Next Article in Journal
Time Series Analysis of Landslide Dynamics Using an Unmanned Aerial Vehicle (UAV)
Next Article in Special Issue
DMSP-OLS Radiance Calibrated Nighttime Lights Time Series with Intercalibration
Previous Article in Journal
Mapping Spatial Distribution of Larch Plantations from Multi-Seasonal Landsat-8 OLI Imagery and Multi-Scale Textures Using Random Forests
Previous Article in Special Issue
Integrating Multiple Source Data to Enhance Variation and Weaken the Blooming Effect of DMSP-OLS Light
Article Menu

Export Article

Remote Sens. 2015, 7(2), 1721-1735; https://doi.org/10.3390/rs70201721

Letter
Dynamics of Urbanization Levels in China from 1992 to 2012: Perspective from DMSP/OLS Nighttime Light Data
1,2,†, 1,†, 1,* and 1,2,†
1
Center for Human-Environment System Sustainability, State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, 19 Xinjiekouwai Street, Beijing 100875, China
2
College of Resources Science & Technology, Beijing Normal University, 19 Xinjiekouwai Street, Beijing 100875, China
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Academic Editors: Christopher Elvidge and Prasad S. Thenkabail
Received: 10 November 2014 / Accepted: 27 January 2015 / Published: 5 February 2015

Abstract

:
The authenticity and reliability of urbanization levels measured by different indicators in China have not reached a consensus, which may impede our understanding of the process of urbanization and its impacts on the environment. The objective of this study was to describe a reliable method of estimating urbanization level based on the Operational Line-scan System (OLS) on the Defense Meteorological Satellite Program (DMSP) nighttime light data and to analyze the dynamics of urbanization levels in China from 1992 to 2012. We calculated the comprehensive urbanization level at the national, provincial, and county scales using a compounded night light index (CNLI) and compared the change rate of CNLI with those of the other two conventional urbanization level indicators, proportion of the nonagricultural population and proportion of built-up area. Our results showed that CNLI derived from the DMSP/OLS data set provided a relatively reliable and accurate measure of the comprehensive urbanization level in China. During the last two decades, China has experienced continued and rapid urbanization with large regional variations. The CNLI increased 3.12 times, from 1.72 × 10−3 to 7.09 × 10−3. The annual increases of CNLI in eastern provinces were much faster than those in western provinces. In addition, we found that the rates of change in these three indicators were consistent for most provinces with the exception of the four municipalities (Beijing, Tianjin, Shanghai, and Chongqing) and a few eastern coastal provinces (Jiangsu, Zhejiang, Fujian, and Guangdong). Because the imbalance among population growth, urban expansion and socioeconomic development may affect cities’ sustainable development, we should pay more attention to these regions with large disparities between different indicators.
Keywords:
urbanization level; DMSP/OLS; compounded night light index; multi-scale; China

1. Introduction

Urbanization level is an indicator that could be used for characterizing urbanization in terms of demographic dynamics, socioeconomic development and spatial expansion simultaneously. It is an important basis for evaluating the development stage [1], assessing the effects of urbanization on society and the environment [2,3], and formulating a regional development strategy [4]. Along with the rapid urbanization and dramatic economic growth in China from late 1978, some environmental problems, such as air pollution [5], water pollution [6], increase in greenhouse gas emissions [7], and enhanced urban heat islands [8], are increasing in many Chinese cities. Monitoring the dynamics of urbanization levels in China accurately and quickly plays a fundamental role in understanding the process of urbanization and evaluating its environmental influence [9].
However, the authenticity and reliability of China’s urbanization level has not reached a consensus [10,11,12]. For example, according to the National Bureau of Statistics, China’s urbanization level was 36.22% in 2000, but Shen [10] claimed that it was actually 37.04% in the same year. The difference in urban population between these two studies was 3.74 × 106. The difference was due to the underreporting of urban populations and changes in the definitions of urban populations in China. The main reason for the controversy is because researchers often simply use a single indicator (i.e., a population index or a land index) to measure the urbanization level, but both of them have deficiencies in terms of comparability and comprehensiveness [13]. For instance, the population index lacks continuity and temporal comparability because the criteria used to define “urban” and “urban population” have changed several times for different censuses in China [14,15]. Specifically, in the fifth national census in 2000, the urban population accounted for 36.09% of the total population, but the percentage would change to 31.39% if we were to use the criteria applied in 1990 [4]. Also, the land index derived from different data sources is incomparable. For example, according to the National Bureau of Statistics of China [16], the built-up area in China increased from 20,465 km2 in 1995 to 32,520 km2 in 2005. However, according to the land use/cover data sets produced by the Chinese Academy of Sciences, the built-up area increased from 31,756 km2 to 43,852 km2 during this period [17]. In addition, revealing the whole picture of the regional urbanization level is difficult if we just use a single index because urbanization is a complicated process involving economic, demographic, and societal changes [9]. Therefore, how to monitor and understand the dynamics of urbanization levels quickly and accurately remains a challenging problem in China.
An Operational Line-scan System (OLS) on the Defense Meteorological Satellite Program (DMSP) provided a valuable data source for elucidating the dynamics of China’s urbanization [18,19,20,21]. This sensor can detect city lights, gas flares, and fires at night with a low-light detecting capability [22]. The DMSP/OLS dates back to the 1970s but was not widely noticed until 1992 when the images became available in digital format. In January 2010, the National Oceanic and Atmospheric Administration (NOAA)/National Geophysical Data Center (NGDC) published the Version 4 DMSP-OLS Nighttime Lights Time Series (V4DNLTS) data set. The files in this data set came from sensors with identical onboard design and continuous space platform, providing a unique and valuable resource for monitoring the long-term dynamics of urbanization [9]. More importantly, the DMSP/OLS nighttime light images have been utilized in several studies for quantitatively estimating and mapping socioeconomic activities related to urbanization processes from regional to global scales [23]. For example, Sutton et al. [24] estimated the global human population using the statistical relationship between nighttime lighted area and urban population, while Doll et al. [25] created regional disaggregated maps of the gross domestic product (GDP) based upon the relationships between nighttime radiance data and the GDP. Also, Zhang et al. [26] adopted multi-temporal DMSP/OLS nighttime lights data to estimate regional and global urban growth based on a linear correlation between night light brightness and the urban population. However, using DMSP/OLS data, there were few studies focused on monitoring the dynamics of urbanization level for the last two decades at multiple scales in China, especially the dynamics after 2000, and at the county scale.
The objective of this study was to describe a reliable method of estimating urbanization level based on DMSP/OLS nighttime light data and to analyze the dynamics of urbanization levels in China from 1992 to 2012. First, we calculated the compound night light index (CNLI) [18], which reflects the regional comprehensive urbanization level. Second, we analyzed the dynamics of urbanization levels at national, provincial, and county scales based on the CNLI. Last, we compared the rate of change in CNLI with that in two other conventional urbanization level indicators, the proportion of the nonagricultural population (PNP) and the proportion of built-up area (PBA), to explore the coordination among these three indicators.

2. Data

Three types of data were used in this research. The first type is nighttime stable light (NSL) data from 1992 to 2012 in the V4DNLTS data set. The data were obtained from the NGDC Web site and included data from six DMSP satellites: F10, F12, F14, F15, F16, and F18 [27]. Annual composites were produced for each satellite using the highest-quality data collected. Several constraints were considered to identify the best-quality nighttime light data to produce the composite [22]. For example, ephemeral events such as wildfires were discarded. In the annual composites, the Digital Number (DN) value of each pixel was the average of the visible-band DN values of lights from urban areas and other sites with persistent lighting. Background noises in the composite were identified and replaced with values of zero, and the final DN values for lit pixels ranged from 1 to 63. Data were in 30 arcseconds (approximately 1 km at the equator and 0.8 km at 40°N), spanning −180° to 180° in longitude and −65° to 75° in latitude. After obtaining the global NSL data for 1992–2012, NSL data were extracted according to the Chinese administrative boundary. Last, data were projected using the Lambert Azimuthal Equal Area projection and resampled to a pixel size of 1 km.
Due to the differences between sensors, differences in crossing time between satellites, and the degradation of sensors, these NSL data collected by sensors onboard different satellites could significantly differ (i.e., discrepancies could exist in DN values between two satellites for the same year, as well as abnormal fluctuations in DN values from the same satellite for different years), even if no real changes occurred on the ground [28]. Therefore, these data cannot be used directly to extract the dynamics of urbanization levels in China [9]. To reduce yearly variations among sensors, we used a systematic correction method proposed by Liu et al. [28] to correct multiyear, multi-satellite NSL data. This cross-calibration method achieved the goal of improving the continuity and comparability of NSL data based on three main steps: intercalibration, intra-annual composition and inter-annual series correction. The process of intercalibration was used to guarantee that DN values among different years and different satellites were comparable. The objective of intra-annual composition was to remove any intra-annual unstable lit pixels. The last step, inter-annual series correction, could reduce the inter-annual discrepancies of DN values to the minimum level. Then, to calculate the CNLI, we adopted the threshold technique developed by Liu et al. [28] to determine the optimal thresholds (Table 1) for extracting lighted urban areas in different regions of China (Equation (1)). For each economic region, optimal thresholds were determined for each period when the urban areas extracted using NSL data matched the ancillary data best (Table 1).
Table 1. Optimal thresholds for lighted urban area extraction in eight economic regions of China. Note: We divided China into eight economic regions: Northeast China (NEC), Northern Coastal China (NCC), Southern Coastal China (SCC), Eastern Coastal China (ECC), the Middle Reaches of the Yellow River (MRYLR), the Middle Reaches of the Yangtze River (MRYTR), Southwest China (SWC), and Northwest China (NWC). We determined the thresholds based on the method provided by Liu et al. [28]
Table 1. Optimal thresholds for lighted urban area extraction in eight economic regions of China. Note: We divided China into eight economic regions: Northeast China (NEC), Northern Coastal China (NCC), Southern Coastal China (SCC), Eastern Coastal China (ECC), the Middle Reaches of the Yellow River (MRYLR), the Middle Reaches of the Yangtze River (MRYTR), Southwest China (SWC), and Northwest China (NWC). We determined the thresholds based on the method provided by Liu et al. [28]
Economic RegionProvinces in the Economic RegionOptimal Threshold
19921993–19971998–20022003–20072008–2012
NECLiaoning, Jilin, Heilongjiang5051545859
NCCBeijing, Tianjin, Hebei, Shandong5156576159
SCCFujian, Guangdong, Hainan6162636161
ECCShanghai, Jiangsu, Zhejiang5054585959
MRYLRShaanxi, Shanxi, Henan, Inner Mongolia4956545959
MRYTRHubei, Hunan, Jiangxi, Anhui4142425251
SWCYunnan, Guizhou, Sichuan, Chongqing, Guangxi4044465857
NWCGansu, Qinghai, Ningxia, Tibet, Xinjiang3950515960
The second type of data is socioeconomic census data, including the nonagricultural population, production of secondary industry and tertiary industry (PST), and built-up area of each province and city. These data were obtained from the China Population and Employment Statistics Yearbook, China Statistical Yearbook, and China City Statistical Yearbook [29,30]. The third type is the GIS auxiliary data. The administrative boundaries for the provinces, cities, and counties in China at a scale of 1:4,000,000 were obtained from the National Geomatics Center of China [31]. The boundaries for the 23 urban agglomerations in China came from Fang et al. [32]. Our study focused only on Mainland China due to the lack of statistical data for Taiwan, Hong Kong, and Macao.

3. Methods

Our method for analyzing the dynamics of urbanization levels in China included two steps: calculating the CNLI index at the national, provincial, and county scales, and analyzing the dynamics of urbanization levels based on the annual increase in the CNLI.

3.1. Calculating the Compounded Night Light Index (CNLI)

We used the CNLI proposed by Zhuo et al. [18] to reflect the regional comprehensive urbanization level. The CNLI takes two parameters, night light brightness and lit urban areas, into account simultaneously. The former is closely correlated with urban population and economic scale [25,33], and the latter is closely correlated with urban area [34,35]. Therefore, changes in the CNLI can reflect the dynamics of urban population size, economic scale, and urban expansion simultaneously. We computed the CNLI at the national, provincial, and county scales using the following formula [18]:
C N L I = I × S
where I is the average night light brightness of all lit pixels in a region. It can be described as follows:
I = 1 N L × D N M × i = P D N M ( D N i × n i )
where DNi is the DN value of the ith gray level, ni is the number of lit pixels belonging to the ith gray level, P is the optimal threshold to extract the lighted urban area from the DMSP/OLS image (Table 1), DNM is the maximum DN value, and NL is the number of lit pixels with a DN value between P and DNM.
S is the proportion of lit urban areas to the total area of a region. It can be described as follows:
S = A r e a N A r e a
where AreaN is the area of lit urban areas in a region and Area is the total area of the region.

3.2. Analyzing the Dynamics of Urbanization Levels

Because the spatiotemporal dynamics of urbanization levels appears to be significantly different at different scales [36], we examined it at the national, provincial, and county scales. We used the annual increase in the CNLI to reflect the dynamics of urbanization levels [17]:
U = C N L I t + n C N L I t n
where U is the annual increase in the CNLI, and CNLIt+n and CNLIt are the CNLI in the t+nth year and tth year, respectively.

3.3. Accuracy Assessment

We used the method proposed by Zhuo et al. [18] to evaluate the credibility of using the CNLI to monitor dynamics of the urbanization level in China. First, we calculated Composite Urbanization Level Indices (CUI) by adding three types of census data, percentage of nonagricultural population, percentage of PST, and percentage of built-up area, with equal weights for 31 provinces and 157 cities in 1992, 1995, 2000, 2005, and 2008. Then, we established a regression equation between the CUI and CNLI. Last, we simulated the CUI in 2011 based on the regression equation, and we calculated the differences between the simulated CUI and the actual CUI from census data in 2011. The result showed that the CNLI was significantly correlated with the CUI at the national, provincial, and county scales in 1992, 1995, 2000, 2005, and 2008. The correlation coefficients were 0.96, 0.71, and 0.69, respectively, and the differences between simulated CUI and actual CUI in 2011 were 0.51%, 10.53%, and 12.26%, respectively, at the three scales (Table 2). Therefore, we believe that the dynamics of urbanization levels over the past two decades derived from DMSP/OLS NSL data were consistent with the actual conditions in China.
Table 2. Validation results between the compounded night light index (CNLI) and the composite urbanization level index (CUI).
Table 2. Validation results between the compounded night light index (CNLI) and the composite urbanization level index (CUI).
ScaleRegression EquationCorrelation CoefficientDifference between Simulated CUI and Actual CUI in 2011 (%)
Nationaly = 1705.43x + 30.910.96 (p < 0.01)0.51
Provincialy = 90.52x + 36.230.71 (p < 0.01)10.53
Countyy = 72.81x + 43.220.69 (p < 0.01)12.26

4. Results

4.1. Dynamics of Urbanization Levels at the National Scale

The urbanization levels in China increased rapidly from 1992 to 2012 (Figure 1). The CNLI increased 3.12 times, from 1.72 × 10−3 to 7.09 × 10−3, with an average annual growth rate of 7.34%. Furthermore, the rate of change in CNLI varied over time. Between 1992 and 1995, it exceeded the average annual growth rate slightly, with an annual growth rate of 10.59%. Then it slowed down during 1996–2000, with an annual growth rate equal to 2.90%. After 2000, it increased quickly, with an annual growth rate of 6.73%.
Figure 1. Dynamics of urbanization levels at the national scale from 1992 to 2012.
Figure 1. Dynamics of urbanization levels at the national scale from 1992 to 2012.
Remotesensing 07 01721 g001

4.2. Dynamics of Urbanization Levels at the Provincial Scale

The rates of change in CNLI showed significant regional differences. The rates in eastern provinces were much faster than those in western provinces (Figure 2). The annual increase in the CNLI in Beijing, Shanghai, and Tianjin were the fastest, all above 5.0 × 10−3, with the CNLI in 2012 was 2.92 times, 2.31 times and 1.82 times that in 1992. The annual increase in the CNLI in other eastern provinces, such as Zhejiang, Jiangsu, Shandong, and Guangdong, was between 1.0 × 10−3 and 5.0 × 10−3, with the CNLI in 2012 was approximately 5.48–10.46 times that in 1992. In contrast, the annual increase in eight provinces in western China (e.g., Xinjiang and Tibet) was much less than in the aforementioned provinces, with an annual increase in the CNLI of less than 0.26 × 10−3. It was slowest in Tibet, which was less than one-twentieth of the national average.
Figure 2. Dynamics of urbanization levels at the provincial scale in China from 1992 to 2012 (Note: We used the “natural breaks (Jenks)” method [37] to set the breaks for the annual increases in compounded night light index (CNLI)).
Figure 2. Dynamics of urbanization levels at the provincial scale in China from 1992 to 2012 (Note: We used the “natural breaks (Jenks)” method [37] to set the breaks for the annual increases in compounded night light index (CNLI)).
Remotesensing 07 01721 g002

4.3. Dynamics of Urbanization Levels at the County Scale

Counties in which the annual increases of CNLI were higher than the national average were mainly located inside urban agglomerations (Figure 3a). The annual increases of CNLI were faster than the national average in 711 counties, and nearly 59.2% of these counties were located within urban agglomerations. Among the 23 urban agglomerations, the Yangtze River delta, Jing-Jin-Ji, and Shandong Peninsula ranked as the top three in the numbers of counties that had higher rate than the national average. Specifically, in the urban agglomeration of the Yangtze River Delta, the annual increases in the CNLI were larger than the national average in 67 counties. Among these counties, Jiading district in Shanghai experienced the highest increase in CNLI, with an annual increase in the CNLI of more than 0.035 (Figure 3d). In the Jing-Jin-Ji urban agglomeration, the number of counties exceeding the national average was 64. In this urban agglomeration, Jinnan district in Tianjin experienced the highest increase in CNLI, nearly 108 times greater than the national average (Figure 3b). For the urban agglomeration of Shandong Peninsula, 37 counties exceeded the national average. Among these counties, Qingdao experienced the highest increase in CNLI, with an annual increase in the CNLI of about 0.016, nearly 58 times the national average (Figure 3c).
Figure 3. Dynamics of urbanization levels at the county scale in China from 1992 to 2012 (a) for the 23 urban agglomerations, (b) in the Jing-Jin-Ji urban agglomeration, (c) in the Shandong Peninsula urban agglomeration, (d) in the Yangtze River delta urban agglomeration. (Note: we used the “natural breaks (Jenks)” method [37] to set the breaks in the annual increases in the compounded night light index (CNLI)).
Figure 3. Dynamics of urbanization levels at the county scale in China from 1992 to 2012 (a) for the 23 urban agglomerations, (b) in the Jing-Jin-Ji urban agglomeration, (c) in the Shandong Peninsula urban agglomeration, (d) in the Yangtze River delta urban agglomeration. (Note: we used the “natural breaks (Jenks)” method [37] to set the breaks in the annual increases in the compounded night light index (CNLI)).
Remotesensing 07 01721 g003

5. Discussion

5.1. We Provided Updated and Accurate Information about the Dynamics of China’s Urbanization Level Based on DMSP/OLS Nighttime Data

First of all, our study provided timely information on the urbanization level based on DMSP/OLS NSL data for China, especially the dynamics after 2000. To our knowledge, almost all studies examining the urbanization level in China based on DMSP/OLS NSL data only covered the time period before 2000. Chen et al. [38] analyzed the dynamics of urbanization levels at the provincial scale in China from 1992 to 1998 using DMSP/OLS nighttime light data for the first time. In the same year, Zhuo et al. [18] constructed the new CNLI index to analyze the dynamics of urbanization levels from 1992 to 1998. Although these studies laid a good foundation for the research on urbanization levels based on DMSP/OLS NSL data, the information on the dynamics of urbanization levels needed to be updated urgently. We updated the information at the national, provincial, and county scales to the year 2012, and the result has provided a new data source for understanding the actual urbanization process in China.
Second, comparing the monitoring accuracy with previous studies (i.e., Chen et al. [38] and Zhuo et al. [18]), we found that our result was more accurate. Specifically, the difference between the simulated CUI and the actual CUI at the provincial scale in our study is 10.53%, lower than the 14% in Chen et al. [38]. In addition, at the county scale, the accuracy of our study is 12.26%, also lower than the 16.21% in Zhuo et al. [18].

5.2. Dynamics of Urbanization Levels Measured by the CNLI Was Consistent with the Result Measured Based on the Population and Land Indicators

After standardizing the data, we compared the annual increase in the CNLI with that of the other two urbanization level indicators, the proportion of nonagricultural population (PNP) and the proportion of built-up area (PBA) from 1992 to 2012. To standardize the data, we calculated the ratio of the annual increase in the CNLI in a certain region relative to that of the whole country [39] (Equation (5)). Likewise, we standardized the annual increase in PNP and PBA in the same way. Then, we compared the disparities between the normalized annual increase in the CNLI and that of PNP ( Δ V P o p ) and PBA ( Δ V L a n d ). Due to the accessibility of data, the comparison analysis was only implemented at the provincial scale,
V m n = U m n U n ( m = 1 , 2 31 )
where V m n is the normalized annual increase of indicator n (i.e., CNLI or PNP or PBA) in the mth province, U m n is the annual increase of indicator n in the mth province, and U n is the annual increase of indicator n in the whole country.
The results showed no significant disparities between the normalized annual increase in the CNLI and that in the PNP in most provinces (Figure 4). The average of Δ V P o p in 31 provinces was 4.33. Twenty-five provinces had Δ V P o p values below the national average difference. However, in some eastern coastal provinces (i.e., Beijing, Tianjin, Shanghai, Shandong, Jiangsu, and Zhejiang), the Δ V P o p values were larger than 4.33 and ranged from 5.07 to 62.87, which indicated that the rates of change in CNLI were much faster than those derived from the PNP in these provinces.
Figure 4. The differences between the standardized annual increase of the compounded night light index (CNLI) and that of the proportion of nonagricultural population (PNP) at the provincial scale (Note: the two dotted lines represent the average differences).
Figure 4. The differences between the standardized annual increase of the compounded night light index (CNLI) and that of the proportion of nonagricultural population (PNP) at the provincial scale (Note: the two dotted lines represent the average differences).
Remotesensing 07 01721 g004
We did not find significant differences between the normalized annual increase in the CNLI and that of the proportion of built-up area in most provinces (Figure 5). The average Δ V L a n d in 31 provinces was 1.05, and only nine provinces had a Δ V L a n d value larger than 1.05. Notably, Δ V L a n d in three municipalities (i.e., Beijing, Tianjin, and Shanghai), Jiangsu, Zhejiang, and Anhui ranged from 1.37 to 21.51, equal to 1.37–21.51 times more than the national average, indicating that the rates of change in CNLI were much faster than those derived from the land indicator in these provinces. In contrast, Δ V L a n d in Guangdong, Chongqing, and Fujian ranged from −2.27 to −5.42, which indicated that the rates of change in CNLI were slower than those in the PBA in these provinces.
In brief, the rates of change in CNLI were basically consistent with those derived from the population indicator (i.e., PNP) and land indicator (i.e., PBA) in most provinces of China from 1992 to 2012. These three indicators showed significant disparities only in four municipalities and a few eastern coastal provinces.
Figure 5. The differences between the standardized annual increase of the compounded night light index (CNLI) and that of the proportion of built-up area (PBA) at the provincial scale (Note: the two dotted lines represent the average differences).
Figure 5. The differences between the standardized annual increase of the compounded night light index (CNLI) and that of the proportion of built-up area (PBA) at the provincial scale (Note: the two dotted lines represent the average differences).
Remotesensing 07 01721 g005

5.3. Limitations and Future Work

Our proposed approach has a few limitations. First, the accuracy of the CNLI needs to be improved. Due to the limited radiometric range of DMSP/OLS, NSL data values in urban centers tend to be saturated and truncated [40]. This saturation problem may influence the accuracy of the CNLI by affecting the calculation of I (Equation (1)) for some regions with well-developed urban cores [26,41]. In addition, the ability of DMSP/OLS NSL data to characterize urbanization levels is limited because of the relatively coarse spatial resolution, especially in some regions with weak nightlight [28]. Second, we identified some hot spots where the rates of change in urbanization were remarkably different by comparing three indicators (i.e., CNLI, the proportion of nonagricultural population, and the proportion of built-up area), but we still need to explore whether it was the different planning strategies between counties or other socioeconomic factors that caused the disparities [42,43,44]. Third, we set the optimal thresholds to extract lit urban information at the economic region scale, which could not consider the regional variation in physical environment and social-economic status within each economic region. Therefore, these thresholds may affect the accuracy at the provincial and county scale. This was also the reason that the accuracy assessment at the provincial scale and county scale were lower than that at the national scale (Table 2).
In future studies, we can try to alleviate the saturation effects of NSL data in urban cores by combining the NSL data with the Normalized Difference Vegetation Index to improve the accuracy of light brightness (Equation (1)) [41]. In addition, the availability of the global radiance calibrated nighttime lights with no sensor saturation is expected to provide a valuable alternative data source for improving the light-detecting capability [45].

6. Conclusions

The study found that the urbanization levels based on DMPS/OLS NSL data are relatively accurate and comprehensive. These data provided a new way to understand the real urbanization process in China. A significant correlation was observed between the CNLI and the CUI derived from socioeconomic census data.
The results showed that China has experienced continued and rapid urbanization during the last two decades. The dynamics of urbanization levels showed significant regional differences, with the annual increase of CNLI in eastern provinces much higher than those in western provinces. The annual increase in the CNLI in the eastern administrative unit, such as Shanghai, Beijing, and Tianjin, was 20 times larger than that of the 10 western provinces. In addition, rapidly urbanized areas were concentrated in the vigor development regions, i.e., urban agglomerations. Nearly 60% of the counties with an annual increase in the CNLI larger than the national average were located inside the 23 urban agglomerations.
In the past two decades, the dynamics of urbanization levels derived from the CNLI were basically consistent with those taken from the other two conventional used indicators (i.e., the proportion of nonagricultural population and the proportion of built-up area) in most provinces. Significant disparities only existed in the four municipalities and a few eastern coastal provinces of China. Specifically, the annual increases in CNLI were much faster than those derived from the other two indicators in Beijing, Tianjin, Shanghai, Jiangsu, and Zhejiang, whereas the annual increases in CNLI were much slower compared with those derived from the land indicator in Fujian, Chongqing, and Guangdong. Researchers and policy makers should pay more attention to these regions with large disparities between different indicators.

Acknowledgments

We would like to express our respects and gratitude to the anonymous reviewers and editors for their valuable comments and suggestions on improving the quality of the paper. This work is supported by the National Natural Science Foundation of China (Grant No. 41222003 and 41321001). It is also supported by the project of “the new century excellent youth teacher” under Grant NO.NCET-13-0053 by Ministry of Education and the 111 project “Hazard and Risk Science Base at Beijing Normal University” under Grant No. B08008 by Ministry of Education and State Administration of Foreign Experts Affairs, PRC.

Author Contributions

Bin Gao collected and processed the data, performed analysis and wrote the paper. Qingxu Huang and Chunyang He conceived and designed the study and the methods. Qun Ma contributed to analysis and interpretation of the data. All authors reviewed and edited the draft, approved the submitted manuscript, and agreed to be listed and accepted the version for publication.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Kilinc, G.; Gulersoy, N.Z. Evaluation of the potential for districts/counties to become provinces with respect to the level of urbanization in Turkey. Eur. Plan. Stud. 2011, 19, 1557–1580. [Google Scholar] [CrossRef]
  2. Bagat, M.; Drakulic, V.; Kauzlaric, K.S.; Vlahusic, A.; Bilic, I.; Matanic, D. Influence of urbanization level and gross domestic product of counties in Croatia on access to health care. Croat. Med. J. 2008, 49, 384–391. [Google Scholar] [CrossRef] [PubMed]
  3. Reis, E.; Lopez-Iborra, G.M.; Pinheiro, R.T. Changes in bird species richness through different levels of urbanization: Implications for biodiversity conservation and garden design in Central Brazil. Landsc. Urban Plan. 2012, 107, 31–42. [Google Scholar] [CrossRef]
  4. Zhou, Y.X.; Ma, L.J.C. China’s urbanization levels: Reconstructing a baseline from the fifth population census. China Q. 2003, 176–196. [Google Scholar]
  5. Chan, C.K.; Yao, X. Air pollution in mega cities in China. Atmos. Environ. 2008, 42, 1–42. [Google Scholar] [CrossRef]
  6. Shao, M.; Tang, X.; Zhang, Y.; Li, W. City clusters in China: Air and surface water pollution. Front. Ecol. Environ. 2006, 4, 353–361. [Google Scholar] [CrossRef]
  7. Zhu, H.M.; You, W.H.; Zeng, Z.F. Urbanization and CO2 emissions: A semi-parametric panel data analysis. Econ. Lett. 2012, 117, 848–850. [Google Scholar] [CrossRef]
  8. Zhou, L.M.; Dickinson, R.E.; Tian, Y.H.; Fang, J.Y.; Li, Q.X.; Kaufmann, R.K.; Tucker, C.J.; Myneni, R.B. Evidence for a significant urbanization effect on climate in China. Proc. Natl. Acad. Sci. USA 2004, 101, 9540–9544. [Google Scholar] [CrossRef] [PubMed]
  9. Ma, T.; Zhou, C.H.; Pei, T.; Haynie, S.; Fan, J.F. Quantitative estimation of urbanization dynamics using time series of DMSP/OLS nighttime light data: A comparative case study from China’s cities. Remote Sens. Environ. 2012, 124, 99–107. [Google Scholar] [CrossRef]
  10. Shen, J.F. Estimating urbanization levels in Chinese provinces in 1982–2000. Int. Stat. Rev. 2006, 74, 89–107. [Google Scholar] [CrossRef]
  11. Fan, J.; Ma, T.; Zhou, C.; Zhou, Y.; Xu, T. Comparative estimation of urban development in China’s cities using socioeconomic and DMSP/OLS night light data. Remote Sens. 2014, 6, 7840–7856. [Google Scholar] [CrossRef]
  12. Xu, T.; Ma, T.; Zhou, C.; Zhou, Y. Characterizing spatio-temporal dynamics of urbanization in China using time series of DMSP/OLS night light data. Remote Sens. 2014, 6, 7708–7731. [Google Scholar] [CrossRef]
  13. Xu, X.Q.; Zhou, Y.X.; Ning, Y.M. Urban Geography; Higher Education Press: Beijing, China, 2009. [Google Scholar]
  14. Shen, J.F. Rural development and rural to urban migration in China 1978–1990. Geoforum 1995, 26, 395–409. [Google Scholar] [CrossRef] [PubMed]
  15. Ma, L.J.C.; Cui, G. Administrative changes and urban population in China. Ann. Assoc. Am. Geogr. 1987, 77, 373–395. [Google Scholar] [CrossRef]
  16. National Bureau of Statistics of China. China Statistical Yearbook; China Statistics Press: Beijing, China, 1993–2012.
  17. Liu, J.; Kuang, W.; Zhang, Z.; Xu, X.; Qin, Y.; Ning, J.; Zhou, W.; Zhang, S.; Li, R.; Yan, C.; et al. Spatiotemporal characteristics, patterns, and causes of land-use changes in china since the late 1980s. J. Geogr. Sci. 2014, 24, 195–210. [Google Scholar]
  18. Zhuo, L.; Shi, P.J.; Chen, J.; Ichinose, T. Application of compounded night light index derived from DMSP/OLS data to urbanization analysis in China in the 1990s. Acta Geogr. Sin. 2003, 58, 893–902. [Google Scholar]
  19. Shao, Z.; Liu, C. The integrated use of DMSP-OLS nighttime light and MODIS data for monitoring large-scale impervious surface dynamics: A case study in the Yangtze River Delta. Remote Sens. 2014, 6, 9359–9378. [Google Scholar] [CrossRef]
  20. Yue, W.; Gao, J.; Yang, X. Estimation of gross domestic product using multi-sensor remote sensing data: A case study in Zhejiang province, East China. Remote Sens. 2014, 6, 7260–7275. [Google Scholar] [CrossRef]
  21. Li, X.; Ge, L.; Chen, X. Detecting Zimbabwe’s decadal economic decline using nighttime light imagery. Remote Sens. 2014, 5, 4551–4570. [Google Scholar] [CrossRef]
  22. Elvidge, C.D.; Ziskin, D.; Baugh, K.E.; Tuttle, B.T.; Ghosh, T.; Pack, D.W.; Erwin, E.H.; Zhizhin, M. A fifteen year record of global natural gas flaring derived from satellite data. Energies 2009, 2, 595–622. [Google Scholar] [CrossRef]
  23. Huang, Q.; Yang, X.; Gao, B.; Yang, Y.; Zhao, Y. Application of DMSP/OLS nighttime light images: A meta-analysis and a systematic literature review. Remote Sens. 2014, 6, 6844–6866. [Google Scholar] [CrossRef]
  24. Sutton, P.; Roberts, D.; Elvidge, C.D.; Baugh, K. Census from heaven: An estimate of the global human population using night-time satellite imagery. Int. J. Remote Sens. 2001, 22, 3061–3076. [Google Scholar] [CrossRef]
  25. Doll, C.N.H.; Muller, J.P.; Morley, J.G. Mapping regional economic activity from night-time light satellite imagery. Ecol. Econ. 2006, 57, 75–92. [Google Scholar] [CrossRef]
  26. Zhang, Q.L.; Seto, K.C. Mapping urbanization dynamics at regional and global scales using multi-temporal DMSP/OLS nighttime light data. Remote Sens. Environ. 2011, 115, 2320–2329. [Google Scholar] [CrossRef]
  27. Elvidge, C.D.; Sutton, P.C.; Baugh, K.E.; Ziskin, D.; Ghosh, T.; Anderson, S. National trends in satellite observed lighting: 1992–2012. Available online: http://http://ngdc.noaa.gov/eog/dmsp/download_national_trend.html (accessed on 9 November 2013).
  28. Liu, Z.; He, C.; Zhang, Q.; Huang, Q.; Yang, Y. Extracting the dynamics of urban expansion in China using DMSP-OLS nighttime light data from 1992 to 2008. Landsc. Urban Plan. 2012, 106, 62–72. [Google Scholar] [CrossRef]
  29. National Bureau of Statistics of China. China Population and Employment Statistics Yearbook; China Statistics Press: Beijing, China, 1993–2012.
  30. National Bureau of Statistics of China. China City Statistical Yearbook; China Statistics Press: Beijing, China, 1993–2012.
  31. Yang, Y.; He, C.; Zhang, Q.; Han, L.; Du, S. Timely and accurate national-scale mapping of urban land in China using defense meteorological satellite program’s operational linescan system nighttime stable light data. J. Appl. Remote Sens. 2013, 7, 073535–073535. [Google Scholar] [CrossRef]
  32. Fang, C.L.; Song, J.T.; Zhang, Q.; Li, M. The formation, development and spatial heterogeneity patterns for the structures system of urban agglomerations in China. Acta Geogr. Sin. 2005, 60, 827–840. [Google Scholar]
  33. Amaral, S.; Câmara, G.; Monteiro, A.M.V.; Quintanilha, J.A.; Elvidge, C.D. Estimating population and energy consumption in Brazilian Amazonia using DMSP night-time satellite data. Comput. Environ. Urban Syst. 2005, 29, 179–195. [Google Scholar] [CrossRef]
  34. Elvidge, C.D.; Imhoff, M.L.; Baugh, K.E.; Hobson, V.R.; Nelson, I.; Safran, J.; Dietz, J.B.; Tuttle, B.T. Night-time lights of the world: 1994–1995. ISPRS J. Photogramm. Remote Sens. 2001, 56, 81–99. [Google Scholar] [CrossRef]
  35. Small, C.; Elvidge, C.D.; Balk, D.; Montgomery, M. Spatial scaling of stable night lights. Remote Sens. Environ. 2011, 115, 269–280. [Google Scholar] [CrossRef]
  36. Jiang, L.; Deng, X.; Seto, K.C. Multi-level modeling of urban expansion and cultivated land conversion for urban hotspot counties in China. Landsc. Urban Plan. 2012, 108, 131–139. [Google Scholar] [CrossRef]
  37. Economic and Social Research Institute (ESRI). ArcGIS Desktop Help 9.2; Economic and Social Research Institute (ESRI): Redlands, CA, USA, 2007. [Google Scholar]
  38. Chen, J.; Zhuo, L.; Shi, P.J.; Ichinose, T. The study on urbanization process in China based on DMSP/OLS data: Development of a light index for urbanization level estimation. J. Rem. Sens. 2003, 7, 168–175. [Google Scholar]
  39. Xiao, P.; Wang, X.; Feng, X.; Zhang, X.; Yang, Y. Detecting China’s urban expansion over the past three decades using nighttime light data. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 2014, 7, 4095–4106. [Google Scholar] [CrossRef]
  40. Elvidge, C.D.; Cinzano, P.; Pettit, D.R.; Arvesen, J.; Sutton, P.; Small, C.; Nemani, R.; Longcore, T.; Rich, C.; Safran, J.; et al. The Nightsat mission concept. Int. J. Remote Sens. 2007, 28, 2645–2670. [Google Scholar] [CrossRef]
  41. Zhang, Q.L.; Schaaf, C.; Seto, K.C. The Vegetation Adjusted NTL Urban Index: A new approach to reduce saturation and increase variation in nighttime luminosity. Remote Sens. Environ. 2013, 129, 32–41. [Google Scholar] [CrossRef]
  42. Bai, X.; Shi, P.J.; Liu, Y.S. Society: Realizing China’s urban dream. Nature 2014, 509, 158–160. [Google Scholar] [CrossRef] [PubMed]
  43. Wu, J.G. Landscape sustainability science: Ecosystem services and human well-being in changing landscapes. Landsc. Ecol. 2013, 28, 999–1023. [Google Scholar] [CrossRef]
  44. Liu, J.; Zhan, J.; Deng, X. Spatio-temporal patterns and driving forces of urban land expansion in China during the economic reform era. Ambio 2005, 34, 450–455. [Google Scholar] [PubMed]
  45. Ziskin, D.; Baugh, K.; Hsu, F. C.; Elvidge, C.D. Methods used for the 2006 radiance lights. Proc. Asia-Pacific Adv. Netw. 2010, 30, 131–142. [Google Scholar] [CrossRef]
Remote Sens. EISSN 2072-4292 Published by MDPI AG, Basel, Switzerland RSS E-Mail Table of Contents Alert
Back to Top