Improving Urban Resilience through Green Infrastructure: An Integrated Approach for Connectivity Conservation in the Central City of Shenyang, China

Abstract

Green infrastructure (GI) as an operational physical framework is being increasingly recognized as the most cost-effective way to mitigate and adapt to social-ecological challenges through multifunctional ecosystem services. Conserving the connectivity of GI is conducive to maintaining biodiversity and facilitating ecological processes, which contributes to promote urban resilience and implies that urban governance has made a conscious effort to prepare for uncertainties. Though important, there are few studies on operating GI practically to navigate urban resilience. Based on interdisciplinary knowledge and multiple techniques, this study provides an integrated approach, in which relationships between GI connectivity, resilience potential, and conservation strategies are better addressed. The results indicate that significant changes have taken place in terms of the composition, layout, and connectivity of GI in the central city of Shenyang between 1995 and 2015. Through pinch point identification and barrier detection, conservation strategies by protecting key structures, eliminating local barriers, and implementing differentiated measures according to land use types are therefore proposed. The strategies may be helpful for future policy formulation, planning, and management by rehabilitating a GI network to increase urban social and ecological resilience in the study area and other similar megacities. This integrated approach based on a generic process of geometric analysis has general applicability to make interdisciplinary contributions toward urban resilience.

Full text

Research Article
Improving Urban Resilience through Green Infrastructure:
An Integrated Approach for Connectivity Conservation in the
Central City of Shenyang, China
Zhimin Liu,
1
Chunliang Xiu
,
2
and Chao Ye
1
1
School of Geographic Sciences & Institute of Eco-Chongming, East China Normal University, Shanghai 200241, China
2
Jangho Architecture College, Northeastern University, Shenyang 110169, China
Correspondence should be addressed to Chunliang Xiu; xiuchunliang@mail.neu.edu.cn
Received 5 March 2020; Revised 28 April 2020; Accepted 30 April 2020; Published 2 July 2020
Guest Editor: Jun Yang
Copyright ©2020 Zhimin Liu et al. is is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Green infrastructure (GI) as an operational physical framework is being increasingly recognized as the most cost-effective way to
mitigate and adapt to social-ecological challenges through multifunctional ecosystem services. Conserving the connectivity of GI
is conducive to maintaining biodiversity and facilitating ecological processes, which contributes to promote urban resilience and
implies that urban governance has made a conscious effort to prepare for uncertainties. ough important, there are few studies
on operating GI practically to navigate urban resilience. Based on interdisciplinary knowledge and multiple techniques, this study
provides an integrated approach, in which relationships between GI connectivity, resilience potential, and conservation strategies
are better addressed. e results indicate that significant changes have taken place in terms of the composition, layout, and
connectivity of GI in the central city of Shenyang between 1995 and 2015. rough pinch point identification and barrier
detection, conservation strategies by protecting key structures, eliminating local barriers, and implementing differentiated
measures according to land use types are therefore proposed. e strategies may be helpful for future policy formulation,
planning, and management by rehabilitating a GI network to increase urban social and ecological resilience in the study area and
other similar megacities. is integrated approach based on a generic process of geometric analysis has general applicability to
make interdisciplinary contributions toward urban resilience.
1. Introduction
Contemporary cities have undergone extensive strains on
their safety, liveability, and sustainability as they experience
rapid urbanization and aggravated global environmental
change [1]. As earth systems are urbanized at an unsus-
tainable pace, ecosystems and biodiversity undergo large-
scale destruction and human well-being and urban resilience
are now experiencing unprecedented challenges [2]. A
spectrum of acute shocks and chronic stress events such as
extreme heat waves, soil contamination, droughts, floods,
and air pollution, which have steadily escalating effects, are
increasing in frequency. ese challenges create serious
vulnerabilities in cities and heighten the risk of losses
causing a divergence from a sustainable trajectory of de-
velopment [3, 4].
Cities need to create positive interactions with their life-
support systems to build a broad resilience capacity while
undergoing stresses, rather than develop at the expense of
them [5, 6]. In the last two decades, there has been a growing
consensus on the essential role of reconnecting people in
urbanized areas to the biosphere to address current envi-
ronmental challenges, achieve long-term survivability, and
thrive [7]. Nature-based solutions that incorporate natural
capital components to address urban environmental issues
create multiple economic, social, and ecological benefits,
which are now widely acknowledged for cities [8, 9]. Green
infrastructure (GI) has therefore emerged and been identified
as a concept closely related to and embedded within the
framework of nature-based solutions. It provides an integrated
and valuable approach for the transition to an inclusive, re-
silient, and sustainable urban environment [10, 11].
Hindawi
Complexity
Volume 2020, Article ID 1653493, 15 pages
https://doi.org/10.1155/2020/1653493

Moving beyond traditional green spaces, which are
good for urban aesthetics and public health, GI is discussed
more as a cost-effective solution to realign urban areas
toward long-term resilience and sustainability [12–14]. As a
constantly evolving concept, GI is generally viewed as a
strategically planned network of natural and seminatural
areas with other environmental features designed and
managed to generate and deliver a wide range of ecosystem
services under the discourse of sustainability [15, 16].
Although GI provides only a small portion of ecosystem
services, in urbanized areas it has strategic significance in
the key arena of human-nature interdependence. Human
interventions may contribute to urban resilience changes in
the current, human-dominated era of the Anthropocene,
the era when increasingly negative effects of human ac-
tivities on natural systems are aggravated [17]. While GI
has received increased attention in both academic and
practical fields in last decades, there is a lack of clarity
around the role that planning and management can play in
operationalizing GI to enhance urban resilience and
translating that effort into future policy.
Regardless of differentiated project types and local set-
tings, GI is firmly characterized by its multifunctionality and
connectivity, which reflects on how it may be used and its
capacity for the provisioning of ecosystem services [18].
Multifunctionality is the primary mechanism for GI con-
tributing to urban resilience, making it possible to perform
in a flexible way through diversified ecosystem services to
respond to uncertainties [19]. Since it is not economically
feasible to cover large sections of urban areas with GI,
coupled with the growing habitat fragmentation by human
activities during the rapid urbanization process, connectivity
plays a critical role in maintaining GI function. Good
connectivity is conducive in alleviating the isolation effect
and maintaining ecological processes, and thus, securing
sustainable ecosystem services effectively [20]. Mounting
evidence for the provision of ecosystem services to improve
resilience suggests the need for better understanding and
sustaining GI connectivity [21, 22].
Connectivity represents the degree to which an activity
impedes or facilitates ecological processes, e.g., movement of
species and gene flow among habitat patches, and deter-
mines the proportion of the total habitat area that may be
reached and is available in a landscape [23]. Studies suggest
that the effects of GI for connectivity may directly influence
the magnitude of species movement and indirectly influence
biodiversity and ecosystem functions that GI contains [24].
Appropriate connectivity can have a positive influence on
biodiversity protection and resilient ecosystem service
provision at each location, while decreased connectivity is
certain to lead to negative effects on the integrity and sta-
bility of ecosystem [25–27]. Two critical aspects of con-
nectivity, structural and functional connectivity, are both of
significance in an urban landscape context. Structural
connectivity describes the physical relationships between the
components of GI networks, while functional connectivity
conversely concerns the behavioural response of biological
species to the physical structure [28]. Although they are not
necessarily related, it is commonly agreed that functional
connectivity will increase when an improvement in the
landscape structure allows for the movement or flow of
species across the landscape. Furthermore, functional con-
nectivity plays a key role in guaranteeing the provision of
ecologically valuable services [29, 30].
Relevant literatures indicate that the significance of
urban GI connectivity is largely reflected in its promotion of
biodiversity within the site, the movement of organisms, and
ecosystem functions [19, 22, 24]. To achieve sustainability in
cities, the research and application of GI should focus on the
overall improvement of GI connectivity [31]. Notwith-
standing, a knowledge gap still exists with respect to con-
nectivity modeling of GI networks and their performance for
optimization in real world practice. is deficiency may
hinder effective intervention in planning and management
of urban GI to improve urban resilience [32]. In terms of
connectivity representation, landscape indices and GIS-
based methods which can only extract GI network elements,
that is, nodes and corridors separately for analysis are
common. ere is no doubt that these methods make it
unlikely for us to intuitively and comprehensively under-
stand the structure and function of a GI network and create
obstacles for comprehensive suggestions to improve GI
connectivity. Due to the complexity of GI networks, an
integrated approach involving multidisciplinary knowledge
and multidomain technology is needed to quantitatively
simulate the connectivity of GI networks and identify
conservation priorities, making reasonable and multiaspect
suggestions for strategic GI network and habitat conser-
vation planning [33].
is study focuses on three specific issues. e first is GI
network construction and connectivity modeling, the sec-
ond is the identification of pinch points and barriers of the
GI network, and the last is proposing a connectivity con-
servation strategy considering the first two. e remainder of
this study is structured as follows. Section 2 describes the
study area, data processing, and the associated methods. e
results of the progression, pinch points, and barriers of
urban GI corridors, as well as the related analysis, are
presented in Section 3. Section 4 proposes some pertinent
strategies for connectivity conservation. Section 5 further
discusses the advances and limitations of this research. e
concluding remarks are summarized in Section 6.
2. Data and Methods
2.1. Study Area. Located in the middle of the Liaohe Plain
from 41°11′N to 43°02′N and from 122°25′E to 123°48′E,
Shenyang is the capital city of the Liaoning Province. e
city has 8.3 million permanent residents (within a total area
of 12,860 km
2
) with 5.3 million residents living in the
central city (1353 km
2
) as of 2015. e elevation of the city’s
landscape decreases from the northeast to southwest, and
2Complexity

the terrain varies from hilly mountains to alluvial plains. e
average elevation is approximately 41.45m above sea level (see
Figure 1). It has a temperate continental climate with an
average annual temperature of 6.2–9.7°C. Nearly 27 rivers,
including the Liaohe and Hunhe, run through its territory.
2.2. Data Source and Processing. e datasets collected for
this study consist of Landsat remote sensing images (at a
resolution of 30 m in 1995 and 2015), Google Earth images
(at a resolution of 14 m and 3 m in 1995 and in 2015), 1:1000
land use maps from the urban master plan (1996–2010), and
its revised version (2016). For data processing, we first
conducted preprocessing of the two-stage Landsat images.
e preprocessed images were then interpreted in the way of
human-computer interaction. According to “e Code of
Current Land Use Classification” (GB/T 21010-2017), the
land use of the study area is classified into five types: ag-
ricultural land, forest land, meadow land, built-up land, and
water area and wetland with verified accuracy (a precision of
over 90%). Second, registration and vectorization were
performed on the Google Earth images and land use maps.
en, the green space in the urbanized area was identified
and extracted by referring to “e Code for Classification of
Urban Land Use and the Planning Standards of Develop-
ment Land” (GB50137-2011). Finally, the subdivided urban
green space (such as urban parks and green belts) was added
to the initial classification result, and a total of five cate-
gories, along with urban green space, including natural and
seminatural elements capable of providing ecosystem ser-
vices were obtained and recognized as GI, while the rest were
designated as background objects. e results of land use
classification of the study area are shown in Figure 2.
2.3. Methods
2.3.1. Morphological Spatial Pattern Analysis (MSPA).
Morphological Spatial Pattern Analysis (MSPA) originated
in mathematical morphology and was later introduced into
the science of image analysis and landscape ecology. is
method is often used to conduct segmentation on an input
binary map and provide a generic pattern analysis frame-
work for detecting and measuring the morphometric aspects
through quantitative analysis of the shape, connectivity, and
spatial arrangement of digital images [34]. Based on geo-
metric algorithms such as corrosion, expansion, opening,
and/or closing operations, it generally divides the fore-
ground objects (objects of interest) of the binary map into
seven mutually exclusive morphological feature classes: core,
edge, perforation, bridge, loop, branch, and islet [35] (see
Figure 3). In this study, all GI elements are defined as a
foreground area, whilst urban construction land (excluding
urban green land) is recognized as the background. Only the
foreground was incorporated into the MSPA segmentation.
In contrast to the traditional landscape pattern analysis
methods (e.g., landscape indices), the results of the MSPA
identify the areas that are important to the flow of material
and energy at the pixel level. Moreover, it has additional
advantages such as avoiding the shortcomings of
redundancy and high repeatability of landscape features and
displaying the segmentation results in a raster format ex-
plicitly, which can further be implemented to inform the
spatial planning practice.
2.3.2. GI Network Building. e GI network in this study
was constructed in four steps (see Figure 4): first, identify
adjacent nodes; second, build a network using adjacent
nodes and distance data between the nodes; third, calculate
cost-weighted distances; and fourth, calculate and mosaic
least-cost corridors. e input data and necessary calcula-
tions involved in the network construction process are
explained below.
(1) Identify the nodes: based on the MSPA results, the
importance of all the cores was evaluated to deter-
mine the nodes of the GI network. Since the per-
centage of integral index of connectivity value loss
(dIIC) and the percentage of probability of con-
nectivity index value loss (dPC) can be used to
quantify the relative variations of the overall con-
nectivity in the case when a particular patch is
missing, the two indicators were thus introduced to
examine the contribution of every single core area in
terms of its significance for maintaining and en-
hancing the overall connectivity of the GI network in
this study. e larger the values of dIIC and dPC, the
more important are the patches that they represent to
the GI network [37]. e equations of the indicators
and corresponding interpretations are shown as
follows [38]:
dIIC �IIC −IIC′
IIC ×100,
IIC �􏽐n
i�1􏽐n
j�1aiaj/ 1 +nlij
􏼐 􏼑􏽨 􏽩
A2
L
,
(1)
where dIIC quantifies the importance of a particular
green patch to maintain the IIC in the GI network, IIC
is the integral index of connectivity, IIC′is the value of
IIC after the removal of the patch area from the
landscape, nis the total number of green patches, aiand
ajare attributes of patch iand j,nlij is the number of
links in the shortest path (topological distance) between
patch iand j, and ALis the total area of GI. In general,
0≪dIIC ≪1. When all the green patches are not
connected, dIIC equals zero. dIIC equals 1 in the case
when the total landscape is occupied by GI:
dPC �PC −PC′
PC ×100%,
PC �􏽐n
i�1􏽐n
j�1aiajp∗
ij
A2
L
,
(2)
where dPC quantifies the importance of a particular
green patch to maintain PC in the GI network, PC is
Complexity 3

probability of connectivity, PC′is the value of PC
after the removal of the patch area from the land-
scape, p∗
ij is the maximum product probability of all
paths between patch iand j, and the meanings of
other parameters are the same as above. e value of
dPC ranges from 0 to 1 [33].
(2) Create a resistance raster: a resistance raster is needed
to model the impacts of landscape surfaces such as the
difficulty, energetic cost, and mortality risk on eco-
logical processes. Ecological processes to provide
ecosystem services are generally affected by the ac-
cumulated resistance of all the landscape types in-
volved. In the study area, the resistance mainly comes
from the artificial and built surfaces, so it is expected to
be large in the main urban area and may become quite
small in the urban periphery. e resistance values of
different landscape types were assigned through re-
ferring to the similar study [39] (see Table 1).
(3) Specify Euclidean distance between the nodes: a text
file specifying Euclidean distances between all core
pairs can be obtained by using the Conefor Inputs
tool, taking the results of the MSPA as input data. e
parameters were set to make sure that only the dis-
tance between the core areas is less than 1000 m which
can be viewed as connected, and not vice versa.
2.3.3. Pinch Point Analysis. Efforts to increase urban
resilience through a GI-based approach emphasize the
importance of the functional connectivity within the GI
network. Applying circuit theory into the research makes it
incorporate functional connectivity analysis in the GI net-
work exploration effectively [40]. rough abstracting the
heterogeneous landscape to the conductive surface and the
species or genes to electrons, ecological processes such as
migration and diffusion of species or genes across complex
landscapes according to the characteristics of the random
walk of electrons in a circuit were simulated [41]. In specific
simulation, some nodes are grounded while the current is
inputted to the other nodes, and the current density between
the nodes may represent the magnitude of diffusion prob-
ability of electrons along a certain path, and the areas with
obviously large current densities are defined as pinch points
[42, 43].
2.3.4. Barrier Detection and Improvement. Barriers are
landscape features that impede ecological processes between
ecological areas, and the removal of whom would signifi-
cantly improve the connectivity potential of the whole
network. Detecting and identifying the influential barriers
would quantitatively provide an alternative restoration
option for practical conservation investments. In addition to
focusing on maintaining pinch points from a point of view,
removing barriers is a more active but neglected way to
promote GI network connectivity. Specifically, based on the
modeled least-cost corridors, the barriers and the corre-
sponding improvement scores are obtained. Improvement
score quantifies the extent to which conservations can be
South China Sea
Diaoyu Islands
DEM
High: 237
Low: 20
123°10′E 123°20′E 123°30′E 123°40′E
123°10′E123°0′E 123°20′E
41°30′E41°40′E 41°50′E 42°0′E
41°40′E 41°50′E 42°0′E
Study area
Figure 1: Location of the study area.
4Complexity

expected to improve connectivity. Restoration of the barriers
according to improvement scores will greatly lead to the
reduction of moving difficulty while increasing the con-
nectivity of the total landscape. A larger value of the im-
provement score indicates higher connectivity improvement
after removing the barriers [44].
3. Results and Analysis
3.1. Pattern and Evolution of Urban GI. Significant changes
have taken place in terms of the quantity, components, and
layout of GI in the central city of Shenyang between 1995
and 2015 (see Table 2 and Figure 5). e overall area of GI,
which accounts for 64.27% of the total study area in 1995 and
for 35.08% of the total study area in 2015, decreased by
nearly 38.76% over last 20 years. Although not significant in
the overall area, cultivated (agricultural) land still represents
the dominant type of land for GI. According to the results of
the MSPA of the study area, specifically, core takes the
largest proportion amongst the associated types for GI.
However, due to apparent scale heterogeneity, most of the
cores are too small, indicating that there are limited core
patches suitable for acting as nodes of a GI network. Except
for cores, the areas of perforation and edge are also in large
proportions. Over the past 20 years of development, the area
of perforation has decreased, while that of edge has sig-
nificantly increased. Additionally, the shares of other
morphological types, i.e., Bridge, Loop, Branch, and Islet, to
varying degrees, have all increased by 2015 despite the fact
that they remain in smaller ratios when compared with the
core, edge, or perforation.
Regarding spatial distribution, cores which are mainly
made up of cultivated land and major urban parks are
situated contiguously in the periphery and scattered across
the central urban area. In terms of variation, other than the
concentrated built-up area, the contiguous GI patches are
either destroyed and/or divided, leading to the core patches
with reduced area and fragmented form. However, moderate
connectivity is still retained because of the growth of loop,
bridge, and branch, which may provide effective connec-
tions. Meanwhile, the number of edge and perforation
obviously increased due to the fragmentation of the cores,
which may limit the ecological processes for the fringe effects
that these morphological types cause to a certain extent.
1995 2015
0 2.5 5 10km
EW
S
N
Industrial land
Commercial and business land
Residential land
Agricultural land
Forest land
Road and transportation land
Administration and public services land
Meadow land
Urban greenspace and square land
Logistics and warehouse land
Public utilities land
Other construction lands
Water area and wetland
Figure 2: Land use patterns of the study area.
Complexity 5

3.2. Features of GI Network Structure. Both the 1136 cores
with unique identification information dataset and a text file
of distances between the cores were imported to recognize
the nodes of the GI network. e output of dIIC and dPC
were sorted by value, and consequently, 31 cores were se-
lected as the nodes of the GI network (see Table 3). All the 31
patches occupy approximately 24.74% of the total study area.
Moreover, based on the cost-weighted distance and least-
cost path calculation, 43 least-cost paths were generated. e
GI network which includes 31 nodes and 43 potential paths
is identified and shown in Figure 6.
e pattern of the GI network illustrates that the spatial
distribution of nodes and paths is uneven with the nodes
primarily located in the northside and southside of the city
area and the corridors concentrated on the middle part
communicating the nodes on both sides. In the central area,
GI elements are insufficient or missing and are also greater
distances apart. Considering the characteristics of the nodes,
the number and size of nodes in the main urban area are
deficient except for uneven distribution, and the fragmen-
tation within the interior of node patches is quite evident. In
terms of the paths, the lengths of the paths are comparatively
short, ranging from 0.03 km to 11.6 km. e paths that are
longer than 5 km account for approximately 12% of the
paths, and those with lengths greater than 1 km approxi-
mately account for 40% of the total paths. Furthermore, the
longest paths are primarily aligned southwest to northeast in
the main urban area where node density is low. ese central
paths link up a majority of the existing important ecological
sources such as the Century Golf Club, Tiexi Forest Park,
Labor Park, Zhao Mausoleum Park, Huanggu Hero Park,
and several other parks along the Hunhe river. However, the
shorter paths connect the smaller nodes in the main urban
area as well as those in larger size near the periphery.
3.3. Pinch Point Identification and Protection. e connec-
tivity characteristics are further explored quantitatively
based on circuit theory, and the pinch points in the corridors
of GI network are presented in Figure 7. In the figure, the
locations with darker yellow have a greater contribution to
functional connectivity and can be determined as being the
pinch points of the GI network. e location where the color
transitions from blue to yellow suggests increased current
density from small (blue) to large (yellow). e pinch points
are mainly located in certain long but narrow corridors in
Input: binary map
Foreground
Background
Output: MSPA segmentation results
MSPA foreground classes
Core: interior area
excluding perimeter
Islet: disjoint and too
small to contain core
Loop: connected to the
same core area
Bridge: connected to
dierent core areas
Perforation: internal
object perimeter
Edge: external object
perimeter
Branch: connected at one
end to edge, perforation,
bridge, or loop
MSPA background classes
Background
Border-opening: along edge
Core-opening: within perforation
Figure 3: e MSPA segmentation (source: [35]).
6Complexity

central locations of the main urban area. ey represent
green belts with good connectivity along the main roads and
rivers. However, in the south-western and north-eastern
parts of the study area, current density is low and pinch
points can rarely be seen, which indicate poorer functional
connectivity.
e identification and analysis of pinch points may offer
informative suggestions for connectivity conservation. e
areas with stronger yellow should be given priority for
protection to promote necessary ecological processes that
provide ecosystem services. Moreover, the analysis also finds
that the nature of a node can affect the pinch point qualities.
If the size and quality of GI nodes on both sides of the
corridors are appropriate, the pinch point in the corridors is
obvious. Conversely, if the nodes are in small scale and poor
quality, the pinch point is quite inconspicuous.
Resistance
raster
Map of core areas
to connect
Text file of distances
between core areas
(1) Identify adjacent
nodes
(2) Construct a network
(3) Calculate cost-weighted
distances and least-cost paths
(4) Mosaic least-cost corridors
Figure 4: Schematic diagram for GI network building (source: reference [36]).
Table 1: Resistance values for different landscape types.
Type Core Bridge Islet Edge Loop Branch Perforation Background
Resistance 1 10 15 30 30 60 80 100
Table 2: Statistical results for classification of GI landscape based on MSPA.
1995 2015
Area (km
2
) Ratio to GI (%) Ratio to the total area (%) Area (km
2
) Ratio to GI (%) Ratio to the total area (%)
Core 746.49 94.73 64.27 407.51 84.45 35.08
Islet 0.63 0.08 0.05 4.75 0.98 0.41
Perforation 24.85 3.15 2.16 6.82 1.41 0.59
Edge 12.76 1.62 1.15 49.29 10.21 4.51
Loop 0.52 0.06 0.03 1.51 0.31 0.07
Bridge 1.01 0.12 0.04 5.19 1.07 0.24
Branch 1.76 0.22 0.15 7.48 1.55 0.64
Complexity 7

3.4. Barrier Diagnosis and Connectivity Improvement.
Investigation through identifying pinch point locations
highlights the crucial role of areas that need to be protected
in order to maintain present connectivity. Moving one step
beyond, detecting current barriers will provide an alternative
restoration option since the reduction of barriers implies a
practical strategy for connectivity optimization. e con-
sequences of barrier analysis, as shown in Figure 8, identify
the areas that can be modified for improved connectivity
within the GI network. Due to the large-scale destruction by
man-made construction, especially by industrial activities
during the process of rapid urbanization, most of the GI
corridors need to be restored to improve the connectivity of
the GI network.
In general, most of the identified barriers are distributed
between the geographically isolated nodes. e area of all the
barriers is 19.01 km
2
accounting for about 3.94% of the total
area of GI. e improvement score which gives expected
reduction in least-cost distance per unit distance restored
suggests the priority of conservation. Based on the im-
provement scores, barriers to be optimized can be divided
into two levels according to the rule of Natural Breaks. e
key improvement areas are those whose improvement scores
range from 53.58 to 99.00 and the general improvement area
are those with improvement scores between 0 and 53.58.
Spatially, the key improvement areas are mainly distributed
in the Shenxi Development Zone and the northeast side of
Beiling Park, while the general improvement areas are
mostly located around the key improvement areas in a band-
shaped manner with the scope depending on the quality and
quantity of GI nodes.
4. Strategies for Connectivity Conservation
Since critical locations for GI connectivity in the study area
were obtained through pinch point identification and barrier
detection, we then proposed the following strategies for
connectivity conservation in our study area.
4.1. Protect Key Areas to Maintain the Overall Connectivity.
Corridors with currently high densities and the connected
node patches deserve attention and should be considered a
priority for conservation (see Figure 7). Moreover, to ensure
overall connectivity, areas with poor connectivity, especially
in the central urban area need to add more dotted green
patches (also called steppingstones) to facilitate the regen-
eration of ecological processes. Meanwhile, attention should
Perforation
Edge
Bridge
Core
02.5 510km
1995 2015
EW
S
N
Loop
Islet
Background
Figure 5: Landscape pattern of GI according to MSPA.
8Complexity

be paid to improve the quality of the GI structure such as
widening the corridors on the western side of the city that
connect the southern and northern farmlands and restoring
vegetation coverage near large non-GI areas. It also should
be noted that the conservation measures advocate scientific
and diversified methods to incorporate the complex
adaptability of natural ecological spaces rather than sug-
gesting a total ban on construction. For example, in pinch
point areas, construction activities and development may be
controlled by limiting the volume ratio and specifying the
green rate. Conversely, in nonpinch point areas, facilities
with different functions can be conditionally constructed to
form a composite green infrastructure network pattern.
4.2. Repair Barriers to Enhance Connectivity. e barriers
should be removed or reduced one by one, according to their
importance and urgency, from the key improvement areas to
the general improvement areas. More importantly, green
vegetation should be added at the locations where barriers
are removed in order to restore their capacity for ecosystem
service provision. In addition, other options to improve
connectivity can also be considered in the light of local
conditions if a barrier is difficult to remove such as by
opening new routes between the nodes near the original
barriers [45]. e removal of barriers and implementation of
GI could improve the use of public space, activate pedestrian
paths, and perform more leisure and health services within
the city in addition to being beneficial to the maintenance of
ecological functions. ese benefits are particularly evident
in the main urban area and the Shenxi Industrial Corridor in
the southwest part, as well as the Huishan Economic and
Technological Development Zone near the north-eastern
part.
4.3. Implement Differentiated Measures to Optimize
Connectivity. To achieve better connectivity, different
measures can be considered for each type of urban land use.
For example, residential lands should have green areas
added to the barrier areas to expand the coverage of GI and
increase the richness of vegetation [46]. As for the barriers
around the municipal roads, increasing the number of green
patches to broaden greening is suggested in combination
with the construction of biological channels combined with
other GI corridors. With respect to the barriers in com-
mercial areas, the degree of vegetation can be appropriately
increased. Additionally, for the barriers near the rivers, the
protective green space on both sides of the rivers should be
widened.
5. Discussion
5.1. e Significance of the Integrated Approach to Promote
Urban Resilience. Traditional urban risk governance prefers
rigid engineering solutions focus on addressing determin-
istic disasters, which are lagging current experience and are
unsustainable, and may even aggravate the sensitivity and
vulnerability of urban systems [47]. e efforts to introduce
(social-ecological) resilience theory into urban systems
provide a new framework for urban disaster governance,
which suggests that cities need to improve their awareness of
the influence and dependence of social factors on ecological
conditions and also increase their attention to the ecological
factors [48, 49]. GI provides a feasible tool to improve urban
resilience in practice and has attained increased importance
for the fields of urban development, planning, and man-
agement. e complexity of urban resilience may be exac-
erbated by the future omission of green solutions and
adopting GI as an adaptive response and a cost-effective
strategy with both short-term and long-term benefits
[15, 50].
To make up for the lack of systematic research on ap-
plication of GI in practice to increase urban resilience, this
study provides a practical and integrated approach for
navigating urban resilience through the operation of GI, in
which the relationships between GI connectivity, resilience
potential, and connectivity conservation are better addressed
within the context of climate change and rapid urbanization.
Intervention in GI through connectivity conservation rep-
resents a consciousness that humans are preparing for
uncertainty. e above is especially relevant in the inte-
gration of multiple technologies to explore GI connectivity,
the involvement of functional connectivity in the quanti-
tative analysis of GI connectivity, and the planning and
Table 3: e importance and ranking of the nodes of GI network.
Patch ID dIIC dPC
1 76.16 77.12
783 42.15 49.51
431 14.92 19.28
758 1.35 2.59
66 1.23 1.67
913 1.12 1.53
873 0.75 1.02
778 0.10 0.87
251 0.59 0.80
809 0.05 0.79
164 0.46 0.61
910 0.30 0.40
4 0.28 0.38
226 0.28 0.37
596 0.26 0.35
390 0.19 0.27
115 0.18 0.24
1113 0.17 0.24
1013 0.12 0.17
489 0.11 0.15
829 0.10 0.14
796 0.00 0.14
914 0.10 0.14
866 0.10 0.13
781 0.10 0.13
490 0.09 0.13
777 0.09 0.12
227 0.09 0.12
593 0.09 0.11
699 0.08 0.10
582 0.08 0.10
Complexity 9

management guidance for Shenyang and other similar
megacities in mitigating and adapting to climate change
through nature-based solutions, thereby expanding the
boundaries of traditional research [51–53].
is study explores an integrated approach that depends
on openly and universally available technological methods to
foster the interdisciplinary contribution to reshape urban
resilience through the extraction and quantification of
properties that are contained in raster maps. e conceptual
basis of the involved analysis is of a generic nature, for it is
based on the process of geometric analysis. Since it is im-
practical to observe functional corridors for movement, the
analysis of functional connectivity through simulations
provides an alternative and objective recognition of con-
nectivity for species movement in real landscapes. e
spatially explicit mapping of the results and the conversions
between the storage formats of them contribute to this in-
tegration research and practical applications of GI network
connectivity preservation. Nevertheless, the mathematical
models offer the possibility for establishing potential con-
nectivity among patches rather than the preferred pathways
used to successfully move between patches [54, 55]. erefore,
this integrated approach can be applied to other empirical
studies, with the same input data types but specific parameter
settings according to targeted research needs.
5.2. e Implication of Rapid Urbanization on the Changes of
GI. As a complex and open system, urban GI is largely
affected by urban dynamics and its spatial pattern is
High
Low
Cost-weighted distance Node
Path
02.5 510km
EW
S
N
Figure 6: GI network based on least-cost paths.
10 Complexity

constantly changing [56–58]. e large-scale artificial de-
velopment and construction during the process of rapid
urbanization between 1995 and 2015 have resulted in a large
amount of GI components in the central city of Shenyang
being swallowed up such that the total area of GI has been
significantly reduced and has become more complicated and
less connected (see Table 4). e agricultural belt in the
suburban areas is seriously damaged, and the GI elements in
the main urban area tend to be gradually homogenized,
while in the peripheral area they are fragmented. is is
particularly evident in the Shenxi Industrial Corridor in the
southwest part and in the Huishan Economic and
Technological Development Zone near the north-eastern
edge. e process of urbanization has significantly changed
the overall pattern and function of GI.
e area and area ratio of the core and perforation are
significantly reduced, while those of the bridge and branch
are increased suggesting that the large ecological “source
area” is divided into small patches and the concentrated
urban construction area is connected with the core area
through the street green belt. ese small-scale GI com-
ponents to a certain extent help to improving the local
environment of the city, provide leisure places for the cit-
izens, and more importantly provide for the possibility of the
0 2.5 5 10km
Node
Current density
High
Low
EW
S
N
Figure 7: Pinch points in the corridors of GI network.
Complexity 11

development of GI network [59, 60]. In this case, the green
belt plays a key role in maintaining the total city’s social-
ecological resilience, and it is also important in conserving
and optimizing the whole GI network. erefore, it can be
seen that the fragmentation of GI components caused by
the rapid urbanization is also the cause for their evolution
from large-scale development and centralization toward
the direction of the network. e urban GI can be guided to
a more resilient trajectory and made conducive for the
improvement of urban resilience if the opportunity can be
seized and scientifically reasonable protection measures
are taken [61]. is radically changes our understanding of
a larger green space being more suitable for an optimized
pattern and function contributing toward the resilient
development of the city. Contrarily, we achieve a new
realization that suggests that, with the development of an
urban GI system, green patches tend to form a network
pattern through “self-organization” with diversified spatial
structures. e network pattern of GI is the more resilient
and desirable structural pattern. erefore, the interven-
tion of urban planning and governance should focus on the
differentiated protection and optimization of the bridge,
branch, and edge.
02.5 510km
Node
Improvement score
0–53.53
53.54–99
EW
S
N
Figure 8: Improvement areas based on barrier analysis of GI network.
Table 4: e overall connectivity of the landscape.
1995 2015
IIC 0.28 0.09
PC 0.54 0.29
12 Complexity

5.3. Limitations. Despite these advances, as a specific work,
there are some limitations. First, the MSPA method is ex-
tremely sensitive to the scale and granularity of the landscape.
With the changes of scale and granularity, the number and
size of the seven types of GI components from the MSPA will
respond in a nonlinear way. In this case, the way in which the
appropriate distance threshold, edge width, and other pa-
rameters are set may affect the scientific rationality of the
results and the feasibility of practical application. Second, this
study only investigates the impact of corridor barriers and
improvement measures, that is, local GI structures and their
impacts on connectivity. e connectivity optimization
measures of the elements’ whole network are worthy of
further consideration. Finally, some parameters in the se-
lection process of GI components, connectivity distance
threshold setting, and pinch point simulation are referenced
to the empirical value of the existing research. erefore, it is
recommended that, in further research and practice the
relevant parameters should be tailored according to the
particularities of specific study areas to improve the overall
effectiveness of GI connectivity conservation.
6. Conclusions
To enhance urban resilience and sustainability within the
framework of the social-ecological system, this study utilizes
interdisciplinary knowledge and multiple analysis techniques
by taking the central city of Shenyang as an example to ex-
plore an integrated approach through navigating GI. During
the past 20 years (1995–2015), the changes in the pattern of GI
have shown an increase in fragmentation and overall decrease
in GI connectivity in the central city of Shenyang. A GI
network based on the least-cost path algorithm is constructed
to identify significant structures in connectivity conservation.
Based on circuit theory, the pinch points are highlighted as
critical to maintain the present connected state of GI com-
ponents, while barrier diagnosis provide an alternative res-
toration option for connectivity optimization through
governing the improvement areas. For the practical conser-
vation strategy, protecting key areas to maintain the overall
connectivity, repairing barriers to enhance connectivity, and
implementing differentiated measurements in the light of
local condition to optimize connectivity are underscored. It is
proposed that, with the development of an urban GI system,
large-scale green patches tend to be fragmentized and re-
formed into a network pattern with a diversified spatial
structure. In this case scientific planning and governance
interventions through protection and optimization of con-
nective structures might guide the development of urban GI
toward a more resilient trajectory. e results may provide the
basis for future policy formulation, planning, and design
practice by installing or rehabilitating a GI network such that
it provides a conducive situation for the improvement of
urban social and ecological resilience. is integrated ap-
proach that is based on a generic process of geometric analysis
can be applied universally to foster its interdisciplinary
contribution toward improving urban resilience through GI
network connectivity conservation.
Data Availability
e land-use data used in this study can be downloaded
from the websites https://www.usgs.gov/ and https://www.
earth.google.com/.
Conflicts of Interest
e authors declare that there are no conflicts of interest
regarding the publication of this paper.
Acknowledgments
is study was funded by the National Natural Science
Foundation of China (Grant nos. 41871162 and 41471141).
References
[1] T. Elmqvist, E. Andersson, N. Frantzeskaki et al., “Sustain-
ability and resilience for transformation in the urban cen-
tury,” Nature Sustainability, vol. 2, no. 4, pp. 267–273, 2019.
[2] Z. Liu, C. Xiu, and W. Song, “Landscape-based assessment of
urban resilience and its evolution: a case study of the central
city of Shenyang,” Sustainability, vol. 11, no. 10, p. 2964, 2019.
[3] S. N. Jonkman, B. Maaskant, E. Boyd, and M. L. Levitan, “Loss
of life caused by the flooding of New Orleans after hurricane
Katrina: analysis of the relationship between flood charac-
teristics and mortality,” Risk Analysis, vol. 29, no. 5,
pp. 676–698, 2009.
[4] J. Colding and S. Barthel, “Exploring the social-ecological
systems discourse 20 years later,” Ecology and Society, vol. 24,
no. 1, p. 2, 2019.
[5] Y. Jabareen, “Planning the resilient city: concepts and strat-
egies for coping with climate change and environmental risk,”
Cities, vol. 31, pp. 220–229, 2013.
[6] N. S. Wigginton, J. Fahrenkamp-Uppenbrink, B. Wible, and
D. Malakoff, “Cities are the future,” Science, vol. 352, no. 6288,
pp. 904-905, 2016.
[7] E. Andersson, S. Barthel, S. Borgstr¨
om et al., “Reconnecting
cities to the biosphere: stewardship of green infrastructure
and urban ecosystem services,” AMBIO, vol. 43, no. 4,
pp. 445–453, 2014.
[8] N. Frantzeskaki, N. Kabisch, and T. McPhearson, “Advancing
urban environmental governance: understanding theories,
practices and processes shaping urban sustainability and
resilience,” Environmental Science & Policy, vol. 62, pp. 1–6,
2016.
[9] W. Steffen, K. Richardson, J. Rockstrom et al., “Planetary
boundaries: guiding human development on a changing
planet,” Science, vol. 347, no. 6223, Article ID 1259855, 2015.
[10] S. T. Lovell and J. R. Taylor, “Supplying urban ecosystem
services through multifunctional green infrastructure in the
United States,” Landscape Ecology, vol. 28, no. 8, pp. 1447–
1463, 2013.
[11] M. R. Masnavi, F. Gharai, and M. Hajibandeh, “Exploring
urban resilience thinking for its application in urban plan-
ning: a review of literature,” International Journal of Envi-
ronmental Science and Technology, vol. 16, no. 1, pp. 567–582,
2019.
[12] M. J. McDonnell and I. MacGregor-Fors, “e ecological
future of cities,” Science, vol. 352, no. 6288, pp. 936–938, 2016.
[13] J. Yang, C. Li, Y. Li, J. Xi, Q. Ge, and X. Li, “Urban green space,
uneven development and accessibility: a case of Dalian’s
Complexity 13

Xigang district,” Chinese Geographical Science, vol. 25, no. 5,
pp. 644–656, 2015.
[14] Y. Wang, X. Li, J. Li, Z. Huang, and R. Xiao, “Impact of rapid
urbanization on vulnerability of land system from complex
networks view: a methodological approach,” Complexity,
vol. 2018, Article ID 8561675, 18 pages, 2018.
[15] P. Romero-Lankao, D. Gnatz, O. Wilhelmi, and M. Hayden,
“Urban sustainability and resilience: from theory to practice,”
Sustainability, vol. 8, no. 12, p. 1224, 2016.
[16] A. A. Zuniga-Teran, C. Staddon, L. de Vito et al., “Challenges
of mainstreaming green infrastructure in built environment
professions,” Journal of Environmental Planning and Man-
agement, vol. 63, no. 4, pp. 710–732, 2020.
[17] S. S. Myers, “Planetary health: protecting human health on a
rapidly changing planet,” e Lancet, vol. 390, p. 10114, 2017.
[18] D. L. Badiu, A. Nita, C. I. Ioj˘
a, and M. R. Nit¸˘
a, “Disentangling
the connections: a network analysis of approaches to urban
green infrastructure,” Urban Forestry and Urban Greening,
vol. 41, 2019.
[19] R. Ramyar and E. Zarghami, “Green infrastructure contri-
bution for climate change adaptation in urban landscape
context,” Applied Ecology and Environmental Research,
vol. 15, no. 3, pp. 1193–1209, 2017.
[20] C. Liquete, S. Kleeschulte, G. Dige et al., “Mapping green
infrastructure based on ecosystem services and ecological
networks: a Pan-European case study,” Environmental Science
& Policy, vol. 54, pp. 268–280, 2015.
[21] B. F. Nero, “Urban green space dynamics and socio-envi-
ronmental inequity: multi-resolution and spatiotemporal data
analysis of Kumasi, Ghana,” International Journal of Remote
Sensing, vol. 38, no. 23, pp. 6993–7020, 2017.
[22] A. Sylwester, Green Infrastructure Supporting Connectivity,
Maintaining Sustainability, European Commision, DG En-
vironment, Brussels, Belgium, 2009.
[23] M. G. E. Mitchell, E. M. Bennett, and A. Gonzalez, “Linking
landscape connectivity and ecosystem service provision:
current knowledge and research gaps,” Ecosystems, vol. 16,
no. 5, pp. 894–908, 2013.
[24] E. Berte and T. Panagopoulos, “Enhancing city resilience to
climate change by means of ecosystem services improvement:
a SWOT analysis for the city of Faro, Portugal,” International
Journal of Urban Sustainable Development, vol. 6, no. 2,
pp. 241–253, 2014.
[25] A. S. Hadley and M. G. Betts, “e effects of landscape
fragmentation on pollination dynamics: absence of evidence
not evidence of absence,” Biological Reviews, vol. 87, no. 3,
pp. 526–544, 2011.
[26] J. Yang, Y. Guan, J. Xia, C. Jin, and X. Li, “Spatiotemporal
variation characteristics of green space ecosystem service
value at urban fringes: a case study on Ganjingzi District in
Dalian, China,” Science of the Total Environment, vol. 639,
pp. 1453–1461, 2018.
[27] J. Yang, J. Sun, Q. Ge, and X. Li, “Assessing the impacts of
urbanization-associated green space on urban land surface
temperature: a case study of Dalian, China,” Urban Forestry &
Urban Greening, vol. 22, pp. 1–10, 2017.
[28] G. Sanesi, G. Colangelo, R. Lafortezza, E. Calvo, and
C. Davies, “Urban green infrastructure and urban forests: a
case study of the metropolitan area of Milan,” Landscape
Research, vol. 42, no. 4, pp. 1–12, 2017.
[29] J. Zhang, X. Qi, and C. Liang, “Tackling complexity in green
contractor selection for mega infrastructure projects: a hes-
itant fuzzy linguistic MADM approach with considering
group attitudinal character and attributes’ interdependency,”
Complexity, vol. 201831 pages, 2018.
[30] J. M. R. Benayas, J. M. Bullock, and A. C. Newton, “Creating
woodland islets to reconcile ecological restoration, conser-
vation, and agricultural land use,” Frontiers in Ecology and the
Environment, vol. 6, no. 6, pp. 329–336, 2008.
[31] S. Pauleit, E. Andersson, B. Anton et al., “Urban green infra-
structure—connecting people and nature for sustainable cities,”
Urban Forestry & Urban Greening, vol. 40, pp. 1–3, 2019.
[32] K. J. Bagstad, G. W. Johnson, B. Voigt, and F. Villa, “Spatial
dynamics of ecosystem service flows: a comprehensive ap-
proach to quantifying actual services,” Ecosystem Services,
vol. 4, pp. 117–125, 2013.
[33] E. Garmendia, E. Apostolopoulou, W. M. Adams, and
D. Bormpoudakis, “Biodiversity and green infrastructure in
Europe: boundary object or ecological trap?” Land Use Policy,
vol. 56, pp. 315–319, 2016.
[34] P. Vogt and K. Riitters, “GuidosToolbox: universal digital
image object analysis,” European Journal of Remote Sensing,
vol. 50, no. 1, pp. 352–361, 2017.
[35] P. Vogt, MSPA Guide, European Commission, Joint Research
Centre, Ispra, Italy, 2012, https://ies-ws.jrc.ec.europa.eu/gtb/
GTB/MSPA_Guide.pdf.
[36] B. H. McRae and D. M. Kavanagh, Linkage Mapper Con-
nectivity Analysis Software, e Nature Conservancy, Seattle,
WA, USA, 2011, http://www.circuitscape.org/linkagemapper.
[37] S. Saura and L. Pascual-Hortal, “A new habitat availability
index to integrate connectivity in landscape conservation
planning: comparison with existing indices and application to
a case study,” Landscape and Urban Planning, vol. 83, no. 2-3,
pp. 91–103, 2007.
[38] S. Saura, C. Estreguil, C. Mouton, and M. Rodr´
ıguez-Freire,
“Network analysis to assess landscape connectivity trends:
application to European forests (1990-2000),” Ecological In-
dicators, vol. 11, no. 2, pp. 407–416, 2011.
[39] X. Shi and M. Qin, “Research on the optimization of regional
green infrastructure network,” Sustainability, vol. 10, no. 12,
p. 4649, 2018.
[40] P. B. Leonard, E. B. Duffy, R. F. Baldwin, B. H. McRae,
V. B. Shah, and T. K. Mohapatra, “Gflow: software for mod-
elling circuit theory-based connectivity at any scale,” Methods
in Ecology and Evolution, vol. 8, no. 4, pp. 519–526, 2017.
[41] B. H. McRae and V. B. Shah, Circuitscape User’s Guide,
University of California, Santa Barbara, CA, USA, 2009.
[42] V. B. Shah and B. McRae, “Circuitscape: a tool for landscape
ecology,” Proceedings of the 7th Python in Science Conference,
vol. 22, pp. 836–851, 2008.
[43] K. A. Zeller, K. Mcgarigal, and A. R. Whiteley, “Estimating
landscape resistance to movement: a review,” Landscape
Ecology, vol. 27, no. 6, pp. 777–797, 2012.
[44] B. H. McRae, S. A. Hall, P. Beier, and D. M. eobald, “Where
to restore ecological connectivity? Detecting barriers and
quantifying restoration benefits,” PLoS One, vol. 7, no. 12,
Article ID e52604, 2012.
[45] T. Elmqvist, H. Set¨
al¨
a, S. N. Handel et al., “Benefits of re-
storing ecosystem services in urban areas,” Current Opinion in
Environmental Sustainability, vol. 14, pp. 101–108, 2015.
[46] M. Benedict, E. McMahon, T. C. Fund et al., “Green infra-
structure: linking landscapes and communities,” Landscape
Ecology, vol. 22, no. 5, pp. 797-798, 2007.
[47] J. Ahern, “From fail-safe to safe-to-fail: sustainability and
resilience in the new urban world,” Landscape and Urban
Planning, vol. 100, no. 4, pp. 341–343, 2011.
14 Complexity

[48] C. Wilkinson, “Social-ecological resilience: insights and issues
for planning theory,” Planning eory, vol. 11, no. 2,
pp. 148–169, 2012.
[49] M. Spaans and B. Waterhout, “Building up resilience in cities
worldwide—Rotterdam as participant in the 100 resilient
cities programme,” Cities, vol. 61, pp. 109–116, 2017.
[50] B. Pizzo, “Problematizing resilience: implications for planning
theory and practice,” Cities, vol. 43, pp. 133–140, 2015.
[51] K. Tzoulas, K. Korpela, S. Venn et al., “Promoting ecosystem
and human health in urban areas using green infrastructure: a
literature review,” Landscape and Urban Planning, vol. 81,
no. 3, pp. 167–178, 2007.
[52] T. Matthews, A. Y. Lo, and J. A. Byrne, “Reconceptualizing
green infrastructure for climate change adaptation: barriers to
adoption and drivers for uptake by spatial planners,” Land-
scape and Urban Planning, vol. 138, pp. 155–163, 2015.
[53] C. Staddon, S. Ward, L. De Vito et al., “Contributions of green
infrastructure to enhancing urban resilience,” Environment
Systems and Decisions, vol. 38, no. 3, pp. 330–338, 2018.
[54] P. Vogt, K. H. Riitters, C. Estreguil, J. Kozak, T. G. Wade, and
J. D. Wickham, “Mapping spatial patterns with morphological
image processing,” Landscape Ecology, vol. 22, no. 2,
pp. 171–177, 2007.
[55] P. Vogt, J. R. Ferrari, T. R. Lookingbill, R. H. Gardner,
K. H. Riitters, and K. Ostapowicz, “Mapping functional
connectivity,” Ecological Indicators, vol. 9, no. 1, pp. 64–71,
2009.
[56] M. K. Hossain and Q. Meng, “A thematic mapping method to
assess and analyze potential urban hazards and risks caused by
flooding,” Computers, Environment and Urban Systems,
vol. 79, Article ID 101417, 2020.
[57] A. Sharifi, “Urban form resilience: a meso-scale analysis,”
Cities, vol. 93, pp. 238–252, 2019.
[58] M. Olazabal, L. Chelleri, and A. Sharifi, “Is connectivity a
desirable property in urban resilience assessments?,” in
Resilience-Oriented Urban Planning, Y. Yamagata and
A. Sharifi, Eds., vol. 65, pp. 197–211, Springer, Cham, Swit-
zerland, 2018.
[59] L. Jiang, B. Xue, Z. Ma, L. Yu, B. Huang, and X. Chen, “A life-
cycle based co-benefits analysis of biomass pellet production
in China,” Renewable Energy, vol. 154, pp. 445–452, 2020.
[60] A. Guo, J. Yang, X. Xiao, J. Xia (Cecilia), C. Jin, and X. Li,
“Influences of urban spatial form on urban heat island effects
at the community level in China,” Sustainable Cities and
Society, vol. 53, Article ID 101972, 2020.
[61] W. Kuang, “Seasonal variation in air temperature and relative
humidity on building areas and in green spaces in Beijing,
China,” Chinese Geographical Science, vol. 30, no. 1, pp. 75–88,
2020.
Complexity 15