Remote Sensing of Environment 193 (2017) 65–75

Contents lists available at ScienceDirect

Remote Sensing of Environment

journal homepage: www.elsevier.com/locate/rse

Environmental degradation in the urban areas of China: Evidence from

multi-source remote sensing data
Chunyang He a,b,⁎,1, Bin Gao a,b,1, Qingxu Huang a,b,1, Qun Ma a,b, Yinyin Dou a,b
a
Center for Human-Environment System Sustainability (CHESS), State Key Laboratory of Earth Surface Processes and Resource Ecology(ESPRE), Beijing Normal University, Beijing 100875, China
b
School of Natural Resources, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China

a r t i c l e i n f o a b s t r a c t

Article history: The rapid and timely evaluation of urban environmental change is highly important for understanding urban sus-
Received 24 July 2016 tainability in China. However, the comprehensive understanding of urban environmental change in China based
Received in revised form 12 February 2017 on multi-source remote sensing data remains inadequate because current studies have mainly focused on a sin-
Accepted 28 February 2017
gle aspect of the urban environment using a specific source of remote sensing data. In this study, we developed a
Available online xxxx
comprehensive evaluation index (CEI) combining the remote sensing data of the fine particulate matter (PM2.5)
Keywords:
concentration, land surface temperature (LST) and vegetation cover (VC) to assess the urban environmental
Environmental conditions change in China at the national scale, among urban agglomerations and across the rapidly urbanized regions.
Multi-source remote sensing We found a trend of environmental degradation in the urban areas of China between 2000 and 2012. Environ-
Multi-scale analysis mentally degraded and moderately degraded urban areas accounted for 48.14% of the total urban area in
Urban expansion China. In particular, the expanded urban areas exhibited the most extensive environmental degradation, with
China 52.33% of the total expanded urban areas from 1992 to 2012 exhibiting environmental degradation or moderate-
ly environmental degradation. The increase in the PM2.5 concentration was one of the main manifestations of the
environmental degradation in the expanded urban areas. We suggest that more attention should be paid to urban
environmental issues during future urban development in China.
© 2017 Published by Elsevier Inc.

1. Introduction
China has experienced unprecedented and rapid urbanization over
the last three decades (Bai et al., 2014). From 1981 to 2012, the urban
population increased by 2.54 times, from 201 million to 712 million
(National Bureau of Statistics of China, 2013). The built-up area in-
creased from 7438 km2 in 1981 to 40,058 km2 in 2010, with an annual
growth rate of 9.26% (Ministry of Housing and Urban-Rural
Development, 2012). In addition to rapid urbanization, urban areas in
China have been facing increasing environmental pressure, and various
environmental issues have emerged. Studies have shown that in the
context of rapid industrialization and urbanization, the quality of natu-
ral resources in urban areas has degraded over the last several decades
(Cumming et al., 2014; McDonnell and MacGregor-Fors, 2016; Seto et
al., 2012). Examples included the decline of vegetation cover (VC) in
China's urban areas (Sun et al., 2011), the loss of natural habitats due
⁎ Corresponding author at: State Key Laboratory of Earth Surface Processes and
Resource Ecology, Beijing Normal University, 19 Xinjiekouwai Street, Beijing 100875,
China.
E-mail addresses: hcy@bnu.edu.cn (C. He), Bing0402@126.com (B. Gao),
qxhuang@bnu.edu.cn (Q. Huang), maqun0127@bnu.edu.cn (Q. Ma),
douyinyin1986@163.com (Y. Dou).
1
Co-first author of this work.
to urban expansion (He et al., 2014), the deterioration of urban soil
quality (Teng et al., 2014; Yao, 2016), the shortage of freshwater sup-
plies in large cities (McDonald et al., 2011), the pollution of urban rivers
(Qu and Fan, 2010) and groundwater (Coulon et al., 2016), and in-
creased urban heat island effects (Yang et al., 2011) and fine particulate
matter (PM2.5) in urban areas (Han et al., 2015). In addition, China has
become the largest urban solid waste generator in the world (Geng,
2012; Yang et al., 2016). Rapid urbanization has also contributed con-
siderably to exacerbating warming in China (Sun et al., 2016). According
to a recent government plan, the proportion of the Chinese population
residing in urban areas will increase from 52.6% in 2012 to 60% by
2020, with an annual growth rate of 1% (Bai et al., 2014). Given the
rapid urban development in this country, a timely and accurate evalua-
tion of urban environmental change is highly important to ensure urban
sustainability.
Urban sustainability is an adaptive process of maintaining balance
between ecosystem services and human wellbeing within and beyond
the urban areas. It includes three dimensions: environment, society
and economy (Wu, 2010). Urban environment, in a narrow sense, refers
to the natural environment within urban areas, including air, soil, water,
habitat and biodiversity (Forman, 2014). Urban environment plays a
fundamental role in affecting urban sustainability because the degrada-
tion of urban environment can be irreversible and cannot be fully

http://dx.doi.org/10.1016/j.rse.2017.02.027
0034-4257/© 2017 Published by Elsevier Inc.
66 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75

Table 1
Studies on evaluating urban environmental change in China based on remote sensing data.

Urban environment Data/sensor Study area Time period Reference

Urban heat island MODIS LST 419 global large cities 2003–2008 Peng et al. (2011)
NOx concentration GOME China 1996–2010 Zhang et al. (2012)
Vegetation cover NOAA/AVHRR NDVI 117 Chinese cities 1982–2006 Sun et al. (2011)
Vegetation cover SPOT/VGT Yangtze River Delta 2012 Li et al. (2014)
Urban heat island Landsat TM thermal band Beijing 1997 Jiang et al. (2006)
Water quality Landsat TM 7 bands Shenzhen 1988–1996 Wang et al. (2004)

substituted by economic development (Wu, 2010). Therefore, a set of
indices has been developed to gauge the urban environmental change
(see Huang et al., 2015 for a review), including theme-based sustain-
ability framework developed by United Nations Commission on Sus-
tainable Development (UNCSD), ecological footprint (EF), green city
index (GCI), environmental performance index (EPI), genuine progress
indicator (GPI), and human development index (HDI). However, most
of these indices need a plethora of data based on census and cannot pro-
vide a spatially explicit pattern of urban environmental change.
Remote sensing data offer an effective and spatially explicit alterna-
tive to evaluate urban environmental change from local to global scales
(Sutton, 2003). Remote sensing can obtain the information of observed
objects or region without physical contacts, usually from sensors aboard
satellites or aircrafts. Recently, a number of studies have been conduct-
ed to evaluate various aspects of urban environmental conditions using
different types of remote sensing data (Table 1, Gong, 2012). However,
these studies mainly focused on a single dimension of the urban envi-
ronmental condition and are insufficient to provide a comprehensive
picture of urban environmental change in China. Therefore, developing
an integrated index which solely needs multi-source remote sensing
data can provide a complementary means to evaluate urban environ-
mental change.
The objective of this study is to comprehensively evaluate the urban
environmental changes, i.e., the changing environmental conditions in
China's urban areas based on multi-source remote sensing data. To
achieve this goal, we first developed a comprehensive evaluation
index (CEI) of the environmental condition that integrated multi-source
remote sensing data of the PM2.5 concentration, land surface tempera-
ture (LST) and VC. Then, we evaluate the performance of this index on
measuring urban environmental change by comparing its results with
other common indices based on census data. Finally, we examined the
urban environmental changes in China between 2000 and 2012. Our
findings represent the latest comprehensive understanding of urban en-
vironmental change in China over the last 13 years and will support sus-
tainable urban development in this country.
2. Study area and data
2.1. Study area
Our study focused on the urban areas in China. In this study, the
urban areas refer to the areas dominated by the built-up land, which
is smaller than a city's administrative boundary but larger than its im-
pervious surface (Liu et al., 2014). We analyzed the environmental
changes in the urban area of China at three scales, the national, the
urban agglomeration and the rapidly urbanized region scales (Fig. 1).
Urban agglomerations are the major engine of urbanization in China.
The rapidly urbanized regions are the counties in which the urban

Fig. 1. Urban expansion in China from 1992 to 2012. Note: Abbreviations for the 14 urban agglomerations: Jing-Jin-Ji (JJJ), Liaozhongnan (LZN), Hachang (HC), Shandong peninsula (SD),
Jianghuai (JH), Yangtze River Delta (YRD), Haixiaxian (HXXA), Changjiangzhongyou (CJZY), Pearl River Delta (PRD), Guangxi (GX), Chengyu (CY), Guanzhong (GZ), Zhongyuan (ZY), and
Tianshanbeipo (TS).
 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75
Table 2
Three remote sensing dataset used in this study.
Indicator Data generation
PM2.5 MODIS + MISR + SeaWIFS observation + GEOS chemical transport
concentration model (GEOS-Chem)
LST Terra-MODIS nighttime 8-day composite
NDVI SPOT/VGT 10-day composite
built-up areas experienced significant growth. Therefore, the multi-
scale analysis can demonstrate an in-depth picture of urban environ-
mental change in China.
According to He et al. (2014), the total urban area in China increased
from 12,192 km2 in 1992 to 65,186 km2 in 2012. The growth of urban
area accounts for 0.55% of total land in China. According to Fang
(2015), there are 14 major urban agglomerations (e.g., Jing-jin-ji, Yang-
tze River Delta and Pearl River Delta urban agglomerations) in China.
Between 1992 and 2012, the urban area of these urban agglomerations
increased by 35,546 km2, accounting for 2.56% of total land in these
urban agglomerations. By 2012, the total urban areas in these urban ag-
glomerations accounted for 71.56% of China's total urban areas (He et
al., 2014).
Eight rapidly urbanized regions can be identified in China: Beijing,
Yangtze River Delta, Pearl River Delta, Ordos-Yulin, Chengdu, Chang-
sha-Zhuzhou-Xiangtan, Kunming-Yuxi, and Fuzhou-Quanzhou-Xiamen
regions (Fig. 1). Between 1992 and 2012, the total urban areas in these
eight regions increased by 14,225 km2, accounting for 18.96% of land in
these counties (He et al., 2014).
2.2. Data
Three remote sensing datasets were used in this study to examine
environment condition in China (Table 2). First, the ground-level fine
particulate matter (PM2.5) concentration dataset with a spatial resolu-
tion of 1 km was downloaded directly from the Atmospheric Composi-
tion Analysis Group of Dalhousie University (http://fizz.phys.dal.ca/
~atmos/martin/). This dataset recorded the estimated annual average
PM2.5 concentration based on MODIS, Multi-angle Imaging
Spectroradiometer (MISR) and Sea-viewing Wide Field-of-view Sensor
(SeaWIFS) observations and the Goddard Earth Observing System
chemical transport model (GEOS-Chem, van Donkelaar et al., 2016).
To reduce inter-annual errors, following the method proposed by Han
et al. (2015), three-year moving averages were calculated for the period
from 2000 to 2012.
Second, the LST dataset consisted of the Terra-MODIS nighttime 8-
day composite imageries at a spatial resolution of 1 km. The data were
obtained from the data archive and distribution system of the National
Aeronautics and Space Administration (NASA, https://modis-land.gsfc.
nasa.gov/temp.html). We only used the nighttime LST dataset since pre-
vious studies have found that the nighttime LST were closely related to
nighttime light pollution (Peng et al., 2011), extraction of urban built
area (He et al., 2014) and extensive warming effect around urban fringe
(Zhou et al., 2015). We didn't consider the Aqua-MODIS retrievals as
they are not available until 2002. The annual mean LST dataset was pro-
duced from the intra-annual time-series data to capture the nighttime
LST characteristics in China from 2000 to 2012.
Table 3
Classification of environmental changes in China.
Types Improved Moderately improved
Dividing criteria bx−1.5 s x−1.5 s~x−0.5 s
CEI b40.80 40.80–44.18
x and s represent the average value and the sample standard deviation of the CEI of the environmental condition, respectively.
67
Spatial Temporal Provider
scale scale
1 km 2000–2012 Atmospheric Composition Analysis Group of
Dalhousie University
National Aeronautics and Space Administration (NASA)
Flemish Institute for Technological Research
Third, the SPOT/VGT 10-day composite NDVI dataset was used to
calculate VC because previous studies have shown that it can successful-
ly monitor vegetation dynamics in China (Xiao et al., 2002; Zhou et al.,
2009). First, this dataset was collected from the Flemish Institute for
Technological Research (http://www.spot-vegetation.com/) at a spatial
resolution of 1 km. Then, the annual mean NDVI values were generated
from the intra-annual time-series NDVI data. Finally, following the
method proposed by Sellers et al. (1996), the VC value of each pixel in
China from 2000 to 2012 was computed using the following formula:
VC i ¼ ðNi −Ns Þ=ðNv −N s Þ ð1Þ
where VCi is the VC value at pixel I, Ni refers to the NDVI value at pixel i,
and Ns and Nv are the minimum and maximum values of the NDVI in the
study area, respectively.
The spatial distribution of urban area from 1992 to 2012 was pro-
duced by He et al. (2014). This dataset provides reliable information
on urban expansion dynamics in China at a spatial resolution of 1 km
(Liu et al., 2012; He et al., 2014, Xu et al., 2016). To validate this dataset,
He et al. (2014) selected eight cities with different level of urbanization
in China and compared the results with the land-use maps derived from
Landsat TM/ETM+ data for 1995, 2000, 2005, and 2012. The compari-
son yielded an average Kappa (i.e., a common metric measuring the
agreement between two land use/cover maps) of 0.66 and an average
overall accuracy of 95.20%. All the spatial data were projected using
the Albers Equal Area projection.
The data used for validation include ecological footprint (EF), natural
resource consumption and waste discharge from census data. The EF
data were collected from previous studies on assessing urban sustain-
ability in 10 Chinese megacities from 2000 to 2010 (Huang et al.,
2016) and regional sustainability in China from 1990 to 2010 (Li et al.,
2016), respectively. Natural resource consumption and waste discharge
in 283 prefecture-level cities across China between 2000 and 2010 were
collected from census data. Specifically, seven indicators were included:
water supply, gas supply, liquefied petroleum gas, electricity consump-
tion, domestic garbage discharge, urban sewage discharge and industri-
al wastewater discharge.
In addition, the administrative boundaries of counties, prefecture-
level cities and provinces were downloaded as Geographical Informa-
tion System (GIS) files from the National Geomatics Center of China
(http://ngcc.sbsm.gov.cn). The boundaries for the 14 major urban ag-
glomerations in China were derived from Fang (2015).
3. Methods
In this study, based on the theme-oriented environment framework
proposed by United Nations, we developed an integrated index (i.e.,
CEI) using remote sensing information of PM2.5 concentration, VC and
Unchanged Moderately degraded Degraded
x-0.5 s~x+0.5 s x+0.5 s~x+1.5 s Nx+1.5 s
44.18–47.56 47.56–50.94 ≥50.94
68 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75

Fig. 2. Normalized indicators of environmental changes between 2000 and 2012.

LST. Then, we used the CEI to measure the urban environmental change
in China at three scales. We also validated the results based on this index
with traditional urban environmental indices based on census data
using correlation analysis.
3.1. Developing the CEI
We developed the CEI to evaluate urban environmental change in
China based upon the theme-oriented framework proposed by United
Nations Commission on Sustainable Development (UNCSD). This frame-
work is one of the most commonly used frameworks for evaluating the
environmental conditions at the national and regional scales (UN,
2007). It identifies five major environmental themes: freshwater
(water quality and water quantity); land (desertification, agriculture,
forests, and urbanization); atmosphere (air quality, ozone layer deple-
tion and climate change); biodiversity (ecosystem and species); and
oceans, seas and coasts (coastal zone and fisheries). This framework en-
sures a holistic and comprehensive assessment of environmental condi-
tions from multiple dimensions (UN, 2007). The third edition of the
framework was published in 2007 by UNCSD and has been used for
evaluating the environmental condition in many countries and regions
(UN, 2007; Wu and Wu, 2012). For example, based on the UNCSD
framework, the environmental conditions in Africa (Brent and
Heuberger, 2005) and South Wales coastline (Phillips et al., 2007)
have been evaluated.
Following this framework, the developed index, CEI, considers
changes in three environmental indicators: PM2.5 concentration, LST
and VC. To some extent, these three remote sensing indicators can rep-
resent three themes of the UNCSD framework, atmosphere, land and
biodiversity in the urban area of China because they are highly related
to air pollution (Watts et al., 2015), land cover (Buyantuyev and Wu,
2010) and natural habitat (Kerr and Ostrovsky, 2003; Turner et al.,

Fig. 3. Environmental changes in China between 2000 and 2012.

 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75 69

Fig. 4. Correlation between the CEI and EF in the megacities and provinces of China.

2003; Wang et al., 2010). The other two themes, freshwater, and oceans, 3.2. Evaluating environmental changes in China's urban areas from 2000 to
seas and coasts are not considered in this study due to a lack of effective 2012
indicators that can be derived at the national scale from remote sensing
data. The CEI is calculated as the geometric average of these three indi- Urban environmental changes were evaluated during 2000–2012
cators (UNDP, 2014): due to the availability of remote sensing data (e.g., LST and NDVI)
from 2000 and a significant environmental change along with the
p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
CEI i ¼ 3
ðΔPMi þ 1Þ  ðΔVCi þ 1Þ  ðΔLSTi þ 1Þ ð2Þ urban expansion from 2000 to 2012 (Han et al., 2015; He et al., 2014;
Sun et al., 2011; Teng et al., 2014). First, the changes in environmental
conditions were divided into five levels according to the mean and stan-
where CEIi is the value of environment change at pixel i. CEIi ranges from
dard deviation of the CEI (He et al., 2012): improved, moderately im-
1 to 101, where a larger value of CEIi indicates a higher level of environ-
proved, unchanged, moderately degraded and degraded (Table 3).
mental degradation. ΔPMi, ΔLSTi and ΔVCi are the normalized changes in
Then, using the following formula, the areal percentages of the five
PM2.5 concentration, LST, and VC at pixel i between 2000 and 2012,
types of environmental change in urban areas were calculated:
respectively.
Specifically, ΔPMi and ΔLSTi are computed using the following for-
Arean
mula: Pn ¼  100% ð5Þ
Area2012
 
V 2012
i −V 2000
i −minv where Arean is the urban area experiencing an environmental change of
ΔV i ¼  100 ð3Þ
maxv −minv level n, and Area2012 is the total urban area in 2012. A larger percentage
of degraded areas suggested more extensive environmental
where ΔVi are the normalized changes in PM2.5 concentration or LST. degradation.
V2000
i and V2012
i are the values of the PM2.5 concentration or LST value Supported by GIS, we used the CEI to investigate the urban environ-
at pixel i in 2000 and 2012, respectively; and minv and minv refer to mental change at three scales: the national, the urban agglomeration
the minimum and maximum changes in the PM2.5 concentration or and the rapidly urbanized region scale. Meanwhile, at each scale, we
LST value between 2000 and 2012 in China, respectively. further evaluated urban environmental changes among three areas
Since higher VC values suggest better environmental condition, (Fig. 1), i.e., the urban areas in 1992 (UA1992), the expanded urban
which is the contrary of PM2.5 concentration and LST, the ΔVCi was com- areas from 1992 to 2012 (EUA1992–2012) and urban fringe
puted using the following formula: (UFA2012). Here, UFA2012 refers to a 1-km buffer zone around the
urban areas in 2012. The objective of setting the UFA2012 is to further
  investigate the spatial heterogeneity of urban environmental change.
maxVC − VC 2012
i −VC 2000
i
ΔVC i ¼  100 ð4Þ
maxVC −minVC 3.3. Validating the CEI

where VC2000
i and VC2012
i are the values of the VC at pixel i in 2000 and Two approaches were used to evaluate the performance of CEI on
2012, respectively; minVC and maxVC refer to the minimum and maxi- measuring urban environmental change. First, the correlation between
mum changes in the VC between 2000 and 2012 in China. CEI and a traditional sustainability index was analyzed. We chose a

Table 4
Correlation between CEI and changes in natural resource consumption and waste discharge between 2000 and 2010 across 283 prefecture-level cities in China.

Natural resource and energy consumption Water supply Gas supply Liquefied petroleum gas Electricity consumption
0.55⁎⁎ 0.0.38⁎⁎ −0.09 0.86⁎⁎

Environmental pollution Domestic garbage discharge Urban sewage discharge Industrial wastewater discharge
0.65⁎⁎ 0.51⁎⁎ −0.14⁎
⁎ Correlation is significant at the 0.05 level (2-tailed).
⁎⁎ Correlation is significant at the 0.01 level (2-tailed).
70 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75

sustainability index, ecological footprint (EF), because it can systemati-

cally evaluate the human appropriation of natural resources and has
been widely adopted to evaluate human impact on environmental sus-
tainability at varying scales (Wackernagel et al., 1999; Wiedmann and
Barrett, 2010; Wu and Wu, 2012). EF is defined as the land (and
water) area that not only provide all the energy/material resources
demanded by a defined human population or economy but also absorb
all the discharged waste (Mayer, 2008; Wackernagel et al., 2002). An in-
crease in the EF denotes an increase in the environmental pressure.
Therefore, a positive correlation between CEI and changes in EF would
suggest the potential of CEI to evaluate urban environmental change
in China. Due to the data availability, we were able to evaluate the cor-
relation between the total CEI in the urban areas of the 10 Chinese
megacities and 29 provinces between 2000 and 2010 and the changes
in EF in these cities and provinces, respectively.
Additionally, Tao et al. (2016) has found that the consumption of
natural resources and waste discharge can also be used to evaluate the
effects of urbanization on environment in China. Therefore, the correla-
tion between CEI and changes in natural resource consumption and
waste discharge between 2000 and 2010 was investigated based on
census data. Following Tao et al. (2016), we collected four indicators
of natural resource consumption (i.e., water, gas, liquefied petroleum
gas, electricity) and three indicators of waste discharge (i.e., domestic
garbage, urban sewage and industrial wastewater) in 283 prefecture-
level cities across China and analyzed their association with CEI.

4. Results

The environmental changes in China between 2000 and 2012 were

obtained based on the normalized changes in PM2.5 concentration, LST
and VC (Figs. 2 and 3).

4.1. Performance of the CEI

The validation results showed that CEI provides a good proxy for
evaluating urban environmental changes. The correlation between CEI
and changes in the EF from 2000 to 2010 over 10 Chinese megacities
yielded a positive coefficient (R = 0.86 P b 0.01, 2-tailed, Fig. 4a). The
correlation coefficient between the CEI and the changes in EF from
2000 to 2010 in 29 Chinese provinces is 0.81 (P b 0.01, 2-tailed Fig. 4b).
Additionally, the correlation between CEI and changes in natural re-
source consumption and waste discharge between 2000 and 2010 in
283 prefecture-level cities across China demonstrated that the CEI was
strongly correlated with some indicators (Table 4), such as electricity
consumption (R = 0.86, P b 0.01), domestic garbage discharge (R =
0.65, P b 0.01), water supply (R = 0.55, P b 0.01), urban sewage dis-
charge (R = 0.51, P b 0.01), and gas supply (R = 0.38, P b 0.01). There-
fore, we believe that the CEI could effectively represent the
comprehensive environment conditions in China's urban areas.

4.2. National-scale urban environmental changes

A trend of environmental degradation was detected in China's urban

areas between 2000 and 2012. At the national scale, the urban areas that
experienced moderately environmental degradation and environmen-
tal degradation accounted for 30.95% (19,882 km2) and 17.19% Fig. 5. Environmental changes in urban areas of China from 2000 to 2010 (a) national
scale, (b) urban agglomerations, and (c) rapidly urbanized regions. Note: Please refer to
(11,046 km2) of the total urban area, respectively (Fig. 5a). By contrast,
Fig. 1 for an explanation of the abbreviations.
the urban areas exhibiting moderately improved and improved envi-
ronmental conditions only accounted for 13.95% (8963 km2) and
2.57% (1650 km2) of the total urban area, respectively.

4.3. Urban agglomeration-scale environmental changes
A clear trend of environmental degradation was found in the urban
areas of these major urban agglomerations. The overall urban areas
that exhibited moderately environmental degradation and
environmental degradation accounted for 49.59% (22,178 km2) of the
total urban areas across the 14 major urban agglomerations (Fig. 5a).
By contrast, the overall urban areas with improved environmental con-
ditions only accounted for 1.91% (852 km2). The urban agglomeration of
Shandong peninsula suffered the most extensive environmental
 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75 71

Table 5
Environmental changes in urban areas across major Chinese urban agglomerations.

Urban Urban area in Expanded urban area Urban Degraded Moderately Unchanged Moderately Improved
agglomerationsa 1992 (km2) 1992–2012 (km2) fringe degraded improved
(km2)
Area Percentage Area Percentage Area Percentage Area Percentage Area Percentage
(km2) (km2) (km2) (km2) (km2)

SD 700 2917 3719 2653 73.35% 701 19.38% 244 6.75% 19 0.53% 0 0.00%
ZY 264 1079 1402 646 48.10% 638 47.51% 56 4.17% 3 0.22% 0 0.00%
JJJ 1564 4021 4696 1732 31.01% 2873 51.44% 959 17.17% 19 0.34% 2 0.04%
CY 286 2898 3492 969 30.43% 1572 49.37% 578 18.15% 46 1.44% 19 0.60%
JH 219 1284 1481 255 16.97% 799 53.16% 443 29.47% 5 0.33% 1 0.07%
CJZY 602 2988 3672 569 15.85% 1475 41.09% 1252 34.87% 263 7.33% 31 0.86%
GX 99 571 748 97 14.48% 358 53.43% 203 30.30% 12 1.79% 0 0.00%
YRD 1425 8248 7715 955 9.87% 4164 43.05% 3724 38.50% 803 8.30% 27 0.28%
GZ 263 969 984 53 4.30% 521 42.29% 603 48.94% 53 4.30% 2 0.16%
HXXA 212 2664 2904 59 2.05% 250 8.69% 1363 47.39% 1154 40.13% 50 1.74%
HC 654 1100 1639 14 0.80% 264 15.05% 1138 64.88% 329 18.76% 9 0.51%
PRD 2043 5050 3978 24 0.34% 458 6.46% 3628 51.15% 2801 39.49% 182 2.57%
TS 127 577 976 0 0.00% 41 5.82% 604 85.80% 58 8.24% 1 0.14%
LZN 583 1317 1992 0 0.00% 38 2.00% 448 23.58% 886 46.63% 528 27.79%
a
Please refer to Fig. 1 for an explanation of the abbreviations.

degradation, affecting approximately 73.35% (2653 km2) of the urban
areas (Table 5, Fig. 5b). Three other urban agglomerations (i.e.,
Zhongyuan, Jing-Jin-Ji, and Chengyu) also experienced relatively exten-
sive environmental degradation, with 30–50% of their urban areas
exhibiting degradation (Fig. 5b). Four urban agglomerations (i.e,
Hachang, Pearl River Delta, Tianshanbeipo and Liaozhongnan) exhibited
smaller extents of environmental degradation. The percentage of urban
areas suffering from environmental degradation in these urban agglom-
erations was b1%.
4.4. Rapidly urbanized region-scale urban environmental changes
A trend of environmental degradation was evident in the urban
areas of the eight rapidly urbanized regions, where the degraded and
moderately degraded urban areas covered 35.49% (6473 km2) of the
total urban area (Fig. 5a). By comparison, the urban areas showing mod-
erately improved and improved environmental conditions only covered
20.75% (3784 km2) and 1.04% (189 km2) of the total urban area, respec-
tively (Table 6). Among the eight rapidly urbanized regions, the Cheng-
du region showed the most extensive environmental degradation, with
45.57% of its urban area being degraded (Table 6). The Changsha-Zhu-
zhou-Xiangtan, Beijing, and Kunming-Yuxi regions also had large pro-
portions of degraded urban area, corresponding to 39.77% (208 km2),
21.15% (376 km2), and 17.36% (146 km2) of their total urban areas, re-
spectively. The Yangtze River delta, Fuzhou-Quanzhou-Xiamen, and
Pearl River Delta regions contained relatively small proportions of
urban areas exhibiting environmental degradation. The Ordos-Yulin re-
gion exhibited the smallest extent of environmental degradation, with
only 0.15% (1 km2) of its urban area being affected (Table 6).
4.5. Expanded urban areas exhibited the most extensive environmental
degradation
At the national scale, approximately 52% of the EUA1992–2012 ex-
perienced environmental degradation or moderately environmental
degradation. These values were 30% for the UA1992 and 42% for the
UFA2012 (Fig. 6). By comparison, urban areas with improved environ-
mental conditions accounted for only 2.42% of the EUA1992–2012,
3.22% of the UA1992, and 4.34% of the UFA2012 (Fig. 6).
Among the 14 urban agglomerations, we also found that the
EUA1992–2012 had a relatively larger rate of environmental degrada-
tion than the UA1992 and the UFA2012. Specifically, 53.96% of the
EUA1992–2012 suffered environmental degradation or moderately en-
vironmental degradation (Fig. 6), whereas these values were only
32.39% and 46.28% for the UA1992 and the UFA2012, respectively.
Among the eight rapidly urbanized regions, the EUA1992–2012 also
showed more extensive environmental degradation than the UA1992
and the UFA2012 (Fig. 6). The environmentally degraded or moderately
degraded urban areas covered 41.05% of the EUA1992–2012. By com-
parison, 15.25% of the UA1992 and 28.29% of the UFA2012 were degrad-
ed or moderately degraded.
5. Discussion
5.1. CEI provides a timely and reliable evaluation of urban environmental
changes
First, the CEI has a concrete conceptual framework, i.e., the theme-
oriented framework proposed by UNCSD on evaluating environmental

Table 6
Environmental changes in urban areas across the eight rapidly urbanized regions of China.

Rapidly urbanized regions Urban Expanded urban Urban Degraded Moderately Unchanged Moderately Improved
area in area from 1992 fringe degraded improved
1992 to 2012 (km2) (km2)
Area Percentage Area Percentage Area Percentage Area Percentage Area Percentage
(km2)
(km2) (km2) (km2) (km2) (km2)

Chengdu 126 890 508 463 45.57% 427 42.03% 126 12.40% 0 0.00% 0 0.00%
Changsha-Zhuzhou-Xiangtan 92 431 194 208 39.77% 274 52.39% 41 7.84% 0 0.00% 0 0.00%
Beijing 698 1080 712 376 21.15% 947 53.26% 450 25.31% 3 0.17% 2 0.11%
Kunming-Yuxi 82 759 633 146 17.36% 464 55.17% 221 26.28% 10 1.19% 0 0.00%
Yangtze River Delta 964 5078 3584 198 3.28% 2323 38.45% 2870 47.50% 634 10.49% 17 0.28%
Fuzhou-Quanzhou-Xiamen 111 1275 1029 31 2.24% 155 11.18% 733 52.89% 453 32.68% 14 1.01%
Pearl River Delta 1869 4105 2545 11 0.18% 370 6.19% 3038 50.85% 2402 40.21% 153 2.56%
Ordos-Yulin 10 668 1182 1 0.15% 79 11.65% 313 46.17% 282 41.59% 3 0.44%
72 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75

of input data, for example, 25 indicators for the theme-based frame-

work proposed by UNCSD. On the contrary, the CEI only needs three re-
motely sensed inputs. In addition, the urban environmental changes
measured by CEI can be visualized in a spatially explicit way. However,
it is difficult to visualized census-based results at the local scale. Addi-
tionally, the CEI provides a potential way to measure urban environ-
mental changes across regions, because the remote sensing datasets
are globally accessible. In contrast, the census-based indices may face
the incomparability problem because the calculation at the local scale
may be compromised by data availability.
Third, the quantitative validation results reported here showed that
the CEI can reliably reflect the environmental changes in the urban areas
of China. The CEI were highly correlated with other widely used indices
(i.e., EF, natural resource consumption and waste discharge) at multiple
scales. In addition, the findings reported here are qualitatively in line
with the conclusions from previous studies on different scales. For in-
stance, at the national scale, Yu and Wen (2010) evaluated the environ-
mental changes in 46 representative cities in China during 2006–2007
and found that the urban environment suffered from degradation and
was unsustainable in 76% of these cities. At the urban agglomeration
scale, Zeng and Chen (2012) evaluated the environmental conditions
of ten urban agglomerations in China and also found that the Shandong
peninsula and the Zhongyuan agglomerations had the highest level of
environmental pollution.
In general, the proposed index is an important application of urban
remote sensing and has the potential to be used to examine urban envi-
ronmental changes on various scales quickly and reliably. In the future,
generating a time-series CEI product in China will facilitate further re-
search and urban sustainable planning. For example, based on the prod-
uct, it is possible to examine the population dynamics exposed to urban
environmental degradation and its effects on human health. It is also
possible to pinpoint the cities that experienced the most significant deg-
radation in China and thereby guide the national policies on protecting
urban environment.

5.2. The increased PM2.5 concentration was one of the major manifestations
of environmental degradation in the expanded urban areas

Unlike previous studies which focus on a single aspect of urban en-

vironmental change based on remote sensing data (Table 1), the CEI
combining multi-source remote sensing data provides a comprehensive
evaluation of urban environmental change. We found that the changes
in PM2.5 concentration, VC, and LST between 2000 and 2012 have a
close relationship with CEI (Table 7). The correlation coefficient be-
tween the changes in PM2.5 and CEI was 0.88, the largest among the
three indicators. The correlation coefficients between the changes in
VC and CEI, and between the changes in LST and CEI were 0.45 and
0.18, respectively.
To explore which environmental indicator contributes more to the
degradation, we compared the normalized changes between 2000 and
2012 in the EUA 1992–2012 among the three indicators. The result indi-
cated that the normalized change in the PM2.5 concentration between
2000 and 2012 was larger than the normalized changes in LST and VC
in the expanded urban areas (Fig. 7). At the national scale, the

Fig. 6. Comparisons of environmental changes among UA1992, EUA1992–2012, and

UFA2012 on different scales (a) national scale; (b) urban agglomeration scale; (c) Table 7
rapidly urbanization region scale. Correlation coefficients among PM2.5 concentration, LST, NDVI and CEI.

ΔPM2.5 ΔLST ΔVC CEI

changes in urban areas. Given the complexity of the urban environmen- ΔPM2.5 1
ΔLST −0.073⁎⁎ 1
tal conditions, the CEI considers three aspects of urban environmental
ΔVC 0.087⁎⁎ 0.010⁎ 1
conditions (i.e., atmosphere, land and biodiversity) simultaneously. CEI 0.881⁎⁎ 0.181⁎⁎ 0.451⁎⁎ 1
Second, the advantages of the remotely sensed data make it possible
ΔPM2.5, ΔLST, and ΔVC represent the changes in PM2.5, LST and VC, respectively. They were
for the CEI to enrich our understanding on urban environmental change. calculated by Eqs. (3) and (4).
This index can gauge urban environmental changes in a cost-effective ⁎ Correlation is significant at the 0.05 level (2-tailed).
way. Commonly, the census-based measurement needs a large number ⁎⁎ Correlation is significant at the 0.01 level (2-tailed).
 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75
Fig. 7. Comparisons of the normalized changes in PM2.5, LST and VC in the EUA1992–2012
on different scales (a) national scale, (b) urban agglomerations, and (c) rapidly urbanized
regions. Note: Please refer to Fig. 1 for an explanation of the abbreviations. The normalized
variation refers to the changes in each indicator divided by its minimum value at the local
scale. For example, ΔPM2.5 was normalized as ΔPM2.5/minΔPM2.5, whereas minΔPM2.5
refers to the minimum ΔPM2.5 among the 14 urban agglomerations and the 8 rapidly
urbanized regions.
73
normalized change in the PM2.5 concentration was 1.38 in the
EUA1992–2012, whereas the normalized changes in LST and VC during
the same period were 1.05 and 1.13, respectively (Fig. 7a).
Similar results can be found at the other two scales. The normalized
change in the PM2.5 concentration from 2000 to 2012 in the EUA1992–
2012 was 22.12% and 32.69% higher than the normalized changes in LST
and VC at the urban agglomeration scale, and 17.27% and 22.86% higher
at the rapidly urbanized region scale (Fig. 7a). The comparison was most
notable in the Shandong peninsula urban agglomeration (Fig. 7b) and
Beijing region (Fig. 7c), where the normalized changes in the PM2.5 con-
centration were 1.88 and 1.62, respectively.
5.3. Limitations and future perspectives
There are several limitations to this study. First, the temporal cover-
age of the remote sensing data was limited. The LST dataset was not
available until 2000. Therefore, the temporal changes of the urban envi-
ronment in China were restricted to the period between 2000 and 2012.
Second, the accuracy of the urban area maps can affect the results. The
urban area maps used in this study were derived from nighttime light
data. The coarse spatial resolution of the nighttime light data and the re-
flectance of light from adjacent land (e.g., water bodies) may influence
the mapping accuracy of the urban area (Huang et al., 2014). Third,
the changes in environmental conditions were divided by a common
approach based on the mean and standard deviation of the CEI. Howev-
er, there are several classification approaches, such as, equal interval,
quantile and natural breaks (Law and Collins, 2013). Fourth, the perfor-
mance of the CEI needs further verification with other indices and its ap-
plicability in other regions also needs verification.
In the future, a time-series approach can be used to investigate
urban environmental change since the remote sensing data used in
this study are available annually. Moreover, additional remote sensing
data and statistical methods can facilitate further understanding urban
environmental changes in China. For example, rapid urban develop-
ment has been shown to result in serious water pollution in many cities
(Shao et al., 2006; Zhao et al., 2013). Thus, integrating remote sensing
data on urban water quality (Olmanson et al., 2013; Singh et al., 2013)
should be a priority to further improve the comprehensive evaluation
of the urban environmental conditions. In addition, statistical methods,
such as principal component analysis, stepwise regression and spatially
explicit regression, could be used to explore the associations of urban
environmental change with potential drivers, i.e., socio-economic vari-
ables, locational factors and government policies (Burchfield et al.,
2006).
6. Conclusions
CEI provides a timely and reliable evaluation of urban environmental
changes because of three advantages. First, it is developed based on a
concrete conceptual framework, the theme-oriented framework pro-
posed by the UNCSD. Second, compared to the traditional indices
based on census data, the CEI can gauge urban environmental changes
in a cost-effective, spatially explicit, and comparable way. Third, the
quantitative validation confirmed that the CEI-based evaluation showed
a high degree of consistency (R N 0.81, P b 0.01) with those based on the
EF.
During 2000–2012, a clear trend of environmental degradation was
observed in the urban areas of China. The environmentally degraded
and moderately degraded urban areas accounted for 48.14% of the
total urban area in China. The EUA1992–2012 experienced more exten-
sive environmental degradation than the UA1992 and UFA2012. The
proportions of urban areas suffering from environmental degradation
or moderately environmental degradation exceeded 41% in the
EUA1992–2012 on all three scales, which were higher than the propor-
tions in both the UA1992 and UFA2012. The increase in the PM2.5 con-
centration was one of the main manifestations of environmental
74 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75
degradation in the EUA1992–2012. It was 1.31 times and 1.22 times
those of LST and VC, respectively. Therefore, effective policies should
be implemented to address these urban environmental issues in the fu-
ture course of urban development in China.
Acknowledgements
We would like to thank Prof. Jianguo Wu from Arizona State Univer-
sity, USA for his professional suggestions on the initiative idea of the ar-
ticle. We also want to express our respects and gratitude to the
anonymous reviewers and editors for their professional comments
and suggestions. This work has been supported in part by the National
Natural Science Foundation of China (Grant No. 41621061 &
41501092) and the National Basic Research Program of China (Grant
No. 2014CB954302 & 2014CB954303). It was also supported by the pro-
ject of “New Century Excellent Teacher (NCET-13-0053)” from Ministry
of Education of China.
References
Bai, X., Shi, P., Liu, Y., 2014. Realizing China's urban dream. Nature 509, 158–160.
Brent, A.C., Heuberger, Manzini D., 2005. Evaluating projects that are potentially eligible
for clean development mechanism (CDM) funding in the South African context: A
case study to establish weighting values for sustainable development criteria. Envi-
ron. Dev. Econ. 10, 631–649.
Burchfield, M., Overman, H.G., Puga, D., Turner, M.A., 2006. Causes of sprawl: A portrait
from space. Q. J. Econ. 121, 587–633.
Buyantuyev, A., Wu, J., 2010. Urban heat islands and landscape heterogeneity: Linking
spatiotemporal variations in surface temperatures to land-cover and socioeconomic
patterns. Landsc. Ecol. 25 (1), 17–33.
Coulon, F., Jones, K., Li, H., Hu, Q., Gao, J., Li, F., et al., 2016. China's soil and groundwater
management challenges: Lessons from the UK's experience and opportunities for
China. Environ. Int. 91, 196–200.
Cumming, G.S., Buerkert, A., Hoffmann, E.M., Schlecht, E., von Cramon-Taubadel, S.,
Tscharntke, T., 2014. Implications of agricultural transitions and urbanization for eco-
system services. Nature 515, 50–57.
Fang, C., 2015. Important progress and future direction of studies on China's urban ag-
glomerations. J. Geogr. Sci. 25, 1003–1024.
Forman, R.T., 2014. Urban Ecology: Science of Cities. Cambridge University Press, Cam-
bridge, UK.
Geng, Y., 2012. Toward safe treatment of municipal solid wastes in China's urban areas.
Environ. Sci. Technol. 46, 7067–7068.
Gong, P., 2012. Remote sensing of environmental change over China: a review. Chin. Sci
Bull 57, 2793–2801.
Han, L., Zhou, W., Li, W., 2015. City as a major source area of fine particulate (PM2.5) in
China. Environ. Pollut. 206, 183–187.
He, C., Ma, Q., Li, T., Yang, Y., Liu, Z., 2012. Spatiotemporal dynamics of electric power con-
sumption in Chinese Mainland from 1995 to 2008 modeled using DMSP/OLS stable
nighttime lights data. J. Geogr. Sci. 22, 125–136.
He, C., Liu, Z., Tian, J., Ma, Q., 2014. Urban expansion dynamics and natural habitat
loss in China: a multi-scale landscape perspective. Glob. Chang. Biol. 20,
2886–2902.
Huang, L., Wu, J., Yan, L., 2015. Defining and measuring urban sustainability: a review of
indicators. Landsc. Ecol. 30, 1175–1193.
Huang, L., Yan, L., Wu, J., 2016. Assessing urban sustainability of Chinese megacities:
35 years after the economic reform and open-door policy. Landsc. Urban Plan. 145,
57–70.
Huang, Q., Yang, X., Gao, B., Yang, Y., Zhao, Y., 2014. Application of DMSP/OLS nighttime
light images: a meta-analysis and a systematic literature review. Remote Sens. 6,
6844–6866.
Jiang, Z., Chen, Y., Li, J., 2006. On urban heat island of Beijing based on Landsat TM data.
Geospatial Info. Sci. 9 (4), 293–297.
Kerr, J.T., Ostrovsky, M., 2003. From space to species: ecological applications for remote
sensing. Trends Ecol. Evol. 18 (8), 299–305.
Law, M., Collins, A., 2013. Getting to Know ArcGIS for Desktop (Third Edition). ESRI Press,
Redlands, CA. USA.
Li, G.Y., Chen, S.S., Yan, Y., Yu, C., 2014. Effects of urbanization on vegetation degradation
in the Yangtze River Delta of China: assessment based on SPOT-VGT NDVI. J. Urban
Plan. Dev. Div. http://dx.doi.org/10.1061/(ASCE)UP.1943-5444.0000249.
Li, J., Liu, Z., He, C., Tu, W., Sun, Z., 2016. Are the drylands in northern China sustainable? A
perspective from ecological footprint dynamics from 1990 to 2010. Sci. Total Environ.
553, 223–231.
Liu, Z., He, C., Zhang, Q., Huang, Q., Yang, Y., 2012. Extracting the dynamics of urban ex-
pansion in China using DMSP-OLS nighttime light data from 1992 to 2008. Landsc.
Urban Plan. 106, 62–72.
Liu, Z., He, C., Zhou, Y., Wu, J., 2014. How much of the world's land has been urbanized,
really? A hierarchical framework for avoiding confusion. Landsc. Ecol. 29, 763–771.
Mayer, A.L., 2008. Strengths and weaknesses of common sustainability indices for multi-
dimensional systems. Environ. Int. 34, 277–291.
McDonald, R.I., Green, P., Balk, D., Fekete, B.M., Revenga, C., Tod, M., et al., 2011. Urban
growth, climate change, and freshwater availability. P. Natl. Acad. Sci. U. S. A. 108,
6312–6317.
McDonnell, M.J., MacGregor-Fors, I., 2016. The ecological future of cities. Science 352,
936–938.
Ministry of Housing and Urban-Rural Development PRC, 2012. China Urban Construction
Statistical Yearbook 2011. China Planning Press, Beijing.
National Bureau of Statistics of China, 2013. China Statistical Yearbook. China Statistics
Press, Beijing.
Olmanson, L.G., Brezonik, P.L., Bauer, M.E., 2013. Airborne hyperspectral remote sensing
to assess spatial distribution of water quality characteristics in large rivers: the Mis-
sissippi River and its tributaries in Minnesota. Remote Sens. Environ. 130, 254–265.
Peng, S., Piao, S., Ciais, P., Friedlingstein, P., Ottle, C., Bréon, F., et al., 2011. Surface urban
heat island across 419 global big cities. Environ. Sci. Technol. 46, 696–703.
Phillips, M.R., Abraham, E.J., Williams, A.T., House, C., 2007. Function analysis as a coastal
management tool: the South Wales Coastline, UK. J. Coast. Conserv. 11, 159–170.
Qu, J., Fan, M., 2010. The current state of water quality and technology development for
water pollution control in China. Crit. Rev. Environ. Sci. Technol. 40, 519–560.
Sellers, P., Los, S.O., Tucker, C.J., Justice, C.O., Dazlich, D.A., Collatz, G.J., et al., 1996. A re-
vised land surface parameterization (SiB2) for atmospheric GCMS. Part II: the gener-
ation of global fields of terrestrial biophysical parameters from satellite data. J. Clim.
9, 707–737.
Seto, K.C., Güneralp, B., Hutyra, L.R., 2012. Global forecasts of urban expansion to 2030 and
direct impacts on biodiversity and carbon pools. P. Natl. Acad. Sci. U. S. A. 109,
16083–16088.
Shao, M., Tang, X., Zhang, Y., Li, W., 2006. City clusters in China: air and surface water pol-
lution. Front. Ecol. Environ. 4, 353–361.
Singh, A., Jakubowski, A.R., Chidister, I., Townsend, P.A., 2013. A MODIS approach to
predicting stream water quality in Wisconsin. Remote Sens. Environ. 128, 74–86.
Sun, J., Wang, X., Chen, A., Ma, Y., Cui, M., Piao, S., 2011. NDVI indicated characteristics of
vegetation cover change in China's metropolises over the last three decades. Environ.
Monit. Assess. 179, 1–14.
Sun, Y., Zhang, X., Ren, G., Zwiers, F.W., Hu, T., 2016. Contribution of urbanization to
warming in China. Nat. Clim. Chang. http://dx.doi.org/10.1038/nclimate2956.
Sutton, P., 2003. An empirical environmental sustainability index derived solely from
nighttime satellite imagery and ecosystem service valuation. Popul. Environ. 24,
293–311.
Tao, Y., Li, F., Crittenden, J.C., Lu, Z., Sun, X., 2016. Environmental impacts of China's urban-
ization from 2000 to 2010 and management implications. Environ. Manag. 57,
498–507.
Teng, Y., Wu, J., Lu, S., Wang, Y., Jiao, X., Song, L., 2014. Soil and soil environmental quality
monitoring in China: a review. Environ. Int. 69, 177–199.
Turner, W., Spector, S., Gardiner, N., Fladeland, M., Sterling, E., Steininger, M., 2003. Re-
mote sensing for biodiversity science and conservation. Trends Ecol. Evol. 18 (8),
306–314.
UN, 2007. Indicators of Sustainable Development: Guidelines and Methodologies. third
ed. United Nations, New York.
UNDP, 2014. Human Development Report 2014. Sustaining Human Progress: Reducing
Vulnerabilities and Building Resilience. United Nations, New York.
van Donkelaar, A., Martin, R.V., Brauer, M., Hsu, N.C., Kahn, R.A., Levy, R.C., et al., 2016.
Global estimates of fine particulate matter using a combined geophysical-statistical
method with information from satellites, models, and monitors. Environ. Sci. Technol.
50, 3762–3772.
Wackernagel, M., Onisto, L., Bello, P., Linares, A.C., Falfán, I.S.L., Garcia, J.M., et al., 1999. Na-
tional natural capital accounting with the ecological footprint concept. Ecol. Econ. 29
(3), 375–390.
Wackernagel, M., Schulz, N.B., Deumling, D., Linares, A.C., Jenkins, M., Kapos, V., et al.,
2002. Tracking the ecological overshoot of the human economy. P. Natl. Acad. Sci.
U. S. A. 99, 9266–9271.
Wang, Y., Hao, X., Fu, J., Sheng, G., 2004. Water quality change in reservoirs of Shenzhen,
China: detection using Landsat TM data. Sci. Total Environ. 328, 195–206.
Wang, K., Franklin, S.E., Guo, X., Cattet, M., 2010. Remote sensing of ecology, biodiversity
and conservation: a review from the perspective of remote sensing specialists. Sen-
sors 10 (11), 9647–9667.
Watts, N., Adger, W.N., Agnolucci, P., Blackstock, J., Byass, P., Cai, W., et al., 2015. Health
and climate change: policy responses to protect public health. Lancet 386,
1861–1914.
Wiedmann, T., Barrett, J., 2010. A review of the ecological footprint indicator—perceptions
and methods. Sustainability 2, 1645–1693.
Wu, J., 2010. Urban sustainability: an inevitable goal of landscape research. Landsc. Ecol.
25, 1–4.
Wu, J., Wu, T., 2012. Sustainability indicators and indices: an overview. In: Christian, N.M.,
Kuei, C. (Eds.), Handbook of Sustainable Management. Imperial College Press, Lon-
don, pp. 65–86.
Xiao, X., Boles, S., Liu, J., Zhuang, D., Liu, M., 2002. Characterization of forest types in North-
eastern China, using multi-temporal SPOT-4 VEGETATION sensor data. Remote Sens.
Environ. 82 (2), 335–348.
Xu, M., He, C., Liu, Z., Dou, Y., 2016. How did urban land expand in China between 1992
and 2015? A multi-scale landscape analysis. PLoS ONE 11, e0154839.
Yang, X., Hou, Y., Chen, B., 2011. Observed surface warming induced by urbanization in
east China. J. Geophys. Res. 116, D14113.
Yang, H., Huang, X., Thompson, J.R., Bright, R.M., Astrup, R., 2016. The crushing weight of
urban waste. Science 351, 674.
Yao, Y., 2016. Pollution: spend more on soil clean-up in China. Nature 533, 469.
Yu, Y., Wen, Z., 2010. Evaluating China' s urban environmental sustainability with data en-
velopment analysis. Ecol. Econ. 69, 1748–1755.
 C. He et al. / Remote Sensing of Environment 193 (2017) 65–75 75

Zeng, P., Chen, F., 2012. Empirical research of “energy-environment-economy” compre-
hensive accounting system of the top ten urban agglomeration in China. Forum. Sci.
Technol. China 8, 107–113 (in Chinese).
Zhang, Q., Geng, G., Wang, S., Richter, A., He, K., 2012. Satellite remote sensing of changes
in NOx emissions over China during 1996–2010. Chin. Sci. Bull. 57, 2857–2864.
Zhao, H., Duan, X., Stewart, B., You, B., Jiang, X., 2013. Spatial correlations between urban-
ization and river water pollution in the heavily polluted area of Taihu lake basin,
China. J. Geogr. Sci. 23, 735–752.
Zhou, H., Rompaey, A.V., Wang, J., 2009. Detecting the impact of the “grain for green” pro-
gram on the mean annual vegetation cover in the Shaanxi province, China using
SPOT-VGT NDVI data. Land Use Policy 26 (4), 954–960.
Zhou, D., Zhao, S., Zhang, L., Sun, G., Liu, Y., 2015. The footprint of urban heat island effect
in China. Sci. Rep. 11160. http://dx.doi.org/10.1038/srep11160.