Operational carbon transition in the megalopolises' commercial buildings

Abstract

This manuscript has been accepted by Building and Environment

Full text

Building and Environment 226 (2022) 109705
Available online 14 October 2022
0360-1323/© 2022 Elsevier Ltd. All rights reserved.
Operational carbon transition in the megalopolises’ commercial buildings
Minda Ma
a
,
1
, Wei Feng
b
, Jingwen Huo
a
, Xiwang Xiang
c
,
2
,
*
a
Department of Earth System Science, Tsinghua University, Beijing, 100084, PR China
b
Building Technology & Urban Systems Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, United States
c
School of Management Science and Real Estate, Chongqing University, Chongqing, 400045, PR China
ARTICLE INFO
Keywords:
Megalopolises
Operational carbon emissions
Commercial buildings decarbonization
Carbon neutrality goal
Generalized divisia index method
ABSTRACT
Megalopolises are important political and economic centers and offer the best opportunities for decarbonizing
commercial building operations. This study estimates the decarbonization level of commercial buildings from
China’s ve major megalopolises (Jing-Jin-Ji, Yangtze River Delta, Pearl River Delta, Yangtze River Middle
Reach, and Cheng-Yu) through the generalized Divisia index method, considering the impacts of socio-economy,
technology evolution, and climate. Results found the following: First, economic growth effects [service industry
added value (42.2%) and gross domestic product (36.5%)] and energy consumption (23.8%) were the main
drivers of the spike in operational carbon emissions of the megalopolises from 2000 to 2018. Second, except for
Cheng-Yu, operational carbon emissions were decoupled from economic growth effects since 2009, with the most
signicant decoupling status occurring in Jing-Jin-Ji, and technical effects being the main factor leading to the
decoupling. Third, the ve megalopolises cumulatively decarbonized 233.1 mega-tons of carbon dioxide
(MtCO
2
), offsetting 4.4% of the operational carbon of commercial buildings, with the highest decarbonization
level in Jing-Jin-Ji (5.7 MtCO
2
per yr, 7.8 kg of carbon dioxide per square meter per yr, and 55.6 kg of carbon
dioxide per capita per yr). Furthermore, current decarbonization strategies for megalopolises are reviewed to
plan for future low-carbon developments. Overall, this study assesses the operational decarbonization change of
commercial buildings in Chinese megalopolises. The ndings help inform building of sector pathways toward
China’s carbon peaking and neutral goals and develop global low-carbon cities.
1. Introduction
Buildings, the largest energy-consuming and emitting sector in
China, are the last mile toward the carbon neutral century [1]. The latest
evidence shows that China’s building sector consumed 1.6 giga-ton of
standard oil equivalent (Gtoe) and emitted nearly ve giga-ton of carbon
dioxide (GtCO
2
) in 2019, which accounts for 50.0% of the anthropo-
genic emissions [2]. Meanwhile, the service industry’s rapid urbaniza-
tion and economic growth have led to a surge in commercial building
(see the denition in Appendix A) energy demand, resulting in a 3.7-fold
increase in operational carbon emissions over almost two decades [3].
Notably, an increasing number of studies have conrmed that the
decarbonization potential of commercial buildings exceeds the previ-
ously believed values [4,5]. Hence, net-zero emissions from the building
sector are signicant to China’s carbon peaking and neutral actions,
with the decarbonization of commercial building operations playing a
crucial role.
Megalopolises are the most densely populated areas for economic
development [6] and core contributors to building energy consumption
and emissions (especially for commercial buildings) [7]. For example,
the Yangtze River Delta produced 19.8% of China’s total GDP with
11.0% of the population on 2.2% of the land area in 2018 [8], and
released nearly 12.4% of operational carbon emissions from China’s
commercial buildings [9]. Moreover, since megalopolises are usually
located in the same climate zones and share similar building energy
structures [10,11]; they are often regarded as the best units for energy
and climate policy implementation [12]. Currently, most countries have
wisely tilted policies toward megalopolises [13]. For China, deep
decarbonization of the operational carbon emissions of megalopolises is
a necessary way of achieving low-carbon development of the building
sector and a step toward achieving carbon peaking and neutral goals.
Nevertheless, few studies have assessed the operational decarbon-
ization levels of commercial buildings, especially among megalopolises
* Corresponding author.
E-mail addresses: maminda2020@tsinghua.org.cn, maminda@tsinghua.edu.cn (M. Ma), xiangxiwang@cqu.edu.cn (X. Xiang).
1
Personal Homepage: https://scholar.google.com/citations?user=240qUyIAAAAJ&hl=en.
2
Personal Homepage: https://scholar.google.com/citations?user=hvuvSLkAAAAJ&hl=en.
Contents lists available at ScienceDirect
Building and Environment
journal homepage: www.elsevier.com/locate/buildenv
https://doi.org/10.1016/j.buildenv.2022.109705
Received 16 August 2022; Received in revised form 23 September 2022; Accepted 11 October 2022

Building and Environment 226 (2022) 109705
2
in China. Therefore, three questions have been raised regarding com-
mercial buildings in China’s megalopolises.
•What drives the decarbonization of commercial buildings in China’s
megalopolises?
•How has the decoupling state of carbon emissions from economic
growth changed?
•What is the level of decarbonization of commercial buildings, and
how can it be facilitated?
To answer the above questions, this study estimates the historical
decarbonization level of commercial building operations in the top ve
emitting megalopolises—Jing-Jin-Ji, Yangtze River Delta, Pearl River
Delta, Yangtze River Middle Reach, and Cheng-Yu (see the denition in
Appendix B)—considering socio-economic, technological evolution, and
climate factors using the generalized Divisia index method (GDIM).
Specically, the GDIM decomposition technique is rst used to identify
the drivers of commercial building decarbonization. Based on the
decomposition results, the decoupling status of decarbonization from
economic growth is further evaluated. Finally, the historical decarbon-
ization levels are evaluated from three emission scales, and the decar-
bonization efciency is compared among megalopolises. In addition,
deep decarbonization strategies are proposed to seek the best practice
pathways by combining historical decarbonization levels with a review
of current decarbonization efforts.
Regarding the most necessary contribution, this study is the rst
to estimate the historical decarbonization of commercial buildings from
the megalopolises perspective to respond to the call for carbon peak and
neutrality in China. For this purpose, this study examines the effects of
socio-economic, technological evolution, and climate on the decarbon-
ization and decoupling from the economic growth of commercial
building operations and assesses and compares the historical decar-
bonization levels of ve megalopolises at three levels: total, per oor
space, and per capita.
The rest of this paper is structured as follows: Section 2 presents a
literature review. Section 3 describes the methods and materials,
covering the emission model, GDIM, the GDIM-based decoupling efforts
index model, and the datasets. Section 4 provides the results of the
drivers and the decoupling effect of operational carbon emissions. Sec-
tion 5 includes three aspects: Section 5.1 shows the operational decar-
bonization estimation of commercial buildings. Section 5.2 compares
the decarbonization efciencies of the ve megalopolises. Section 5.3
discusses the strategies of deep decarbonization toward net-zero emis-
sions. Finally, Section 6 concludes the core ndings and proposes future
studies.
2. Literature review
In recent decades, megalopolises have gradually emerged worldwide
and offer the best chance but also the greatest challenge for decarbon-
ization [14,15]. According to the latest statistics, the operational carbon
emissions from commercial buildings in the ve selected megalopolises
have soared more than 3-fold since 2000 and as a whole contribute
approximately half of the total energy consumption and operational
carbon emissions in China’s commercial buildings in 2018 [9] (see
Fig. 1). Many studies on major Chinese emitting cities have highlighted
the importance of megalopolises for decarbonizing commercial build-
ings [16], for instance, Jing-Jin-Ji [17,18], Yangtze River Delta [19],
Pearl River Delta [20], Guangdong-Hong Kong-Macao Greater Bay Area
[21], and other important pilot cities [22,23]. However, to our knowl-
edge, existing studies only cover specic megalopolises or regions and
lack comprehensive analysis and comparison. More importantly, most of
these efforts have neglected the assessment of the historical operational
decarbonization levels of commercial buildings.
Regarding operational carbon emissions of buildings, index decom-
position analysis (IDA), especially the log-mean decomposition index
(LMDI) [24], is a key tool for measuring the driving forces of economic
variables over time based on the Kaya identity [25]. Because of its ad-
vantages of simple operation and perfect decomposition, IDA has been
widely used to analyze the inuence of socioeconomic factors on oper-
ational carbon emissions and to further assess the decarbonization po-
tential of the buildings sector [26,27], including commercial buildings
[28,29], urban residential buildings [30], and rural residential buildings
[31]. Moreover, with the continuous development of decomposition
methods, increasing evidence suggests that the decomposition method
based on Kaya identity is too dependent on the relationship between
factors to output stable decomposition results [32]. This has led to better
alternatives being proposed, such as the GDIM [33], decomposing
structure decomposition [34] and Marshall-Edgeworth with structural
effects decomposition [35].
Considering the existing research on carbon emissions released by
commercial buildings described above, two gaps should be noted.
Regarding decarbonization assessment, although a large number
of studies on decarbonization have been reported in the classical liter-
ature, they have mainly focused on the international or national level
[36,37]. For instance, Zhang et al. [38] compared the carbon mitigation
of commercial buildings in China and the United States and concluded
that the mitigation efciency in China was 1.5 times that in the United
States. Lin et al. [39] rst used the LMDI to investigate the drivers of
carbon change in the commercial buildings sector of China and provided
corresponding policies to reduce emissions. Zhou et al. [40] and Tang
et al. [41] further analyzed the low-carbon development trajectory of
commercial buildings by the mid-century. Recently, Li et al. [2]
reviewed the provincial decarbonization levels of commercial buildings
in China and compared them with those at the national level. However,
megalopolises, as major carriers of operational carbon released by
commercial buildings, have not received the necessary attention to date.
Regarding the analytical approach to assessing building
Abbreviation notation
GDIM – Generalized Divisia Index Method
GDP – Gross Domestic Product
IBED – International Building Emission Dataset
IDA – Index decomposition analysis
LMDI – Log-Mean Divisia Index
Nomenclature
C – Carbon emissions
E – Energy consumption
F – Gross oor space
G – Gross Domestic Product
G
s
– Gross Domestic Product of service industry
Gtoe – Giga-ton of standard oil equivalent
GtCO
2
– Giga-ton of carbon dioxide
kgCO
2
– Kilograms of carbon dioxide
MtCO
2
–Megatons of carbon dioxide
P – Population
φ – Decarbonization effort index
ΔC – Change in operational carbon emissions
ΔD – Decarbonization effort
ΔDC – Total decarbonization
ΔDF – Decarbonization per oor space
ΔDP – Decarbonization per capita
M. Ma et al.

Building and Environment 226 (2022) 109705
3
decarbonization, GDIM is the preferred decomposition method for
assessing historical decarbonization. Compared with other conventional
decomposition techniques, GDIM not only overcomes the inherent aws
in the traditional index decomposition analysis framework but also en-
ables a multi-perspective analysis of the nonlinear interrelationships
among potential driving forces [42,43]. Namely, the GDIM can identify
both absolute and relative indicators [44]. In addition, a large number of
empirical analyses from the energy and emission sectors conrm the
validity of this method [45,46]. Therefore, GDIM has been used to reveal
the main contributors and decarbonization processes in major emission
sectors [47,48], including industry [49], transportation [50], and elec-
tricity [51,52]. However, to our knowledge, no studies on GDIM in the
area of building operation emissions have yet been conducted.
Therefore, to ll these gaps, the historical evolution of carbon
emissions of commercial buildings in ve Chinese megalopolises is
investigated using the GDIM decomposition method. The contributions
cover two aspects.
•This is the rst time the operational carbon change of com-
mercial buildings is reviewed via GDIM. As mentioned earlier, the
existing literature is mainly based on the classical index decompo-
sition method, and analysis via GDIM in the building sector has not
been reported. Therefore, this study rst uses an advanced GDIM to
identify the factors affecting operational carbon, then reviews the
decoupling effect based on the decomposition results, and nally
evaluates the historical decarbonization levels of commercial
buildings.
•The historical decarbonization levels of the top ve mega-
lopolises’ commercial buildings are assessed across three scales
for the rst time. Currently, only a few studies have conducted
decarbonization assessments for commercial building operations,
and none have emerged at the megalopolis level. This study allows
for the vital role of China’s megalopolises’ buildings in carbon
neutral actions. Operational decarbonization levels of commercial
buildings are assessed on different scales: total decarbonization,
decarbonization per oor space, and decarbonization per capita.
Furthermore, the decarbonization efciency of the ve mega-
lopolises is compared to explore the operational decarbonization
potential of commercial buildings.
3. Methods and materials
The GDIM was adopted to explore the driving factors, decoupling
effects, and mitigation of operational carbon emissions from commercial
buildings in ve megalopolises in China. The emission model for com-
mercial building operations is described in Section 3.1. The GDIM
approach used to determine the drivers of operational decarbonization
of commercial buildings is presented in Section 3.2, and the GDIM-based
decoupling effort index model is described in Section 3.3. Finally, data
collection is presented in Section 3.4.
3.1. Operational carbon emission model
Ehrlich and Holden [53] developed an environmental impact model
with population, afuence, and technology (IPAT) in 1971, which pro-
vides an important analytical framework for exploring the effects of
environmental change [54]. The classical IPAT model is expressed as
follows:
I=P×A×T(1)
where I on the left-hand side of Eq. (1) represents the environmental
impact caused by the inuencing factors, and P, A, and T on the right-
hand side of Eq. (1) represent population size, afuence level, and
technology level, respectively. Correspondingly, the impact of carbon
emissions can be described by the IPAT model as follows:
C=P×G
P

A
×C
G

T
(2)
For the building sector, based on the emission characteristics and
inherent properties of the commercial building sector, the emissions
model [55] in Eq. (2) is redened as:
C=P×G
P

A
×Gs
G×F
Gs
×E
F×C
E

T
(3)
In previous studies, Eq. (3) was generally decomposed according to
the index decomposition analysis method (e.g., LMDI [56]) as follows:
ΔC|0→T=C|T−C|0=ΔC[P] + ΔCG
P+ΔCGs
G+ΔCF
Gs+ΔCE
F
+ΔCC
E(4)
where ΔC|0→T denotes the change in the total carbon emissions over the
period [0,T]. The terms on the right-hand side (e.g., ΔC[P]) of Eq. (4)
represent the contribution level of population size to the operational
carbon in buildings.
However, as stated in the literature review, the decomposition re-
sults (Eq. (4)) calculated using the conventional method have many
drawbacks compared to the GDIM. Hence, a new emission model was
proposed in this study to quantify the drivers of carbon emissions from
Fig. 1. (a) Operational carbon change of commercial buildings in ve megalopolises (2000–2018) and (b) energy and emission levels from commercial buildings
in 2018.
M. Ma et al.

Building and Environment 226 (2022) 109705
4
commercial building operations based on the classical IPAT model,
which is expressed as follows:
C=P×C
P

P
=G×C
G

A
=Gs×C
Gs
=F×C
F=E×C
E

T
(5)
The GDIM was then applied to decompose the emission model pro-
posed in this study.
3.2. Generalized divisia index method (GDIM)
Following the theory and content of the GDIM, a commercial
building operation emission model (Eq. (5)) proposed in this study can
be rewritten as follows:
C=x1x2=x3x4=x5x6=x7x8=x9x10 (6)
where x1,x3,x5,x7,x9 are absolute factors and x2,x4,x6,x8,x10 are
relative factors. Then, the other four relative indicators can be further
derived by





























x11 =x6
x10
=C
GsC
E=E
Gs
x12 =x6
x4
=C
GC
Gs
=Gs
G
x13 =x2
x4
=C
PC
G=G
P
x14 =x8
x10
=C
FC
E=E
F
(7)
Let the vector x= [x1,x2,x3,⋯,x14], which represents the inuencing
factors that determine the carbon emissions of commercial building
operations. The denitions and explanations of the inuencing factors in
vector x are summarized and tabulated in Table 1.
Simultaneous Eqs. (6) and (7) are obtained as follows:



























C=x5x6
θ1=x5x6−x1x2=0
θ2=x5x6−x3x4=0
θ3=x5x6−x7x8=0
θ4=x5x6−x9x10 =0
θ5=x9−x5×x11 =0
θ6=x5−x3×x12 =0
θ7=x3−x1×x13 =0
θ8=x9−x7×x14 =0
(8)
Eq. (8) can be expressed in matrix form as:
Θ(x) = 0
C=f(x) = f(x1,x2,⋯,x14)(9)
According to [33], the change in carbon emissions from commercial
buildings can be decomposed as
ΔC[X|Θ]T=L
∇CTI−ΘxΘ+
xdX(10)
where L is the time interval, I is the identity matrix, ∇C=
(x5,x6,0,0,0,0,0,0,0,0,0,0,0,0)T is the gradient of the function f(x)
with respect to the inuence factor, Θx is the Jacobi matrix of the matrix
function Θ(x)and satises (Θx)ij =
∂
θj
∂
xi. In this study, Θx reects the
marginal impact of different drivers on carbon emissions from com-
mercial building operations and is dened as follows:
In addition, superscript +in Eq. (10) represents the generalized in-
verse matrix operator, and Θ+
x= (ΘT
xΘx)−1ΘT
x is satised when the Ja-
cobian matrix column is full-rank. At this time, Eq. (10) can be further
transformed into
ΔC[X|Θ]T=L
∇CTI−ΘxΘ+
xΘx−1ΘT
xdX(12)
The Simpson integral rule was used to integrate Eq. (12), and the
change in commercial building emissions can be perfectly decomposed
into the sum of the contributions of the 14 factors:
ΔC|0→T=C|T−C|0=14
i=1
ΔC[xi](13)
Furthermore, the decarbonization (ΔDC|0→T) of commercial building
operations during period ΔT can be calculated using the emission
drivers with negative contributions from 0 to T, which is dened as
follows [57]:
ΔDC|0→T=ΔCxj(14)
where ΔC[xj] = {ΔC[xi],i=1,2,⋯,14 |ΔC[xi] ≤ 0}. Accordingly, the
Table 1
The denition and explanation of driving factors in GDIM.
Driving factors Denition Explanation
x1 P Population size
x2 C
P
Carbon emission per capita
x3 G GDP
x4 C
G
Carbon emission per GDP
x5 Gs Service industry added value
x6 C
Gs
Economic efciency of emissions
x7 F Floor space
x8 C
F
Carbon intensity
x9 E Energy consumption
x10 C
E
Emission factor
x11 E
Gs
Economic efciency of energy
x12 Gs
G
Industrial structure
x13 G
P
GDP per capita
x14 E
F
Energy intensity
Θx=












−x2−x10 0 x6x50 0 0 0 0 0 0 0
0 0 −x4−x3x6x50 0 0 0 0 0 0 0
0 0 0 0 x6x5−x8−x70 0 0 0 0 0
0 0 0 0 x6x50 0 −x10 −x90 0 0 0
0 0 0 0 −x11 0 0 0 1 0 −x50 0 0
0 0 −x12 0 1 0 0 0 0 0 0 −x30 0
−x13 0 1 0 0 0 0 0 0 0 0 0 −x10
0 0 0 0 0 0 −x14 0 1 0 0 0 0 −x7












T
8×14
(11)
M. Ma et al.

Building and Environment 226 (2022) 109705
5
decarbonization per oor space (ΔDF|0→T) and decarbonization per
capita (ΔDP|0→T) can also be dened according to Eq. (14), as follows:







ΔDF|0→T=ΔDC|0→T
F
ΔDP|0→T=ΔDC|0→T
P
(15)
3.3. The GDIM-based decoupling efforts index model
According to the decomposition results obtained through the GDIM,
the change in operational carbon due to the change in GDP (ΔG) and
GDP per capita change (ΔG
P) can be regarded as economic growth effects
[58,59]. In general, decarbonization efforts are the sum of the factors
(excluding economic growth effects) that lead to the reduction of
operational carbon emissions from commercial buildings, and represent
all efforts that directly or indirectly lead to operational decarbonization.
Thus, the operational decarbonization effort from the base year to year T
is characterized as [60]
ΔD|0→T=ΔC|0→T− (ΔG+ΔGS) = ΔC|0→T− (Δx3+Δx13)(16)
Furthermore, the decarbonization effort offset by the economic
growth effect was used to assess the coupling effect between economic
growth and the decarbonization of commercial buildings [61]. As
illustrated in Eq. (17):
φ|0→T=ΔD|0→T
Δx3+Δx13
(17)
where φ|0→T denotes the degree of decoupling of commercial buildings
decarbonized during period [0, T]. The denitions of the different
decoupling states are summarized in Table 2.
Evidently, a decoupling indicator can be determined for each
decomposition factor based on the additivity and comparability of the
decomposition factors. To reveal the inuencing factors driving the
change in decoupling status, the decoupling indexes related to energy
intensity and emission intensity are attributed to technical effects in this
study, while the remaining decoupling indexes that do not directly
reect technological progress are identied as non-technical effects [62,
63]. Thus, the decoupling index for the decarbonization of commercial
building operations can be decomposed as
For convenience, Eq. (18) is rewritten as follows:
where φxi denotes the relative contribution level of the ith inuencing
factor to the overall decoupling effort index φ|0→T, and φNon−tech and
φTech reect the inuence of technical and non-technical effects,
respectively, on the decoupling of carbon emissions from commercial
building operations.
3.4. Datasets
In this study, historical data from ve megalopolises in China from
2000 to 2018 were selected as the research sample for use in a com-
mercial building emission model. Fig. 2 and Tables C1-C5 (see Appendix
C) show the descriptive statistics of all raw data, including population,
GDP, service industry added value, building stock, and energy and
emissions of commercial buildings. Building-related data were sourced
from the IBED database (www.researchgate.net/project/International-B
uilding-Emission-Dataset-IBED), which is a multi-regional dataset that
mainly serves the building sector. Indicators related to population and
economy were collected from http://data.stats.gov.cn/english/.
4. Results
4.1. Drivers of operational decarbonization in commercial buildings
Based on the emission model established in Eq. (6), the change in
carbon emissions from building operations was decomposed by the
GDIM into the contributions of 14 driving factors, including ve abso-
lute drivers and nine relative drivers. The heat maps in Fig. 3 present the
changes in the operational carbon emissions of commercial buildings
and the decomposition results for the ve megalopolises in China from
2000 to 2018.
Fig. 3 shows the total effect of the drivers. Operational carbon of
commercial buildings in megalopolises continued to rise from 106.5
MtCO
2
in 2000 to 444.9 MtCO
2
in 2018, with an average annual growth
rate of 17.7%. The most signicant rise occurred in the Yangtze River
Delta (ΔCY−R−D|2000→2018 =94.7 MtCO
2
), followed by the Yangtze River
Middle Reach (ΔCY−R−M−R|2000→2018 =80.1 MtCO
2
). The carbon
Table 2
The denition of decoupling state.
Decoupling state Δx3+Δx13 ΔD|0→T φ|0→T
Strong decoupling >0 <0 φ|0→T<−1
Weak decoupling >0 <0 −1<φ|0→T<0
Expansive negative decoupling >0 >0 φ|0→T>0
Strong negative decoupling <0 >0 φ|0→T<−1
Weak negative decoupling <0 >0 −1<φ|0→T<0
Recessive decoupling <0 <0 φ|0→T>0
φ|0→T=Δx1+Δx5+Δx7+Δx9+Δx12
Δx3+Δx13
+Δx2+Δx4+Δx6+Δx8+Δx10 +Δx11 +Δx14
Δx3+Δx13
=Δx1+Δx5+Δx7+Δx9+Δx12
Δx3+Δx13

Non−technical effect
+Δx2+Δx3+Δx2+Δx3+Δx2+Δx3+Δx2
Δx3+Δx13

Technical effect
(18)
φ|0→T=φx1+φx5+φx7+φx9+φx12

Non−technical effect
+φx2+ +φx4+φx6+φx8+φx10 +φx11 +φx14

Technical effect
=φNon−tech +φTech (19)
M. Ma et al.

Building and Environment 226 (2022) 109705
6
emissions of all megalopolises had a brief peak around 2014, and most
megalopolises started to rebound strongly after peaking. Notably, in
Jing-Jin-Ji, the operational carbon emissions of commercial buildings
maintained a steady decline after peaking at 134.2 MtCO
2
in 2016. In
addition, the effect of climate on changes in operational carbon is
striking. Following the building climate zones provided by standard GB
50352–2019 (see Fig. 1 b) [64], the operational carbon emissions in hot
summer/cold winter zones (Yangtze River Middle Reach: 55.6%,
Cheng-Yu: 30.9%, Yangtze River Delta: 21.7%) had substantially higher
average increase rates than those in hot summer/warm winter zones
Fig. 2. Descriptive statistics of historical data for ve megalopolises in China (2000–2018). Notes: the short vertical lines under the density plot indicate the sample,
and the dots and diamonds in the density plot indicate the sample mean and 25th (75th) percentile, respectively.
Fig. 3. Contribution of drivers to change in operational carbon emissions in ve Chinese megalopolises.
M. Ma et al.

Building and Environment 226 (2022) 109705
7
(Pearl River Delta: 19.6%) and cold zones (Jing-Jin-Ji: 7.5%).
As shown in the heat map in Fig. 3, the impact of drivers on the
operational carbon emissions of commercial buildings from different
megalopolises is similar. For the drivers that contribute to operational
carbon emissions, the two largest contributors were the service industry
added value (42.2%) and GDP (36.5%), both of which contributed
consistently to the operational carbon of commercial buildings, espe-
cially for commercial buildings in Jing-Jin-Ji (59.4% and 68.9%).
Another key factor driving carbon emissions was energy consumption
(23.8%), which contributed to increases of 28.0% (Jing-Jin-Ji), 23.0%
(Cheng-Yu), 22.7% (Pearl River Delta), 22.9% (Yangtze River Delta),
and 22.4% (Yangtze River Middle Reach) in the ve megalopolises. In
addition, oor space (14.1%) also had a small positive impact on the
carbon emissions of commercial buildings. For drivers that contribute to
decarbonization, the economic efciency of emissions (−20.4%) and
carbon emissions per GDP (−15.1%) played a critical role in driving the
Fig. 4. (a) The decoupling state of operational carbon in commercial buildings and (b) the drivers of changes in the decoupling state.
M. Ma et al.

Building and Environment 226 (2022) 109705
8
decarbonization of commercial building operations. In addition, the
emission factor (−3.5%) was also an important driver of the decarbon-
ization of commercial buildings, especially after 2014 (before 2014:
−1.2%, after 2014: −4.2%). In contrast, the contribution of carbon
emissions per capita and carbon intensity to decarbonization was not
stable. For instance, carbon intensity had a positive contribution in Jing-
Jin-Ji and Yangtze River Delta, whereas it was negative in the rest of the
megalopolises. Additionally, the effect of carbon emission per capita on
carbon emissions varied greatly from one period to another and shows
large uctuations. The remaining drivers had insignicant effects on
carbon emissions and were, therefore, ignored.
Overall, the above decomposition results portray the major drivers
driving the operational decarbonization of commercial buildings in
megalopolises in China and respond to the rst question raised in
Fig. 5. Operational decarbonization change of commercial buildings in megalopolises and share of decarbonization at different stages (2000–2018). Note: the
bubbles on the map indicate the total cumulative decarbonization from 2000 to 2018; the error bands in the line graph indicate one standard deviation.
M. Ma et al.

Building and Environment 226 (2022) 109705
9
Section 1.
4.2. Decoupling effect of operational carbon in commercial buildings
After identifying the drivers of commercial building decarbon-
ization, based on the decomposition results from the GDIM, the decou-
pling effort index model was used to investigate the decoupling effect of
operational carbon of commercial buildings, and the results are illus-
trated in Fig. 4. For ease of analysis and comparison, the study period
was divided into four phases: 2000–2005, 2006–2009, 2010–2014, and
2015–2018. Since the 21st century, China’s service economy has always
maintained rapid development, resulting in the economic growth of the
ve megalopolises showing an increasing trend. Therefore, the decou-
pling results (see the map in Fig. 4 a) only cover three decoupling states:
expansion negative decoupling, weak decoupling, and strong decou-
pling (see Fig. 5).
Fig. 4 a presents the evolution of the decoupling of operational
carbon emissions from 2000 to 2018, and the results show that the
average decoupling indexes of the ve megalopolises from 2000 to 2018
were: 1.96 (2000–2005), 0.63 (2006–2009), 0.02 (2010–2014), and
−0.08 (2015–2018). This indicates that the decoupling efforts of carbon
emissions from commercial building operations have been increasing,
from the initial expansion negative decoupling to weak decoupling.
Specically, before 2009, the decoupling effort index of the decarbon-
ization of commercial buildings in the ve megalopolises was greater
than zero, showing an expansion negative decoupling state, indicating
that the carbon emissions of commercial building operations and eco-
nomic growth effects in these regions have not been decoupling, that is,
Fig. 6. (a) Operational decarbonization per oor space and (b) per capita of commercial buildings in the ve Chinese megalopolises.
M. Ma et al.

Building and Environment 226 (2022) 109705
10
the decarbonization efforts during this period could not offset the in-
crease in carbon emissions driven by economic effects. Since 2009,
except for the Cheng-Yu region, which still maintains the negative
decoupling state, the rest of the megalopolises started to show a state of
less than zero or even less than −1, indicating that these regions have
entered a deep decoupling state, and the decarbonization effect caused
by the constraints has been gradually offset or even exceed the driving
effect brought by the economic growth effect. In other words, the
operational carbon in these regions decreased with the growth of eco-
nomic effects, including GDP and the service industry added value. To
be more specic, from 2009 to 2014, operational carbon in the Jing-Jin-
Ji region (φJ−J−J|2009→2014 = − 0.48), the Pearl River Delta
(φP−R−D|2009→2014 = − 0.15)and the Yangtze River Middle Reach
(φY−R−M−R|2009→2014 = − 0.04)were the rst to show a weak decoupling
state. During 2015–2018, the Yangtze River Delta
(φY−R−D|2015→2018 = − 0.08)also entered a weak decoupling state, and
the Pearl River Delta (φP−R−D|2015→2018 =0.18)and Yangtze River Delta
(φY−R−D|2015→2018 =0.57)returned to the expansion negative decou-
pling state, and the Jing-Jin-Ji region entered a strong decoupling state.
Fig. 4b further provides the drivers of change in the decoupling of
carbon emissions from commercial buildings. For all megalopolises, the
decoupling efforts index of technology effects decreased continuously
throughout the study period (φTech|2000→2005 =0.84,φTech |2006→2009 =
0.12,φTech|2010→2014 = − 0.35,φTech|2015→2018 = − 0.43), while non-
technology effects changed insignicantly and always led to an in-
crease in the decoupling efforts index (φNon−Tech|2000→2005 =0.78,
φNon−Tech|2006→2009 =0.46,φNon−Tech |2010→2014 =0.50,
φNon−Tech|2015→2018 =0.52), indicating that changes in decoupling of
operational carbon in megalopolises were mainly driven by technolog-
ical effects. From the decoupling efforts index of the drivers, emission
factor and carbon emission per GDP were key technical factors that led
to the decoupling of operational carbon emissions from economic
growth effects. Additionally, although some of the remaining technical
factors (such as carbon emission per capita and economic efciency of
energy) had a dampening effect on the decoupling of operational carbon
emissions, this effect gradually diminished or even reversed. Energy
consumption and population size determined the non-technical effects.
Overall, the above results present the change in decoupling effect of
operational carbon from economic growth in megalopolises and the
reason behind this transformation, answering the second question posed
in Section 1.
5. Discussion
The deep decarbonization of megalopolises is an important prereq-
uisite for China to accomplish its ambitious blueprint for net-zero
emissions. Therefore, to close the gap between the current carbon
mitigation progress of commercial buildings and the existing targets,
megalopolises should make more prominent contributions. In this sec-
tion, the historical decarbonization levels from 2000 to 2018 are dis-
cussed. Section 5.1 traces the dynamic evolution of the operational
decarbonization of commercial buildings in three aspects (including
total decarbonization, decarbonization per oor space, and decarbon-
ization per capita), and Section 5.2 compares the decarbonization ef-
ciency of megalopolises at different stages. Finally, Section 5.3 reviews
the energy efciency and emission reduction policies of commercial
buildings in China’s megalopolises, and the corresponding policy in-
sights are stated based on historical decarbonization levels.
5.1. Operational decarbonization assessment of commercial buildings
As described in Section 3.2, based on the GDIM decomposition re-
sults, the historical decarbonization level of commercial buildings can
be evaluated using Eqs. 13–15, and the results are presented in Fig. 6.
The size of the bubbles on the map reects the total cumulative
decarbonization in the Chinese megalopolises’ commercial buildings.
From 2000 to 2018, the ve megalopolises have decarbonized a total of
233.1 MtCO
2
, which is equivalent to 4.4% of the total operational car-
bon emissions; the megalopolis with the highest level of total decar-
bonization was the Jing-Jin-Ji region with 102.0 MtCO
2
(equivalent to
the total emissions of the ve megalopolises in 2000 or the sum of the
cumulative decarbonization in Cheng-Yu, Yangtze River Delta, and
Yangtze River Middle Reach), followed by the Yangtze River Delta (55.4
MtCO
2
), the Yangtze River Middle Reach (29.9 MtCO
2
), and the Pearl
River Delta (29.5 MtCO
2
), whereas the megalopolis with the lowest total
decarbonization was the Cheng-Yu region, with 16.3 MtCO
2
(less than
1/6 of Jing-Jin-Ji’s decarbonization).
From the perspective of the specic historical trajectory of decar-
bonization, during the entire study period, the decarbonization trajec-
tory of commercial building operations in China’s megalopolises
generally showed an upward trend. Specically, as the top two mega-
lopolises with the largest emissions, the decarbonization in the Jing-Jin-
Ji and the Yangtze River Delta regions was always signicantly higher
than that of the other three megalopolises during the same period and
uctuated more signicantly, with annual average decarbonization of
5.7 MtCO
2
/yr and 3.1 MtCO
2
/yr respectively, peaking in 2004 (22.7
MtCO
2
) and 2014 (10.8 MtCO
2
), respectively. In contrast, the changes in
the decarbonization of commercial buildings located in the remaining
three megalopolises were relatively stable and very similar. From 2000
to 2012, the total decarbonization in these regions was relatively low,
and the total decarbonization between 2012 and 2018 showed an M-
shape with all peaks around 2014. The total annual average decarbon-
ization of Yangtze River Middle Reach, Pearl River Delta, and Cheng-Yu
was 1.7 MtCO
2
/yr, 1.6 MtCO
2
/yr, and 0.9 MtCO
2
/yr, respectively. With
respect to the change in stage decarbonization, from 2000 to 2018, the
share of total decarbonization in the ve megalopolises across the four
stages was 15.8%, 18.6%, 32.6%, and 33.0%, respectively, indicating
that the operational decarbonization process of commercial buildings
was gradually accelerating.
Fig. 6 further depicts the change of decarbonization per oor space
and per capita in China’s megalopolises. From 2001 to 2018, the average
annual decarbonization per oor space of the ve megalopolises is
ranked as follows: Jing-Jin-Ji (7.8 kgCO
2
/m
2
/yr) >Pearl River Delta
(2.3 kgCO
2
/m
2
/yr) >Yangtze River Delta (2.0 kgCO
2
/m
2
/yr) >Yangtze
River Middle-Reach (1.5 kgCO
2
/m
2
/yr) >Cheng-Yu (1.3 kgCO
2
/m
2
/
yr). Further, the ranking of average annual decarbonization per capita
is: Jing-Jin-Ji (55.6 kgCO
2
per capita/yr) >Yangtze River Delta (19.5
kgCO
2
per capita/yr) >Pearl River Delta (15.3 kgCO
2
per capita/yr) >
Yangtze River Middle Reach (9.8 kgCO
2
per capita/yr) >Cheng-Yu (8.1
kgCO
2
per capita/yr). In general, the megalopolis with the highest level
of decarbonization from the three scales was Jing-Jin-Ji (especially in
terms of per oor space and per capita, which both exceed the combined
total of the remaining megalopolises), while the lowest was Cheng-Yu.
Although there is a huge gap between the total amount of decarbon-
ization in Cheng-Yu and the other megalopolises, the decarbonization
per oor space and decarbonization per capita in Cheng-Yu are close to
other megalopolises, especially the Yangtze River Middle Reach. In
addition, the decarbonization per oor space and decarbonization per
capita in the Pearl River Delta and in the Yangtze River Delta are very
similar, but the total amount of decarbonization in the Yangtze River
Delta is almost twice that of the Pearl River Delta.
Overall, the above discussion estimates the operational decarbon-
ization of commercial buildings in ve megalopolises at three scales
over the last decade, tentatively answering the third question posed in
Section 1.
5.2. Comparison of decarbonization efciency of the ve megalopolises
Although Section 5.1 assesses the operational decarbonization level
of commercial buildings in China from three emission scales, this
assessment framework, which relies only on absolute indicators, is
M. Ma et al.

Building and Environment 226 (2022) 109705
11
incomplete. To gain insight into the decarbonization potential and thus
allocate future emission shares for each megalopolis [65]. In this sec-
tion, decarbonization efciency is introduced into the assessment
framework of this study, which is dened as the ratio of total decar-
bonization to carbon emissions. Similarly, the intensity-based decar-
bonization efciency can also be calculated. Fig. 7 a–c shows the
decarbonization efciency of China’s megalopolises at three emission
scales: total amount, per capita, and per oor space, respectively.
Meanwhile, the decarbonization efciency is very close at different
scales, which is caused by insignicant changes in population size and
oor space within each period. Therefore, the following discussion takes
the total amount of decarbonization as an example to compare the
decarbonization efciency between megalopolises.
As shown in Fig. 7, the decarbonization efciency in Jing-Jin-Ji and
Yangtze River Delta was consistently higher than that of the other
megalopolises between 2000 and 2018. From 2000 to 2005, the com-
mercial building operations in Jing-Jin-Ji reduced carbon emissions by
29.2 MtCO
2
, accounting for 28.7% of the total decarbonization, which is
equivalent to 8.4% of the operational carbon emissions of Jing-Jin-Ji
during the same period. The decarbonization efciency during this
time period was the highest of all megalopolises during the entire study
period. After 2005, the decarbonization efciency in Jing-Jin-Ji steadily
increased (3.9%, 4.5%, and 6.1%), whereas the opposite was true for the
Yangtze River Delta (4.8%, 4.1%, 3.7%). For commercial building op-
erations in Cheng-Yu, Pearl River Delta, and Yangtze River Middle
Reach, decarbonization efciency increased steadily during 2000–2014
and then decreased. Although total decarbonization in the Yangtze River
Middle Reach and Pearl River Delta were very close and had similar
trends, the decarbonization efciency in the Pearl River Delta was
higher than that in the Yangtze River Middle Reach (except for the
period 2006–2018). In addition, although the total decarbonization of
Cheng-Yu was much smaller than that of the Pearl River Delta region,
the decarbonization efciencies of both were very similar from 2000 to
2018. The overall decarbonization efciency of commercial buildings in
the ve megalopolises from to 2000–2018 was Jing-Jin-Ji (5.6%) >
Yangtze River Delta (4.0%) >Cheng-Yu (3.8%) >Yangtze River Middle
Reach (3.5%) =Pearl River Delta (3.5%).
In summary, the above discussion compares the decarbonization
efciency of different megalopolises, further answering the third ques-
tion posed in Section 1.
Fig. 7. Operational decarbonization efciency of commercial buildings at different scales: (a1-a5) total, (b1-b5) per oor space, and (c1-c5) per capita.
M. Ma et al.

Building and Environment 226 (2022) 109705
12
5.3. Deep decarbonization strategies toward net-zero emissions
Sections 5.1 and 5.2 report the operational decarbonization levels of
commercial buildings in megalopolises. Although the results show that
the pace of decarbonization in megalopolises is accelerating, the overall
decarbonization efciency is still relatively low (ranging from 3.5% to
5.6%). The deep decarbonization of building operations is closely
related to improvements in energy conservation levels. Since 1980, the
Chinese government has proposed a series of energy conservation
standards to improve the energy efciency of commercial building op-
erations and thus reduce operational carbon emissions, as well as the
concepts of ultralow, near-zero, and zero energy buildings. Following
these policies, the regions in which megalopolises are located have
proposed corresponding policies that consider climatic conditions and
specic situations. Fig. 8 outlines the roadmap of decarbonization
strategies for commercial buildings over the last few decades.
Combining the proposed ndings and existing energy conversion
policies, future decarbonization strategies for China’s megalopolises can
be implemented in two ways.
From the perspective of megalopolises, the agglomeration effect of
megalopolises should be maximized, thereby promoting the develop-
ment of emerging industries and industrial structure upgrading [66].
The Chinese government has already launched three batches of
low-carbon pilot cities in seven provinces and 80 cities since 2010 and
accomplished initial achievements [67,68], and more attention needs to
be paid simultaneously to resource allocation and the spatial layout of
megalopolises to combat environmental threats caused by rapid ur-
banization (e.g., urban heat island effect [69]). In addition, unied and
more advanced energy efciency standards and evaluation systems for
buildings can help improve the energy efciency of the construction
industry within megalopolises; for example, the Jing-Jin-Ji region
jointly promulgated the Design Standard Of Green Buildings in 2021 [70].
In addition, some advanced experiences from abroad are worth learning
from, such as the vertical greening system implemented in Singapore
[71,72] and the net-zero energy building goal proposed in California
[73].
From the perspective of building operations, end-use electrication
should be promoted, and the proportion of renewable energy in building
operations should be signicantly increased [74,75], especially for
megalopolises located in cold northern regions (e.g., the Jing-Jin-Ji re-
gion). Meanwhile, strengthening the supervision and management of
energy consumption monitoring and auditing in commercial buildings is
recommended (e.g., Shanghai [76] and Shenzhen [77]), where the
provision of large amounts of online monitoring data will effectively
constrain the energy intensity of the end-use equipment [78,79]. In
addition to these, the government should guide the development of
low-carbon technologies by improving the scal incentive system or
using economic instruments (e.g., carbon taxes [80] and carbon markets
[81]) and digital nancial inclusion [82].
Overall, the policy recommendations for the deep decarbonization of
commercial buildings address the third question posed in Section 1.
6. Conclusion
This study investigated the progress of decarbonization of commer-
cial building operations in China’s megalopolises using the GDIM. First,
the GDIM was applied to identify the driving forces of operational car-
bon emissions from 2000 to 2018. Based on the decomposition results,
the decoupling state of operational carbon emissions was characterized
by the decoupling efforts index. Second, the study assessed the historical
decarbonization level on the three emission scales. Third, decarbon-
ization efciency was investigated to determine the potential for deep
decarbonization. Finally, policies to decarbonize commercial building
operations in megalopolises were reviewed to identify potential path-
ways to net-zero emissions. The ndings are summarized as follows:
6.1. Main ndings
•Operational carbon emissions from commercial buildings
continued to rise at an annual level of 17.7%, with economic
growth effects and energy use being the key drivers contrib-
uting to the rise. Carbon emissions from commercial building op-
erations in ve megalopolises continued to rise from 106.5 MtCO
2
in
2000 to 444.9 MtCO
2
in 2018. In addition, the economic growth
effect [service industry added value (42.2%) and GDP (36.5%)] was
the largest contributor promoting operational carbon emissions
(especially for commercial buildings in the Jing-Jin-Ji region), fol-
lowed by energy consumption (23.8%). In contrast, the economic
efciency of emissions (−20.4%), carbon emissions per GDP
(−15.1%), and emission factor (−3.5%) were the key drivers for
decarbonizing building operations.
•Since 2009, operational carbon emissions were decoupled from
economic growth effects in most megalopolises (except the
Cheng-Yu). From 2000 to 2009, there was a clear coupling effect
between operational carbon emissions and economic growth in
commercial buildings. After 2009, operational carbon in Jing-Jin-Ji,
Fig. 8. The roadmap of decarbonization strategies for the commercial buildings in China.
M. Ma et al.

Building and Environment 226 (2022) 109705
13
Pearl River Delta, and Yangtze River Middle Reach decoupled from
economic growth effects and displayed a weak decoupling in
2010–2014. During 2015–2018, the Yangtze River Delta also dis-
played a weak decoupling state, and Jing-Jin-Ji presented a strong
decoupling state. Following the decoupling effort index of drivers,
technical effects were key contributors to the decoupling of opera-
tional carbon from economic growth, whereas non-technical effects
had always prevented it.
•The operational decarbonization of commercial buildings of
megalopolises was gradually accelerating, with cumulative
decarbonization of 233.1 MtCO
2
from 2000 to 2018, with the
highest decarbonization level in the Jing-Jin-Ji region (5.7
MtCO
2
/yr, 7.8 kgCO
2
/m
2
/yr, 55.6 kgCO
2
per capita per yr). This
study assessed and compared the decarbonization level and ef-
ciency of commercial building operations in megalopolises on three
scales. From 2000 to 2018, the total decarbonization of the ve
megalopolises of Jing-Jin-Ji, Yangtze River Delta, Yangtze River
Middle Reach, Pearl River Delta, and Cheng-Yu was 5.7, 3.1, 1.7, 1.6,
and 0.9 MtCO
2
/yr, which corresponds to 5.6%, 4.0%, 3.5%, 3.5%,
and 3.8% of their respective total carbon emissions. Moreover, the
decarbonization per oor space for these regions was 7.8, 2.0, 1.5,
2.3 and 1.3 kgCO
2
/m
2
/yr, and the decarbonization per capita for
these regions was 55.6, 19.5, 9.8, 15.3 and 8.1 kgCO
2
per capita per
yr respectively.
6.2. Future direction
To further deepen the theme of decarbonization and accelerate the
process of moving commercial buildings toward net-zero emissions,
several other issues deserve future investigation. Because commercial
buildings rely heavily on fossil energy for their energy consumption and
serve end-use activities such as space heating and water heating, the
inuence of end-use behavior and energy structure on carbon emissions
from commercial buildings should be further investigated in detail in
future studies. Although the decarbonization assessment in this work
provides a baseline level of historical emissions for commercial build-
ings, the projected decarbonization potential of commercial buildings is
still unclear. Therefore, an assessment framework for evaluating future
decarbonization of building operations should be developed to clarify
the roadmap for deep decarbonization. In addition, the remaining
budget and allocation of carbon emissions from commercial buildings at
the megalopolis level should also be addressed as a priority in future
studies, which would help control the total amount of carbon emissions
from commercial building operations to ease the pressure on China to
achieve its carbon peak by 2030.
CRediT authorship contribution statement
Minda Ma: Writing – review & editing, Writing – original draft,
Validation, Supervision, Resources, Project administration, Methodol-
ogy, Funding acquisition, Data curation, Conceptualization. Wei Feng:
Validation, Supervision, Resources, Writing – review & editing. Jing-
wen Huo: Validation. Xiwang Xiang: Writing – original draft, Visual-
ization, Validation, Software, Investigation.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgment
This study was supported by the Graduate Research and Innovation
Foundation of Chongqing, China (CYS22071), the National Planning
Ofce of Philosophy and Social Science Foundation of China
(21CJY030), the Beijing Natural Science Foundation (8224085), the
China Postdoctoral Science Foundation (2020M680020), and the
Shuimu Tsinghua Scholar Program of Tsinghua University
(2019SM139).
Appendix A-C. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.buildenv.2022.109705.
References
[1] B. Probst, S. Touboul, M. Glachant, A. Dechezleprˆ
etre, Global trends in the
invention and diffusion of climate change mitigation technologies, Nat. Energy 6
(2021) 1077–1086.
[2] K. Li, M. Ma, X. Xiang, W. Feng, Z. Ma, W. Cai, et al., Carbon reduction in
commercial building operations: a provincial retrospection in China, Appl. Energy
306 (2022), 118098.
[3] X. Xiang, X. Ma, Z. Ma, M. Ma, Operational carbon change in commercial buildings
under the carbon neutral goal: a LASSO–WOA approach, Buildings 12 (2022) 54.
[4] Y. Qiu, M.E. Kahn, Better sustainability assessment of green buildings with high-
frequency data, Nat. Sustain. 1 (2018) 642–649.
[5] S. Azimi, W. O’Brien, Fit-for-purpose: measuring occupancy to support commercial
building operations: a review, Build. Environ. 212 (2022), 108767.
[6] W. Yu, J. Shi, Y. Fang, A. Xiang, X. Li, C. Hu, et al., Exploration of urbanization
characteristics and their effect on the urban thermal environment in Chengdu,
China, Build. Environ. 219 (2022), 109150.
[7] T. Ruan, D. Rim, Indoor air pollution in ofce buildings in mega-cities: effects of
ltration efciency and outdoor air ventilation rates, Sustain. Cities Soc. 49 (2019),
101609.
[8] L. Dong, L. Longwu, W. Zhenbo, C. Liangkan, Z. Faming, Exploration of coupling
effects in the economy-society-environment system in urban areas: case study of
the Yangtze River Delta urban agglomeration, Ecol. Indicat. 128 (2021), 107858.
[9] M. Ma, X. Ma, W. Cai, W. Cai, Low carbon roadmap of residential building sector in
China: historical mitigation and prospective peak, Appl. Energy 273 (2020),
115247.
[10] Z. Ye, K. Cheng, S.-C. Hsu, H.-H. Wei, C.M. Cheung, Identifying critical building-
oriented features in city-block-level building energy consumption: a data-driven
machine learning approach, Appl. Energy 301 (2021), 117453.
[11] W. Ji, K. Zhao, B. Zhao, The trend of natural ventilation potential in 74 Chinese
cities from 2014 to 2019: impact of air pollution and climate change, Build.
Environ. 218 (2022), 109146.
[12] Y. Zhou, H. Zhao, S. Mao, G. Zhang, Y. Jin, Y. Luo, et al., Studies on urban park
cooling effects and their driving factors in China: considering 276 cities under
different climate zones, Build. Environ. 222 (2022), 109441.
[13] J. Zhang, Z. Yu, Y. Cheng, C. Chen, Y. Wan, B. Zhao, et al., Evaluating the
disparities in urban green space provision in communities with diverse built
environments: the case of a rapidly urbanizing Chinese city, Build. Environ. 183
(2020), 107170.
[14] H. Lu, M. de Jong, Y. Song, M. Zhao, The multi-level governance of formulating
regional brand identities: evidence from three Mega City Regions in China, Cities
100 (2020), 102668.
[15] E. Taveres-Cachat, S. Grynning, J. Thomsen, S. Selkowitz, Responsive building
envelope concepts in zero emission neighborhoods and smart cities - a roadmap to
implementation, Build. Environ. 149 (2019) 446–457.
[16] C. Chen, L. Bi, Study on spatio-temporal changes and driving factors of carbon
emissions at the building operation stage- A case study of China, Build. Environ.
219 (2022), 109147.
[17] Y. Wang, D. Hu, C. Yu, Y. Di, S. Wang, M. Liu, Appraising regional anthropogenic
heat ux using high spatial resolution NTL and POI data: a case study in the
Beijing-Tianjin-Hebei region, China, Environ. Pollut. 292 (2022), 118359.
[18] H. Liu, D.L. Mauzerall, Costs of clean heating in China: evidence from rural
households in the Beijing-Tianjin-Hebei region, Energy Econ. 90 (2020), 104844.
[19] C. Yu, D. Hu, S. Wang, S. Chen, Y. Wang, Estimation of anthropogenic heat ux and
its coupling analysis with urban building characteristics - a case study of typical
cities in the Yangtze River Delta, China, Sci. Total Environ. 774 (2021), 145805.
[20] Y. Lv, H. Peng, M. He, Y. Huang, J. Wang, Denition of typical commercial
building for South China’s Pearl River Delta: local data statistics and model
development, Energy Build. 190 (2019) 119–131.
[21] J. Geng, J. Wang, J. Huang, D. Zhou, J. Bai, J. Wang, et al., Quantication of the
carbon emission of urban residential buildings: the case of the Greater Bay Area
cities in China, Environ. Impact Assess. Rev. 95 (2022), 106775.
[22] M. Cai, Y. Shi, C. Ren, Developing a high-resolution emission inventory tool for
low-carbon city management using hybrid method - a pilot test in high-density
Hong Kong, Energy Build. 226 (2020), 110376.
M. Ma et al.

Building and Environment 226 (2022) 109705
14
[23] W. Pan, J. Du, Impacts of urban morphological characteristics on nocturnal
outdoor lighting environment in cities: an empirical investigation in Shenzhen,
Build. Environ. 192 (2021), 107587.
[24] Y. Sun, Q. Hao, C. Cui, Y. Shan, W. Zhao, D. Wang, et al., Emission accounting and
drivers in East African countries, Appl. Energy 312 (2022), 118805.
[25] Y. Lu, P. Cui, D. Li, Carbon emissions and policies in China’s building and
construction industry: evidence from 1994 to 2012, Build. Environ. 95 (2016)
94–103.
[26] X. Chen, C. Shuai, Y. Wu, Y. Zhang, Analysis on the carbon emission peaks of
China’s industrial, building, transport, and agricultural sectors, Sci. Total Environ.
709 (2020), 135768.
[27] Y. Lu, P. Cui, D. Li, Which activities contribute most to building energy
consumption in China? A hybrid LMDI decomposition analysis from year 2007 to
2015, Energy Build. 165 (2018) 259–269.
[28] X. Zhong, M. Hu, S. Deetman, J.F.D. Rodrigues, H.-X. Lin, A. Tukker, et al., The
evolution and future perspectives of energy intensity in the global building sector
1971-2060, J. Clean. Prod. 305 (2021), 127098.
[29] J. He, Q. Yue, Y. Li, F. Zhao, H. Wang, Driving force analysis of carbon emissions in
China’s building industry: 2000-2015, Sustain. Cities Soc. 60 (2020), 102268.
[30] B. Lin, H. Liu, CO2 mitigation potential in China’s building construction industry: a
comparison of energy performance, Build. Environ. 94 (2015) 239–251.
[31] M. Zhang, Y. Song, P. Li, H. Li, Study on affecting factors of residential energy
consumption in urban and rural Jiangsu, Renew. Sustain. Energy Rev. 53 (2016)
330–337.
[32] S. Shao, J. Liu, Y. Geng, Z. Miao, Y. Yang, Uncovering driving factors of carbon
emissions from China’s mining sector, Appl. Energy 166 (2016) 220–238.
[33] A. Vaninsky, Factorial decomposition of CO2 emissions: a generalized Divisia index
approach, Energy Econ. 45 (2014) 389–400.
[34] J. Boraty´
nski, Decomposing structural decomposition: the role of changes in
individual industry shares, Energy Econ. 103 (2021), 105587.
[35] N. Roux, T. Kastner, K.-H. Erb, H. Haberl, Does agricultural trade reduce pressure
on land ecosystems? Decomposing drivers of the embodied human appropriation of
net primary production, Ecol. Econ. 181 (2021), 106915.
[36] X. Yang, M. Hu, C. Zhang, B. Steubing, Key strategies for decarbonizing the
residential building stock: results from a spatiotemporal model for Leiden, The
Netherlands, Resour. Conserv. Recycl. 184 (2022), 106388.
[37] E. Ohene, A.P.C. Chan, A. Darko, Prioritizing barriers and developing mitigation
strategies toward net-zero carbon building sector, Build. Environ. 223 (2022),
109437.
[38] S. Zhang, M. Ma, K. Li, Z. Ma, W. Feng, W. Cai, Historical carbon abatement in the
commercial building operation: China versus the US, Energy Econ. 105 (2022),
105712.
[39] B. Lin, H. Liu, CO2 emissions of China’s commercial and residential buildings:
evidence and reduction policy, Build. Environ. 92 (2015) 418–431.
[40] N. Zhou, N. Khanna, W. Feng, J. Ke, M. Levine, Scenarios of energy efciency and
CO2 emissions reduction potential in the buildings sector in China to year 2050,
Nat. Energy 3 (2018) 978–984.
[41] B.-J. Tang, Y.-Y. Guo, B. Yu, L.D.D. Harvey, Pathways for decarbonizing China’s
building sector under global warming thresholds, Appl. Energy 298 (2021),
117213.
[42] Y. Yu, X. Zhou, W. Zhu, Q. Shi, Socioeconomic driving factors of PM2.5 emission in
Jing-Jin-Ji region, China: a generalized Divisia index approach, Environ. Sci.
Pollut. Control Ser. 28 (2021) 15995–16013.
[43] R. Yan, X. Xiang, W. Cai, M. Ma, Decarbonizing residential buildings in the
developing world: historical cases from China, Sci. Total Environ. 847 (2022),
157679.
[44] Y. Pan, F. Dong, Dynamic evolution and driving factors of new energy
development: fresh evidence from China, Technol. Forecast. Soc. Change 176
(2022), 121475.
[45] J. Wang, Q. Jiang, X. Dong, K. Dong, Decoupling and decomposition analysis of
investments and CO2 emissions in information and communication technology
sector, Appl. Energy 302 (2021), 117618.
[46] Z. Miao, S. Liu, X. Chen, Driving factors and spatio-temporal features underlying
industrial SO2 emissions in “2+26” in North China and extended cities, Chin. J.
Popu. Res. Environ. 18 (2020) 296–318.
[47] B. Li, S. Han, Y. Wang, Y. Wang, J. Li, Y. Wang, Feasibility assessment of the carbon
emissions peak in China’s construction industry: factor decomposition and peak
forecast, Sci. Total Environ. 706 (2020), 135716.
[48] X. Zhang, Y. Geng, S. Shao, J. Wilson, X. Song, W. You, China’s non-fossil energy
development and its 2030 CO2 reduction targets: the role of urbanization, Appl.
Energy 261 (2020), 114353.
[49] H-x Wen, Z. Chen, Q. Yang, J-y Liu, P-y Nie, Driving forces and mitigating
strategies of CO2 emissions in China: a decomposition analysis based on 38
industrial sub-sectors, Energy 245 (2022), 123262.
[50] Y. Wang, Y. Zhou, L. Zhu, F. Zhang, Y. Zhang, Inuencing factors and decoupling
elasticity of China”s transportation carbon emissions, Energies 11 (2018) 1157.
[51] D. Fang, P. Hao, Q. Yu, J. Wang, The impacts of electricity consumption in China’s
key economic regions, Appl. Energy 267 (2020), 115078.
[52] Q. Yan, Y. Wang, T. Baleˇ
zentis, D. Streimikiene, Analysis of China’s regional
thermal electricity generation and CO2 emissions: decomposition based on the
generalized Divisia index, Sci. Total Environ. 682 (2019) 737–755.
[53] R. Ehrlich Paul, P. Holdren John, Impact of population growth, Science 171 (1971)
1212–1217.
[54] R. York, E.A. Rosa, T. Dietz, STIRPAT, IPAT and ImPACT: analytic tools for
unpacking the driving forces of environmental impacts, Ecol. Econ. 46 (2003)
351–365.
[55] S. Zhang, M. Ma, X. Xiang, W. Cai, W. Feng, Z. Ma, et al., Potential to decarbonize
the commercial building operation of the top two emitters by 2060, Resour.
Conserv. Recycl. 185 (2022), 106481.
[56] X. Xiang, X. Ma, Z. Ma, M. Ma, W. Cai, Python-LMDI: a tool for index
decomposition analysis of building carbon emissions, Buildings 12 (2022) 83.
[57] X. Xiang, M. Ma, X. Ma, L. Chen, W. Cai, W. Feng, et al., Historical decarbonization
of global commercial building operations in the 21st century, Appl. Energy 401
(2022), 119401.
[58] Y. Wang, X. Su, L. Qi, P. Shang, Y. Xu, Feasibility of peaking carbon emissions of
the power sector in China’s eight regions: decomposition, decoupling, and
prediction analysis, Environ. Sci. Pollut. Control Ser. 26 (2019) 29212–29233.
[59] C. Shuai, B. Zhao, X. Chen, J. Liu, C. Zheng, S. Qu, et al., Quantifying the impacts of
COVID-19 on Sustainable Development Goals using machine learning models,
Fundament. Res. (2022).
[60] D. Diakoulaki, M. Mandaraka, Decomposition analysis for assessing the progress in
decoupling industrial growth from CO2 emissions in the EU manufacturing sector,
Energy Econ. 29 (2007) 636–664.
[61] X. Chen, C. Shuai, Y. Wu, Y. Zhang, Understanding the sustainable consumption of
energy resources in global industrial sector: evidences from 114 countries, Environ.
Impact Assess. Rev. 90 (2021), 106609.
[62] B. Yu, D. Fang, Decoupling economic growth from energy-related PM2.5 emissions
in China: a GDIM-based indicator decomposition, Ecol. Indicat. 127 (2021),
107795.
[63] J. Liu, Q. Yang, S. Ou, J. Liu, Factor decomposition and the decoupling effect of
carbon emissions in China’s manufacturing high-emission subsectors, Energy 248
(2022), 123568.
[64] Z. Wang, J. Xu, Y. Wang, L.W. Wang, R.Z. Wang, Annual energy simulation for the
air conditioning of Fuxing high speed trains, Appl. Therm. Eng. 188 (2021),
116591.
[65] L. Gan, H. Ren, W. Cai, K. Wu, Y. Liu, Y. Liu, Allocation of carbon emission quotas
for China’s provincial public buildings based on principles of equity and efciency,
Build. Environ. 216 (2022), 108994.
[66] X. Chen, B. Zhao, C. Shuai, S. Qu, M. Xu, Global spread of water scarcity risk
through trade, Resour. Conserv. Recycl. 187 (2022), 106643.
[67] H. Zhang, X. Sun, C. Bi, M. Ahmad, J. Wang, Can sustainable development policy
reduce carbon emissions? Empirical evidence from resource-based cities in China,
Sci. Total Environ. 838 (2022), 156341.
[68] P. Tang, H. Zeng, S. Fu, Local government responses to catalyse sustainable
development: learning from low-carbon pilot programme in China, Sci. Total
Environ. 689 (2019) 1054–1065.
[69] L. Zhang, X. Yang, Y. Fan, J. Zhang, Utilizing the theory of planned behavior to
predict willingness to pay for urban heat island effect mitigation, Build. Environ.
204 (2021), 108136.
[70] Beijing_Municipal_Commission_of_Housing_and_Urban-Rural_Development, Design
Standard of Green Buildings, May 20, 2021. in Chinese, http://scjgj.beijing.gov.
cn/zwxx/gs/dfbzgg/202104/t20210408_2349427.html.
[71] P.Y. Tan, J. Wang, A. Sia, Perspectives on ve decades of the urban greening of
Singapore, Cities 32 (2013) 24–32.
[72] J. Coma, G. P´
erez, A. de Gracia, S. Bur´
es, M. Urrestarazu, L.F. Cabeza, Vertical
greenery systems for energy savings in buildings: a comparative study between
green walls and green facades, Build. Environ. 111 (2017) 228–237.
[73] M. Stadler, M. Kloess, M. Groissb¨
ock, G. Cardoso, R. Sharma, M.C. Bozchalui, et al.,
Electric storage in California”s commercial buildings, Appl. Energy 104 (2013)
711–722.
[74] R.C. Johnson, M. Royapoor, M. Mayeld, A multi-zone, fast solving, rapidly
recongurable building and electried heating system model for generation of
control dependent heat pump power demand proles, Appl. Energy 304 (2021),
117663.
[75] R. Li, A.J. Satchwell, D. Finn, T.H. Christensen, M. Kummert, J. Le Dr´
eau, et al.,
Ten questions concerning energy exibility in buildings, Build. Environ. 223
(2022), 109461.
[76] L. Wang, D. Zheng, Integrated analysis of energy, indoor environment, and
occupant satisfaction in green buildings using real-time monitoring data and on-
site investigation, Build. Environ. 182 (2020), 107014.
[77] J.-J. Ma, G. Du, Z.-K. Zhang, P.-X. Wang, B.-C. Xie, Life cycle analysis of energy
consumption and CO2 emissions from a typical large ofce building in Tianjin,
China, Build. Environ. 117 (2017) 36–48.
M. Ma et al.

Building and Environment 226 (2022) 109705
15
[78] L. Zhang, J. Wen, Y. Li, J. Chen, Y. Ye, Y. Fu, et al., A review of machine learning in
building load prediction, Appl. Energy 285 (2021), 116452.
[79] H. Yin, X. Zhai, Y. Ning, Z. Li, Z. Ma, X. Wang, et al., Online monitoring of PM2.5
and CO2 in residential buildings under different ventilation modes in Xi’an city,
Build. Environ. 207 (2022), 108453.
[80] Y. Zhang, L. Qi, X. Lin, H. Pan, B. Sharp, Synergistic effect of carbon ETS and
carbon tax under China’s peak emission target: a dynamic CGE analysis, Sci. Total
Environ. 825 (2022), 154076.
[81] X. Ren, Y. Li, C. yan, F. Wen, Z. Lu, The interrelationship between the carbon
market and the green bonds market: evidence from wavelet quantile-on-quantile
method, Technol. Forecast. Soc. Change 179 (2022), 121611.
[82] X. Wang, X. Wang, X. Ren, F. Wen, Can digital nancial inclusion affect CO2
emissions of China at the prefecture level? Evidence from a spatial econometric
approach, Energy Econ. 109 (2022), 105966.
M. Ma et al.