.

Values of Global Moran’s I Index and Z Scores

Year	1999	2004	2009	2014	2018
Moran’s I	0.79	0.83	0.82	0.81	0.82
Z score	60.06	62.74	61.69	61.09	62.08

Source: The authors.

Analysis with Local Moran’s I Index

The urbanization process in each local area of Chongqing is correlated to that of the nearby regions in one way or another. In order to understand how spatial divergences across regions have evolved, local Moran’s I index, which revealed the spatial differences and correlations among observation values, was calculated and analysed here.

Using the GeoDa software (Anselin et al., 2006), the scatter plots of local Moran’s I for Chongqing city in 1999, 2004, 2009, 2014 and 2018 were yielded (see Figure 9). The scatter plots depict the local spatial autocorrelation of urban construction in each of the gridded regions through the urbanization process (Negreiros et al., 2010). The x-axis value of each point represents the normalized score of urbanization for each region, whereas the y-axis value represents the normalized score of the spatial lag of urbanization. The four quadrants of the Moran’s I scatter plots relate to four types of local spatial connections between each region and nearby regions:

The first quadrant, ‘high-high’ (HH), means that the attribute is quite high for a gridded region and for the nearby regions.

The second quadrant, ‘low-high’ (LH), means that the attribute is low for a gridded region but is high for the nearby regions.

The third quadrant, ‘low-low’ (LL), means that the attribute is low for a gridded region and for the nearby regions.

The fourth quadrant, ‘high-low’ (HL), means the attribute is quite high for a gridded region but is quite low for the nearby regions.

Figure 9.

Source: The authors.

While ‘HH’ and ‘LL’ indicate a strong spatial positive correlation and a high spatial homogeneity for a region, ‘HL’ and ‘LH’ indicate a strong spatial negative correlation and a high spatial heterogeneity for a region.

From the Moran’s I scatter plot, it can be observed that:

Most of the gridded regions fall into the first or third quadrants. This means that within Chongqing, highly urbanized regions were adjacent to one another and lowly urbanized regions were also adjacent to one another. This indicates that urbanization has been strongly clustered and fast-urbanizing regions have driven development in the nearby regions. Slowly urbanizing regions, which normally lied on the suburbs, also have slowly urbanizing neighbours. This gives rise to the ‘HH’ or ‘LL’ clustering of the city.

The number of gridded regions that fall into the first quadrant is far larger than that in the third quadrant. This suggests that fast-urbanizing regions were more convergent, as exemplified by the spatial concentration of Yuzhong district.

The number of gridded regions in both the first and the third quadrants has been increasing year by year. This suggests that urbanization in the city has given rise to city zones, wherein urbanization has driven the development of the nearby regions. In the suburb regions away from the city zones and in sections connecting highways between administrative districts or counties, urbanization has been stagnant and the nearby regions were less developed.

Discussions on Urban Evolution Pattern

Mapping the local features of urban sprawl into the geographic space, LISA plots (Anselin, 1995) and Getis’ G heat maps (ESRI, 2010) were generated to study the characteristics and hot spots of built-up area, with which the spatial evolution pattern of Chongqing was recognized.

Characteristic Study with LISA Plots

According to the LISA agglomerative plots (see Figure 10), there was a distinct spatial distribution pattern for the urbanization of the gridded regions in central Chongqing. It is characterized by the following:

The number of ‘HH’ gridded regions has been increasing year by year, suggesting that not only have these regions been developing rapidly themselves, but their neighbours have also been growing fast and urban groupings have taken shape. This is a significant indicator for the soaring rate of urbanization of Chongqing over the past few years.

The northern, eastern and southern outskirts of the study area, which are typically covered by sparsely populated tall mountains and, therefore, do not favour land development, have not been much affected by urbanization over the past two decades.

Some ‘LH’ gridded regions also appeared in the urban groupings. This is suggestive of the timing sequence and the unbalanced nature of the urbanization process—despite their rapid-growing neighbours, these regions have been developed at a slower pace for one reason or another, thus remaining well behind surrounding areas.

Figure 10.

Source: The authors.

As analysed above, after two decades of development, the spatial difference among urbanized regions in Chongqing has been expanding. An inter-regional polarization has been emerging and the urban sprawl in the main city remains unbalanced. While playing the role of clustering during urbanization, the core zone of the city, centring round Yuzhong district, also showed some diffusion effects by driving the development of the nearby regions. On the other hand, the outskirts and mountainous areas away from the cities have been progressing at a very low pace. In the backward regions lying next to one another, farmers were self-sufficient and do not search for anything more than fulfilling their basic everyday needs. The urbanization rate of the main city, including Yuzhong, Dadukou, Nanan, Shapingba, Jiangbei and Jiulongpo districts, has exceeded 90%. Due to its favourable geographical location and historical importance, Yuzhong district has always developed faster than any other district or county.

Hotspot Study with Getis’ G Heat Maps

According to the heat maps generated with ArcGIS (see Figure 11), there was an obvious spatial difference in the urbanization levels of Chongqing among different periods of time, which was characterized by the following:

The built-up area of the city has gained remarkable development over the past 20 years, with a focus on different regions during different periods of time. Overall, urban development has gradually expanded outward, taking the Yuzhong district as the centre.

From 1999 to 2004, the urbanization of the city was most remarkable along the Yangtze River in a south-to-north orientation, around Yuzhong district as the centre. In 2009, development was the fastest on the northern side encircled by the Yangtze and Jialing Rivers. Different levels of urban expansion also occurred in Shapingba and Jiangbei districts.

From 2009 to 2018, urbanization in central Chongqing was basically finalized. Urban building land expansion typically occurred in Yubei, Beibei, Shapingba and Jiulongpo districts, with pronounced expansion speeds and efficiency.

Figure 11.

Source: The authors.

Conclusion

In the central and western regions of China, Chongqing is the only municipality that is under the direct administration of the central government. Although the hills and mountains account for over 70% of the city’s administrative area, it is the biggest city in China with a population of over 30 million residents. It has witnessed the rapid social and economic development of urban China ever since 1997.

Although the built-up area increased more than 5 times in the north-south and then the east-west directions over the latest 20 years, it still accounts for less than a seventh of the whole area of central Chongqing. While the area of vegetation and water body did not decrease noticeably, the bare land underwent enormous developments. With the exception of the land near the suburbs, the plain ground in the city centre has almost all been concretized.

Urban expansion in Chongqing has experienced a course of slow to fast change and has been clearly unbalanced in the main city. Along the Yangtze and the Jialing Rivers, the city has expanded rapidly. However, due to the terrain and historical reasons, Yuzhong district located in the centre actually developed much faster than other areas. Restricted by the mountains and infrastructure constructions, northwest was the most dominant direction of developments.

There has been a clear tendency toward spatial convergence and development polarization in central Chongqing. Urbanization in the city has been clustered at a number of regions and urban groupings formed across different districts. With full urban functions, the urban groups are all relatively independent but also connected to each other. When compared with the outskirt regions, the core area surrounding Yuzhong district exhibited the stronger driving force of development.

As for the sustainable development of Chongqing city in the long run, there are still many issues to be addressed. One aspect of this concerns the prevention of disorderly urban expansion (Brueckner, 2000) and potential environmental destruction (Uttara et al., 2012), towards which policies should be formulated to strictly control and manage land acquisition and sales in the future. Another is the coordinated development of different regions (Rodríguez-Pose, 2008), for which it would be recommendable to employ more efforts in the urbanization of surrounding counties (Shao et al., 2005) and avoid only focusing on the central districts. In addition, faced with overcrowding in the built-up area in the main city, it is necessary to take a more reasonable and scientific approach to land use planning according to spatial distribution and growth of the population.

Footnotes

Declaration of Conflicting Interests

The authors declare no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

This article was written as part of the work of the Centre for Sustainable, Healthy and Learning Cities and Neighbourhoods (SHLC), which is funded via UK Research and Innovation, and administered through the Economic and Social Research Council, as part of the UK Government’s Global Challenges Research Fund. Project Reference: ES/P011020/1.

References

Anselin

(1995). Local indicators of spatial association—LISA. Geographical Analysis, 27(2), 93–115.

Anselin

, Syabri

, & Kho

(2006). GeoDa: An introduction to spatial data analysis. Geographical Analysis, 38, 5–22.

Bai

, Chen

, & Shi

(2012). Landscape urbanization and economic growth in China: Positive feedbacks and sustainability dilemmas. Environmental Science & Technology, 46(1), 132–139.

Brueckner

J. K.

(2000). Urban sprawl: Diagnosis and remedies. International Regional Science Review, 23(2), 160–171.

Bruzzone

, Roli

, & Serpico

S. B.

(1995). An extension of the Jeffreys-Matusita distance to multiclass cases for feature selection. IEEE Transactions on Geoscience and Remote Sensing, 33(6), 1318–1321.

Cao

, Bai

, Zhou

, & Ai

(2015). Gradient analysis of urban construction land expansion in the Chongqing urban area of China. Journal of Urban Planning and Development, 141(1), 05014009.

Chen

, Gong

, Lu

, & Ye

(2019). Build a people-oriented urbanization: China’s new-type urbanization dream and Anhui model. Land Use Policy, 80, 1–9.

Dadashpoor

, Azizi

, & Moghadasi

(2019). Land use change, urbanization, and change in landscape pattern in a metropolitan area. Science of the Total Environment, 655, 707–719.

ESRI (2010). Hot spot analysis. Environmental Systems Research Institute. Retrieved June 24, 2020, from http://downloads2.esri.com/gallery/videos/HotSpotPart2.wmv

10.

Getis

(2008). A history of the concept of spatial autocorrelation: A geographer’s perspective. Geographical Analysis, 40(3), 297–309.

11.

Gong

, Hu

, Chen

, Liu

, & Wang

(2018). Urban expansion dynamics and modes in metropolitan Guangzhou, China. Land Use Policy, 72, 100–109.

12.

Hao

(2008). Opening up, market reform, and convergence clubs in China. Asian Economic Journal, 22(2), 133–160.

13.

Kamusoko

(2017). Importance of remote sensing and land change modeling for urbanization studies. In Urban development in Asia and Africa (pp. 3–10). Springer.

14.

, Liu

, Zhang

, & Gao

(2011). Evaluation on the human settlements environment suitability in the Three Gorges Reservoir Area of Chongqing based on RS and GIS. Journal of Geographical Sciences, 21(2), 346–358.

15.

Lin

, Li

, & Ma

(2018). Exploring the interactive development between population urbanization and land urbanization: Evidence from Chongqing, China (1998–2016). Sustainability, 10(6), 1741.

16.

Lin

, Wang

, & Wang

(2015). Spatial differences and driving forces of land urbanization in China. Journal of Geographical Sciences, 25(5), 545–558.

17.

Liu

, Qi

, Cao

, & Liu

(2015). Spatial patterns, driving forces, and urbanization effects of China’s internal migration: County-level analysis based on the 2000 and 2010 censuses. Journal of Geographical Sciences, 25(2), 236–256.

18.

Lopez

J. M. R.

, Heider

, & Scheffran

(2017). Frontiers of urbanization: Identifying and explaining urbanization hot spots in the south of Mexico City using human and remote sensing. Applied Geography, 79, 1–10.

19.

Melchiorri

, Florczyk

A. J.

, Freire

, Schiavina

, Pesaresi

, & Kemper

(2018). Unveiling 25 years of planetary urbanization with remote sensing: Perspectives from the global human settlement layer. Remote Sensing, 10(5), 768.

20.

Mohan

, & Kandya

(2015). Impact of urbanization and land-use/land-cover change on diurnal temperature range: A case study of tropical urban airshed of India using remote sensing data. Science of the Total Environment, 506, 453–465.

21.

Negreiros

J. G.

, Painho

M. T.

, Aguilar

F. J.

, & Aguilar

M. A.

(2010). A comprehensive framework for exploratory spatial data analysis: Moran location and variance scatterplots. International Journal of Digital Earth, 3(2), 157–186.

22.

Ren

, Li

, Shen

, Zhang

, Chen

, & Wang

(2018). What is the efficiency of fast urbanization? A China study. Sustainability, 10(9), 3180.

23.

Rodríguez-Pose

(2008). The rise of the ‘city-region’ concept and its development policy implications. European Planning Studies, 16(8), 1025–1046.

24.

Rozenstein

, Karnieli

(2011). Comparison of methods for land-use classification incorporating remote sensing and GIS inputs. Applied Geography, 31(2), 533–544.

25.

Shao

, Wei

, & Xie

(2005). Mountain land use planning of metropolitan suburbs: the case of the Jinyun mountain and its surrounding area, Chongqing, China. Journal of Mountain Science, 2(2), 116–128.

26.

Taubenböck

, Standfuß

, Wurm

, Krehl

& Siedentop

(2017). Measuring morphological polycentricity: A comparative analysis of urban mass concentrations using remote sensing data. Computers, Environment and Urban Systems, 64, 42–56.

27.

Uttara

, Bhuvandas

, & Aggarwal

(2012). Impacts of urbanization on environment. International Journal of Research in Engineering and Applied Sciences, 2(2), 1637–1645.

28.

Wang

X. R.

, Hui

E. C. M.

, Choguill

, & Jia

S.-H.

(2015). The new urbanization policy in China: Which way forward? Habitat International, 47, 279–284.

29.

Wang

X. R.

, Hui

E. C. M

, & Sun

J. X.

(2017). Population migration, urbanization and housing prices: Evidence from the cities in China. Habitat International, 66, 49–56.

30.

Wang

, Lin

, Glendinning

, & Xu

(2018). Land-use changes and land policies evolution in China’s urbanization processes. Land Use Policy, 75, 375–387.

31.

, & Min

(2013). Quantifying spatiotemporal patterns of urban expansion in China using remote sensing data. Cities, 35, 104–113.

32.

, Zheng

, & Zheng

(2019). Do urban planning policies meet sustainable urbanization goals? A scenario-based study in Beijing, China. Science of the Total Environment, 670, 498–507.

33.

, & Wu

A. M.

(2014). Urbanization, land development, and land financing: Evidence from Chinese cities. Journal of Urban Affairs, 36(supl. 1), 354–368.

34.

You

(2016). Quantifying the coordinated degree of urbanization in Shanghai, China. Quality & Quantity, 50(3), 1273–1283.

35.

Zeng

, Liu

, Stein

, & Jiao

(2015). Characterization and spatial modeling of urban sprawl in the Wuhan Metropolitan Area, China. International Journal of Applied Earth Observation and Geoinformation, 34, 10–24.

36.

Zhang

(2008). Conceptualizing China’s urbanization under reforms. Habitat International, 32(4), 452–470.

37.

Zhang

, & Wang

M. Y.

(2018). Spatial-temporal characteristics and determinants of land urbanization quality in China: Evidence from 285 prefecture-level cities. Sustainable Cities and Society, 38, 70–79.

38.

Zhu

(2017). Change detection using Landsat time series: A review of frequencies, preprocessing, algorithms, and applications. ISPRS Journal of Photogrammetry and Remote Sensing, 130, 370–384.