Energy Policy 82 (2015) 233–248

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

China's transportation energy consumption and CO2 emissions from a

global perspective
Xiang Yin a, Wenying Chen a,n, Jiyong Eom b,c, Leon E. Clarke b, Son H. Kim b, Pralit L. Patel b,
Sha Yu b, G. Page Kyle b
a
Institute of Energy, Environment and Economy, Tsinghua University, China
b
Joint Global Change Research Institute, Pacific Northwest National Laboratory, USA
c
Korea Advanced Institute of Science and Technology, South Korea

H I G H L I G H T S

Transport sector in China are analyzed from a global perspective.

Passenger transport turnover reduction and modal shifts is less sensitive to carbon price.

Bio-fuel, electricity and H2 will play an important role for carbon mitigation in transport sector.

The transport sector is more difficult to decarbonize than other sectors.

art ic l e i nf o a b s t r a c t

Article history: Rapidly growing energy demand from China's transportation sector in the last two decades have raised
Received 14 October 2014 concerns over national energy security, local air pollution, and carbon dioxide (CO2) emissions, and there
Received in revised form is broad consensus that China's transportation sector will continue to grow in the coming decades. This
31 January 2015
paper explores the future development of China's transportation sector in terms of service demands, final
Accepted 21 March 2015
Available online 2 April 2015
energy consumption, and CO2 emissions, and their interactions with global climate policy. This study
develops a detailed China transportation energy model that is nested in an integrated assessment model
Keywords: —Global Change Assessment Model (GCAM)—to evaluate the long-term energy consumption and CO2
Integrated assessment emissions of China's transportation sector from a global perspective. The analysis suggests that, without
Transport sector
major policy intervention, future transportation energy consumption and CO2 emissions will continue to
CO2 mitigation
rapidly increase and the transportation sector will remain heavily reliant on fossil fuels. Although carbon
GCAM-China
price policies may significantly reduce the sector's energy consumption and CO2 emissions, the asso-
ciated changes in service demands and modal split will be modest, particularly in the passenger trans-
port sector. The analysis also suggests that it is more difficult to decarbonize the transportation sector
than other sectors of the economy, primarily owing to its heavy reliance on petroleum products.
& 2015 Elsevier Ltd. All rights reserved.

1. Introduction
The International Energy Agency (IEA) estimates that the
transportation sector accounts for approximately 19% of global
energy consumption and 23% of energy-related carbon dioxide
(CO2) emissions (IEA, 2012). The sector's energy consumption and
CO2 emissions, and their contributions relative to other sectors, are
all projected to grow substantially over the course of the current
century (Kyle and Kim, 2011). The IEA projects that global trans-
portation energy use and CO2 emissions will increase by
n
Corresponding author. Fax: þ 86 10 62772759.
E-mail address: chenwy@tsinghua.edu.cn (W. Chen).
approximately 50% by 2030 and by over 80% by 2050 (IEA, 2009).
China's transportation sector has grown rapidly in recent years,
and it is probable that this will continue in the coming decades
given the country's current low vehicle ownership level. As of
2010, the transportation sector accounts for approximately 15.6%
of total final energy use in China with gasoline and diesel con-
sumed in the secondary and tertiary for transport included (Na-
tional Bureau of Statistics (NBS, 2011; Wang, 2010), considerably
below the average figure of 32% for Organization for Economic
Cooperation and Development (OECD) countries (IEA, 2012).
A number of studies have provided scenarios for energy con-
sumption and CO2 emissions in China's transportation sector.
There are energy consumption and CO2 emissions scenarios for

http://dx.doi.org/10.1016/j.enpol.2015.03.021
0301-4215/& 2015 Elsevier Ltd. All rights reserved.
234 X. Yin et al. / Energy Policy 82 (2015) 233–248
transportation and other sectors from 2005 to 2050 with MARKAL
(MARKet ALlocation) models, including MARKAL-MACRO (Chen,
2005), MARKAL-ED (Chen et al., 2007), Western China MARKAL
(Chen et al., 2010), and China TIMES (Chen et al., 2013). These
studies disaggregate the transportation sector into 10 modes (air,
railway, bus, car and waterway for passenger transport, and air,
railway, truck, waterway and pipeline for freight transport). Zhou
et al. (2013) provided projections through to 2050 using the Long
range Energy Alternative Planning (LEAP) system model. This
study calculated freight transportation service demand as a func-
tion of economic activity, and passenger transport on assumed
average kilometer traveled by individual modes (e.g., bus, train,
and car). Fu (2011) analyzed medium- and long-term energy sav-
ing potentials in different energy efficiency improvement scenar-
ios. Other studies analyzed the sector's energy consumption and
CO2 emissions from a global perspective (Koljonen and Lehtila,
2012; Luderer et al., 2012; Pietzcker et al., 2014).
Based on the historical trend during 2000–2005, Yan and
Crookes (2009) projected road energy consumption and CO2
emissions from 2005 to 2050 in both business-as-usual and best-
case scenarios. In the best case scenario, a series of policy mea-
sures such as private vehicle controls and fuel tax are assumed.
Huo et al. (2012a) developed the Fuel Economy and Environmental
Impacts (FEEI) model for road transportation, and projected the
well-to-wheels and tank-to-wheels energy use and greenhouse
gas emissions up to 2050. Specifically, the model presented var-
ious vehicle types, such as private light-duty passenger vehicles
(LDVs), taxis, business LDVs (owned by the government and
companies), public buses (for urban and rural transport), intercity
buses, and light-and heavy-duty trucks. Hao et al. (2011, 2012) also
separated passenger and freight vehicles for their analysis. Several
studies analyzed energy consumption and CO2 emissions of rural
vehicles (Sperling et al., 2005; Yao et al., 2011), and railways (He
et al., 2010). These bottom-up studies analyzed passenger and
freight sectors in detail, providing scenario analyses at various
scales informed by disaggregated technological and socioeconomic
representations.
The major limitation of these models is the inability of re-
presenting how various transportation service demands and their
modal split may evolve over time and interact with long-term
climate mitigation target to limit warming to about 2 degrees.
Moreover, some previous studies do not separately represent ur-
ban, rural and intercity passenger transport despite significant
differences in travel patterns, vehicle speed, and fuel economy,
because of different road infrastructure and traffic conditions
(Brand et al., 2012; Girod et al., 2012; Zhang et al., 2007). Trans-
portation service is considered to increase with income (Kim et al.,
2006; Zhou et al., 2013), hence urban/rural income inequalities
may contribute to different development paths for per capita
travel and transportation energy use in these areas.
This study attempts to address the literature gap by nesting a
highly disaggregated Chinese transportation sector in the Global
Change Assessment Model (GCAM) and the improved model is
named as GCAM-China. This paper explores the development of
different transportation services and fuel consumption using re-
ference and various carbon price scenarios in China with appli-
cation of the improved GCAM-China model. Section 2 of this paper
describes the disaggregated China transportation model and the
basic assumptions for the reference scenario. Section 3 presents
the reference scenario results, focusing on the service demands by
mode, final energy use and CO2 emissions. Section 4 explores the
effect of carbon policy on the sector's development. The final
section summarizes the results and concluds with policy
implications.
2. Methods
2.1. Current transportation energy use in China
Characterizing the current patterns of transportation energy
consumption in China is important because it reveals consumer
preferences for competing modes and technologies and thereby
shapes the extent that fuel substitution and modal shifts can occur
in response to external drivers. There are official energy con-
sumption statistics (NBS, 2011) for the Chinese transportation
sector. However, the definition and coverage of the statistics differ
from other countries (Mao et al., 2009; Wang, 2010). For example,
road transportation energy consumption statistics only covers
business vehicle fuel consumption, while other transportation
energy consumption is assigned to residential and commercial
sectors (Jia et al., 2010). Thus, an engineering approach estimates
approximately a 75% higher energy demand in the Chinese
transportation sector than the statistics indicate (Mao et al., 2009).
However, such an approach requires assumptions relating to cur-
rent fuel consumption (measured in liters or MJ of fuel per
100 km) and vehicle use intensity (measured in kilometers of
travel per vehicle per year), neither of which are currently fully
known in China (Huo et al., 2012b).
To estimate final transportation energy by fuel, Wang (2010)
applied an oil allocation method. His study suggests that, in ad-
dition to oil consumed by business vehicles, 95% of the gasoline
and 35% of the diesel used in the industry and commercial sectors,
and all the gasoline and 95% of the diesel used in residential and
agriculture sectors are attributable to the transportation sector.
Since this simple method can estimate transportation energy use
by fuel, it is used in various studies (Qi, 2012).
Our study uses an engineering approach to calculate trans-
portation energy consumption by fuel and mode, all calibrated to
fuel-level transportation energy consumption estimated by the oil
allocation method. This method reveals that total transportation
energy consumption has increased from 4.165 EJ in 2000 to
9.428 EJ in 2010, with an annual average growth rate of 8.5%
(Fig. 1). Oil products were the main energy sources in the trans-
portation sector, accounting for over 92% of total transportation
energy consumption. Coal consumption decreased in conjunction
with the phase-out of steam locomotives during this period, being
substituted by oil and electricity. The share of natural gas and
electricity remained relatively low.
2.2. Transport policies in China
In the transportation sector, there are four key policy me-
chanisms: fuel economy standards and labeling, vehicle and fuel
taxation, public transportation, and subsides for energy-efficient
vehicles, electric and H2 vehicles (Lo, 2014).
2.2.1. Fuel economy standards and labelling
Fuel economy standards (FES) are regarded to be the most
important measure to improve automobile energy efficiency. FES
regulate fuel consumption or carbon emissions per distance tra-
veled of vehicle, and require automakers to design and produce
more efficient vehicles. The phase 1 standards were introduced in
July 2005. They set up maximum allowance fuel consumption
limits by weight category rather than fleet average. After the im-
plementation of the phase 1 standards, average fuel economy of
new vehicles had decreased from 9.11 L/100 km to 8.06 L/100 km
during 2002 and 2006. The phase 2 standards were introduced in
January 2008, and put forward even higher requirements.
Unlike the phase 1 and phase 2 standards, the fleet average
system was introduced in the phase 3 standards. Compared to the
phase 2 standards, the phase 3 standards require average energy
 X. Yin et al. / Energy Policy 82 (2015) 233–248 235

Fig. 1. Energy consumption in China's transport sector 2000–2010.

intensity to decrease to 7 L/100 km by 2015. The phase 3 standards different studies, but they all believe it will increase constantly,
also present the requirement of average fuel economy for vehicle with the growth rate declining over time. IEA (2013) believes the
producers. Considering producers might not able to meet the re- average annual growth rate will 6.6% and 4.8% during 2010–2020
quirement in the short period. The new standards proposed a and 2020–2030 respectively. ERI (2010) and He (2013) are much
progressive goal, and will take full effect in 2015. more optimistic, believing that the average growth rate of GDP in
In 2012, the Chinese government introduced the phase 4 stan- China will reach 8.4% and 5.1% during 2010–2020, 7.5% and 5.0%
dards, which will take full effect in 2020, and achieve the target of during 2020–2050 respectively. Considering the Chinese govern-
5.0 L/100 km in 2020. ment is now shifting its focus from high economic growth to more
sustainable economic development, this study assume the Chinese
2.2.2. Vehicle and fuel taxation economy growing at an average rate of 8% from 2005 to 2020
In China, vehicles are subjected to four taxes: vehicle purchase (considering the average growth rate of 11.2% from 2005 to 2010,
tax, vehicle and vessel tax, consumption tax and a 17% value-ad- economic growth rate is about 6.5% from 2010 to 2020), 4.6% from
ded tax. Vehicle purchase tax was a universal 10% of the vehicle 2020 to 2050, and 2.6% from 2050 to 2095, as shown in Table 1.
value, but for new energy vehicles, including electric vehicles Population and income in urban and rural areas, major drivers
(EVs) and fuel cell vehicles (FCVs), it is reduced to be 0% from of passenger transport demands, are assumed to develop con-
September 1st 2014 to December 31st 2017. The vehicle and vessel sistently with the above national-level development pathways.
tax standards of passenger vehicles are divided into seven levels Following the assumptions made by Chen et al. (2013), the urba-
according to displacement. The annual tax baseline is between 60 nization level in China increases from 43.4% in 2005 to 72.1% in
and 360 Yuan under 1.0 L, and 3600 Yuan to 5400 Yuan over 4.0 L. 2050 and 82.4% in 2095. China's urban-to-rural per capita income
The vehicle consumption rate is also calculated based on engine ratio that had increased from 2.4 in 1997 to 3.1 in 2011 (CEG, 2008;
capacity, from 1% to 40%, but 0% for EVs, FCVs and plug-in hybrid NBS, 2012), starts to decline as the pace of urbanization slows. By
vehicle (PHEV). 2050 and 2095, the per capita income in urban areas will be ap-
proximately 1.9 and 1.5 times that of rural areas, respectively.
2.2.3. Public transportation
Energy savings and CO2 mitigation can also be achieved by
2.4. The China transportation energy model
modal shifts, typically by improving public transportation. China
starts to develop high speed rail (HSR) from 2004, and total op-
2.4.1. Representing transportation services and technologies
erating railway mileage has increased to 15,000 km by 2014, about
Our model divides passenger and freight transportation sectors
50% of global operating railway mileage. Nevertheless, bus rapid
into five passenger services and four freight services. Business LDV
transit (BRT) has also experienced rapid growth over the past few
passenger services owned by government bodies and companies
years, though it has not received as much attention as metro. To-
are differentiated from private passenger services as they do not
day, bus rapid systems are operating in 18 Chinese cities, serving
4.0 million people every day, with total length over 563 km. generally compete. The private passenger transport sector is dis-
aggregated into four distinct inland passenger transport services
2.2.4. Subsides to electric vehicles (EVs) and fuel cell vehicles (FCVs)
Table 1
In order to promote the development of EVs and FCVs, many
Assumptions for population, per capita income and urbanization rate.
policies and regulations have been implemented in China. For
example, the Shanghai municipal government announced a plan 2005 2020 2035 2050 2065 2080 2095
in May 2014 to pay subsides of 40,000 Yuan to each sold battery
electric vehicle (BEV) and 30,000 Chinese Yuan to each PHEV, Population (million)
National 1427 1545 1575 1524 1450 1377 1294
200,000 and 500,000 Chinese Yuan to each sold fuel cell light- Urban 620 865 1034 1099 1115 1107 1067
duty passenger vehicle (LDV) and bus. Rural 807 680 541 425 334 270 227
Urbanization level (%) 43.4 56.0 65.7 72.1 76.9 80.4 82.5
2.3. Socioeconomic development in the future Per capita income (thousand 1990$)
National 1.1 3.3 7.1 13.0 21.8 33.3 48.9
Urban 1.8 4.6 8.8 15.0 24.1 35.8 51.8
China's population will peak around 2035 and then start to Rural 0.6 1.7 4.0 7.9 14.0 23.1 35.2
decline. For economic development, there are large variances from
236 X. Yin et al. / Energy Policy 82 (2015) 233–248
Fig. 2. (a) Representation of passenger transportation services and modes. (b) Representation of freight transportation services and modes.
based on geographical coverage and travel purposes (Fig. 2a): in-
tercity (pass_intercity), urban (pass_urban), rural (pass_rural), and
business (government- and company-owned LDVs, pass_business),
and one international passenger service: international aviation
(pass_intl_aviation). The freight transportation sector (Fig. 2b)
encompasses two inland freight transportation services: domestic
freight (freight_domestic) and rural freight (freight_rural); and
two international freight transportation services: international
ship (freight_intl_ship) and international aviation (freight_-
intl_aviation). Multiple subsectors representing available modal
choices are created within each service type. Passenger transport
includes motorbikes, walking, bicycles, LDV, buses, rail, high speed
rail and air: freight transportation modes are truck, rail, ship, air
and rural vehicles (three- and four-wheel).
We made detailed assumptions on the characteristics of various
transportation modes based on literature. Huo et al. (2012a) pro-
vided the 2009 average vehicle kilometer values by vehicle type,
as well as fuel consumption rates. We used Harvey's (2013) esti-
mates for urban, rural and intercity shares in LDV travel, as well as
assumptions for urban, rural and intercity fuel consumption rates
and speed (Clarke et al., 2007; Edmonds et al., 2006). The total
vehicle population was obtained from China statistics (NBS, 2012),
and the category-level vehicle populations were from previous
studies (Hao et al., 2012; Huo and Wang, 2012). To characterize
rail, ship, and air we use transportation service demand—in pas-
senger-kilometer (pkm) or ton-kilometer (tkm)—and its associated
energy intensity estimates from the survey conducted by Germa-
ny's Institute for Energy and Environmental Research (IFEU, 2008)
and Mao et al. (2009). Girod et al. (2012) provided data on various
transportation mode speeds. For figures obtained from surveys
conducted pre-2010, we extrapolated the figures assuming a
continuation of earlier trends through 2010. Table 2 shows the
speeds, vehicle use intensities, fuel consumption rates, and load
factors in 2010 assumed in this study. These mode-specific as-
sumptions are used to construct transportation service and energy
consumption by mode and fuel in 2010, later proportionately ca-
librated to the estimates obtained by the oil allocation method
(see Fig. 1).
As freight transportation in China mainly comprises hauling
bulk materials such as coal, minerals, and building materials
(Zhou, 2010), it should be noted here that long-distance trucks in
China are often seriously overloaded (IFEU, 2008; Zhou, 2010). This
means that actual transportation energy intensity is lower than
designed standards and the global average (Girod et al., 2012,
 X. Yin et al. / Energy Policy 82 (2015) 233–248 237

Table 2 FPt × FI t + NFPt Wt

Speed, energy intensity and technology improvement assumptions used in this Pt = + ,
LFt St
study.
where FP is the fuel price ($/MJ), FI is the vehicle fuel intensity (MJ/
Sector Mode Speed Energy in- Energy efficiency
(km/h) tensity improvement
VKM), NFP is the non-fuel price ($/VKM), LF is the load factor
(MJ/pkm, (average % change (persons or tons per vehicle), W is the wage rate [$/(person-h)]
MJ/tkm) per year) that is related to economic behavior, and S is the vehicle speed
2010– 2050– (km/h). Fuel prices, including carbon prices, are endogenous and
2050 2095
all other variables are exogenous for different technologies. For
Freight_domestic Truck – 0.58 0.1 0.0 freight transportation modes where hauling time may not incur
Rail – 0.14 0.1 0.1 direct opportunity costs, the time value part is excluded.
Ship – 0.29 1.0 0.5
Air – 1.47 0.1 0.1 2.4.3. Modal split
Freight_intl_aviation Air – 1.47 0.1 0.1
The market share captured by each mode (and transportation
Freight_intl_shipping Ship – 0.09 1.0 0.5
Freight_rural 3W – 2.96 0.1 0.1 technology within the mode) is endogenous and determined using
4W – 2.85 0.1 0.1 a logit formulation. The market share of each transportation mode
Pass_intercity LDV 60 1.25 0.4 0.2 (or technology) is given by
Bus 40 0.45 0.1 0.1
Rail 50 0.28 0.1 0.1 SW iP i, tλ
Air 225 1.07 0.1 0.1 S i, t =
Pass_business LDV 40 1.20 0.4 0.2
∑in SW iP i, tλ
Pass_rural Walk 5 – – –
Cycle 11 – – – where Si,t is market share of each transportation mode (technol-
Motorcycle 21 0.48 0.4 0.2 ogy), SW is the share weight, P is cost of transportation service, λ is
LDV 50 1.23 0.4 0.2 cost distribution parameter, n is the number of transportation
Public bus 20 0.28 0.1 0.1 modes/technologies for each transportation service/mode. The
Pass_urban Walk 4 – – –
Cycle 11 – – –
share weight captures consumer preferences for modal and tech-
Motorcycle 21 0.48 0.4 0.2 nology choice, and rate of technology diffusion that is not ex-
LDV 40 1.33 0.4 0.2 plicitly modeled (Kim et al., 2006). Cost distribution parameter
Public bus 20 0.30 0.1 0.1 regulates the reflection of service price changes on modal shifts.
Metro 22 0.12 0.1 0.1
Then, turnover of each transportation mode and technology is
Pass_intl_aviation Air 340 0.99 0.1 0.1
given by the following equation:
D i, t = D t × S i, t
2013). Importantly, Chinese bus load factors are considerably
higher than developed countries. Such overloading problems are where Di,t is the turnover of transportation mode/technology, Si,t is
likely to be alleviated in the future, hence the energy intensity of its share, Dt is the turnover of transportation service/mode.
buses and trucks are assumed to decrease more slowly than LDVs
in this study (Table 2). 2.5. Global change assessment model (GCAM)

2.4.2. Demands for transportation services
In our model, the demand for transportation services is de-
termined by income, population, and the weighted average service
cost of various transportation modes and technologies (Kim et al.,
2006). Passenger transport services (passenger kilometers) in
period t is given by
D t = kItα Ptβ N t.
where k represents a calibration coefficient; I is per capita income
or regional income (for urban and rural transport); P is the
weighted-average service price; N is total population or regional
populations (for urban and rural transport); α and β are income
and price elasticities, respectively. Note that urban and rural pas-
sengers have separate socioeconomic drivers: rural passenger
transport service is driven by rural population and rural per capita
income, while urban passenger service is driven by urban popu-
lation and urban per capita income. Freight transportation service
demands, which rely on the aggregate level of economic activity,
are similarly formulated but are scaled by total income, Y, rather
than per capita income and population. Freight transportation
service in period t is given as follows:
D t = kYtα Ptβ.
The transportation service price as expressed by the cost per
unit of service demand is the weighted average of technology-
specific service prices that includes the technology's capital and
fuel costs and the time value of using it. The service price (Kyle and
Kim, 2011) of using a transportation mode in period t is given by
The GCAM is a dynamic-recursive model with the representa-
tion of numerous energy technologies and agriculture and land
use decisions. The emissions and concentrations of 16 greenhouse
gases and short-lived species are tracked in the model (Brenkert
et al., 2003; Edmonds et al., 1997; Kim et al., 2006). The current
version of GCAM runs in a 5-year time step from 2005 to 2095 for
energy and land use systems for 14 regions including China. GCAM
allows for enhanced detail where needed, while ensuring feedback
and interaction between all model components. Therefore, it has
been widely used to develop energy consumption and CO2 emis-
sion scenarios for individual regions, sectors and technologies,
such as the development of concentrated solar power (Smith et al.,
2010) and biomass (Luckow et al., 2010), and building energy
consumption in the United States (Kyle et al., 2007), China (Eom
et al., 2012a) and India (Chaturvedi et al., 2013).
The most recent representation of the transportation sector in
GCAM is featured by three disaggregated services for global re-
gions—passenger, freight and international shipping services—
with multiple modal choices and vehicle technologies (Kim et al.,
2006). However, the transportation model still requires major
improvements particularly for developing regions including China,
in terms of the representation of sectoral structure, technology
efficiencies, non-energy costs, load factors and transit speeds.
Because of the lack of local authority information, the GCAM
transportation model has not been customized for developing
regions including China (Wise et al., 2010).
In this paper, GCAM is customized for China by nesting the
China Transportation Energy Model in it. The improved model
238 X. Yin et al. / Energy Policy 82 (2015) 233–248
(GCAM-China) can reflect the interactions of the transportation
sector with other sectors in China and other global regions.
3. Results
3.1. Results without climate policy
3.1.1. Service output and modal split
In the reference scenario, both passenger and freight trans-
portation service demands continue to expand over this century
(Fig. 3). Passenger transportation service saturates around 2085 as
a result of rapidly decreasing population and continued increase in
travel costs, although it still increases over 3-fold by the century
end. Meanwhile, per capita motorized travel increases approxi-
mately 5 times from 4.9 thousand pkm to 24.1 thousand pkm. As a
comparison, the estimated U.S. per capita travel in 2008 was
25.8 thousand pkm (International Transport Forum, 2012). The per
capita inland freight service grows nearly 4 times from
6.3 thousand tkm to 22.9 thousand tkm, which is comparable to
Australia, the U.S. and Canada in 2007 (Eom et al., 2012b; Inter-
national Transport Forum, 2012).
Intercity transport and urban transport continue to account for
a majority of passenger travel in China (Fig. 4). In 2010, these two
services accounted for approximately 40% of total passenger
transport, followed by rural (17%), business (3%) and international
aviation (1%) services. Intercity travel increases faster than urban
travel, with intercity travel accounting for the largest share (54%)
in total passenger transport activity by the century end. Interna-
tional passenger service experiences the fastest increase albeit
from a smaller base: its share in total passenger transport in-
creases from 0.9% to 3.1% over the century. By contrast, rural
passenger transport continues to decline after 2025 because of
rapid urbanization and the resulting decrease in rural population.
Domestic freight transportation continues to account for the lar-
gest proportion of freight transportation with its share decreasing
from 66% to 61%. The slight decrease is because of the increasing
roles played by international freight shipping and to a lesser ex-
tent international freight aviation, although the share of interna-
tional freight aviation remains tiny with no more than 0.2% by the
century end.
The increase of passenger travel service demand can also be
manifested by people spending more time on traveling; the
average daily travel time increases from 1.1 h in 2010 to 1.8 h in
2095. Walking and cycling time decreases in the future, leading to
a rapid increase in motorized travel time from 0.3 to 1.2 h per day;
this is comparable to the assumed 1.1 h by 2100 in the global
multi-regional EDAF-TIMES model (Muehlich and Hamacher,
2009). The average time spent on traveling grows at a slower rate
Fig. 3. Total passenger and freight transportation service demands.
than the per capita service demands because of a gradual shift
from low speed modes (e.g., walking and cycling) to high speed
modes (e.g., LDVs and airplanes) (Fig. 4).
From a modal split perspective, LDVs and buses account for an
increasingly greater share of urban and rural passenger transport,
eroding the previously dominant walking, bicycles and motorbikes
(Fig. 4). Non-motorized modes, notably walking and cycling, that
dominated urban and rural passenger transport in China (Yan and
Crookes, 2010) are already waning because of increasing LDV
ownership and increasing travel demand; this trend is expected to
continue. The combined share of LDVs and buses in the urban
passenger transport sector increases from 44% to 79% during the
century, while the combined share of motorbikes, bicycles and
walking decreases from 52% to 17%. The modal shift is more dra-
matic in the rural passenger transport sector: the combined share
of LDVs and buses is 62% of total rural passenger transport by 2095
(up from 5% in 2010) and rapidly displaces the conventional pas-
senger transport modes of motorbikes, bicycles and walking.
The intercity passenger transport sector exhibits a different
modal shift (Fig. 4). The most pronounced change is the dominant
role of buses being replaced by LDVs, airplanes and high speed rail.
The share of buses, including heavy- and light-duty buses, de-
creases gradually from 57% in 2010 to 16% over the century, while
the combined share of LDVs and airplanes increase from 21% to
61%. The relatively high population density in China means that
rail travel accounted for approximately 20% of intercity passenger
transport in 2010, much higher than many other countries; its
share remains stable in the future. Taking all passenger transport
services together: walking, bicycles, motorbikes and buses play an
increasingly less important role with a pronounced expansion
expected in LDVs and air transport. Their travel share increases
from 19% and 4% to 42% and 21%, respectively, during the century.
However, the fraction of LDV in annual travel is still lower than
that in developed countries (Harvey, 2013), which is about 49.5%
of Pacific Asia OECD, 78.5% of North America and 63.5% of Western
Europe in 2005. Considering vehicle-use intensity and load factors,
the stock of private LDVs in China will increase from 49.19 million
in 2010 to 571.20 million in 2095, and car ownership in China will
increase from 33.5 to 441.6 per 1000 people, reaching current le-
vels of European (450–600) and Japan (440), but lower than the U.
S. level of 800 per 1000 people.
The relative contributions of various modes to freight trans-
portation service vary over time, but at a modest rate (Fig. 5). Rail
share steadily increases from 30% to 41%, with a corresponding
decrease in road and domestic shipping shares. The currently
negligible air freight will expand, reaching 1% by the century end.
Taking domestic and international freight together, shipping con-
tinues to account for approximately half of the freight transpor-
tation service in China.
 X. Yin et al. / Energy Policy 82 (2015) 233–248 239

Fig. 4. Passenger transport service demand by mode.

3.1.2. Energy consumption and CO2 emissions
Our model shows that both transportation energy use and CO2
emissions in China will increase 5-fold during the current century
(Fig. 6). This growth is faster than the anticipated increase in
transportation service (Fig. 3). The difference is mainly owed to the
rapid switch toward more energy intensive LDVs and air transport
away from less energy intensive modes such as walking and cy-
cling that more than offset the general effect of energy efficiency
improvements. In addition, the proportions of individual trans-
portation services in CO2 emissions remain largely the same as
those in energy consumption as the fuel mix does not vary sig-
nificantly across different transportation services. Particularly,
passenger transport and freight transport continue to account for
approximately half each in terms of energy use and CO2 emissions,
with intercity plus urban travel and domestic freight being the
predominant sources for each sector: the proportion of other
services is no more than 4%. Note that international shipping,
which is responsible for approximately 35% of total freight service,
now accounts for no more than 10% because of its very low freight
energy intensity level (MJ/tkm).

Fig. 5. Freight transport service demand by mode.

240 X. Yin et al. / Energy Policy 82 (2015) 233–248
Fig. 6. Transportation final energy use and direct CO2 emissions by service.
With no policy intervention, liquids will remain the pre-
dominant source of transportation energy in China throughout the
century (Fig. 7). The currently small use of coal will phase out over
the coming decades, whereas electricity, gas, and molecular hy-
drogen gas (H2) use continues to expand, although their combined
share of final energy use remains less than 10%. However, fuel
substitution is relatively pronounced at the liquid production level.
Although crude oil remains the major source of transportation li-
quids, coal-to-liquid (CTL), biofuel, and to a less extent gas-to-li-
quid (GTL) enter the system in sizeable proportions, as crude oil
prices increase faster than the price of other domestic fuels with
synthetic fuel technologies becoming increasingly viable. By 2095,
CTL and biofuel account for 17 and 7% of total transportation liquid
use respectively, indicating a higher carbon intensity of transpor-
tation liquids. The H2 production processes and the electricity used
for transportation also release CO2. When these indirect CO2
emissions are incorporated, total CO2 emissions will grow by over
6 times from 2010 to 2095, and the share of indirect CO2 emissions
will increase from 13% to 33%.
3.1.3. Comparison with previous studies
A detailed comparison of transportation service and energy
consumption with previous studies poses major challenges as they
employ different data sources and sectoral coverage. For instance,
while domestic and international freight shipping are treated se-
parately in this study, those undertaken by China's Energy Re-
search Institute (ERI) (ERI, 2010) and Chen et al. (2013) consider
them together, and Fu (2011) covers only domestic shipping. In the
passenger transport sector, this study considers both motorized
and non-motorized travel, with other studies considering motor-
ized travel only. In addition, motorbikes are represented in this
Fig. 7. Transportation final energy use by fuel and final liquids production by technology.
study and Hu (2012) but not in other studies. Thus, this study
presents the highest base year passenger transport service level of
all the studies, at approximately 10 trillion pkm, followed by
6.5 trillion pkm in Hu (2012) and 5.2 trillion pkm in ERI (2010). For
energy consumption, the IEA (2012), which uses NBS data (NBS,
2011), reports the transportation energy consumption of 5.4 EJ in
2010. Other studies (Chen et al., 2013; Hu, 2012) that use the oil
allocation method—including this study—report the base year
transportation energy consumption of 9.4 EJ. To remove the effects
of coverage and estimation differences, this section will compare
modeling results relative to 2010, focusing on their future trends
in the baseline scenario. Additionally, as most current studies on
China's transportation sector focus on 2010–2050, this sector will
only compare modeling results for that period.
All of the discussed studies show upward trends in passenger
and freight transportation service demands (Fig. 8). Passenger
service demands in this study and Hu (2012) that account for in-
come and price responsiveness tend to increase less dramatically
than those in Fu (2011), ERI (2010) and Chen et al. (2013) that are
all based on mode-specific bottom-up methods. For freight
transportation service demand, ERI (2010) and Chen et al. (2013)
exhibit a faster increase than other studies. However, all studies
show similar results for the final transportation energy con-
sumption, projecting it to increase by 130–240% by 2030, and 210–
320% by 2050).
3.2. Impact of carbon policy
3.2.1. Carbon policy scenarios
The GCAM-China is capable of analyzing the impacts of differ-
ent carbon polices, such as carbon tax and constrained emissions
 X. Yin et al. / Energy Policy 82 (2015) 233–248 241

Fig. 8. Results comparison for service demand and total final energy use.

scenarios. This study analyzed three constrained emissions sce-
narios that are consistent with Representative Concentration
Pathways RCP6.0, RCP4.5 and RCP2.6 that achieve the end-of-the-
century radiative forcing levels of 6.0, 4.5 and 2.6 Wm  2 with
approximate CO2 equivalent concentration levels being 850, 650
and 450 ppm, respectively. With the emissions constraints in
place, shadow prices as expressed by a schedule of carbon tax that
starts from 2020 are derived with the assumption that carbon tax
rises over time at the rate of interest plus the removal rate of CO2
from the atmosphere: a Hotelling–Peck–Wan pathway (Peck and
Wan, 1996). An annual growth rate of 5% is assumed for this
purpose, so that the price reaches approximately 40 times the
price in 2020 by the end of the century. In the most ambitious
scenario RCP2.6, carbon tax is about 17 1990 $/tC in 2020, nearly
17 times of that in RCP6.0, suggesting greater mitigation effort at
the global level. Consequently, global CO2 emissions will start to
decline in 2070, 2050 and 2030 in RCP6.0, RCP4.5 and RCP2.6
scenarios, respectively. By 2095, mitigation scenarios RCP6.0,
RCP4.5 and RCP2.6 will reduce emissions by 40%, 90% and 120%,
respectively (Fig. 9).

Fig. 9. Carbon price and global CO2 emissions.

242 X. Yin et al. / Energy Policy 82 (2015) 233–248
3.2.2. Service output and modal share with carbon policy
Freight transportation service is more responsive to carbon
price than passenger transport service, although the magnitude of
the response differs by the type of services delivered (Fig. 10). The
greater price responsiveness in freight service is because energy
costs (including carbon costs) generally constitute a larger portion
of total service costs in freight transportation than in passenger
transport,where travel time cost, independent of carbon tax, plays
a major role in service cost accounting (Kyle and Kim, 2011).Em-
pirical analysis (Litman, 2006) and the TREMOVE model (Delsalle,
2002) also find similar differences: freight transportation exhibits
greater responsiveness than passenger transport.
Our results indicate that carbon policy does not have a pro-
nounced effect on energy service demand reductions in the pas-
senger transport sector (Fig. 11). For example, total passenger
transport turnover will only decrease by 10.8% and intercity pas-
senger transport turnover decrease by 14.4% by 2095 in the RCP2.6
scenario. Nevertheless, an ambitious carbon policy, such as the RCP
2.6 scenario, results in a smaller share of air and slightly greater
shares of railway and bus than the other cases. Among various
passenger services—that is, intercity, urban, and rural services—
intercity passenger transport exhibits the greatest modal shifts,
mainly because (i) it has the highest average travel speed, (ii) re-
latively low capital cost [$/(pkm/y)], and (iii) the energy intensity
of airplanes is much higher than other modes, all of which in-
crease the extent to which the change in energy cost can influence
total service cost. Importantly, carbon policy causes non-motor-
ized travel (i.e., walking and cycling) to persist in a greater pro-
portion as these modes remain unaffected by the change in energy
cost. Taken together, however, carbon policy and its associated
modal shift contribute limit roles to CO2 mitigation in the pas-
senger transport sector. This finding is consistent with Girod et al.
(2012) and Kim et al. (2006), both of whom suggest that carbon
policy will not have major effects on total passenger service de-
mand or its modal splits. Girod et al. (2013) compared five dif-
ferent global transport models with harmonized input assump-
tions on income and population, and found the contribution of
modal shift to emission reduction is estimated to be minor com-
pared to be monior compared to other changes in the transport
system. Kim et al. (2006) used ObjECTS-MiniCAM model, and
found, total transport services provided by automobile and light
trucks do not vary much across scenarios, and the quantity and
type of fuel consumed does depend on the composition of vehicles
and the rate of advanced vehicle penetration.
Freight transportation, however, exhibits greater sensitivity of
modal share to carbon policy than passenger transport (Fig. 12), as
energy cost accounts for a greater share of total service cost in
freight transportation. A more ambitious stabilization target would
Fig. 10. Passenger and freight service demand reductions in the carbon policy scenarios.
require a more drastic change in modal shares in the freight
transport sector. Delsalle (2002) studied the effect of the change in
fuel price on the transport sector in nine European countries and
reported similar findings: when fuel price increases by 86%, the
share of trucks will decrease by 0.9% while that of LDVs will only
decrease by 0.4%. Our results point out that the dominant position
of shipping, which possesses limited technology options for low
cost mitigation, is eroded largely by the rapid expansion of rail.
3.2.3. Final energy consumption and CO2 emissions with carbon
policy
Carbon policy does reduce final energy consumption in the
Chinese transportation sector (Fig. 13). The sector's final energy
consumption decreases by approximately half by 2095 when car-
bon policy achieving RCP2.6 is implemented. Because of larger
service demand reductions (Fig. 10), energy consumption de-
creases faster in the freight transportation sector than in the
passenger transport sector as the carbon price increases. In the
RCP2.6 scenario, passenger and freight energy consumption will
decline by 41% and 56%, respectively, from the reference scenario
by 2095, with the share of passenger transport energy consump-
tion continually increasing. The domestic freight share change is
the largest, from 40% in the reference scenario to 34% in the RCP2.6
scenario. However, there is little change in the RCP6.0 and RCP4.5
scenarios where the share of domestic freight will only decrease
by 0.6% and 2.1%, respectively, by 2095.
At the fuel level, the reduction involves a rapid shift toward
biofuel, electricity, and hydrogen away from the conventional oil
liquids that currently account for the largest proportion of trans-
portation fuel demand in China (Fig. 14). Particularly, transporta-
tion liquids experience a major transformation. Biofuel, one of the
most promising mitigation options, is increasingly used: with the
most ambitious climate policy (RCP2.6) it accounts for over half of
total liquid used and approximately 35% of the century's trans-
portation energy demand. By contrast, the share of synthetic li-
quids from coal and gas decreases with the introduction of carbon
policy as these processes are carbon intensive.
With carbon policy in place, both passenger and freight trans-
portation sectors continue to be reliant on liquids over the course
of the century (Fig. 15). However, various transportation services
exercise different mitigation options to achieve the stabilization
target. For urban and rural passenger transport services, because
there are more LDVs and buses, H2 and electricity will account for
approximately 30% and 15%, respectively, by the century end. The
role of H2 is relatively modest in intercity passenger transporta-
tion, although the sector becomes increasingly reliant on elec-
tricity owing to the expansion of the electric rail system. Electricity
will become the major mitigation option in the freight
 X. Yin et al. / Energy Policy 82 (2015) 233–248 243

Fig. 11. Comparison of passenger transport service by mode in different scenarios.

transportation sector, as H2 is not considered as a fuel for freight
transportation.
The decreased final energy consumption, combined with the
rapid switch toward low-carbon fuels, results in substantial CO2
emission reductions in response to carbon policy (Fig. 16). The
abatement of transportation CO2 emissions can be attributed to:
(i) reductions in tail-pipe emissions from the direct combustion of
fossil fuels and (ii) reductions in indirect upstream emissions
generated to deliver transportation fuels, including hydrogen,
electricity, synthetic liquids and gases. Direct CO2 emissions with a
carbon policy either level out (RCP6.0) or exhibit peak-and-decline
trajectories (RCP4.5 or RCP2.6) because of the reductions in
transportation final energy consumption and the overall carbon
intensity of fuels. Because of the deployment of non-fossil energy
and carbon capture and storage (CCS) technologies in the up-
stream sectors (e.g., electricity, refinery, and hydrogen), indirect

Fig. 12. Comparison of freight transportation service by mode in different scenarios.

244 X. Yin et al. / Energy Policy 82 (2015) 233–248
Fig. 13. Total final energy and final energy use by service.
Fig. 14. Total final energy and energy use by fuel.
CO2 emissions exhibit even greater sensitivities to carbon policy
than direct CO2 emissions: the RCP6.0 scenario is large enough to
reduce indirect CO2 emissions by over 50%, and the RCP2.6 sce-
nario gives nearly zero indirect CO2 emissions by the century end.
That is, carbon policy effects are greater from decarbonizing up-
stream sectors than from direct responses from end-use demand
and technologies. The rapid decarbonization of transformation
sectors is a well-established dynamic in previous mitigation stu-
dies (Clarke et al., 2007; Edmonds et al., 2006).
3.2.4. Comparison of the transportation sector with other end-use
sectors in China
As the transportation sector develops, its share in total final
energy consumption will also increase rapidly from 13.1% in 2010
to 20.7% in 2050 and 24.6% in 2095; similar trends are observed in
CO2 emissions. When carbon policy is implemented, both energy
consumption and CO2 emissions decrease significantly at the na-
tional level. To analyze differences between the transportation and
other end-use (industry and buildings) sectors, Fig. 17 shows the
percentage reductions of final energy use and CO2emissionsin
three carbon tax scenarios.
Fig. 17 demonstrates that the non-transportation sectors gen-
erally show larger energy consumption reduction and CO2 miti-
gation potential than the transportation sector at least in the near-
or mid-term but as carbon price continues to rise, final transpor-
tation energy consumption decreases at an augmented rate. As a
consequence, to achieve a certain level of energy reduction (10%
and 20%), the transportation sector requires 25–30 years longer
than non-transportation sectors. In terms of CO2 mitigation levels
(10% and 30%), the transportation sector takes even longer (40–45
years).
Differences in the responsiveness to carbon policy of final en-
ergy consumption and CO2 emissions between the transportation
sector and the non-transportation sectors can be to some extent
explained by very different shares of ‘fossil fuels’ in final energy
consumption as seen in Fig. 18 (the fossil fuels here refer to liquids
including bio-liquids fuel, coal and gas that are CO2 emitters in
final sectors). In the reference scenario, fossil fuels remain as the
dominant source with the shares nearly unchanged in both sec-
tors. When carbon price is imposed, however, the fossil fuels share
starts to decrease, quickly in the case of the non-transportation
sectors and followed later by the transportation sector. Although
the transportation sector substantially decreases its reliance on
fossil fuels by the century end, in the RCP2.6 scenario it remains
above 60% as compared with 20% for the non-transportation
sector.
4. Discussions
This study explores the potential impacts of different carbon
price policies on transportation services, energy consumption and
CO2 emissions. The assessment was conducted by incorporating a
detailed representation of the transportation sector into the
GCAM. This study improves our understanding of the develop-
ment of China's transportation sector, and allows us to gain rich
insights into the transition of the transportation energy system
with and without carbon prices.
In the reference scenario, all services except rural passengers
will increase rapidly in the future, and gradually saturate by the
end of this century. Because of the continuous urbanization in
China, the transportation services of the rural population will ra-
pidly peak, and then start to decline. As a result of modal shifts
towards more energy intensive modes (such as private LDVs and
aviation), energy consumption and CO2 emissions will grow faster
than transportation services.
Both passenger and freight transportation energy consumption
increase rapidly in the future and are dominated by liquids
throughout the century. The rapid increase in liquids use will
undoubtedly place tremendous pressure on China's energy se-
curity. CTL and biomass-to-liquids (BTL) will be of great
 X. Yin et al. / Energy Policy 82 (2015) 233–248 245

Fig. 15. Final transportation energy use by fuel in scenario RCP 2.6.

Fig. 16. Direct and Indirect CO2 emissions in transport sector.

246 X. Yin et al. / Energy Policy 82 (2015) 233–248
Fig. 17. Final energy use reduction and CO2 mitigation in transportation and other end-use sectors.
significance in the future energy system. However, CTL is carbon-
intensive and requires considerable water resources so there is
uncertainty on its policy in China.
Passenger transport service demands and modal shifts are not
particularly sensitive to carbon prices. Previous studies suggest
service reductions and modal shifts could contribute significantly
to energy saving and CO2 mitigations in the transportation sector
(Cuenot et al., 2012; IEA, 2009; Ong et al., 2011). However, since
time value is considered in energy service cost, our model results
indicate that a carbon price would lead to only a small reduction in
service demand and slight modal shifts, and contribute little to
CO2 mitigation unless a very high carbon price (RCP2.6) is im-
posed. The freight transportation sector is much more sensitive to
CO2 prices than the passenger transport sector. This is because
time value is not considered, and fuel cost (including carbon price)
contributes to a larger share in the service costs. With a high level
of carbon tax in the RCP2.6 scenario, we can see significant
changes in service demand and modal shifts.
It is difficult to make the transportation sector liquid-free, and
the share of liquid fuels will be over 60% even in the RCP2.6 sce-
nario. As a result, the transportation sector is more difficult to
decarbonize in carbon policy scenarios than other end-use sectors
(industry and building). Energy consumption reduction and CO2
mitigations in the transportation sector are not only less than
those for the non-transportation sectors but also start later, find-
ings consistent with Pietzcker et al. (2014). Although this study
suggests that the transportation sector is difficult to decarbonize,
it still needs to be a major part of mitigation portfolios, particularly
Fig. 18. The share of fossil energy in final energy use in transportation and other end-use sectors.
to achieve aggressive stabilization targets (e.g., RCP2.6). In the
RCP2.6 scenario, the transportation sector shows larger energy
reduction potential than the non-transportation sectors post-2075.
There are some research limitations. First, we do not dis-
aggregate China into different regions, and this paper provides
average numbers only. Since there are large differences in the
economic development, industry structure and population dis-
tribution among different provinces in China, a disaggregated
model can help to better understand the development of trans-
portation services and modal shifts. Second, we do not consider
the uncertainties of future global carbon budgets and carbon
permit allocation mechanisms; both are highly uncertain and will
significantly impact future energy consumption and CO2
emissions.
5. Conclusions and policy implications
This study suggests, due to lower load factors, and rapid modal
shifts from less energy intensive walking and cycling towards
more energy intensive modes such as LDVs and air transport,
transportation energy consumption and CO2 emissions will con-
tinue to increase over the century, even faster than the increase in
transportation services. However, alternative carbon policies could
help reduce the sector's energy consumption and CO2 emissions.
Chinese transportation CO2 emissions will reach the peak at 2080,
2065 and 2060, and will decrease by 23%, 44% and 73% in 2095,
when compared with the reference scenario with the climate
 X. Yin et al. / Energy Policy 82 (2015) 233–248
forcing target as 6.0 W/m2, 4.5 W/m2, and 2.6 W/m2 respectively.
The reduction comes from decrease in service demands, modal
shifts towards less energy intensive technologies and less carbon
intensive fuels. These results highlight the importance of future
reasonable service demands growth, modal shifts and technology
development, but also that the transport sector can play an im-
portant role in CO2 mitigation.
Our study also indicates, although industry and building sectors
have greater energy saving and CO2 mitigation potentials than the
transportation sector in the short-term, the transportation sector
will be of considerable significance for the long-term carbon mi-
tigation, with the use of advanced electric and fuel cell vehicle
technologies in combination with carbon free primary/secondary
energy sources. Decarbonizing electricity, hydrogen and liquids by
introducing low- even negative-carbon fuel production technolo-
gies will be essential to meet ambitious CO2 stabilization target.
Although this study focuses on carbon policy simulation, more
direct policy instruments like improving and labelling fuel econ-
omy standards, subsidies for electric vehicles and fuel cell vehicles,
encouraging public transport, license bidding to avoid fast growth
of private car and etc. directly or indirectly considered in the
scenarios are of great importance to CO2 mitigation. For freight
transportation, it's possible to increase the production of locally
produced products, and to locate industry facilities and freight
infrastructures, which determine where goods are delivered and
demanded, along the railways and rivers, thus reducing freight
transportation demands, and increasing the share of railways and
shipping. Given the rail and ship infrastructures requirement and
the existing investment in infrastructures and vehicles, most of
these policies are slow to change, and might take decades, and we
should take actions right way to reduce short- and mid-term
transportation CO2 emissions.
Due to the high cost, electric and fuel cell vehicles will not
come into use largely with low carbon prices, transport policies to
remove barriers to improving transport energy efficiencies, such as
strengthen the R&D to break through core technology, lower
purchase price, increase subsides and adjust subsidy mechanisms,
need to be taken immediately to promote their deployment. Since
public opinion and consumer attitudes are influenced by media
and networks, it's also important for the government to spread the
non-financial benefits of adopting electric and fuel cell vehicles,
like energy security and reduce carbon footing print.
Since the transportation sector is not as sensitive to carbon
prices as other sectors, especially without aggressive carbon po-
licies like the RCP 2.6 scenario, direct policy options to reduce
transportation service demands, promote modal shifts towards
low energy intensive modes, and the deployment of low carbon
intensive fuels will be of great importance. Though the transpor-
tation sector seems harder to be carbon-free, the transportation
sector will be of great importance in the long-term CO2 mitigation,
considering its growing share in total CO2 emissions, and sig-
nificant reduction with aggressive carbon policies.
Acknowledgments
This study is supported by the Ministry of Science and Tech-
nology of China (2012BAC20B01), Ministry of Education of China
(12JJD630002), National Development and Reform Commission,
China (2012011), and China Scholar Council
References
Brand, C., Tran, M., Anable, J., 2012. The UK transport carbon model: an integrated
life cycle approach to explore low carbon futures. Energy Policy 41, 107–124.
247
Brenkert, A., Kim, S., Smith, A., Pitcher, H., 2003. Model Documentation for the
MiniCAM. Pacific Northwest National Laboratory, Richland, Washington.
CEG, 2008. China Energy Data Book v7.0. China Energy Group, Beijing, China.
Chaturvedi, V., Eom, J., E. Clarke, L., R. Shukla, P., 2013. Long term building energy
demand for India: disaggregating end use energy services in an integrated
assessment modeling framework. Energy Policy.
Chen, W., 2005. The costs of mitigating carbon emissions in China: findings from
China MARKAL-MACRO modeling. Energy Policy 33, 885–896.
Chen, W., Li, H., Wu, Z., 2010. Western China energy development and west to east
energy transfer: application of the Western China Sustainable Energy Devel-
opment Model. Energy Policy 38, 7106–7120.
Chen, W., Wu, Z., He, J., Gao, P., Xu, S., 2007. Carbon emission control strategies for
China: a comparative study with partial and general equilibrium versions of the
China MARKAL model. Energy 32, 59–72.
Chen, W., Yin, X., Zhang, H., 2013. Towards low carbon development in China: a
comparison of national and global models. Clim. Change . http://dx.doi.org/
10.1007/s10584-013-0937-7.
Clarke, L., Edmonds, J., Jacoby, H., Pitcher, H., Reilly, J., Richels, R., 2007. Scenarios of
Greenhouse Gas Emissions and Atmospheric Concentrations. Department of
Energy, Office of Biological & Environmental Research, U.S. Washington D.C.
Cuenot, F., Fulton, L., Staub, J., 2012. The prospect for modal shifts in passenger
transport worldwide and impacts on energy use and CO2. Energy Policy 41,
98–106.
Delsalle, J., 2002. The Effects of Fuel Price Changes on the Transport Sector and its
Emissions—Simulations With TREMOVE. Directorate General for Economic and
Financial Affairs, Brussels, Belgium.
Edmonds, J., Wilson, T., Wise, M., Weyant, J., 2006. Electrification of the economy
and CO2 emissions mitigation. Environ. Econ. Policy Stud. 7, 175–203.
Edmonds, J., Wise, M., Pitcher, H., Richels, R., Wigley, T., Maccracken, C., 1997. An
integrated assessment of climate change and the accelerated introduction of
advanced energy technologies. Mitig. Adapt. Strateg. Glob. Change 1, 311–339.
Eom, J., Clarke, L., Kim, S.H., Kyle, P., Patel, P., 2012a. China's building energy de-
mand: long-term implications from a detailed assessment. Energy 46, 405–419.
Eom, J., Schipper, L., Thompson, L., 2012b. We keep on truck in: trends in freight
energy use and carbon emissions in 11 IEA countries. Energy Policy 45,
327–341.
ERI, 2010. China Low Carbon Development Pathways by 2050 Scenario Analysis of
Energy Demand And Carbon Emissions. Science Press, Beijing, China.
Fu, Z., 2011. Medium & Long Term Development of Policies For Energy Saving in
Transportation for China. The China Sustainable Energy Program, Beijing, China
(in Chinese).
Girod, B., Vuuren, D., Deetman, S., 2012. Global travel within the 2 °C climate target.
Energy Policy 45, 152–166.
Girod, B., Vuuren, D., Grahn, M., Kitous, A., Kim, S.H., Kyle, P., 2013. Climate impact
of transportation: a model comparison. Clim. Change 118, 595–608.
Harvey, L., 2013. Global climate-oriented transportation scenarios. Energy Policy 54,
87–103.
Hao, H., Wang, H., Ouyang, M., 2011. Fuel conservation and GHG (Greenhouse gas)
emissions mitigation scenarios for China's passenger vehicle fleet. Energy 36,
6520–6528.
Hao, H., Wang, H., Ouyang, M., 2012. Fuel consumption and life cycle GHG emis-
sions by China's on-road trucks: future trends through 2050 and evaluation of
mitigation measures. Energy Policy 43, 244–251.
He, J., Wu, W., Xu, Y., 2010. Energy consumption of locomotives in China railways
during 1975–2007. J. Transp. Syst. Eng. Inf. Technol. 10, 22–27.
He, J., 2013. Analysis of CO2 emissions peak. China Popul. Resour. Environ. 23 (12),
1–9 (in Chinese).
Hu, G., 2012. Research and application of long term transport energy consumption
and carbon emission model in China. School of Public Policy and Management,
Tsinghua University, Beijing, China (in Chinese).
Huo, H., Wang, M., Zhang, X., He, K., Gong, H., Jiang, k, Jin, Y., Shi, Y., Yu, X., 2012a.
Projection of energy use and greenhouse gas emissions by motor vehicles in
China: policy options and impacts. Energy Policy 43, 37–48.
Huo, H., He, K., Wang, M., Yao, Z., 2012b. Vehicle technologies, fuel-economy po-
licies, and fuel consumption rates of Chinese vehicles. Energy Policy 43, 30–36.
Huo, H., Wang, M., 2012. Modeling future vehicle sales and stock in China. Energy
Policy 43, 17–29.
IEA, 2009. Transport, Energy and CO2: MOVING TOWARD Sustainability. Interna-
tional Energy Agency (IEA), Paris, France.
IEA, 2012. World Energy Outlook 2012, ,. International Energy Agency, Paris, France.
IEA, 2013. World Energy Outlook 2013. International Energy Agency, Paris, France.
IFEU, 2008. Transport in China: Energy Consumption and Emissions of Different
Transport Modes. Institute for Energy and Environmental Research, Heidelberg,
Germany.
International Transport Forum, 2012. OECD. StatExtracts, p. 〈http://stats.oecd.org/
index.aspx〉.
Jia, S., Mao, B., Liu, S., Sun, Q., 2010. Calculation and analysis of transportation en-
ergy consumption level in China. J. Transp. Syst. Eng. Inf. Technol. 10, 22–27.
Kim, S.H., Edmonds, J., Lurz, J., Smith, J., Wise, M., S., 2006. The ObjECTS Framework
for integrated assessment: hybrid modeling of transportation. Energy J. 2,
63–92.
Koljonen, T., Lehtila, A., 2012. The impact of residential, commercial, and transport
energy demand uncertainties in Asia on climate change mitigation. Energy
Econ. 34, 410–420.
Kyle, P., Clarke, L., Rong, F., Smith, S., 2007. Climate Policy and the long-term evo-
lution of the U.S. building sector. Energy J. 31, 145–172.
248 X. Yin et al. / Energy Policy 82 (2015) 233–248
Kyle, P., Kim, S.H., 2011. Long-term implications of alternative light-duty vehicle
technologies for global greenhouse emissions and primary energy demands.
Energy Policy 39, 3012–3024.
Litman, T., 2006. Transportation Elasticities: How Prices and Other Factors Affect
Travel Behavior. Victoria Transport Policy Institute, Victoria BC, Canada.
Lo, K., 2014. A critical review of China's rapidly developing renewable energy and
energy efficiency policies. Renewable and Sustainable Energy Reviews 29,
508–516.
Luckow, P., Wise, M.A., Dooley, J.J., Kim, S.H., 2010. Large-scale utilization of bio-
mass energy and carbon dioxide capture and storage in the transport and
electricity sectors under stringent CO2 concentration limit scenarios. Int. J.
Greenhouse Gas Control 4, 865–877.
Luderer, G., Pietzcker, C., Kriegler, R., Haller, E., Bauer, N., M., 2012. Asia's role in
mitigating climate change: a technology and sector specific analysis with Re-
MIND-R. Energy Econ. 34, 378–390.
Mao, B., Jia, S., Sun, Q., Jia, Y., 2009. Energy consumption, emissions and their
comparison among different transport modes. Beijing Jiaotong University,
Beijing, China (in Chinese).
Muehlich, P., Hamacher, T., 2009. Global transportation scenarios in the multi-re-
gional EFDA-TIMES energy model. Fusion Eng. Des. 84, 1361–1366.
NBS, 2011. China Energy Statistical Yearbook 2011. China Statistics Press, Beijing (in
Chinese).
NBS, 2012. China Statistical Yearbook 2012. China Statistical Press, Beijing (in
Chinese).
Ong, H.C., Mahlia, T.M.I., Masjuki, H.H., 2011. A review on emissions and mitigation
strategies for road transport in Malaysia. Renew. Sustain. Energy Rev. 15,
3516–3522.
Peck, S.C., Wan, Y.S., 1996. Analytic solutions of simple optimal greenhouse gas
emission models. Econ. Atmos. Pollut. 14, 113–121.
Pietzcker, R., Longden, T., Chen, W., Fu, S., Kriegler, E., Kyle, P., Luderer, G., 2014.
Long-term transport energy demand and climate policy: alternative visions on
transport decarbonization in energy-economy models. Energy 64, 95–108.
Qi, Y., 2012. Annual Review of Low Carbon Development in China (2011–2012).
Social Sciences Academic Press, Beijing, China (in Chinese).
Smith, S., Arias, S., Volke, A., 2010. Enhancement of Solar Energy Representation in
the GCAM Model. Pacific Northwest National Laboratory, USA.
Sperling, D., Lin, Z., Hamilton, P., 2005. Rural vehicles in China: appropriate policy
for appropriate technology. Transp. policy 12, 105–119.
Wang, Q., 2010. 2010 Energy Data. The Energy Foundation, Beijing, China (in
Chinese).
Wise, M., Kyle, G.P., Dooley, J., Kim, S.H., J., 2010. The impact of electric passenger
transport technology under an economy-wide climate policy in the United
States: carbon dioxide emissions, coal use, and carbon dioxide capture and
storage. Int. J. Greenh. Gas Control 4, 301–308.
Yan, X., Crookes, R.J., 2009. Reduction potentials of energy demand and GHG
emissions in China's road transport sector. Energy Policy 37, 658–668.
Yan, X., Crookes, R.J., 2010. Energy demand and emissions from road transportation
vehicles in China. Prog. Energy Combust. Sci. 36, 651–676.
Yao, Z., Huo, H., Zhang, Q., Zhang, D., Streets, G., He, K., D., 2011. Gaseous and
particulate emissions from rural vehicles in China. Atmos. Environ. 45,
3055–3061.
Zhang, S., Jiang, K., Liu, D., 2007. Passenger transport modal split based on budgets
and implication for energy consumption: approach and application in China.
Energy Policy 35, 4434–4443.
Zhou, N., Fridley, D., Khanna, N.Z., Ke, J., McNeil, M., Levine, M., 2013. China's energy
and emission outlook to 2050: perspectives from bottom-up energy end-use
model. Energy Policy 53, 51–62.
Zhou, X., 2010. Present situation and trend of energy consumption in traffic and
transportation industry. Sino-Glob. Energy 15, 9–18.