Greenhouse Gas Emission Baselines for Global Cities and Metropolitan Regions
Kennedy, C.A.1, Ramaswami, A.2, Carney, S.3, and Dhakal, S4.
1. Professor, Department of Civil Engineering, University of Toronto, Canada;
[email protected]
2. Professor, Department of Civil Engineering, University of Colorado Denver, USA;
[email protected]
3. Tyndall Centre, University of Manchester, UK; [email protected]
4. Executive Director, Global Carbon Project, National Institute for Environmental
Studies, Japan; [email protected]
Summary
Greenhouse gas (GHG) emissions baselines are presented for 44 urban areas (cities or
metropolitan regions). The types of methodology that have been used to attribute GHGs
to urban areas are reviewed. All are essentially adaptations or simplifications of the IPCC
guidelines, and incorporate the WRI/WBCSD concepts of Scope 2 and 3 “crossboundary” emissions. Analysis of previous studies shows where specific differences in
methodology exist. Some Scope 3 emissions such as those embodied in materials, food,
and fuel consumed in cities, have only been quantified in a few studies, and should be
included in further studies. Baseline emissions are presented with and without emissions
from: industrial processes; and agriculture, forestry and other land-use (AFOLU) (both of
which may be incomplete); as well as: waste; and aviation / marine (for which there are
differences in methodology). Despite these often minor differences, the potential clearly
exists to establish an open, global protocol for quantifying GHG emissions attributable to
urban areas.
1. Introduction
Increasing urbanization, forces of globalization, and risks due to climate change will
necessitate new forms of urban management in the 21st century. As the governance and
management of urban areas (cities and metropolitan regions) becomes increasingly
sophisticated, new urban metrics will be required. These include measures of urban
competitiveness (Duffy, 1995; Llewelyn-Davies et al., 2004), gross metropolitan product
(Bureau of Economic Analysis, 2009), urban greenhouse gas (GHG) emissions (Harvey,
1993, Kates et al. 1998, Satterthwaite 2008, Dodman, 2009), material flows (Kennedy et
al., 2007) and vulnerability to climate change (Rosenzweig et al, 2009). Such measures
will also inform assessment of risks that may be used to guide investment in cities. In
short, many of the metrics that are currently recorded for nations are now needed for
urban areas.
1
This paper is particularly concerned with the establishment of baseline measures of GHG
emissions attributable to urban areas. Over the past two decades, several entities have
been active in establishing methodologies for estimating urban GHG emissions. One such
example is ICLEI (International Coalition for Local Environmental Initiatives, now
known as Local Governments for Sustainability), which is a worldwide coalition of local
governments (ICLEI, 2006). Over 500 of ICLEI’s member cities have established GHG
baselines using software developed by Torrie-Smith Associates, under the Partners for
Climate Protection program. Several larger cities, including for example, London, Paris,
Tokyo and others, have developed their own baselines using their own methodologies.
Eighteen European urban areas, including eight capital regions, have been studied using
the Greenhouse Gas Regional Inventory Protocol (GRIP; Carney et al., 2009) – GRIP has
also been used for Scotland and Sacramento, California. Additional urban areas have
been studied in the academic literature (Baldasano et al., 1999; Dubeux and La Rovere,
2007; Ramaswami et al. 2008; Kennedy et al. 2009b; Dhakal 2009), while other
methodologies have been brought together in meetings such as those hosted by
IGES/APN (2002); and Nagoya University / GCP / NIES (2009). The approaches used to
establish GHG emissions in these studies are essentially adaptations or simplifications of
the Intergovernmental Panel on Climate Change (IPCC) guidelines. There are, however,
minor differences in methodology that need to be resolved – and clearer reporting
mechanisms established.
For urban areas to become more effective at tackling climate change through GHG
reductions there are two key requirements:
1. An open, global protocol for quantifying GHG emissions attributable to urban
areas must be established. Such a protocol must be sufficiently robust and
compatible with the United Nations Framework Convention on Climate Change
(UNFCCC), i.e., the IPCC guidelines. (Such compatibility should include sectoral
methodologies, and emissions factors, but not necessarily a boundary limited
scope)
2. Comparable baseline measures of GHG emissions for urban areas are needed.
The emissions attributable to urban areas may be considered from different perspectives.
Emissions can be strictly based on spatially-limited geographic boundaries of an urban
area, or based on a broader consideration that also includes significant cross-boundary
embodied energy flows occurring in cities. Emission attribution can also be made based
on ’producer‘ approaches and ‘consumer’ approaches. A consumer approach is one
where the emissions from the production of a good or service are allocated to the
consumer of that good or service. A simple example of this would be the allocation of
emissions associated with electricity production to the consumer of the electricity
(regardless of where the electricity is produced). A more complicated example would be
the allocation of the emissions associated with the manufacture of a product in
developing world city being allocated to the consumer of that product in the urban area
studied. A ’producer’ approach is one where the emissions associated with an activity are
allocated to where that activity takes place. A simple example of this would be the
2
emissions associated with the production of electricity being allocated to the area that
plays host to the power station – regardless of who consumes the electricity. A producer
approach is highly dependent on geographic boundaries, while a consumer approach does
not consider any geographic boundary to the location of emissions. A hybrid approach
(Ramaswami et al., 2008) combines the two approaches, accounting for GHG emissions
occurring from all activities (whether producer or consumer) within the geographic
boundary of an urban areas, and focusing on a few most relevant cross-boundary energy
flows critical for human activities to be sustained in cities. As will be discussed further,
care must be taken in applying a hybrid approach to avoid “double counting.”
There are issues of equity associated with each of these approaches. It is important that an
emissions baseline is produced to meet its purpose. It may be of interest to local
government, of interest to wider urban policy makers, or both. It may, if it is for policy
purposes, need to provide data that enables a region to help deliver national and
international commitments on emissions reduction. Public communication about GHG
emissions is also an implicit goal in developing baselines; the role of a GHG baseline to
be effectively used in public engagement for mitigation actions can be very important.
This paper first reviews the types of methodology that have been used to attribute GHGs
to urban areas. We begin by broadly describing the approaches used to determine GHG
emissions for nations (IPCC, 2006) and for corporations (WRI/WBCSD, 2009), both of
which inform the attribution of emissions to urban areas. We then discuss in more detail
the specific differences in methodology between various studies of urban GHG
emissions. The approaches used to establish emissions for over 40 global urban areas
(Table 1) are used to demonstrate where differences in methodology occur (Table 2).
This is followed by an extended discussion of critical cross-boundary emissions most
relevant to urban areas. A few cities have, independently, quantified their cross-boundary
emissions, so-called because their emission occurs outside the geographic boundary of
the city of interest, but is directly caused by activities occurring within the geographic
boundary of a city (such as with ecological footprinting). For example: airline travel has
been included in GHG accounting for Aspen and Seattle, USA; some foods (rice and
milk) and cement have been included in emissions for Delhi and Calcutta, India (Sharma
et al., 2002 a-c); food, cement and freight transport have been included for Paris, France
(Mairie de Paris, 2009); and key urban materials such as food, water, transport fuels and
cement are accounted for in Denver, USA (Ramaswami et al, 2008). We discuss how
such emissions can play a role in augmenting baselines for urban area, the policy
implications, and the methodological approaches that have been used.
Baseline GHG emissions are then presented for 44 urban areas, including those in
developed and developing nations. While total emissions have been reported for urban
areas since the late 1980s (Harvey, 1993; Baldasano et al. 1999), this paper primarily
presents baselines from recent studies (e.g., Kennedy et al., 2009b; Ramaswami et al.
2009; Dhakal, 2004; Dhakal, 2009; Carney et al., 2009). These emission baselines reflect
the methodologies employed and emissions sources considered. Therefore the baselines
3
are presented either with or without emissions from: industrial processes1 (which may be
incomplete); waste (where methods differ), and aviation /marine (which is subject to
debate).
2. Overview of National, Corporate and Sub-national GHG Inventorying
Procedures
The IPCC (2006) Guidelines for National Greenhouse Gas Inventories are the
international standard for national reporting under the UNFCCC. The Guidelines describe
procedures for determining annual (calendar year) inventories of over ten categories of
greenhouse gas emissions (and removals) that occur as a result of human activities. The
aim with national inventories is to include GHG emissions that occur within the territory
and offshore areas under each nation’s jurisdiction, although there are some special issues
with transportation emissions as discussed later. Emissions are categorized under the five
broad sectors of: Energy; Industrial Processes and Product Use; Agriculture, Forestry and
Other Land Use; Waste; and Other (which includes precursor and indirect N2O
emissions).
The methodology for determining most emissions entails multiplication of data on a level
of human activity by an emissions factor. The IPCC Guidelines include substantial
guidance on data collection, managing uncertainty in calculations, conducting quality
assurance procedures and identifying the key categories of emissions. With respect to the
accuracy of calculations, the concept of tiers is particularly important. The tier indicates
the level of complexity in methodology, with Tier 1 being basic, Tier 2 intermediate and
Tier 3 the most complex. Higher tier methods have greater data requirements and are
generally more accurate. The tier concept can apply to both activity data and to emissions
factors. Where for example an emissions factor may be nationally specific, or a general
one.
Volumes 2 to 5 of the IPCC Guidelines provide detailed procedures for determining
emissions from various sub-sectors, using Tier 1, 2 and 3 methods. In the next section,
we will highlight a few specific procedural details from the IPCC guidelines, where they
differ from approaches used to determine GHG baselines for urban areas.
First, however, we outline procedures that corporations have adopted for reporting GHG
emissions since many municipal governments, given their level jurisdiction, have
resorted to tackling their corporate emissions. (Street lighting, for example, has emerged
as one possible area of intervention in municipalities.)
The WRI / WBSCD (World Resources Institute / World Business Council for Sustainable
Development) procedures have arguably become the best practice for reporting of GHGs
1
This is one of the main emissions sectors in UNFCCC inventories. This sector refers to
emissions at industrial sites from non-combustion activities. The majority of emissions
from Industry come from energy combustion which are considered under energy in
international reporting.
4
by corporations (and other institutions). The WRI / WBCSD procedure applies standard
accounting principles of relevance, completeness, consistency, transparency, and
accuracy. Business goals served by conducting GHG inventories include managing GHG
risk and identifying reduction opportunities. Although the standards are in themselves
policy neutral, they have been adopted by many GHG programs, including: voluntary
reduction programs; GHG registries; national and regional industry initiatives; GHG
trading programs; and sector specific protocols (WRI / WBSCD, 2009).
Two different approaches for attributing GHG emissions to a corporation are provided.
By the equity share approach, a company accounts for emissions based on its share of
equity in operations. By the control approach, a company accounts for all (100%) of the
emissions from operations over which it has control, whether financial or operational.
(WRI / WBSCD, 2009).
The WRI / WBCSD procedures make particular efforts to be supportive of national level
reporting programs. First, the procedures use emission factors that are consistent with the
IPCC. The WRI / WBSCD also recognize that official government reporting often
requires GHG data to be reported at a facility level, rather than at a corporate level. So
whether a company uses an equity share approach or a control approach to establish its
corporate inventory, it is also encouraged to itemize emissions from facilities that it
operates. Governments typically require reporting on the basis of operational control,
either at the facility–level, or at some consolidation over geographic boundaries.
The WRI / WBCSD also introduce the concept of scope of emissions, enabling
companies to distinguish between emissions from facilities that they own or control, and
emissions that result from broader company activities (Table 3). Scope 1 emissions are
those from sources such as boilers, furnaces and vehicles that are owned or controlled by
the company (producer). Emissions from electricity consumed by the company are in
Scope 2 (consumer); while other emissions that are a consequence of the company’s
activities, such as: extraction and production of purchased materials; transportation; and
product use, are in Scope 3 (consumer). These Scope 3 emissions do not necessarily
entail a full life cycle assessment, rather a practical determination of the main indirect
emissions attributable to the company’s activities.
The WRI / WBCSD Scopes 1-2-3 framework has been adopted widely, with small
variations, by several organizations that seek to establish standards for carbon accounting
with a view on future carbon trading. Examples of some of these organizations include
the California Climate Action Registry (CCAR), The Chicago Climate Exchange (CCX),
The Colorado Carbon Fund (CCF) and the North American Climate Registry (CR). Many
cities and states are participating in one or more of these registries, although it is often the
municipal government and not the community-wide emissions that are being reported.
For example, participants in the Chicago Climate Exchange include cities such as Aspen,
Boulder, Chicago, Melbourne (Australia), Portland, etc., and states such as Illinois and
New Mexico. The North American Climate Registry notes that its participants include
several large privately owned utilities, as well as local governments from San Francisco,
Austin, Seattle, and the provinces in Canada.
5
Thus, as we seek to develop community-wide GHG accounting protocols at the cityscale, adapting the WRI / WBCSD Scope 1-2-3 framework (already consistent with
IPCC) with relevant modifications necessitated by the smaller spatial scale of cities,
would provide consistency with other GHG accounting protocols. Ramaswami et al.
(2008), in developing a hybrid demand based method for GHG emissions accounting in
Denver, articulated a small set of five relevant Scope 3 items that provide a more holistic
account of the material and energy demand in cities (discussed further in Section 4).
Procedures for attributing GHG emissions to urban areas lie somewhere between those
used for national inventories and those for corporate inventories. Like the IPCC’s
national guidelines, the procedures for urban areas aim to attribute emissions to a
spatially defined area, such as that within a municipal boundary in the case of a city’s
(community) emissions. The ownership of land within the area, i.e., public or private, is
of no relevance. Similar to the WRI / WBCSD Scope 2 and 3 emissions, however, GHG
emissions attributed to urban areas can include those that occur outside of the area, as a
consequence of activities within the area. The main challenge in developing a single
global methodology for urban areas is deciding which (if any) emissions that occur
outside of urban boundaries should be allocated to the urban area (Satterthwaite, 2008).
ICLEI’s recently revised (draft) protocol for local government (community) emissions
adopts the concept of scopes, similar to the WRI/WBCSD. Under Scope 2 emissions,
ICLEI (2009) includes indirect emissions from consumption of electricity, district
heating, steam and cooling. All other indirect or embodied emissions resulting from
activities within the geopolitical boundary are classified under Scope 3, although a
consistent set of relevant Scope 3 activities are not yet explicitly defined by ICLEI for the
city-scale.
Some care has to be taken in interpreting Scope 1, 2 and 3 emissions under ICLEI’s
protocol. Some emissions from utility-derived electricity and heat combustion may be
accounted as both Scope 1 and Scope 2 emissions, if they occur within the geo-political
boundary. Similarly, emissions from landfill waste within the boundary are accounted for
under Scope 1 and Scope 3. To avoid double counting, ICLEI’s final reporting standard
includes: all Scope 1 emissions, plus additional emissions from electricity, heat, steam,
solid waste and waste water that occur outside of the geo-political boundary.
Moving to a slightly larger scale, the GRIP methodology, developed at the University of
Manchester, has primarily been applied to European regions (although it is also being
applied in the USA), typically consisting of a large urban centre with surrounding
industrial and agricultural lands (see definitions in Table 1). GRIP reports emissions
from the six main GHGs (the Kyoto basket): carbon dioxide (CO2), methane (CH4),
nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs) and sulphur
hexafluoride (SF6). The methodology closely follows the IPCC guidelines, by reporting,
for example, energy and industrial process emissions by detailed sub-sectors. Indeed,
results are developed so as to be comparable with national inventories, as well as other
regions. The GRIP methodology is also consistent with approaches used to study other
6
cities, or city-regions. For example, it does assign electricity emissions associated with
electricity generation to the end user (i.e., GRIP reports Scope 1 and 2 emissions; and
some Scope 3).
A particular strength of the GRIP methodology is its ability to recognize and manage
differences in data quality. GRIP has a three level reporting scheme, where level 1
(green) is for the most certain data; level 2 (orange) is for intermediate quality data; and
level 3 (red) is lower quality data, the latter example usually scaled from information in
national inventories. (The levels have some similarities with IPCC tiers, but are not the
same.) The colour coding is used in the reporting procedure to provide a clear indication
of uncertainty in the results.
Overall, the urban GHG methodologies used by ICLEI and GRIP, as well as the
academic studies, are fairly consistent with one another. All draw upon IPCC guidelines,
with many incorporating out of area Scope 2 and Scope 3 emissions. The main
differences lie with which emissions, particularly Scope 3, are included in final reporting.
3. Review of Methodology for Urban Baselines
This section identifies the specific differences in methodology between selected urban
GHG studies; and explains how the approaches taken relate to the IPCC and WRI /
WBCSD procedures. Table 2 shows the emissions sub-categories for which there are
differences between studies. Emissions are discussed under the four main categories of
the IPCC: Energy; Waste; Industrial Processes and Product Use; and Land-use,
Agriculture and Forestry.
3.1 Energy
The Energy sector, including stationary combustion, mobile combustion, and fugitive
sources, is by far the greatest contributor to GHG emissions from urban areas.
The determination of emissions from stationary combustion in urban areas follows the
IPCC guidelines closely, with the exception of emissions from electricity use and district
heating systems. Of the sectors considered under stationary combustion, the residential
and commercial/institutional sectors are consistently of importance for urban areas.
Emissions from these sectors can be determined with high certainty where fuels are
metered, such as with natural gas. There may be some uncertainty with fuels that are
delivered by multiple market participants, e.g., fuel oils, or where many different fuel
types are used2.
2
For example, stationary combustion in Bangkok, including industry, is comprised of the
following fuels: 21% Bagasse; 20% Fuel Oil; 14% Lignite; 13% Coal /coke; 10%
Natural Gas; 8% LPG; 5% Rice Husk; 5% Wood and 4% Diesel (Phdungsilp, 2006;
Kennedy et al., 2009b). For CO2 emissions from biofuel combustion, care has to be
taken to distinguish between biological and fossil carbon (IPCC 2006).
7
The extent of emissions from stationary combustion in the industrial sector varies
considerably by urban region. In some studies, fuel use is not explicitly distinguished by
sector. Under GRIP, however, emissions from energy combustion in the manufacturing
industries is reported according to IPCC’s subcategories. In the inventory for Glasgow
and the Clyde, for example, emissions from combustion are reported for the following
industries: iron & steel; non-ferrous metals; chemicals; pulp, paper & print; food
processing, beverages & tobacco; non-metallic minerals industries; and other. (In GRIP,
emissions from industrial energy combustion may be presented under ‘other industry’
where data is not sufficient to distinguish between different industrial types.) Such
detailed reporting is perhaps more important in wider metropolitan regions for which
industrial energy use it typically more prevalent than in central cities.
For GHGs from electricity and heat production, all the urban areas considered in Table 2
include Scope 2 emissions. From our studies, it appears to be the conventional to allocate
emissions from electricity consumption to the consumer of that electricity. Moreover, in
most studies, the transmission and distribution (T&D) line losses have been included in
the determination of emissions attributable to urban areas (Table 2). The motivation for
including emissions from electricity production is that the size of these emissions is
dependent upon the activity in the urban area (as well as the emissions factor). In
Shanghai and Beijing 30% and 71% of total electricity, respectively, are imported across
their boundaries in 2006 (Dhakal, 2009). The same argument also applies to some heating
systems. Greater Prague, for example, has a district energy system which provides 17%
of the heat used in the urban region; the GHG emissions attributable to Prague include
those from a coal fired power plant at Melnik, 60 km away, which generates steam for the
heat pipes (Kennedy et al., 2009).
Determining GHG emission from mobile sources poses different challenges than with
stationary consumption. For road transportation, there are questions as to whether travel
outside of the urban region, i.e., by commuters, should be included or not. This is a mute
issue for metropolitan regions, but significant when determining emissions from central
cities. In the City of Paris, for example, internal automobile trips generate emission of
3,670 kt CO2e, while trips with origins or destinations outside of the City contribute a
further 2,862 kt CO2e (Mairie de Paris, 2009). (Note that these are life-cycle emissions,
discussed further in section 4.) Nevertheless, for all the urban regions considered in Table
2, it is tailpipe emissions within the urban region that were quantified, so there is
consensus here. A further issue, however, is the means by which travel activity data is
determined. This is an important issue to address given that GHGs from road
transportation can account for over 30% (50% in Sacramento) of emissions in some
North American urban regions.
The IPCC Guidelines on mobile combustion recognize two approaches for quantifying
emissions for road transportation: i) based on quantity of fuel sold; and ii) from vehicle
kilometers travelled (VKT). Approach i) is preferred for CO2 emissions, as it is far more
accurate. Indeed, for reasons of data availability, consistency and the typically small size
of cross-border traffic, the use of fuel sales to calculate CO2 emissions prevails over the
8
strict application of the national territory (IPCC, 2006 Vol 2., section 1.28). Emissions of
CH4 and N2O from road transportation are, however, dependent on the age and
technology of vehicles, as well as the number of cold starts; hence approach ii) is
preferred for CH4 and N2O.
To quantify GHG emissions from road transportation in urban areas, both of the above
approaches have been used (for the 3 GHGs associated with energy: CO2, CH4 & N2O)
and a third approach involving scaling of fuel use from wider regions, e.g., states or
provinces (Table 2). There are several potential pitfalls here. Fuel sales data is not
always available for urban areas – and even if it is, there is an implicit assumption that
the fuel purchased within the region is representative of the activity within the region.
This approach may be considered compatible with the IPCC – which suggests the use of
fuel sales (although this may be more appropriate on a national basis). Meanwhile means
of determining VKT may be inconsistent between cities, due to differences in computer
modelling, surveying or vehicle counting techniques. Nevertheless, by using multiple
approaches for Bangkok, New York City and Greater Toronto, differences between the
three approaches have been shown to be less than 5% (Kennedy et al. 2009a).
Moving to emissions from air transportation, three distinct alternatives have been used
for urban areas:
i)
ii)
iii)
Exclude airplane emissions. In several of the studies in Table 2, no emissions
from combustion of airplane fuels has been counted (or in the case of Tokyo,
just operations within the area). Other than through fuel consumption on takeoff and landing, airplane emissions occur outside of urban regions and so are
not counted in Scope 1. It might also be argued that emissions from air travel
are outside of the control of local government, and so it is appropriate to
exclude them.
Include emissions from domestic aviation, but only include take-off and
landings for international aviation. This approach has primarily been used in
the 18 GRIP studies (Carney et al., 2009). It is consistent with the GRIP
philosophy, in that aviation emissions from all regions could be added
together to give the same national total as reported under IPCC guidelines.
Emissions from cruising on international flights are excluded in accordance
with the UNFCCC.
Include all emission from domestic and international aviation. Both London
(Mayor of London, 2007) and New York City (2007) report GHG emissions
based on all fuels loaded at airports within their boundaries. This approach
was adopted in the study of 10 cities by Kennedy et al. (2009), with
modification for Denver to account for transfers, as per Ramaswami et al.
(2008). This approach is consistent with the notion of world cities as the
headquarters, financial centers, and key gateways between national/regional
economies and the global economy, or as global service centres (Friedman,
1986; Sassen, 1991; Taylor, 2004).
9
Three different approaches have similarly been used for emissions from marine
transportation, where these apply. In some cases marine travel is excluded. For the
GRIP studies, only emissions on inland waters, or within 12 miles of shore are included.
While the studies of Cape Town, Los Angeles and New York City included international
marine emissions based on fuels loaded onto ships at these cities ports.
It is worth noting that there is no international agreed methodology for allocating
emissions from international aviation and marine activities. In national emissions
inventories the fuel sales and associated emissions are reported, but are not included in
the total. On an urban scale this is further complicated by the fact that their airports may
be located outside of their jurisdiction. Also passengers may be using the airport to
transfer to another region, or the airport / port may handle a lot of freight destined for
other areas. All of these issues make the allocation of emissions to the urban scale a
rather difficult task.
Nevertheless, Wood et al (2008) suggests a method by which to allocate these emissions.
The emissions associated with the LTO are allocated to the area in which the airport is
based (this is the same approach as is adopted in air quality emissions), the emissions
associated with the cruise phase are allocated to the region in which the passenger
resides. More complicated issues surrounding tourists, passengers transferring and freight
are also discussed in this paper.
3.2 Waste
The determination of GHG emissions associated with waste is where the greatest
discrepancies in methodology are apparent. In particular, emissions from the land filling
of solid waste have been calculated using at least three different techniques (Table 2):
i)
ii)
iii)
scaling from national inventories,
a total yields gas approach; and
the EPA’s WARM model
Two further techniques could also have been used:
iv) measurement from waste in-place; and
v) local application of IPCC’s first order decay approach.
The divergence of approaches for determining emissions from waste is perhaps partly due
to the complexity of emissions from landfills. The biodegradation of solid waste to form
methane and other landfill gases occurs over time-scales extending beyond a single year.
Hence, it is challenging to assign GHG emissions from waste to a particular year.
The IPCC’s recommended approach (v), involves calculation of emissions in the
inventory year, based on historical waste deposited over previous years. An alternative
(iv) would be to actually monitor and measure emissions in the inventory year, but this
10
requires considerable monitoring and may be challenging for commercial and industrial
waste streams if they are managed by the private sector.
Scaling solid waste emissions from national inventories (i), should give results that
approximate those from approaches (iv) and (v). Such scaling has been used in the GRIP
studies (using its aforementioned level 2 and 3 methods) (Carney et al. 2009).
The total yields gas approach (ii) was formerly recommended by the IPCC (1997).
Essentially it takes the total amount of waste produced by an urban area in a given year,
and then determines the total emissions released from this waste, regardless of how many
years transpire before the full release occurs. This approach has been used by Dubeux
and La Rovere (2007) and Kennedy et al. (2009a,b).
The US EPA’s WARM model (iii) uses a life-cycle accounting approach, which is ideal
in some respects, but not in others. The model recognizes, for example, that the recycling
of waste reduces emissions from the harvesting of raw materials; hence a credit can be
applied. The problem is that emission associated with material flows of paper and plastics
into cities are not currently counted in the GHG emissions for most urban areas. So use
of the WARM model is not consistent with current means of determining urban GHG
emissions, although the life-cycle methodology is indicative of the direction cities should
be headed as consumption based inventory procedures develop (see 3.5).
A few other inconsistencies in reporting of emissions from waste can be made with
reference to Table 2. First, waste emissions were not determined for the Chinese cityprovinces. Second, the GRIP studies and those of Barcelona, Geneva, Prague and
Toronto include emissions from waste incineration within the waste category, although
where such incineration includes energy recovery the IPCC recommends that the
emissions be included under stationary combustion. Finally, emissions from
wastewater/sewerage were omitted in many studies, although these are relatively minor.
3.3 Industrial Processes and Product Use
GHG emissions from Industrial Processes and Product use only include emissions that
are not primarily for energy use purposes (IPCC, 2006). A wide range of industrial
processes and products emit GHGs that are not the result of intended combustion. The
three broad categories of non-energy use are: i) feedstocks; ii) reducing agents; and iii)
non-energy products, such as lubricants, greases, waxes, bitumen and solvents.
Emissions from these types of uses can assigned to various industrial sectors:
•
•
•
•
•
•
•
Mineral Industry (including cement, lime, glass, and other)
Chemical industry
Metal industry
Non-energy Products from Fuels and Solvent Use
Electronics Industry
Product Uses as Substitutes for Ozone Depleting Substances
Other Product Manufacture and Use
11
•
Other (including pulp & paper; food & beverages)
Given the diversity of these non-energy industrial processes and products, reporting of
emissions is recognized to be challenging (IPCC, 2006).
The reporting of industrial process emissions for urban areas is somewhat mixed. Other
than the GRIP studies, which have carefully recorded these emissions, other studies have
been less consistent. For many of the urban areas in Table 2, no emissions are recorded.
This could be because there are no industrial process emissions, or the emissions are
unknown. The GRIP studies perhaps record more industrial process emissions, because
they are regional studies, including industrial areas on edges of central cities (although
emissions are often reported as zero to indicate no activity takes place). There again, it is
clear that industrial process emissions are missing from some urban areas in Table 2.
However, the emissions associated with ‘refilling’ air conditioning units may require
more attention then many studies currently adopt.
The magnitude of industrial process emissions is usually small, but these emissions are
quite city specific. For most of the European regions reported by Carney et al (2009), the
industrial process emissions are typically 1-2% of total emissions. There are exceptions
to this, however: Athens (11%); Torino (11%); Frankfurt (10%); Hamburg (7%); Napoli
(7%); Veneto (7%); Paris (6%); and Madrid (5%). In absolute terms, Frankfurt has the
largest industrial process emissions with 4,987 kt CO2e. Amongst the other urban areas
studied, Toronto has 3,185 kt CO2e of emissions just from two cement plants and a
lubricant facility. Given that some of these emissions are quite substantial, then better
reporting of industrial process emissions is generally required for urban areas.
3.4 Agriculture, Forestry and Other Land-use
GHG emissions and removals in the agriculture, forestry and other land-use (AFOLU)
category are typically small, often negligible, for most urban regions. These emissions
only become significant if the regional boundary is large, including substantial rural area
in addition to the urban core, or where agricultural activities are particularly intense; this
applies to a few cases in the GRIP studies (Carney et al., 2009).
For many of the urban areas in Table 2, AFOLU emissions have not been quantified,
because they have been taken to be negligible. This may be a reasonable assumption for
many urban regions. In the study of Calgary, for example, urban forestry sequesters 13 kt
CO2e, but this is less than 0.1% of total emissions (reported as 16,370 kt CO2e).
Even for cities in the developing world with relatively low total emissions, the
contribution of AFOLU is small. Sharma et al. (2002) determined the methane emissions
from rice cultivation and livestock (dairy cattle, non-dairy cattle and buffaloes) for
Calcutta and Delhi. The emissions from 300,000 hectares of paddy fields in Calcutta, in
1997-98, were 0.45 kt CO2e. This is negligible compared to Calcutta’s CO2 emissions for
the energy sector, which for year 2000 were reported to be 17,270 kt (Mitra et al., 2003).
Delhi’s methane emissions from paddy fields were even smaller than those for Calcutta,
12
but it had substantially more livestock, which emitted 15.16 kt CO2e in 1992 (Sharma et
al. 2002). Here again though, these methane emissions from livestock are negligible
compared to Delhi’s 19,800 kt of CO2 emissions in the energy sector (Mitra et al., 2003).
Agricultural emissions of 13 kt CO2e for Rio de Janeiro were also considered negligible,
although emissions of 256 kt CO2e for land-use change were reported (Dubeux and La
Rovere, 2007). This represents 2% of Rio’s emissions, which is small, but still large
enough to be counted.
Amongst the 18 European regions in the GRIP study, several of the larger regions do
have substantial emissions for the AFOLU sector (Carney et al., 2009). In Hamburg
Metropolitan Region, the agricultural emissions of 4,463 kt CO2e also represent 11% of
the regions total emissions. Over half of these emissions were from agricultural soils;
Hamburg is situated in the largest fruit growing region of Europe. For some other urban
areas in the GRIP study, emissions from agriculture were found to be negligible, e.g., for
Brussels, Oslo, and Helsinki. So although AFOLU emissions are usually small, or
negligible for many urban areas, there are exceptions to this, and so this category needs to
be carefully considered.
4. Inclusion of Scope 3 “Cross-Boundary” GHG Emissions Relevant to Cities
4.1 Why Include Scope 3 Items?
In this section we discuss methods for including the GHG impact of activities that occur
within urban areas that spur production (and associated GHG emissions) elsewhere, often
outside the geographical boundaries of the city of interest. Before we discuss which
Scope 3 items to include, it is useful to articulate why Scope 3 items should be included
in the first place. There is fairly wide acceptance that end-use of electricity in urban
areas should be systematically tracked back to GHG emissions occurring at powerplants
located outside city boundaries, such that these emissions are explicitly counted as Scope
2 emissions for that urban area. The same logic could apply for example to transport fuels
such as diesel and gasoline used for transport in cities – the GHG emissions associated
with refining these fuels should also be included just as is the impact of generating
electricity. The GHG emissions associated with fuel refining (termed Wells to Pump
emissions, WTP) are 20% to 25% of the emissions associated with the combustion of the
refined products in vehicles (termed Pump-to-wheel, PTW emissions), and thus are a
significant contributor to global GHG emissions. Likewise, agriculture and food
production contribute 20% to 25% of global and national GHG inventories, yet are
usually negligible in city-scale Scope 1-2 GHG accounting because much of food
production occurs outside the geographic boundaries of cities; at the same time, life in
cities would not be possible with food consumption.
4.2 What Scope 3 items to include?
The above discussion suggests two criteria for inclusion of cross-boundary GHG
emission motivated by activities occurring within urban areas. First, these activities
13
should be critical for the functionality of cities, and, second, the resulting GHG emission
should be significant contributors at larger spatial scales, such that their exclusion at the
city-scale creates a discontinuity in GHG accounting across spatial scales. We suggest
that Scope 1-2-3 emissions accounting not be used in place of Scope 1-2 accounting
which may be preferable for carbon trading schemes, but rather be used to augment
Scope 1-2 reporting.
Several urban area have included various cross-boundary emissions on an ad hoc basis,
e.g., GHG emissions associated with food consumption in cities have been included for
Paris (Hidalgo et al., 2007), Delhi and Calcutta (Sharma et al, 2002 a-c); with Delhi and
Calcutta focusing on non-energy emission from rice and milk production only. Embodied
emissions from producing various construction materials such as cement, steel and
asphalt have been included in inventories for Denver, Seattle, Paris, Delhi and Calcutta.
In pioneering a holistic emissions inventory for Denver, Ramaswami et al. (2008)
articulated a small set of Scope 3 inclusions critical for functionality of cities; these
included:
• Energy associated with transport of good and people outside city boundaries,
essential for trade and commuter travel to/from cities, allocated equally to origindestination locations.
• Embodied energy and associated GHG emissions associated with production of
key urban materials critical for life in cities such as:
o Food
o Transport fuels (other fuels already being accounted for)
o Water/Wastewater (if such production occurs outside city boundaries)
o Materials for shelter – chiefly cement as it is the single second largest
CO2 emitter following fossil fuel combustion.
The de minimus rule (CCAR 2007) can be applied yielding a stopping rule, i.e., no
further Scope 3 activities need be included unless they show more than a 1% increase in
the GHG accounting of a city. (Note that cities may also export CO2 embodied in goods
and services that needs to be accounted too; the “net” is of interest) Thus, applying a full
Scope 1-2-3 accounting can be made practical with the small and relevant list of Scope 3
activities listed above. Indeed, a small list of relevant Scope 3 items is the
recommendation of WRI and US EPA Climate Leaders Program. The Scope 3 items
listed above focus on critical service provisions required for life in cities that often occur
outside city boundaries; other material flows are assumed to be balanced out in the
trade/exchange of goods and services between cities (Ramaswami et al., 2008).
4.3 Measurement Impact
Inclusion of additional Scope 3 items has been shown to increase the GHG accounting of
cities. Incorporating primarily the impacts of fuel refining was shown to increase GHG
emissions associated with 8 global cities by as much as 24% (Kennedy et al., 2009).
Incorporating all five Scope 3 items increases the GHG accounting by an average of 45%
for eight US cities studies by Hillman and Ramaswami (2009). Further, incorporating all
five Scope 3 activities (Ramaswami et al., 2008) created consistency both in inclusions
14
and in the numeric per capita GHG emission computed at the city-scale for Denver versus
the larger national scale, both of which converged to about 25 Mt CO2e/capita (Table 4).
A similar analysis repeated for eight US cities showed remarkable consistency between
per capita city-scale Scope 1-2-3 emissions and national per capita emissions in the US
(Hillman & Ramaswami, 2009), suggesting that inclusion the specific list of five Scope 3
items proposed by Ramawami et al may help cities develop a more holistic GHG
emissions footprint that shows the overall impact of a city’s activities on global GHG
emissions.
4.4 Policy Impact
Incorporating a full scope 1-2-3 accounting provides city residents with a visualization
tool that helps connect their everyday activities with GHG emissions. Important activities
like food consumption and airline travel that appear in personal GHG calculators and in
national accounts also now appear in city-scale GHG accounts. This facilitates public
understanding of GHG emissions, and can also spur the development of win-win
strategies for GHG mitigation that link demand for materials and energy in cities with
their production. For example, accounting for embodied emissions associated with
cement in Denver resulted in the city adopting green concrete policies that require 15%
fly ash inclusion in concrete to reduce cement consumption at the city-scale (GreenPrint
Denver 2005). Inclusion of airline emission resulted in proposals to offer air travel off-set
programs directly at Denver International Airport.
Furthermore, a Scope 1-2-3 emissions assessment can avoid unintended credit being
given to policies that may merely shift emissions “out-of-boundary”. For example, large
scale use of hydrogen powered cars within city boundaries may result in zero PTW Scope
1 emissions, but significant WTP GHG emissions can occur outside city-boundaries
(Scope 3) if the hydrogen is produced from coal or natural gas. Conversely, many crosssector strategies may be used to reduce our overall Scope 1-2-3 footprint. For example,
information and communication technologies like teleconference and telepresence may
increase energy use in buildings while displacing airline travel. Once again, a full Scope
1-2-3 GHG accounting protocol would support such innovative cross-sector, crossboundary GHG reduction policies and strategies in a manner than boundary limited
(Scope 1-2) accounting does not.
4.5 Methodology, Challenges and a Proposed Framework for Scope 3
Inclusions
In the few studies that have included out-of-boundary impacts, the methodology has
varied. In estimating GHG emission for Delhi and Calcutta, we estimate that Sharma et al
(2002 a-c) used national average consumption data for some of the urban materials
studied, and coupled these with non-energy GHG emission factors for these materials as
specified by the IPCC. Thus a city-specific material flow analysis was not conducted and
only partial emission (non-energy) associated with these products (rice, milk, cement,
steel) were incorporated. For the study of Paris, it is reported that national data from the
15
food industry was used to determine average per capita consumption; however, methods
used for estimating cement and steel consumption were not fully detailed nor were data
sources for the requisite emission factors. In Paris, detailed analysis of resident airline
travel and freight transport was conducted; both incoming and outgoing trips were
counted and fully allocated to Paris. In contrast, a 50% allocation to destination and
origin locations was applied by Ramaswami et al for allocating both airline and surface
transport in Denver; such an origin-destination allocation procedure ensures that the same
trip not double counted at both ends of a trip.
For key urban materials in US cities, Ramaswami et al (2008) used tools from industrial
ecology – material flow analysis (MFA) and Life cycle assessment (LCA). MFA for
Scope 3 material flows in cities often using monetary consumption data available at the
metropolitan spatial scale. These flows are calibrated with national material consumption
data to ensure no methodological double counting for materials occurs. Emission factors
for the various materials considered for Denver – food, cement and transport fuels
(gasoline and diesel) were obtained from nationally calibrated LCA tools such as EIOLCA (www.eio-lca.net), the US National Renewable Energy Labs Life Cycle Inventory
database (NREL) and the US Argonne national Labs GREET model (ANNL) for
transportation fuels. With the exception of food – which is a complex supply chain –
embodied energy and GHG emissions from industrial production of materials such as
cement, steel and petroleum fuels could be readily computed for the US; IPCC provides
specific guidance on these parameters for international applications, particularly for nonenergy related industrial emissions from cement production, etc. (IPCC).
The above review indicates that methods that avoid double counting exist and have been
applied to assess upstream impacts of key Scope 3 consumption activities in cities.
International GHG emissions data for most of these materials exist or can be researched
(e.g., IPCC non-energy emissions), with the exception of food. See also the work of
Birch et al (2004) in the UK. When food production activities or cement factories occur
within city-boundaries, case studies in Delhi and Calcutta (Sharma et al. 2002)
demonstrate that the emissions from these in-boundary activities can be allocated to avoid
double counting between in-boundary and out-of-boundary activities.
Thus, careful Scope 1-2-3 accounting of GHGs at the city-scale is indeed possible. To be
consistent with other GHG accounting protocols (WRI, ICLEI, CCAR), we propose that
all cities and metropolitan regions are required to report Scope 1 and Scope 2 emissions
in their baselines (as well as Scope 3 waste emissions, where applicable). In addition, it is
highly recommended that cities also report on the five specific Scope 3 items listed above
(transport fuel production, food production, cement-steel production, water production,
and origin-destination-allocated external transport emissions), yielding a Scope 1-2-3
emissions footprint.
The city-scale emissions (Scope 1-2) and a broader emissions foot-print (Scope 1-2-3)
can be used together with application of two rules, as described in Hillman and
Ramaswami (2009):
16
•
•
Credit be given for strategies that reduce a city’s Scope 1-2 GHG emissions only
if they also reduce the city’s broader Scope 1-2-3 GHG emissions footprint; credit
is given for the smaller of the two reductions.
Policymakers allow flexibility to award cities credit for innovative strategies that
demonstrate “additionality3”and can quantifiably be shown to reduce their Scope
1-2-3 GHG footprint, even if the Scope 1-2 emissions do not show reductions.
With these rules, Scope 3 accounting can be used in conjunction with existing protocols
for Scope 1-2 accounting to develop more holistic and policy-relevant GHG management
at the city-scale.
5. Baseline Emissions
GHG emissions for 44 urban areas can now be presented using a methodology that is
consistent other than differences in industrial processes, waste, AFOLU and aviation
/marine emissions, discussed in Section 3. For these sectors where there are differences in
approach it is necessary to refer to Table 2.
The baselines presented in Tables 5 and 6 are for the same set of urban areas shown in
Table 2, other than the following changes:
•
•
•
Baselines for Beijing, Shanghai and Tianjin have been revised, including new
calculations for emissions from aviation, marine and waste.
For Delhi and Kolkata, the AFOLU and waste emissions for year 2000 have been
calculated using per capita emissions taken from the study by Sharma et al.
(2002).
Paris I (City of Paris) is excluded due to its unique life-cycle approach, but Paris
II (Isle de France) is included.
Note that the baselines reported are largely for the years 2005 or 2006. Only in the cases
of London, Glasgow, Calgary, Los Angeles, Washington DC, Mexico City Rio de
Janeiro, Sao Paulo, Delhi, Kolkata and Seoul are emissions given for earlier years. As
2005 is the reporting year for most of the studies, it could become a standard baseline
year for reporting emissions for further urban areas.
A precautionary note on the accuracy of baselines should be made. The results for total
emissions are reported to an accuracy of 10 kt in Table 5, but this accuracy is only to
facilitate the calculation of per capita emissions in Table 6. Baselines reported in both
tables are accurate at best to two significant figures.
3
Demonstrating additonality means to show that these reductions would not have
occurred on their own, in the absence of the specific policy of program.
17
6. Conclusion and Recommendations
The primary contribution of this paper has been to present GHG emissions for over 40
urban areas (cities and metropolitan regions) from five continents. This has been
achieved by assembling and assessing previous studies of urban GHG emissions, and
adding further analysis where necessary and where data permits. There have been
discrepancies between previous studies in the methodology for determining emissions
from waste, and in the reporting of emissions for aviation, marine, agriculture, and
industrial processes. Our results have been presented (in Tables 5 and 6) so that these
differences can be recognized (through reference to Table 2).
Despite these often minor differences, this work has shown that the potential clearly
exists to establish an open, global protocol for quantifying GHG emissions attributable to
urban areas. Emissions from waste should be determined using the IPCC (2006)
guidelines; this has primarily been hindered by the significant data requirements.
Agricultural and industrial process emissions, though small for many urban areas, need to
be more carefully accounted for; again, the IPCC guidelines can be followed. Whether
emissions from aviation and marine are excluded or included, primarily depends on
whether the baseline GHG measures are only to inform local government policy, or are to
be a wider reflection of the carbon dependence of urban economies. If aviation and
marine emissions are included, then they should reflect the global connections that exist
between cities – and thus include all emissions from international transportation. (Data to
support such calculations is already collected at national levels, but not reported in
national totals as per the UNFCCC) The methodology of Ramaswami et al. (2008)
addresses issues with assigning emissions when passengers transfer between flights and
incorporates most relevant cross-boundary energy flows critical for functioning of cities.
Overall, resolution of these differences seems tractable.
The baseline emissions include those for some cities and some wider metropolitan
regions. There are merits to developing baseline emissions for both. Cities have a single
administrative authority (albeit subject to national, provincial and state governments),
enabling them to have potentially greater control over emissions reductions.
Metropolitan regions sometimes have more fragmented political authority, yet these
regions typically have higher per capita emissions than cities, due to low density suburbs
(VandeWeghe and Kennedy, 2007; Glaeser and Kahn, 2008), airports, and often higher
concentrations of industry. A strong point of the methodology reviewed in this paper is
that it applies equally well to cities and to metropolitan regions.
Aside from differences with emissions from waste, air, marine, etc (discussed above), the
greatest uncertainty in urban GHG baselines lies with emissions from road transportation.
In Table 2, we distinguished between three techniques for estimating gasoline
consumption in urban areas (sales; models/surveys; and scaling). Differences between
these techniques may be less than 5% (Kennedy et al., 2009a). This uncertainty might be
reduced further, however, if new urban transportation models and surveys were
developed specifically for determining GHG emissions, rather than urban transportation
planning in general. For example, fuels sales data could explicitly be used in model
18
calibration. Such improvements in quantifying urban GHG emissions would also likely
support calculations made for national inventories.
Further assessment of the uncertainty in quantifying urban GHG emissions is warranted.
Other than the colour-coded scheme used in the GRIP studies (Carney et al 2009), there
has been little formal analysis of uncertainty in urban emissions, e.g., using Monte Carlo
simulation. Volume 1 of the IPCC (2006) Guidelines provides recommended approaches
for uncertainty assessment and quality assurance.
Improvements in the reporting of urban GHG emissions might also be made. Under ISO
Standard 14064, emissions should be reported for six individual greenhouse gases. This
has not been common for cities (hence we have not done so here). Perhaps more
importantly for quality assurance purposes, urban areas should always report the activity
data and emissions factors used to determine emissions.
A further recommendation of this work is the continued development of consumptionbased measures of GHG emissions for urban areas. By including Scope 2 and 3
emissions, as per the WRI / WBCSD, the overall methodology of this work recognizes
that emissions should be assigned to urban areas based on end-use activities. There are
further consumption based emissions that can also be attributed to urban areas, beyond
those presented in our results. These include those embodied in food and materials
consumed in cities, and upstream emission associated with the mining and refining of the
fuels combusted in cities. Some of these emissions have been quantified in a few studies
(Table 2), but not in enough to be consistently applied to all cities at this point.
Methodologies for determining further consumption-based emissions have been
developed (Ramaswami et al., 2008; Hillman and Ramaswami, 2009; Hidalgo et al.,
2005; Mairie de Paris 2009), though perhaps need to be compared. The main barrier is
lack of data on material flows into urban areas. Standard approaches for applying
principles of industrial ecology to urban areas need to be developed.
Acknowledgements
We are grateful to the Global Environment Facility for supporting this work. We also
wish to thank Lorraine Sugar, David Bristow and Abel Chavez for their help in compiling
the results tables.
References
Baldasano, J.M., C. Soriano, L. Boada (1999) Emission inventory for greenhouse gases
in the City of Barcelona, 1987-1996. Atmospheric Environment 33, 3765-3775.
Birch, R., Barrett, J., Wiedmann, T (2004) Exploring the consumption and related
environmental impacts of socio-economic groups within the UK, International
Workshop on Sustainable Consumption,
http://homepages.see.leeds.ac.uk/~leckh/leeds04/5.4%20leeds%20conference%20paper.p
df
19
Bureau of Economic Analysis (BEA) Gross Domestic Product by Metropolitan Area,
Regional Economic Accounts, http://www.bea.gov/regional/gdpmetro/ accessed May 21,
2009. For global meaures, also see: http://www.citymayors.com/statistics/richest-cities2005.html
City of Calgary (2003) Calgary Community Greenhouse Gas Emissions Inventory,
accessed on-line
http://www.calgary.ca/docgallery/bu/environmental_management/2003_community_emi
ssions_report.pdf
Carney, S., N. Green, R. Wood, R. Read (2009) Greenhouse Gas Emissions Inventories
for Eighteen European Regions, EU CO2 80/50 Project Stage 1: Inventory Formation.
The Greenhouse Gas Regional Inventory Protocol (GRIP)
District of Columbia, Department of Health (2005) District of Columbia Greenhouse Gas
Emissions Inventories and Preliminary Projections, prepared by Air Quality Division,
Bureau of Environmental Quality, Environmental Health Administration.
Dhakal, S. (2004) Urban Energy Use and Greenhouse Gas Emissions in Asian Megacities: Policies for a Sustainable Future, Institute for Global Environmental strategies,
Hayama, Japan..
Dhakal, S. (2009) Urban energy use and carbon emissions from cities in China and policy
implications, Energy Policy; doi:10.1016/j.enpol.2009.05.020
Dodman, D. (2009) Blaming cities for climate change? An analysis of urban greenhouse
gas emission inventories, Environment and Urbanization; 21; 185
Duffy, H. (1995) Competitive cities: succeeding in the global economy – E&F Spon.
Dubeux, C.B.S. and E.L. La Rovere (2007) Local perspectives in the control of
greenhouse gas emissions – The case of Rio de Janeiro, Cities, Vol. 24, No. 5, p. 353–
364.
Estrategia Local de Accion Cliamtica (ELAC) 2000
Friedmann, J. (1986) The world city hypothesis, Development and Change, 17, 69-83
Glaeser, E.L.; Khan, M.E. (2008) The Greenness of Cities: Carbon Dioxide Emissions
and Urban Development, National Bureau of Economic Research, Working Paper 14238.
August 2008.
Harvey, L.D.D. (1993) Tackling urban CO2 emissions in Toronto, Environment 35(7),
16-20 and 33-44.
20
Hillman, T; Ramaswami, A. (2009) Greenhouse Gas Footprints and Energy Use
Benchmarks for Eight US Cities, submitted to Environmental Science & Technology.
ICLEI (2006) International Progress Report, Cities for Climate Protection
ICLEI (2009) International Local Government GHG Emissions Analysis Protocol Draft
Release Version 1.0. Accessed May 2009:
http://www.iclei.org/fileadmin/user_upload/documents/Global/Progams/GHG/LGGHGE
missionsProtocol.pdf
IGES/APN (2002) Policy Integration & Industrial Transformation towards Sustainable
Urban Energy Use for Cities in Asia, Proceedings of the IGES/ASPN Mega-city Project
Workshop; and the International Symposium on Sustainable Urban Development in Asia,
Kitakyushu, Japan, January 23-25, 2002
IPCC (1997). Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories.
IPCC (2006). Guidelines for National Greenhouse Gas Inventories, Volumes 1 to 5
http://www.ipcc-nggip.iges.or.jp/public/2006gl/index.html
Kates, R.W., M.W. Mayfield, R.D. Torrie, and B. Witcher (1998) Methods for estimating
greenhouse gases from local places, Local Environment 3:3, 279-297.
Kennedy, C.A., J. Cuddihy, and J. Engel Yan (2007). The changing metabolism of cities,
J. Industrial Ecology, 11(2), 43- 59.
Kennedy, C., J. Steinberger, B. Gasson, T. Hillman, M. Havránek, D. Pataki, A.
Phdungsilp, A. Ramaswami, G. Villalba Mendez,. (2009a) Methodology for
Inventorying Greenhouse Gas Emissions from Global Cities, submitted to Energy Policy.
Kennedy, C., J. Steinberger, B. Gasson, T. Hillman, M. Havránek, D. Pataki, A.
Phdungsilp, A. Ramaswami, G. Villalba Mendez, (2009b) Greenhouse Gas Emissions
from Global Cities, submitted to Environmental Science and Technology.
Llewelyn-Davies, David Banister, and Sir Peter Hall (2004) Transport and City
Competitiveness: Literature Review, Department of Transport and Office of the Deputy
Prime Minister.
Mairie de Paris (2009) Le Bilan Carbone de Paris: Bilan des émissions de gaz à effet de
serre
Mayor of London (2007). Action Today to Protect Tomorrow - The Mayor’s Climate
Change Action Plan, London, UK: Mayor of London. [Online]. Available HTTP:
http://www.london.gov.uk/mayor/environment/climate-change/docs/ccap_fullreport.pdf
(Accessed June 9, 2008)
21
Mitra, A.P., C. Sharma and M.A.Y. Ajero (2003). Energy and Emissions in South Asian
Mega-cities: Study on Kolkata, Delhi and Manila, Proceedings of International Workshop
on Policy Integration Towards, Sustainable Urban Energy Use for Cities in Asia,
Honolulu, Hawaii, 4-5 February 2003.
Nagoya University / NIES (National Institute for Environmental Studies, Japan) / GCP
(Global Carbon Project). (2009) International Symposium: Realizing Low Carbon Cities
- Bridging Science and Policy; and International Workshop: Towards Low Carbon Cities
- Understanding and Analyzing Urban Energy and Carbon, Nagoya Japan, February 1618, 2009.
New York City (2007) Inventory of New York City Greenhouse Gas Emissions, Office
of Long-term Planning and Sustainability.
Phdungsilp, A. (2006) Energy Analysis for Sustainable Megacities, Licentiate of
Engineering Thesis, Department of Energy Technology, Royal Institute of Technology,
Sweden.
Ramaswami, A. Hillman, T, Janson, B. Reiner, M. Thomas, G. (2008). A demandcentered, hybrid life cycle methodology for city-scale greenhouse gas emissions, in
press., Environ Sci Technol. 42, 6455-61.
Rio Prefeitura Meio Ambiente (2003), Inventário de Emissões de Gases do
Efeito Estufa da Cidade do Rio de Janeiro, Centro de Estudos Integrados sobre
Meio Ambiente e Mudanças Climáticas, Rio de Janeiro;
Rosenzweig, C., S. Mehrotra, C.E. Natenzon, A. Omojola, R. Folorunsho, J. Gilbride
(2009). Creating Risk-Based Climate Change Information for Urban Areas, 5th Urban
Research Symposium Cities and Climate Change: Responding to an Urgent Agenda,
Marseille, France, June 28 - 30, 2009.
Sassen, S. (1991) The Global City, Princeton University Press, NJ.
Satterthwaite, D. (2008). Cities' contribution to global warming: notes on the allocation
of greenhouse gas emissions. Environment and Urbanization. 20(2): 539-549.
Secretaria Municipal do Verde e do Meio Ambiente de São Paulo (SVMA) (2005),
Inventário de Emissões de Efeito Estufa do Município de São Paulo, Centro de Estudos
Integradossobre Meio Ambiente e Mudanças Climáticas (Centro Clima) da Coordenação
dos Programas de Pós-graduação de Engenharia (COPPE) da Universidade Federal do
Rio de Janeiro (UFRJ).
Sharma, C., A. Dasgupta, and A.P. Mitra (2002) Inventory Of GHGs And Other Urban
Pollutants From Agriculture and Waste Sectors In Delhi And Calcutta, Proceedings of
IGES/APN Mega-City Project, Kitakyushu Japan), 23-25 January 2002.
22
Taylor, P.J. (2004) World City Network: A Global Urban Analysis, Routledge: New
York.
Tokyo Metropolitan Government (2006) Environmental White Paper; available at
http://www.kankyo.metro.tokyo.jp/kouhou/english/
VandeWeghe, J., and Kennedy, C.A. (2007) A Spatial Analysis of Residential
Greenhouse Gas Emissions in the Toronto Census Metropolitan Area. J. Industrial
Ecology, 11(2), 133-144.
WRI / WBCSD (World Resources Institute and World Business Council for Sustainable
Development) The Green house Gas Protocol: A Corporate Accounting and Reporting
Standard: Revised Edition. Accessed May 2009: http://www.ghgprotocol.org/
23
Table 1: Definition and Population of Cities and Metropolitan Regions in this Paper
Abbreviated
name used in
this paper
Athens
Barcelona
Bologna
Brussels
Frankfurt
Geneva
Glasgow
Hamburg
Helsinki
London
Ljubljana
Madrid
Naples
Oslo
Paris I
Paris II
Porto
Prague
Rotterdam
Stockholm
Stuttgart
Torino
Veneto
Definition
Study Year
Population
Metropolitan Region
City
Province
Capital region
Frankfurt/Rhein Main
Canton
Glasgow and the Clyde Valley
Metropolitan Region
Capital Region
Greater London
Osrednjeslovenska Region
Comunidad de Madrid
Province
Metropolitan Region
City
Ile de France
Metropolitan Region
Greater Prague
City
Metropolitan Region
Metropolitan Region
Metropolitan Region
Province
2005
2006
2005
2005
2005
2005
2004
2005
2005
2003
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
3,989,000
1,605,602
899,996
1,006,749
3,778,124
432,058
1,747,040
4,259,670
988,526
7,364,100
500,021
5,964,143
3,086,622
1,039,536
2,125,800
11,532,398
1,666,821
1,181,610
592,552
1,889,945
2,667,766
2,243,000
4,738,313
Austin
Calgary
Denver
Los Angeles
Minneapolis
New York City
Portland
Seattle
Toronto
Washington DC
City
City
City and County
County
City
City
City
City
Greater Toronto Area
District of Columbia
2005
2003
2005
2000
2005
2005
2005
2005
2005
2000
672,011
Mexico City
Rio de Janeiro
Sao Paulo
City
City
City
2000
1998
2000
922,315
579,744
9,519,338
387,711
8,170,000
682,835
575,732
5,555,912
571,723
8,669,594
5,633,407
10,434,252
24
Table 1: cont.
Abbreviated
name used in
this paper
Bangkok
Beijing
Kolkata
Delhi
Shanghai
Seoul
Tianjin
Tokyo
Cape Town
Definition
City
Beijing Government Administered
Area (Province)
Metropolitan Area
National capital territory
Shanghai Government
Administered Area (Province)
Seoul City
Tianjin Government Administered
Area (Province)
Tokyo Metropolitan Government
Administered Areas (Tokyo-to)
City
Study Year
Population
2005
2006
5,658,953
15,810,000
2000
2000
2006
15,700,000
13,200,000
18,150,000
1998
2006
10,321,496
10,750,000
2006
12,677,921
2006
3,497,097
25
?
?
?
?
?
√
√
√
√
?
?
?
?
?
?
?
?
?
?
?
?
?
?
?
√
?
?
√
?
?
?
?
?
?
?
√
?
?
?
?
?
?
NA
√
√
√
√
?
√
?
?
neg
√
√
√
√
?
√
√
√
?
?
?
?
?
NA
√
√
√
√
√
?
√
?
?
√
?
√
√
√
?
√
√
√
√
√
√
√
√
?
√
√
?
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
?
√
√
√
√
√
?
√
√
√
√
√
√
√
√
EMBODIED FOOD OR
MATERIALS
√
√
√
√
√
√
?
√
√
UPSTREAM FUELS
Wastewater
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
Landfill: Total Yield Gas
√
Landfill: EPA WARM
Model
√
Landfill: scaled from
national data
Biofuels (fuel wood, dung
cakes)
Railways
Marine: inland or nearshore (12mile) only
?
?
WASTE
?
?
?
?
?
?
?
?
?
?
√
√
AFOLU
?
?
?
√
?
?
?
?
√
?
?
?
Marine: all fuels loaded
at ports
?
Aviation: all domestic;
international LTO only
?
Aviation: all fuels loaded
at airports
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
Gasoline use from model
or traffic counts
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
INDUSTRIAL
PROCESSES
Carney et al. 2009
Kennedy et al. 2009
Carney et al. 2009
Carney et al. 2009
Carney et al. 2009
Kennedy et al. 2009
Carney et al. 2009
Carney et al. 2009
Carney et al. 2009
Kennedy et al. 2009
Carney et al. 2009
Carney et al. 2009
Carney et al. 2009
Carney et al. 2009
Mairie de Paris 2009
Carney et al. 2009
Carney et al. 2009
Kennedy et al. 2009
Carney et al. 2009
Carney et al 2009
Carney et al. 2009
Carney et al. 2009
Carney et al. 2009
Gasoline use scaled
Athens
Barcelona
Bologna
Brussels
Frankfurt
Geneva
Glasgow
Hamburg
Helsinki
London
Ljubljana
Madrid
Naples
Oslo
Paris I
Paris II
Porto
Prague
Rotterdam
Stockholm
Stuttgart
Torino
Veneto
Electrical line losses
Source
(see Table 1 for
definition)
ENERGY
City or
Metropolitan
Region
Gasoline use from sales
data
Table 2: Comparison of greenhouse gas studies for selected cities and metropolitan regions
√
√
√
√
√
√
√
26
√
√
√
√
√*
√
√
√
√*
√
√
√
?
√
?
√
√
√
?
√
?
√
√
√
√
?
NA
NA
NA
√
√
neg
√
Wastewater
Landfill: Total Yield Gas
Landfill: EPA WARM
Model
Landfill: scaled from
national data
WASTE
AFOLU
√
?
√
√
INDUSTRIAL
PROCESSES
√
#
√
√
√
√
√
√
EMBODIED FOOD OR
MATERIALS
√
√
√
√
UPSTREAM FUELS
√
√*
√*
√
√*
Biofuels (fuel wood, dung
cakes)
ELAC 2000
Dubeux LaRovere 2007
SVMA 2005
?
√
√
Railways
México City
Rio de Janeiro
Sao Paulo
Seattle
?
Marine: inland or nearshore (12mile) only
√
√
New York City
Portland
?
√*
Marine: all fuels loaded
at ports
?
√
√
√
√
√
Aviation: all domestic;
international LTO only
√
√
√
√
√
Calgary
Denver
Denver
Los Angeles
Minneapolis
Aviation: all fuels loaded
at airports
√
Austin
Gasoline use from model
or traffic counts
Electrical line losses
√
Toronto
Washington DC
Hillman &
Ramaswami, 2009
City of Calgary 2009
Ramaswami et al 2008
Kennedy et al. 2009
Kennedy et al. 2009
Hillman &
Ramaswami, 2009
Kennedy et al. 2009
Hillman &
Ramaswami,2009
Hillman &
Ramaswami,2009
Kennedy et al. 2009
DC Dept. Health 2005
Source
(see Table 1 for
definition)
Gasoline use scaled
ENERGY
City or
Metropolitan
Region
Gasoline use from sales
data
Table 2: cont. i
√
√
√
√
√
√
√
?
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
?
?
√
√
?
√
?
?
√
√
√
27
Bangkok
Beijing
Delhi
Kolkata
Shanghai
Seoul
Tianjin
Tokyo
Kennedy et al. 2009
Dhakal 2009
Mitra et al. 2003
Mitra et al. 2003
Dhakal 2009
Dhakal 2004
Dhakal 2009
Tokyo Metropolitan
Government, 2009
√
√
√
√
√
√
√
√
Cape Town
Kennedy et al. 2009
√
√
√
?
?
√
√
√
?
?
√
?
?
?
?
?
?
?
?
√
√
?
?
?
?
√
√
√
√
√
√
♪
√
√
√
?
√
√♦
√♦
√♦
√♦
√?
√
√
EMBODIED FOOD OR
MATERIALS
UPSTREAM FUELS
Wastewater
Landfill: Total Yield Gas
Landfill: EPA WARM
model
Landfill: scaled from
national data
WASTE
AFOLU
INDUSTRIAL
PROCESSES
Biofuels (fuel wood, dung
cakes)
Railways
Marine: inland or nearshore (12mile) only
Marine: all fuels loaded
at ports
Aviation: all domestic;
international LTO only
Aviation: all fuels loaded
at airports
Gasoline use from model
or traffic counts
Gasoline use scaled
Gasoline use from sales
data
Source
(see Table 1 for
definition)
ENERGY
City or
Metropolitan
Region
Electrical line losses
Table 2: cont. ii
√
√
√
√♠
√
Note:
The table only displays emissions sub-categories for which there are differences between studies.
NA = not applicable
neg = negligible
* Aviation emissions are apportioned across co-located cities in the larger metropolitan area.
#AFOLU emission were estimated and found to be less than 0.1% and hence not reported. Data on carbon sequestration in urban soils
was used (from another study).
♦ AFOLU and Waste emissions for Delhi and Kolkata are given in an earlier study by Sharma et al. (2002).
♠ Also includes electricity
♪ only includes aviation emissions within the urban region
28
Table 3 Definition of Scope 1, 2, and 3 GHG Emissions (adapted from p.25 of WRI /
WBCSD)
Scope 1: Direct GHG Emissions
Direct GHG emissions occur from sources that are owned or controlled by the company,
e.g., emissions from combustion in owned or controlled boilers, furnaces, vehicles etc., or
emissions from chemical production in owned or controlled process equipment. (Direct
CO2 emissions from combustion of biomass; and greenhouse gases not covered by the
Kyoto protocol are not included in scope 1)
Scope 2: Electricity indirect GHG emissions
These are emissions from the generation of purchased electricity consumed by the
company. Scope 2 emissions physically occur at the facility where electricity is
generated.
Scope 3: Other indirect emissions
Emissions in this optional reporting capacity are a consequence of the activities of the
company, but occur from sources not owned or controlled by the company. Examples of
scope 3 activities are extraction and production of purchased materials; transportation of
purchased fuels; and use of sold products and services.
29
Table 4 Denver’s Average Per Capita GHG Emissions Compared to the National
Average, State of Colorado Average and to other Colorado Cities (Adapted from
from Ramaswami et al. 2008)
Inclusions
Denver’s Per
capita GHG
emissions (mt
CO2e/cap)
[Scope 1+2 & Waste] plus
airline travel and key
25.3
urban materials
Scope 1 + 2 & Waste (No
airline travel or
embodied energy of jey
urban materials)
18
National, State or Other
City per capita GHG
emissions (CO2e/cap)
National: 24.5
Colorado: 25.2
Other Colorado Cities:
17.8 to 18.4
30
Athens
Barcelona
Bologna
Brussels
Frankfurt
Geneva
Glasgow
Hamburg
Helsinki
London
Ljubljana
Madrid
Naples
Oslo
Paris II
Porto
Prague
Rotterdam
Stockholm
Stuttgart
Torino
Veneto
2005
2006
2005
2005
2005
2005
2004
2005
2005
2003
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
Austin
Calgary
Denver
Los Angeles
Minneapolis
New York City
Portland
Seattle
Toronto
Washington DC
2005
2003
2005
2000
2005
2005
2005
2005
2005
2000
3.68
2.45
2.67
0.74
0.00
46.31
23.00
0.00
9.33
1.06
0.00
10.37
15.94
10.11
83.36
7.00
62.65
8.37
7.75
54.60
10.31
*
35.72
6.35
8.93
7.34
44.40
3.19
13.77
33.96
6.70
69.31
4.31
36.25
10.66
3.35
47.01
11.14
10.39
17.08
6.35
40.89
17.60
39.55
4.38
0.45
0.07
0.14
4.99
0.72
0.02
1.61
0.25
2.82
0.17
0.00
0.07
2.23
0.87
0.11
3.53
0.07
0.51
0.15
0.17
0.36
2.51
3.29
0.72
4.46
0.03
0.00
0.19
0.36
0.49
0.06
6.91
0.38
0.00
0.13
0.24
0.87
1.47
3.28
10.37
-0.01
0.86
17.83
14.19
*
*
4.62
16.10
6.20
0.00
10.97
117.29
7.00
83.04
8.37
7.75
59.22
10.31
2.10
3.19
0.01
-0.02
1.02
0.39
0.25
0.04
0.61
0.16
0.56
0.28
0.04
1.53
0.20
2.15
0.46
0.11
2.20
0.56
0.13
0.28
0.12
0.43
0.28
1.18
0.11
0.44
0.11
4.65
0.03
2.83
0.10
0.07
1.81
0.74
4.07
2.61
47.84
9.97
10.48
16.37
10.22
90.11
7.03
65.48
8.47
7.82
59.60
31
TOTAL
TOTAL (excluding
aviation and marine)
WASTE
AFOLU
INDUSTRIAL
PROCESSES
ENERGY
(see Table 1 for
definition)
Marine
Year
Aviation:
City or
Metropolitan
Region
ENERGY (excluding
aviation and marine)
Table 5: Greenhouse gas emissions for cities and metropolitan regions
(in million t CO2 e)
41.57
6.74
9.97
7.55
51.61
3.35
15.30
41.52
6.94
70.84
4.77
40.98
12.49
3.63
59.64
12.14
11.03
17.64
6.88
42.57
21.86
47.29
*
11.08
124.04
85.87
*
*
64.22
11.04
Mexico City
Rio de Janeiro
Sao Paulo
2000
1998
2000
31.68
5.78
9.57
Bangkok
Beijing
Delhi
Kolkata
Shanghai
2005
2006
2000
2000
2006
42.61
146.93
17.31
13.83
10.85
7.15
0.00
Tianjin
Seoul
Tokyo
1998
2006
2006
197.07
116.43
42.03
55.88
8.55
0.41
5.00
1.23
Cape Town
2006
20.91
3.11
0.00
0.86
0.94
TOTAL
3.59
4.96
3.70
35.27
11.25
13.28
53.46
154.09
6.98
4.92
3.34
3.97
49.59
151.85
20.65
17.80
60.44
159.00
210.62
118.08
1.36
1.17
211.98
119.25
0.56
198.43
117.60
42.03
57.86
6.22
27.13
40.43
6.64
10.51
60.04
10.19
TOTAL (excluding
aviation and marine)
WASTE
AFOLU
INDUSTRIAL
PROCESSES
ENERGY
(see Table 1 for
definition)
Marine
Year
Aviation:
City or
Metropolitan
Region
ENERGY (excluding
aviation and marine)
Table 5: cont.
0.24
0.01
0.27
0.00
1.42
34.21
* Airline data for these cities pending review (Hillman & Ramaswami, 2009)
32
12.11
14.22
62.02
Athens
Barcelona
Bologna
Brussels
Frankfurt
Geneva
Glasgow
Hamburg
Helsinki
London
Ljubljana
Madrid
Naples
Oslo
Paris I
Porto
Prague
Rotterdam
Stockholm
Stuttgart
Torino
Veneto
2005
2006
2005
2005
2005
2005
2004
2005
2005
2003
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
2005
Austin
Calgary
Denver
Los Angeles
Minneapolis
New York City
Portland
Seattle
Toronto
Washington DC
2005
2003
2005
2000
2005
2005
2005
2005
2005
2000
2.3
1.7
0.0
5.7
1.7
0.0
0.0
6.3
3.1
0.0
0.0
0.0
0.0
7.9
0.9
0.0
0.0
15.4
17.3
17.69
8.8
18.28
7.7
12.26
13.5
9.8
18.0
*
1.5
1.9
*
1.7
*
*
0.8
15.4
0.0
19.19
1.7
0.8
12.26
13.57
0.0
0.0
9.0
4.0
9.9
7.3
11.8
7.4
7.9
8.0
6.8
9.4
8.6
6.1
3.5
3.2
4.1
6.7
8.8
28.8
3.4
15.3
7.8
8.3
1.1
0.0
0.1
0.1
1.3
0.0
0.1
0.7
0.2
0.0
0.1
0.4
0.3
0.1
0.3
0.0
0.4
0.2
0.1
0.1
1.1
0.7
0.1
0.0
0.8
0.0
0.4
0.0
0.4
1.0
0.0
0.0
0.4
0.1
0.2
0.1
0.6
0.2
0.0
0.2
0.1
0.3
0.7
0.7
0.0
0.0
12.3
0.2
0.0
10.2
0.0
0.0
10.7
0.6
0.0
0.0
0.0
0.3
0.2
0.3
0.0
0.2
0.4
0.3
0.1
0.0
0.2
0.4
0.4
0.1
0.1
0.2
0.3
0.1
0.5
0.1
0.2
0.1
0.2
0.17
0.5
0.19
0.5
0.06
0.3
0.15
13.68
0.3
1.3
2.5
6.0
6.5
8.4
15.57
17.7
17.88
9.5
18.34
8.0
12.41
13.68
10.7
19.3
33
TOTAL
TOTAL (excluding
aviation and marine)
WASTE
AFOLU
INDUSTRIAL
PROCESSES
ENERGY
(see Table 1 for
definition)
Marine
Year
Aviation:
City or
Metropolitan
Region
ENERGY (excluding
aviation and marine)
Table 6: Per capita greenhouse gas emissions for cities and metropolitan regions (t
CO2 e
10.4
4.2
11.1
7.5
13.7
7.8
8.8
9.7
7.0
9.6
9.5
6.9
4.0
3.5
5.2
7.3
9.3
29.8
3.6
16.0
9.7
10.0
*
19.38
13.0
*
10.5
*
*
11.6
Mexico City
Rio de Janeiro
Sao Paulo
2000
1998
2000
3.7
1.0
0.9
Bangkok
Beijing
Delhi
Kolkata
Shanghai
Tianjin
Seoul
Tokyo
2005
2006
2000
2000
2006
1998
2006
2006
7.5
9.3
1.3
0.9
10.9
10.8
4.1
4.4
Cape Town
2006
6.0
0.0
0.2
0.1
1.2
1.0
1.9
0.5
0.0
9.4
9.7
0.5
0.0
0.3
0.1
11.6
11.0
0.0
0.0
4.7
0.9
2.9
9.8
0.0
0.0
0.0
0.0
0.0
0.0
0.1
0.4
0.9
0.4
4.1
2.0
1.3
1.2
0.3
0.3
0.3
0.1
0.1
0.0
8.8
9.6
1.6
1.1
10.9
10.9
4,1
4.6
1.8
7.8
* Airline data for these cities pending review (Hillman & Ramaswami, 2009)
34
TOTAL
TOTAL (excluding
aviation and marine)
WASTE
AFOLU
INDUSTRIAL
PROCESSES
ENERGY
(see Table 1 for
definition)
Marine
Year
Aviation:
City or
Metropolitan
Region
ENERGY (excluding
aviation and marine)
Table 6: cont.
2.1
1.4
10.7
10.1
11.7
11.1
4.9
11.6