Ecological Indicators 16 (2012) 9199

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Carbon footprints in a bipolar, climate-constrained world

G.R. Cranston a, , G.P. Hammond a,b
a b

Department of Mechanical Engineering, University of Bath, Bath BA2 7AY, UK Institute for Sustainable Energy and the Environment (ISEE), University of Bath, Bath BA2 7AY, UK

a r t i c l e

i n f o

a b s t r a c t
Carbon footprints represent the amount of carbon (or carbon dioxide equivalent) emissions associated with a given activity or community, and are closely related to ecological or environmental footprints. But, unlike the latter, they are generally presented in terms of units of mass or weight (kilograms per functional unit), rather than in spatial units (such as global hectares). These carbon footprints have become the currency of debate in a climate-constrained world. They are increasingly popular ecological indicators, adopted by individuals, businesses, governments, and the media alike. The inuence of economic wealth, population density, and pollutant emission intensity upon national carbon footprints has been correlated using a powerlaw equation. This analysis provides a greater understanding of the carbon footprint concept and its national determinants in a bipolar world: one embracing the countries of the prosperous or industrialised North of the planet and the other covering the developing, Majority South, where some 80% of the global population resides. Such a bipolar world is depicted by highlighting the countries that make up some of the largest economies in the form of the G-20 nations. This international grouping encompasses nation states from both the North and South. Emerging countries from the South with large populations (such as China and India) now have economies whose overall size rivals those of the North. Here the factors that are principally responsible for driving national carbon footprints are illustrated in the context of their potential consequences for global warming. An insight into these parameters is important in a period when the international community is seeking to agree a successor regime to the Kyoto Protocol on climate change mitigation. 2011 Published by Elsevier Ltd.

Keywords: Carbon footprints Carbon weights Carbon dioxide emissions Nation states International development Sustainability

1. Introduction Carbon footprints represent the amount of carbon [or carbon dioxide (CO2 ) equivalent] emissions associated with a given activity or community, and are closely related to ecological or environmental footprints. But, unlike the latter, they are generally presented in terms of units of mass or weight (kilograms per functional unit), rather than in spatial units (such as global hectares). These carbon footprints have become the currency of debate in a climate-constrained world. They are increasingly popular ecological indicators, adopted by individuals, businesses, governments, and the media alike. The carbon footprint is the largest contributor to humanitys total environmental footprint1 ; approximately 50% (Ewing et al., 2008). The global carbon footprint increased by more than ten times between 1961 and 2005; representing the largest anthropogenic demand on the biosphere (Hails et al., 2008). The burden upon Natures resources varies between countries (as has

been demonstrated by Cranston et al., 2010), as does their ability to sequester emissions. In the present study the inuence of economic wealth, population density, and pollutant emission intensity upon national carbon footprints have been correlated using a powerlaw equation.2 This analysis provides a greater understanding of the carbon footprint concept and its national determinants in a bipolar world: one embracing the countries of the prosperous or industrialised North of the planet and the other covering the developing, Majority South, where some 80% of the global population resides (Hammond, 2006; Cranston and Hammond, 2010a, 2010b). This notion is an abstraction, but one that those in the international development community nd to be a useful tool for stimulating the discourse (Hammond, 2006). Such a bipolar world is exemplied by looking at the countries that make up some of the largest economies in the form of the G-20 nations. This international grouping encompasses nation states from both the North and South. Emerging countries from the South with large populations (such as China and India) now have economies whose

Corresponding author. Present address: Global Footprint Network, Switzerland. E-mail addresses: G.R.Cranston@bath.ac.uk, gemma@footprintnetwork.org (G.R. Cranston). 1 Other contributions/factors may be found to dominate in the case of developing countries. 1470-160X/$ see front matter 2011 Published by Elsevier Ltd. doi:10.1016/j.ecolind.2011.03.011

2 The relationship between size and rank in a sample, represented by an equation in which the independent variables are raised to powers (see Hammond, 2006).

92

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

Fig. 1. The G-20 grouping of nations identied on a world map [(l r) USA, Canada, Mexico, Argentina, Brazil, France, UK, Germany, Italy, South Africa, the Russian Federation, Turkey, Saudi Arabia, India, China, Indonesia, South Korea, Japan, and Australia].

statistical data. It was established in 1999 in the aftermath of the 1997 Asian nancial crisis, with the aim of bringing together the major advanced and emerging economies with a view to tackling common economic and related problems. This grouping encompasses 19 nation states, plus the European Union (EU) as a separate entity. These 19 countries account for approximately 90% of global economic product, 80% of world trade (including EU intra-trade), and two-thirds of the global population. The G-20 grouping grew in status following the 2008 Washington summit, where it was announced that it would replace the G8 (industrialised) grouping as the main vehicle for achieving sustainable and balanced economic income. The membership includes countries from all the continents with the exception of Antarctica. The environmental footprint analysis of Cranston et al. (2010) based on 2003 datasets was initially revisited for the present study, using the most up-to-date data for the year 2005. Many of the potential determinants were eliminated from the original correlating equation, because they were found to be very weak. These factors were therefore not included in this updated study. Thus, the initial expression was simplied to the following form for the per capita eco-footprint [in global hectares (gha)]: ef = constant [(GNI)]p (PD)q (EI)r (CR)s ] (1)

overall size rivals those of the North (Hammond, 2006; Cranston et al., 2010). Here the factors that are principally responsible for driving national carbon footprints are illustrated in the context of their potential consequences for global warming. An insight into these parameters is important in a period when the international community is seeking to agree a successor regime to the Kyoto Protocol on climate change mitigation. The interconnections and differences between environmental footprints (ef), carbon footprints (cf), and carbon weights (CW ) of nation states are clearly distinguished. 2. The eco-footprints of nation states revisited In order to devise an expression that reects the relationship between national carbon footprints and their dependencies, a methodology is adopted similar to that employed by Hammond (2006) and Cranston et al. (2010) to correlate the ecological or environmental footprints (ef; sometimes referred to as eco-footprints) of nations from around the world. It was therefore possible to establish the dependence of the carbon footprint on factors such as economic wealth. The situation in the G-20 nations (see Fig. 1) has been highlighted. The earlier eco-footprint analysis has been updated below, using more recent (2005) data. It is therefore comparable to the carbon footprint equation established here. Cranston et al. (2010) conrmed that per capita national environmental footprints (ef) are strongly dependent upon per capita national income and only weakly on population density. They demonstrated that a large variety of other parameters were insignicant drivers of environmental burdens. But the need to improve the environmental performance of countries is currently focused on the reduction of CO2 emissions, and the consequent mitigation of global warming (Hammond, 2000), in both the developed and developing nations. It is seen as a major component of the transition to a sustainable global economy. In the present study a dataset containing a total of 107 countries were analysed for the year 2005. This did not include all the nation states of the world, due to the limited availability of data (particularly for some of the small, low-income developing countries). However, this still reects a good representative sample of nations from the North and South. [It is also similar in size to the earlier datasets studied by Hammond (2006) 113 countries and Cranston et al. (2010) 109 countries in their analysis of national environmental footprints.] The G-20 group of countries (see Fig. 1) has been highlighted in the graphical presentation of the

where GNI represents economic well-being (Gross National Income measured on a purchasing power parity basis {in international dollars [per capita $]}); PD is the population density [national population numbers per hectare]; EI is the energy intensity (primary energy consumption per unit of GNI [MJ/$]), and CR is the carbon emissions ratio (emissions per unit of energy consumed [gC/J]). The relationship between economic growth and per capita environmental footprint was once again found to be the strongest (Fig. 2) with the exponent p = 2/3. In the current study, the energy intensity (EI) and carbon ratio (CR) values were thought to be strong enough drivers to warrant inclusion in the nal correlation equation. This enabled the earlier powerlaw correlation (Hammond, 2006; Cranston et al., 2010) to be re-tested using more recent data. It also provided a common basis for comparison between the nal correlation equations for both the carbon and environmental footprints. Population density was again found to have only a small inuence on ef, although its inclusion gave some consistency with previous work. The nal equation, obtained following the methodology used by Hammond (2006) and Cranston et al. (2010), was found to be: ef = 0.0020 GNI2/3 PD1/10 EI1/4 CR1/5 (2)

This correlation will vary over time as changes in the energy (or fuel) mix, improvements in energy end-use, and other international developments occur. Relevant scenarios or futures were discussed by Cranston and Hammond (2010a, 2010b), who ultimately adopted the so-called marker scenarios suggested by the Intergovernmental Panel on Climate Change (IPCC; Nakicenovic et al., 2000) as the framework for their projections to 2100. Multi-linear regression was used to validate the above result, and this yielded Eq. (3): ef = 0.0042 GNI0.5986 PD0.1158 EI0.2398 CR0.1859 (3)

The resultant exponents from the multi-linear regression [Eq. (3)] were considered to have too many signicant gures. They are not justied by the accuracy of the physico-economic data and associated scatter (Hammond, 2006). Instead the fractional exponents in Eq. (2) were adopted, in line with the methods of dimensional analysis and powerlaw correlation employed in the engineering sciences (Hammond, 2006; Cranston et al., 2010). The close similarities between these values found in evaluating Eqs. (2) and (3) are displayed in Table 1.

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

93

Fig. 2. The relationship between per capita environmental footprint and national income for the year 2005 [G-20 nation states highlighted].

Table 1 Comparison of correlating powerlaw equation exponents. Parameter GNI PD EI CR Multi-linear regression exponents 0.5986 0.1158 0.2398 0.1859 Exponents (to appropriate signicant gures) 0.6 2/3 0.1 1/10 0.25 1/4 0.2 1/5 Dimensional analysis exponents 2/3 1/10 1/4 1/5

The correlation equation for per capita national environmental footprint (ef) [Eq. (2)] is represented graphically in Fig. 3. Nations above the powerlaw correlation curve are generally proigate in terms of their use of natural capital, whilst those below the curve are relatively frugal (Hammond, 2006). Countries such as Australia, France, China, Germany, Indonesia, Italy, Japan, and the UK (from amongst the G-20 grouping) fall almost exactly on the powerlaw curve. In order to illustrate the use of the correlating equation depicted in Fig. 3, consider a populous, emerging economy of China. Its position, on the powerlaw data correlation, implies that it has a footprint typical (that is, close to the mean value) of nation states with a similar combination of income, population density, energy intensity, and carbon ratio. [However, over time it is likely

that such emerging economies will grow rapidly with greater use of natural resources and consequences for environmental impact (Cranston and Hammond, 2010b)]. The same arguments apply to France, Indonesia, Italy, Germany, Japan, and the UK in regard to their characteristics at the present time. In contrast, Argentina, Canada, Russia, Saudi Arabia, and South Africa (which fall below the correlating equation) are seen to be more sparing in their use of natural capital, whereas the Brazil, Mexico, Turkey, and the USA conversely exhibit relative proigacy. But the result for Brazil is thought to be biased by the fact that the calculation for the carbon footprint in the Living Planet Report 2008 is most likely an underestimate (Reed, 2008). This impacts the results for carbon ratio, as well as the nal ef correlating equation. The carbon emissions estimate

Fig. 3. Per capita national environmental footprint correlating equation [G-20 nation states highlighted].

94

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

Fig. 4. Variation of per capita national carbon footprints with economic wealth across nation states of the world in 2005 [G-20 nation states highlighted].

for Brazil has not signicantly changed between the 2006 and 2008 editions of the National Footprint Accounts, so the source of the discrepancy is likely to be associated with the calculated embodied energy (see Hammond and Jones, 2008) in trade ows [see the description of the trade corrections adopted here in Appendix A]. That is the conclusion drawn, for example, by Reed (2008). Nation states that are placed above and below the curve (Fig. 3) are distributed across the spectrum of socio-economic groups; national consumption patterns of biophysical resources do not appear to be related to wealth or poverty per se. The dominance of fossil fuel consumption in the footprints of (particularly) high-income countries obviously means that these states rely on resources laid down over geological timescales (Hammond, 2006). Such energy technologies will need to be replaced by nuclear or renewable energy sources, coupled with measures aimed at signicantly improving energy efciency, over the longer term as fossil fuels are depleted. 3. Deriving a carbon footprint correlation for nation states A similar analysis to that presented by Cranston et al. (2010) was undertaken for the carbon footprint in light of its growing popularity as an indicator in the public domain. Four determinants were considered, many others having been eliminated in line with the arguments of Cranston et al. (2010). Therefore the carbon footprint was postulated to be a function of economic wealth (GNI), population density (PD), energy intensity (EI), and carbon ratio (CR). [The units are as indicated in connection with Eqs. (1)(3).] This can again be written in terms of a correlating powerlaw equation [like Eq. (1)] as: cf = constant[(GNI)a (PD)b (EI)c (CR)d ] (4)

Economic wealth in terms of Gross National Income [based on purchasing power parity] and the primary energy data, used to determine the energy intensity, were both obtained from the World Banks World Development Indicators (World Bank, 2006). The corresponding population density was based upon the population size of each country in millions from the WWF Living Planet Report 2008 (Hails et al., 2008), whilst the carbon ratio was calculated using CO2 emissions in M tonnes from the International Energy Agency (IEA, 2008). The latter production gures were then corrected to give national consumption levels using import and export data tables from the UN Statistical Division (UNSD, 2009): as discussed

in Appendix A below. The carbon footprint (cf) represented by Eq. (4) has units of global hectares (gha) per capita. The variables in Eq. (4) were all shown to be independent of one another, such that there was no double counting or cross-correlation between parameters. Multi-linear regression was again [as in the case of the ef-relation above {Eq. (2)}] applied to verify the determination of the individual powerlaw exponents. The signicance of the four dependent variables in Eq. (4) for the per capita national carbon footprint of 107 countries in the current (2005) dataset was evaluated using powerlaw correlation approach previously employed by Hammond (2006) and Cranston et al. (2010) for their national environmental footprinting. In order to determine the strongest dependency, each of the dependent variables (i.e., GNI, PD, EI, and CR) were plotted against cf. It was found that the carbon footprint was most dependent on economic wealth (or per capita GNI), where it was shown that the exponent a = 1 in Eq. (4); see Fig. 4. This was to be expected as the carbon footprint is a purely consumption driven indicator. Those nation states with larger incomes and successful economies inevitably emit a larger magnitude of emissions. Some countries are more environmentally conscious than others; see for example Fig. 4, where Germany and Australia (amongst the high-income countries) can be seen to be below the correlation line and therefore more frugal in terms of their use of natural capital. The remaining three parameters in Eq. (4) were then plotted sequentially against a combination (product) of per capita carbon footprint and economic prosperity (per capita GNI) in order to avoid cross-correlation. The carbon footprint and carbon ratio (carbon emitted per joule) of a country are linked, and the former could be lowered by either a decrease in income or a decrease in the energy used per unit of income. Technological innovation is a potential route to improved efciencies and lower energy intensities. This will be one energy saving option as humanity transitions towards a low carbon future. Having established the relationship between the per capita carbon footprint and per capita GNI (see Fig. 4), it was necessary to determine the dependencies of the remaining exponents in Eq. (4) in a comparable manner. The energy intensity and carbon ratio were both found to be important drivers of the per capita carbon footprint (cf); resulting in the exponents c = 1 and d = 3/4 in Eq. (4); see Fig. 5. Population density displayed a very low dependency, with its exponent effectively b 0. It was therefore eliminated from the nal powerlaw correlation. Substituting all the exponents for the dependent variables into the original cf expression, Eq. (4)

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

95

Fig. 5. Carbon footprint relationship to economic wealth, energy intensity and carbon ratio for nations from around the world [G-20 nation states highlighted].

tal impact as a consequence of its production and export trade in coal. The developing nations were further down the plot and relatively more spread out. Of the G-20 nations, those considered to be developing were far more sparing of their natural resources, particularly Indonesia and Brazil. Those more transitional countries were placed in the middle, whilst those in early development stages were lower down according to their economic state. It could be seen that those developing countries in the G-20 group were, on balance, more frugal than the industrialised regions. Research has shown that this may not always be the case as they move towards industrialisation and urbanisation (Cranston and Hammond, 2010b). It is therefore crucial that green technologies and good practice are passed onto these nation states to prevent carbon emissions escalating to intolerable and irreparable levels. Countries such as China and India may not yet have the economic afuence of the developed world on a per capita basis, but their consumption and waste emission patterns (as reected in their per capita carbon footprints), are slightly more proigate than other nations (Fig. 5). Nevertheless, their total carbon footprint is amongst the largest in the world due to their large populations sizes. It is important that industrialised world leaders are willing to encourage the developing countries to strive towards higher environmental standards, and support them by the sharing of knowledge, technologies, and experience.

above, yields the following expression: cf = 2 106 GNI EI CR3/4 (5) 4. A comparison between carbon and environmental footprints The updated environmental footprint expression [Eq. (2)] is compared to the nal carbon footprint counterpart [Eq. (5)] in Fig. 6. Both equations yield similar nal results: economic wealth (per capita GNI) dominates the footprint in each case. However, the powerlaw exponents of each equation vary between the two cases. The exponent for afuence is larger for the carbon footprint, which illustrates that this energy-based footprint is even more dependent than the environmental footprint upon economic welfare. Ewing et al. (2008) noted that the carbon (or energy) footprint represents a major segment of the total environmental footprint. The dependency of the carbon footprint on both energy intensity and carbon ratio appears stronger than with the environmental footprint.

This suggests that the per capita carbon footprint (in gha) is heavily dependent upon economic wealth (per capita GNI) and energy intensity, and is also quite strongly related to carbon ratio; see Fig. 5. The G-20 nations are again highlighted on this gure, and it can be seen which countries are frugal or proigate in terms of their per capita carbon emissions. Those above the correlating line produce an excessive amount of CO2 emissions, whilst those below it are more sparing in terms of their potential global warming impact. The developed or industrialised countries were clustered towards the top right of Fig. 5, because they are high-income (and per capita GNI) nations. They are mainly clustered close to the powerlaw correlation line. Australia is the only developed nation that falls below the correlation line. It is relatively prudent in domestic fossil fuel consumption terms, but it obviously has a high environmen-

Fig. 6. Comparison of per capita carbon and environmental footprints, and their powerlaw correlating equations [G-20 nation states highlighted].

96

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

Fig. 7. Carbon overshoot and undershoot the imbalance between the per capita carbon footprint of selected nation states and globally available biocapacity.

The carbonenvironmental footprint comparison in Fig. 6 displays some distinguishing features. For both of these ecological indicators, the industrialised world is grouped about the upper right-hand portion of the correlating line, representing a similar high-level consumption of resources and efciencies. The emerging or transitional economies of the G-20 group tend to be more scattered. Brazil, for example, is seen to exhibit a high overall environmental impact, but to have a relative low carbon footprint due to a low dependency on fossil fuels (or relative abundance of renewable energy sources, including biofuels). It is critically important that nation states aim to reduce their resource use and pollutant emissions in line with their share of the worlds biocapacity (Hammond, 2006). Through improving energy use and decreasing carbon emissions many more countries could nd themselves below their so-called overshoot ratio. Fig. 7 illustrates the disparity between a selection of countries in terms of their carbon footprint and available national biocapacity. The Earthshare, or total biocapacity available per person, for 2005 was 2.1 gha. Given that the average per capita carbon footprint represents approximately 50% of this associated value for ef [based on the Living Planet Report 2008 (Hails et al., 2008)], this leaves roughly 0.8 gha that an individual may use after the sequestration of carbon. Many populations can be seen to use well above this level. The world cannot continually support this level of ecological debt (Cranston et al., 2010) as its natural resource overdraft rises. Such demands upon the ecosystem are unsustainable as the planet can only regenerate its resources at a certain rate. It is therefore important to implement strategies aimed at reducing humanitys impact on the biosphere. It is generally recognised that the industrialised world needs to take the lead in this endeavour, in order to permit the developing nations scope to improve their economic well-being. Cranston and Hammond (2010b), and others, have devised future scenarios and projections into the 21st Century that illustrate alternative pathways towards sustainability. 5. Carbon footprints versus carbon weights The concept of the earlier carbon footprint is rooted within the framework used to determine the eco-footprint. However, Hammond (2007) noted that a footprint would normally be measured in spatial units [such as global hectares (gha)], but that the carbon footprint is typically presented in kilograms per person or

activity. He therefore argued that it should perhaps be termed a carbon weight (CW ) or something similar. The term carbon weight was independently adopted by Ghazi and Lewis (2007) in their programme to aid individuals and families in reducing carbon emissions in the home, on the road, and at play. Wiedmann and Minx (2008) reviewed various suggestions [including that of Hammond (2007)], and then proposed a denition for the carbon footprint as including the total amount of CO2 emissions that is directly and indirectly caused by an activity. Unfortunately, no denition has been formally adopted in a standard with the agreement of the communities involved. Nevertheless, the differing ways in which the expression carbon footprint is employed is clearly misleading. Even the WWFs Living Planet Report 2008 (Hails et al., 2008) has acknowledged that the term carbon footprint has been misused by many, using tonnes of carbon or tonnes of carbon per Euro, rather than the demand on bioproductive area. Indeed, many organisations have adopted the use of the term carbon footprint when assessing the carbon dioxide emissions released during associated processes or activities, although these are again measured in tonnes of carbon dioxide [for example, BP (2009) and Directgov (2009)]. This is highly confusing and detracts from the important, fundamental meaning of the footprint concept. Kitzes and Wackernagel (2009) suggested that the carbon footprint measured in global hectares adds value to carbon emissions in two ways: the rst being that it puts emissions into a context that is far more meaningful to the general public in spatial units that compares our consumption to the available resources supplied by Nature. The second benet is that using the correct terminology also enables a comparison to other demands on productive land. This is far more useful as it allows assessment of other energy-based solutions, such as the growth of biofuels, with respect to the effects this change will have upon the biosphere and its biocapacity. The carbon weight or carbon consumption (CW ) for 107 nations was calculated and compared with their respective carbon footprint (cf). The carbon weight values were calculated from carbon dioxide data per country, based upon consumption and waste assimilation for its population. It excluded the emissions released by producing goods that are exported and used in other countries (see Appendix A). The adjustment procedure from carbon production to carbon weight (or consumption) is illustrated by Eq. (6). This carbon balance equation ensures compatibility between the carbon footprint detailed in the WWF Living Planet Report 2008 (Hails et al., 2008), the carbon weight discussed here, and also the carbon ratio

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

97

(CR) utilised in Section 3. CW = CP + CI CE + CB (6)

where CW is the carbon weight, CP is the carbon production, CI is the carbon from imports, CE is the carbon from exports, and CB is the carbon from bunkers. International Marine and International Aviation Bunkers contain emissions from fuels burned by sea-fairing ships under all ags and aircraft that are engaged in international transport. This amount was added to the consumption value to take account of each nations contribution to international transport; following the practice of Kitzes et al. (2008). They found that a percentage between 3% and 4% of total carbon dioxide emissions is commonly observed, and in the present case a value of 3.27% was adopted from the 2008 Global Footprint Accounts (Global Footprint Network, 2008). The trade corrected data was collected for 627 commodities for the imports and exports from each country (see also Appendix A) was obtained via the UN Statistical Division UN Comtrade (UNSD, 2009).3 These traded commodities were established by their Statistics of International Trade Classication (SITC) revision 1 codes. This classication system offers the longest historical time series presently available, and is used by the leading footprint companies, such as the Global Footprint Network and Best Foot Forward (Simmons et al., 2006). UN Comtrade is the largest repository of international trade data, and has over 140 reporter countries providing annual international trade statistics detailing commodities and partner countries (Vempaty, 2007). All commodities for imports and exports were summed appropriately. The trade data was converted from mass in kg to M tonnes of CO2 emissions, using specic world average embodied energy data provided by the Best Foot Forward report for the WWF One Planet Business Programme (Simmons et al., 2006) and conversion factors for world average carbon intensity of electricity and heat production were applied from the Global Footprint Network (2008) spreadsheets. A nal conversion was made from CO2 to carbon using the molecular weight balance, i.e., 12/(12 + 32) = 0.27. The relationship between the carbon footprint (cf) and carbon weight (CW ) measures were compared, having been veried with respect to the values given in the WWF Living Planet Report 2008 (Hails et al., 2008). The conversion factor between these two parameters, as dened in this study, was found to be approximately one: cf 1 CW where the units for cf are per capita gha, whilst that for CW are kg per capita. But it should be noted that not all companies use trade corrected data (see Appendix A) and may consider only the direct carbon emissions. Others may adopt carbon equivalent values, or try to include a host of GHGs within the indicator (Wiedmann and Minx, 2008). The correlating powerlaw equation for the carbon footprint (cf) that was derived for the present purposes can be reformulated in terms of the carbon weight (CW ): CW = 1 (2 106 ) GNI EI CR3/4 (7)

Fig. 8. Comparison of the carbon footprint and carbon weight for nations around the world [G-20 nation states highlighted].

weight. Those more intermediary economies are in the mid-range, whilst the less developed countries, such as Indonesia and Brazil, are towards the bottom left-hand end of the gure. 6. Concluding remarks It has been noted that carbon footprints represent the amount of carbon emissions associated with a given activity or community, and are closely related to ecological or environmental footprints. But, unlike the latter, they are generally presented in terms of units of mass or weight (kilograms per functional unit), rather than in spatial units (such as global hectares). These carbon footprints have become the currency of debate in a climate-constrained world. They are increasingly popular ecological indicators, adopted by individuals, businesses, governments, and the media alike. It has been shown that the carbon footprint and carbon weight estimated here are very similar in magnitude (despite the difference in units), with a conversion factor of approximately one (unity). Determination of the exponents in the powerlaw correlation (based on dimensional analysis techniques from the engineering sciences) has shown that the per capita national carbon footprint is strongly dependent upon economic wealth (per capita GNI), as well as being inuenced signicantly by the carbon ratio and energy intensity. The resulting powerlaw expression from the present analysis provides a means of assessing the relative contribution of nation states in terms of whether they are proigate or frugal in the emission of carbon dioxide [the main greenhouse gas (GHG) viewed as contributing to global warming]. The correlated data was illustrated graphically with the situation for the G-20 nations highlighted. That provides an indication of the carbon contributions being made by countries in the industrialised North of the planet and those of the developing, populous South. A recent companion piece by the authors (Cranston and Hammond, 2010a) provided estimates of relative contribution of population and economic growth to future carbon emissions from the Northern and Southern regions of our bipolar world out to 2100. The sustainability (or IPAT) equation was employed to assess these contributions, along with historic data (for backcasting) and future scenarios from the IPCC Special Report on Emissions Scenarios (Nakicenovic et al., 2000). This facilitated a decomposition of the impact of likely changes in emissions caused by varying regional

Both the cf and CW values are compared in Fig. 8. This gure shows how closely related these functions are. The G-20 nation states are again highlighted, and it can be seen that the prosperous countries of the North are located at the top right-hand end of the plot in terms of both the per capita carbon footprint and carbon

3 The data from UN Comtrade was unavailable CDIAC values (Marland et al., 2008) were applied in accordance with Global Footprint Network guidelines (Kitzes et al., 2008).

98

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

demographics and per capita afuence over time. It was shown (Cranston and Hammond, 2010a) that population and economic wealth both contribute to the rise in carbon concentrations in the atmosphere. For the industrialised world, it was found that economic wealth was the most signicant driver of CO2 emissions from both historic trends and future scenarios. In the South, regional population and economic growth during the 21st Century are each likely to play a signicant role in affecting the levels of current or year-on-year carbon emissions. In order to achieve global sustainability a serious commitment to GHG emissions reduction is required, and a greater dedication to environmental protection in both the industrialised North and the populous South. However, there is an issue of intergenerational equity in relation to the contribution of the developed world. CO2 has a residence time in the atmosphere of around 100 years (Hammond, 2000). The build-up of atmospheric concentrations is therefore largely a consequence of emissions released by countries of the North since the start of their so-called industrial revolution around 1850 (Cranston and Hammond, 2010a). Thus, the industrialised nations should take a lead in mitigating the global release of GHGs, because they are principally responsible for their lifetime concentrations. Developing countries clearly desire to grow and provide a better standard of living and enhanced well-being for their populations on equity or equality grounds. They typically nd suitable climate change mitigation measures prohibitively expensive. In addition, many rapidly emerging economies (like China and India) are currently more dependent on dirty fuels, like wood and coal, and this results in their relatively high carbon ratios (Hammond, 2006; Cranston et al., 2010). Consequently they could benet from the assistance of industrial countries to promote their economic growth [which will, in time, induce a demographic transition (Nakicenovic et al., 2000)], as well as leading to an improvement in the resource efciency of their energy systems. Environmental sustainability could thereby be aided via the transfer of best practice (or leapfrog) energy technologies from the richer (Northern) to poorer (Southern) regions of our bipolar world. These issues are at the heart of international negotiations aimed at providing a successor to the Kyoto Protocol, originally established to stabilize GHG concentrations at a level that would prevent dangerous climate change. A new agreement will hopefully come into being post-2012 that will ensure a sustainable future for the people and wildlife on Spaceship Earth (Hammond, 2000). This will ultimately be in the interests of all its citizens and species.

inuential in focusing his attention on low carbon development into the 21st Century, albeit principally in the UK context. Appendix A. Trade balance corrections to the footprint calculations The embodied energy of each commodity was taken from the Global Footprint Networks 2008 edition of National Accounts (Global Footprint Network, 2008). Embodied energy according to this source is the energy required throughout the complete life cycle of a product, from manufacture to transportation to use and disposal. In contrast, the Inventory of Carbon and Energy (ICE) produced by Hammond and Jones (2008) contains a wide range of embodied energy data for different materials on a cradleto-(factory) gate basis. The latter includes upstream energy requirements, including those for resource extraction, fabrication and assembly. To convert this data to the denition employed by the Global Footprint Network (2008) would therefore necessitate the addition of dowmstream energy requirements, such as product manufacture, transportation to use and disposal. Footprint studies often use embodied energy data when tracking the trade of goods (Global Footprint Network, 2009). The net weight in kg of each commodity was established for the present purposes via the UN Comtrade database. This could thus be converted to an import energy or export energy via the embodied energy values obtained from: Import Energy [GJ/yr] = EmbEn [GJ/t] NetWeight [kg/yr] (A1)

Import or Export Energy was converted to a carbon weight using the world electricity and heat carbon intensity from the Global Footprint Network (2008), and the ratio of Gigajoules (GJ) of energy to Terawatt Hours (TWh): Import CO2 [Mt CO2 /yr] = Import Energy [GJ/yr] Gigajoule to Terawatthour conversion [TWh/GJ] World Elec&Heat CorbonIntensity [Mt CO2 /TWh] (A2)

Acknowledgements This is a revised and extended version of a paper originally presented by the rst author (GRC) at the Footprint Forum 2010, Colle Val dElsa, Sienna, Italy over 910 June 2010. The University of Baths Strategic Investment Fund has supported her contribution to this work. The second authors research (GPH) on energy, environment, and sustainable development has recently been supported by research awards and contracts in the United Kingdom (UK) from the Biotechnology and Biological Sciences Research Council (BBSRC) [as part of the BBSRC Sustainable Bioenergy Centre (BSBEC) LACE Programme], the Carbon Trust, e.on UK (the electricity generator), the Engineering and Physical Sciences Research Council (EPSRC), the Environment Agency, the Great Western Research (GWR) Alliance, and the UK Energy Research Centre (UKERC). He is presently the coleader of a consortium of eight university partners (jointly with Prof. Peter Pearson, an energy economist, now Director of the Low Carbon Research Institute in Wales) funded via the e.on/EPSRC Strategic Partnership to study the role of electricity within the context of Transition Pathways to a Low Carbon Economy [under Grant EP/F022832/1]. The work of this consortium has been particularly

Finally the values of Import and Export CO2 were converted to carbon on a molecular weight basis. CO2 production from the burning of fossil fuels was taken from the International Energy Agency data (IEA, 2008). A Sectoral Approach was chosen to ensure comparability with the WWFs Living Planet Report 2008 (Hails et al., 2008) and accepted footprinting methodologies (Simmons et al., 2006). This approach requires the calculation of CO2 emissions from fuel combustion based upon the main fuel combustion activities of a specic country; broken down in terms of sectors. This ensures more detailed calculations than would be used for the so-called Reference Approach, where the total CO2 emissions are established purely from fuels supplied to a given nation. Once the carbon weight had been calculated for a particular country, it was then converted to a carbon footprint. This was achieved using Eq. (A3), which is based upon the Global Footprint Networks National Footprint Accounts 2008 (Global Footprint Network, 2008): cf = =
(CW )[MtCO2 /yr](1Ocean Sequestration) [%]ForestEQF [gha/ha] Carbon Sequestration [tCO2 /hayr]

(CW ) (1 0.25) 1.33 0.97

(A3)

These values were veried by assessing the differences between the resulting carbon footprint and that given in the WWF Living Planet Report 2008 (Hails et al., 2008). The calculated results had

G.R. Cranston, G.P. Hammond / Ecological Indicators 16 (2012) 9199

99

an average percentage difference of only 4.8% compared to those reported by the WWF. Carbon sequestration, or the carbon uptake rate, is the annual rate of carbon uptake per hectare of world average bioproductive forest land. It was derived from data on the net annual growth of forests. A forest equivalence factor is then used to convert the forest land to the common unit in global hectares. Carbon uptake rates and forest equivalence values were both taken from the National Footprint Accounts 2008. These gures were 0.97 tC ha1 yr1 and 1.33 gha ha1 respectively (Global Footprint Network, 2008). References
BP, 2009. BP Energy Calculator [online]. Available from: http://www.bp.com/ iframe.do?categoryId=9023118&contentId=7045317 [accessed March 2009]. Cranston, G.R., Hammond, G.P., 2010a. Egalit, fraternit, sustainabilit: evaluating the signicance of regional afuence and population growth on carbon emissions. International Journal of Global Warming 2 (3), 189210. Cranston, G.R., Hammond, G.P., 2010b. North and south: regional footprints on the transition pathway towards a low carbon, global economy. Applied Energy 87 (9), 29452951. Cranston, G.R., Hammond, G.P., Johnson, R.C., 2010. Ecological debt: exploring the factors that effect national footprints. Journal of Environmental Policy & Planning 12 (2), 121140. Directgov, 2009. Act on CO2 Calculator [online]. Available from: http://carboncalculator.direct.gov.uk/index.html [accessed March 2009]. Ewing, B., Reed, A., Rizk, S., Galli, A., Wackernagel, M., Kitzes, J., 2008. Calculation Methodology for the National Footprint Accounts, 2008 Edition. Global Footprint Network, Oakland, USA. Ghazi, P., Lewis, R., 2007. The Low Carbon Diet. Short Books, London, UK. Global Footprint Network, 2009. Glossary [online]. Available from: http://www.footprintnetwork.org/en/index.php/GFN/page/glossary/ [accessed May 2010]. Global Footprint Network, 2008. National Footprint Accounts 2008 Edition. www.footprintnetwork.org. Hails, C., Humphrey, S., Loh, J., Goldnger, S. (Eds.), 2008. Living Planet Report 2008. WWF International, Gland, Switzerland.

Hammond, G.P., 2000. Energy, environment and sustainable development: a UK perspective. Transactions of the IChemE Part B: Process Safety and Environmental Protection 78 (4), 304323. Hammond, G.P., 2006. People, planet and prosperity: the determinants of humanitys environmental footprint. Natural Resources Forum 30 (1), 27 36. Hammond, G., 2007. Correspondence: time to give due weight to the carbon footprint issue. Nature 445 (7125), 256. Hammond, G.P., Jones, C.I., 2008. Embodied energy and carbon in construction materials. Proceedings of the Institution of Civil Engineers: Energy 161 (2), 87 98. IEA, 2008. CO2 Emissions from Fuel Combustion 2008 Edition. OECD/IEA, Paris, France. Kitzes, J., Galli, A., Rizk, S., Reed, A., Wackernagel, M., 2008. Guidebook to the National Footprint Accounts: 2008 Edition. Global Footprint Network, Oakland, USA. Kitzes, J., Wackernagel, M., 2009. Answers to common questions in ecological footprint accounting. Ecological Indicators 9 (4), 812817. Marland, G., Boden, T.A., Andres, R.J., 2008. Global, regional, and national fossil-fuel CO2 emissions. In: Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, TN, USA. Nakicenovic, N., Alcamo, J., Davis, G., de Vries, B., Fenham, J., Gafn, S., Gregory, K., Grubler, A., Jung, T., Kram, T., 2000. Special Report on Emissions Scenarios: A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. Reed, A. 2008. Question for Global Footprint Network Regarding Brazils Carbon Footprint. Personal Communication. Simmons, C., Gonzalez, I., Lewis, K., 2006. Methodology for determining global sectoral material consumption, carbon dioxide emissions and Ecological Footprints. In: A Report for the WWF One Planet Business Programme. Best Foot Forward, Oxford, UK. UNSD, 2009. UN Commodity Trade Statistics Database [online]. Available from: http://comtrade.un.org/ [accessed February 2009]. Vempaty, A., 2007. First Time User [online]. UN Commodity Trade Statistics Database. Available from: http://comtrade.un.org/kb/article.aspx?id=10062 [accessed February 2009]. Wiedmann, T., Minx, J., 2008. A denition of carbon footprint. In: Pertsova, C.C. (Ed.), Ecological Economics Research Trends. Nova Science Publishers, Hauppauge, NY, USA, pp. 111 (Ch. 1). World Bank, 2006. World Development Indicators 2006. World Bank, Washington, DC, USA.