Anthropogenic sulfur dioxide emissions: 1850–2005

Atmos Chem Phys 11 1101ndash1116 2011wwwatmos-chem-physnet1111012011doi105194acp-11-1101-2011copy Author(s) 2011 CC Attribution 30 License

AtmosphericChemistry

and Physics

Anthropogenic sulfur dioxide emissions 1850ndash2005

S J Smith1 J van Aardenne2 Z Klimont 3 R J Andres4 A Volke1 and S Delgado Arias1

1Joint Global Change Research Institute Pacific Northwest National Laboratory 5825 University Research Court Suite 3500College Park MD 20740 USA2European Commission Joint Research Centre European Commission Via E Fermi 2749 TP 290 21027 Ispra (VA) Italy3International Institute for Applied Systems Analysis Schlossplatz 1 2361 Laxenburg Austria4Environmental Sciences Division Oak Ridge National Laboratory Oak Ridge TN 37831-6335 USA current address European Environment Agency Kongens Nytorv 6 1050 Copenhagen K Denmark

Received 28 May 2010 ndash Published in Atmos Chem Phys Discuss 30 June 2010Revised 28 January 2011 ndash Accepted 1 February ndash Published 9 February 2011

Abstract Sulfur aerosols impact human health ecosystemsagriculture and global and regional climate A new annualestimate of anthropogenic global and regional sulfur dioxideemissions has been constructed spanning the period 1850ndash2005 using a bottom-up mass balance method calibrated tocountry-level inventory data Global emissions peaked inthe early 1970s and decreased until 2000 with an increasein recent years due to increased emissions in China inter-national shipping and developing countries in general Anuncertainty analysis was conducted including both randomand systemic uncertainties The overall global uncertaintyin sulfur dioxide emissions is relatively small but regionaluncertainties ranged up to 30 The largest contributors touncertainty at present are emissions from China and interna-tional shipping Emissions were distributed on a 05 grid bysector for use in coordinated climate model experiments

1 Introduction

Anthropogenic emissions have resulted in greatly increasedsulfur deposition and atmospheric sulfate loadings near mostindustrialized areas Sulfuric acid deposition can be detri-mental to ecosystems harming aquatic animals and plantsand damaging to a wide range of terrestrial plant life Sul-fur dioxide forms sulfate aerosols that have a significant ef-fect on global and regional climate Sulfate aerosols re-flect sunlight into space and also act as condensation nu-clei which tend to make clouds more reflective and changetheir lifetimes causing a net cooling The radiative forcing

Correspondence toS J Smith(ssmithpnlgov)

change wrought by sulfate aerosols may be second only tothat caused by carbon dioxide albeit in the opposite direc-tion (Forster et al 2007)

Sulfur is ubiquitous in the biosphere and often occursin relatively high concentrations in fossil fuels with coaland crude oil deposits commonly containing 1ndash2 sul-fur by weight The widespread combustion of fossil fuelshas therefore greatly increased sulfur emissions into theatmosphere with the anthropogenic component now sub-stantially greater than natural emissions on a global basis(Smith et al 2001)

Historical reconstructions of sulfur dioxide emissions arenecessary to access the past influence of sulfur dioxide onthe earth system and as base-year information for future pro-jections This paper presents a new estimate of global andcountry-level sulfur dioxide anthropogenic emissions overthe 1850ndash2005 period This work represents a substantialupdate of previous work (Smith et al 2001 Smith et al2004) with newer data and improved methodologies and wasthe basis for the sulfur emissions in Lamarque et al (2010)The emissions reconstruction presented here accounts for re-gional differences in the pace and extent of emission con-trol programs has annual resolution includes all anthro-pogenic sources and provides global coverage Fuel-basedand activity-based (Eyring et al 2010) estimates of shippingemissions were reconciled for recent decades and then ex-trapolated to 1850 A global mass balance for sulfur in crudeoil was calculated as an independent estimate of petroleumemissions Finally a regional and global uncertainty analy-sis was conducted

Published by Copernicus Publications on behalf of the European Geosciences Union

1102 S J Smith et al Anthropogenic sulfur dioxide emissions 1850ndash2005

2 Methodology

Sulfur emissions from combustion and metal smelting can inprinciple be estimated using a bottom-up mass balance ap-proach where emissions are equal to the sulfur content of thefuel (or ore) minus the amount of sulfur removed or retainedin bottom ash or in products Data limitations howevermake the bottom-up approach uncertain since sulfur contentsvary and information on sulfur removals is not always re-ported Countries with air pollution policies in place gener-ally require detailed reporting of emissions including directmeasurement of emissions for many large sources Thesedata are likely to be more accurate than bottom-up estimatestherefore we constrain our calculations to match country-level inventories where these data are available and judgedto be reliable This method produces an emissions estimatethat is consistent with the available country inventory datacontains complete coverage of all relevant emissions sources(assuming the country inventory data are complete) and isconsistent across years

Emissions were estimated annually by country for the fol-lowing sources coal combustion petroleum combustionnatural gas processing and combustion petroleum process-ing biomass combustion shipping bunker fuels metal smelt-ing pulp and paper processing other industrial processesand agricultural waste burning (AWB) The approach is sum-marized in Fig 1 and further detailed in the Supplement(S2 S3) The first step was to develop a detailed inventoryestimate by sector and country for three key years 19902000 and 2005 These years were chosen due to data avail-ability and because these years span a time period of sig-nificant change in emissions controls Initial estimates forother years were constructed by interpolating emissions fac-tors with final estimates by source and country found af-ter calibrating to country-level emissions inventories wherethose were available (S3) This methodology accounts forchanging patterns of fuel consumption and emission controlsin a context of limited detail for earlier years

Emissions by end-use sector for decadal years were esti-mated for a set of standard reporting sectors (energy indus-try transportation domestic AWB) before downscaling to a05 spatial grid Descriptions of each element in this calcu-lation are given below

21 Fossil fuel combustion

Emissions from coal and petroleum combustion were es-timated starting with the regional emissions factors fromSmith et al (2001) A composite fossil fuel consumptiontime series was constructed (see Supplement S2) using datafrom IEAOECD (2006) UN energy statistics (1996) andAndres et al (1999) Country-level emissions estimates forEurope North America Japan Australia and New Zealandwere compiled from UNFCCC (2009) for 1990ndash2005 En-vironment Canada (2008) the EEA (2002) Fujita (1993)

(1)13Emissions by

fuel and sector13

1990 2000 and 200513

(3)13Default

emissions by country and

fuel13

1850-200513

Fuel consumption

and other drivers13

1850-200513

(2)13Estimates for

smelting (mass balance) and international

shipping13emissions13

1850-200513

Emissions inventory estimates13

where available13

(4)13Final emissions

by fuel amp process13

1850-200513

Sectoral split percentages13

1850-200513

(5)13Emissions by

fuel and sector13

1850 1860 hellip 1990 2000

200513

Fig 1 Calculation summary The key steps in the calculation are(1) development of an inventory by sector and fuel for three keyyears(2) development of detailed estimates for smelting and inter-national shipping(3) calculation of a default set of emissions byinterpolating emissions factors from the key years(4) calculationof final annual emissions values by fuel that match inventory valuesand(5) estimate sectoral emissions

Mylona (1996) Gschwandtner et al (1986) UK NationalAtmospheric Emissions Inventory (2009) USEPA (1996)and Vestreng et al (2007) as detailed in the Supplement(S4 S5) Fossil-fuel emissions factors were scaled suchthat total emissions matched inventory values for those yearswhere inventory values are available While total emissionsfor these countries are therefore constrained by inventoryvalues emissions by fuel are not as well constrained by theavailable data

The ratio of default to inventory emissions in 1975 rangedfrom 03 to 2 for countries in Europe for example givingan indication of the differing rates at which emissions factorschange in different countries These changes can be due todifferent fuel sulfur contents and emission control policiesA similar range was found when comparing 1975 to 1960values (see Supplement S3)

The emission estimates for other Asian countries devel-oped here were compared to a number of regional estimates(Ohara et al 2007 Streets et al 2003 Zhang et al 2009NIER 2008 Klimont et al 2009) and the initial estimateshere were adjusted where necessary to better match thesestudies (see Supplement S4) These estimates differ fromeach other in some countries and sectors highlighting a needfor improved inventories in this rapidly changing regionSpecific assumptions for China and the Former Soviet Unionare described in the Supplement (S6 S7) including the op-eration of SO2 scrubbers in China (Xu et al 2009)

Petroleum emissions are particularly difficult to estimatein the absence of detailed country-level inventory data For

Atmos Chem Phys 11 1101ndash1116 2011 wwwatmos-chem-physnet1111012011

S J Smith et al Anthropogenic sulfur dioxide emissions 1850ndash2005 1103

regions without detailed emission inventory data default val-ues for sulfur contents were used Emissions factors forpetroleum were adjusted downward in recent years to ac-count for increased stringency of sulfur content standards asreported in UNEP (2008)

Petroleum mass-balance

As an overall check on the petroleum assumptions used aglobal petroleum mass balance was constructed by calculat-ing the total amount of sulfur in crude oil and subtractingthe amount of sulfur removed in refineries (see supplemen-tal information S8) The mass balance estimate for globalsulfur emissions from petroleum as compared to the inven-tory estimate is shown in Fig 2 for the period 1950ndash2005where petroleum emissions are a substantial fraction of theglobal total We estimate that in the early portion of the21st century the world passed a threshold where slightlyover half of the sulfur contained in crude oil is removed atthe refinery Thus global emissions from petroleum havebeen relatively flat since 1985 despite increases in petroleumconsumption Mass balance estimates using two assump-tions for the amount of sulfur that is retained in non-energyuse of petroleum products and non-combusted productssuch as asphalt are shown in order to reflect uncertainty inthis parameter

The resulting estimate of global petroleum sulfur is con-sistent with the global inventory value for petroleum emis-sions for most years The mass-balance estimate indicatesthat there may be an underestimate on the order of 5000 GgSO2 in the inventory data during the late 1970s The 10 re-tention mass-balance case results in a lower global estimatethan the inventory data past 1990 The 5 retention casematches well with the inventory during this time perhaps in-dicating the impact of increased refinery efficiency Whilethis comparison validates the overall approach significanterrors could still be present at the country level The globalmass balance estimate is also subject to significant uncertain-ties particularly in regard to the estimates of the amount ofsulfur removed at refineries which is not always reportedseparately from sulfur removed from natural gas (S8)

As noted in Smith et al (2004) the aggregate emissionsfactor from petroleum products in the US as implied by cur-rent estimates (Gschwandtner et al 1986 USEPA 1996) be-fore about 1980 appeared to be lower than those for otherworld regions In order to better estimate petroleum-relatedemissions in the United States a mass balance estimate forthe United States was constructed accounting for importsand exports of crude oil by region and imports and exportsof petroleum products (EIA 2009) This mass balance in-dicated that sulfur emissions from petroleum were substan-tially underestimated prior to 1980 by 30ndash50 (see Supple-ment S8) The increased emissions are in part due to im-ports of residual fuel and a decrease prior to 1980 in reported

013

1000013

2000013

3000013

4000013

5000013

6000013

195013 196013 197013 198013 199013 200013

Sulfu

r as S

O2 (

Gg)13

Year13

Global Sulfur Emissions From Petroleum13

S Emissions Petr13Pet Mass Bal-10ret13Pet Mass Bal-5ret13S Removals (as SO2)13

Fig 2 Global petroleum emissions from the present inventory com-pared to estimates from a global crude oil mass balance (see text) us-ing two different assumptions for the fraction of total sulfur retainedin non-combusted products The inventory data include terrestrialand international shipping combustion and process emissions fromrefineries Also shown is the estimated global amount of sulfur re-moved at refineries (as Gg SO2)

bunker fuel use which is in effect an export of sulfur fromthe continent Emissions from petroleum prior to 1980 wereincreased to be within 20 of the total sulfur as indicated inthe mass balance estimate allowing conservatively for re-tention of sulfur in non-combustion uses This increased USemissions from petroleum combustion by around 2000 GgSO2 in 1972 for example

22 International Shipping

Shipping emissions have been estimated using differentmethodologies some focusing on reported fuel consumptionwhile others use bottom-up estimates using shipping statis-tics (Corbett et al 1999 Corbett and Kohler 2003 2004Endresen et al 2007 Eyring et al 2010) The estimate hereuses both reported data and bottom-up estimates of shippingfuel consumption A composite global estimate of shippingfuel consumption by fuel type (eg coal residual distillateand other) was constructed by combining data from Eyringet al (2010) EIA (2008) Endresen et al (2007) Fletcher(1997) and UN (1996) with additional assumptions as de-scribed in the Supplement (S9) We find that the fraction ofresidual fuel used in shipping has fallen steadily over timefrom an estimated value of 78 in 1971 to 59 in 2005 Thishas a substantial impact on emissions as the sulfur content ofresidual oil can be much higher than that of distillate fuels

These are global values for fuel consumption whichwould ideally need to be reconciled with regional energydata As noted by Eyring et al (2010) and Endresen etal (2007) the IEA energy data used for terrestrial coal andoil consumption here substantially underreport shipping fuelconsumption Comparison of the country and regional level

wwwatmos-chem-physnet1111012011 Atmos Chem Phys 11 1101ndash1116 2011


time series developed here with IEA data finds that the dif-ference is often within the IEA ldquostatistical errorrdquo categoryThere is generally no other consumption category in the IEAdata that is large enough to include the difference betweenour regional fuel consumption estimate and the IEA reportedbunker fuel use We presume therefore that the difference isunreported consumption and no adjustment to the IEA con-sumption data has been made

In order to estimate sulfur dioxide emissions we cali-brate to the year 2000 value from Eyring et al (2010) of11 080 Gg SO2year which is the mean of estimates fromCorbett and Kohler (2003) Eyring et al (2005) and Endresenet al (2007) Using the global division between fuels derivedabove we match this value using the following sulfur con-tents residual 29 and distillateother 13 The averagevalue for sulfur in bunker coal is assumed to be 11 Forsimplicity these values are kept constant over time Thesevalues and their time trends are not known with precision

The most authoritative data on marine fuel sulfur contentsis from the IMO (2007) however the uncertainty in thesevalues difficult to estimate (see Supplement S9) While theestimate here is similar to other estimates in the literature(Corbett and Kohler 2003 Eyring et al 2005 2010 En-dresen et al 2007) as presented in the Supplement theseemissions are particularly uncertain The estimate here fallswithin the 2001 uncertainty of 8400ndash13 100 Gg as estimatedby Corbett and Kohler (2003 2004)

The fossil fuel consumption data from IEA used to es-timate combustion emissions (Sect 21) were processed toexclude fuels reported as international bunkers but includedfuels used for domestic shipping and fishing using sectoraldefinitions as discussed further in Sect 25 In parallel withthis assumption domestic shipping and fishing emissions inthe inventory data were included in the surface transportationsector (see Supplement S2 S9) Emissions due to domes-tic shipping and fishing emissions from inventory data weresubtracted from the total shipping estimate in the tables andfigures reported below to avoid double counting

23 Metal amelting

Smelting emissions were estimated using a mass balance ap-proach where emissions are equal to the gross sulfur con-tent of ore minus reported smelter sulfuric acid productionwith estimates adjusted to match inventory data where avail-able Smelter production of copper zinc lead and nickelwas tabulated by country primarily using USGS mineralyearbooks and predecessor publications for 1930ndash2005 anda variety of supplemental sources particularly for earlieryears Some mineral sources and production technologiesemit minimal sulfur and these were accounted for wherethey could be identified Further details are available in theSupplement (S10)

A number of data biases can affect the estimate of sul-fur emissions from metal smelting To evaluate this possibil-ity we can compare emissions from inventories to emissionsestimated purely using the mass balance approach For theperiod 1990ndash2005 emissions by sector were available fora number of countries When compared to emissions fromthe mass balance approach the inventory values for fifteencountries in 1990 appear to be significantly lower than thoseestimated by the mass balance approach although the oppo-site was the case in the United States This indicates that ei-ther the sulfur content of ore was overestimated or that theamount of sulfur removal was underestimated The lattermay be the more likely possibility given that the sulfur re-covery tabulations from USGS are not necessarily completesince data will be more readily available for commoditiessuch as ores which are internationally traded than for sulfu-ric acid which might be used locally While these changesimpact estimates of emissions from smelting this has onlya small impact on total emissions for recent years since to-tal emissions for most of these countries are constrained byinventory values

These comparisons indicate that emissions from areaswithout inventory data could be overestimated if sulfur re-moval data are underreported In contrast there are severalregions with inventory data where the sulfur content of oreappears to be larger than default values which could also bethe case in regions without inventory data potentially lead-ing to an underestimate These comparisons emphasize theimportance of site-specific information in order to accuratelyestimate metal smelting emissions

24 Other emissions

Natural gas deposits often contain significant amounts of sul-fur compounds particularly hydrogen sulfide that are eitherflared thus producing sulfur dioxide or removed and con-verted to a salable product Natural gas production emissionswere estimated over time for the United States the FormerSoviet Union and other regions without inventory data (seeSupplement S11)

While natural gas distributed for general use has mini-mal sulfur natural gas containing larger amounts of sulfurknown as sour gas can be combusted for industrial applica-tions resulting in sulfur emissions The US EPA inventorycontains an estimate for emissions of 376 Gg SO2 in 2000and the US inventory estimates were used for US emissionsfrom this source This is the only explicit inclusion of thissource in our estimate Any sour gas emissions in industri-alized countries were presumed to be included in country in-ventories but no data were available to explicitly account forthese emissions which would be included indirectly thoughour calibration procedure

Petroleum production emissions were taken either fromcountry level inventories or where those were not availablefrom the EDGAR 32 (Olivier and Berdowski 2001) and 32



FT inventories (Olivier et al 2005) Emissions were scaledwith petroleum production over time where inventory datawere not available

Emissions from pulp and paper operations were estimatedusing emissions factors from Mylona (1996) and inventorydata combined with wood pulp production statistics (seeSupplement S11)

Remaining process emissions originate from a variety ofsources with sulfuric acid production one of the largestsources particularly in earlier years Process emissions weretaken from the above sources and scaled over time prior to1990 where inventory data were not available by the regionalHYDE estimate (van Aardenne et al 2001) Where updated2005 data were not available year 2000 values were used

Emissions from biomass combustion (exclusive of openburning) were estimated using a historical reconstruction ofbiomass consumption based on the estimate of Fernandes etal (2007) combined with IEA data (see Supplement S2)

Emissions from agricultural waste burning on fields werefrom EDGAR v40 (JRCPBL 2009) Production statisticsfor 24 crop types from FAO (FAOSTAT Crop Production)were combined with information on the fraction burned onthe fields (Yevich and Logan 2003 Eggleston et al 2006UNFCCC 2008) and emission factors from Andreae andMerlet (2001) Emissions from waste burning were calcu-lated in a similar manner although these are small and thedata for this sector are incomplete EDGARv40 emissionsfrom agricultural waste burning and waste were included inthe Representative Concentration Pathway (RCP) emissionsrelease described in Lamarque et al (2010)

25 Emissions by sector and grid

The emissions estimates developed here were mapped to astandard set of reporting sectors and then downscaled to a05 spatial grid as part of the production of historical datafor the new RCP scenarios (Moss et al 2010 Lamarqueet al 2010) The reporting sectors for the RCP histori-cal data were the following energy transformation residen-tialcommercial industry surface transportation agriculturalwaste burning on fields waste burning solvent use and agri-cultural activities (non-combustion) There are no apprecia-ble SO2 emissions from the last two sectors

Emissions from smelting and other industrial processeswere mapped to the industrial sector biomass fuel emissionsto the domestic sector and fossil fuel extraction and process-ing emissions to the energy sector Emissions from coal andpetroleum combustion were split into the first four sectorsabove by using inventory data sector-specific emissions fac-tors and IEA fuel use data where these data were availableand additional information from van Aardenne et al (2001)and Bond et al (2007) see Supplement for details (S12)

The emissions estimate was distributed onto a 05 reso-lution global grid for each decade from 1850 to 2000 Thesub-national split within a grid cell was estimated by using

the 25 min national boundary information from the GriddedPopulation of the World dataset (CIESIN and CIAT 2005)From 1960 through 2000 emissions were distributed usinga preliminary version of the year 2000 emissions distribu-tion from the EDGAR 40 project separated into energy sec-tor combustion industrial combustion and other industrialtransportation combustion and agricultural waste burning onfields within each country (Supplement S12)

For 1850ndash1900 emissions from combustion and other in-dustrial sectors for each country were distributed using theHYDE gridded population distribution (Goldewijk 2005)The emissions distribution for each country was interpolatedfrom the ldquomodernrdquo grid in 1960 to the population-basedgrid in 1900 by linearly increasing the weighting for thepopulation-based distribution in each year 1950 to 1910 anddecreasing the weighting for the ldquomodernrdquo grid until a purepopulation-based grid is used in 1900

The emissions grids were produced to facilitate the use ofthese data in global modeling experiments In most casesthe distribution of emissions within each country is deter-mined by proxy data not by actual emissions data Alter-native methods of downscaling these emissions estimates toa spatial grid (van Vuuren et al 2010) including incorpo-ration of emissions measurements could produce improvedemissions distributions No consideration of country bound-ary changes was made during the emissions gridding proce-dure Incorporation of these changes over time was beyondthe scope of this project

3 Uncertainty

It is useful to examine uncertainty in emissions by sourceand region To our knowledge this is the first consistentestimate of global and regional uncertainty in sulfur diox-ide emissions For this estimate we apply a relatively sim-ple approach to uncertainty analysis whereby a set of un-certainty bounds are applied to broad classes of countriesThis is warranted in large part since as noted by Schoppet al (2005) limited data are available to specify parame-ter uncertainty bounds leading to bounds that are generallyspecified through expert judgment This is particularly truefor developing countries In addition sulfur dioxide emis-sions are principally determined by fuel sulfur content andnot technology-specific emissions factors at least in the ab-sence of emissions controls Data on fuel sulfur content aresparse in general and those that contain uncertainty informa-tion even rarer These considerations make a more complexassessment of global uncertainty unwarranted at this time

We consider two sources of uncertainty random and sys-temic uncertainties as summarized in Eq (1)



Table 1 Uncertainty bounds (as 95 confidence interval) by country category and emissions type The uncertainty bounds shown in thetable are used for random effects in Eq (1) An additional systemic uncertainty was added (see text) with a magnitude of 25 for countrieswith category I inventories and 5 for all other countries (and all countries prior to 1970)

Category Coal Petroleum Smelting Other ProcessBiomass

I Recent-Country-Inventory plusmn11 plusmn21 plusmn14 plusmn22II Older Inventory plusmn18 plusmn27 plusmn25 plusmn38

IIa OECD (pre inventory) plusmn25 plusmn43 plusmn25 plusmn52III Other Countries plusmn28 plusmn45 plusmn36 plusmn54IV Int Shipping plusmn28IV Int Shipping (earlier) plusmn42

uncertainty=

radicsumr

sums

(Emissionsrs middotCI randomr

s

)2

+

sumr

radicsums

(Emissionsrs middotCI systemicrs

)2 (1)

where CI is the assumed 5ndash95 confidence level in percentfrom Table 1 for a given region and category The sumss

andr are conducted over the source categories and regionslisted in the Supplement (S1 S15)

The first component of the uncertainty analysis considerserrors in the individual components of the emissions calcu-lation The set of uncertainty bounds given in Table 1 areapplied to countries categorized depending on the estimatedquality of the data used to construct the inventory values (seeSupplement S15) Uncertainties are applied separately ineach country to emissions from the following sources coalpetroleum biomass fuel processing smelting and other pro-cess Uncertainties in each of these categories are assumedto be independent and are combined in quadrature Con-ceptually aggregate uncertainty can be divided into uncer-tainty in driving forces such as fuel consumption or smeltermetal output and uncertainty in sulfur content (and assump-tions such as ash retention) such as shown in the Supple-ment (S15) Only the total values however are used in thiscalculation

The values in Table 1 are based on the authorsrsquo judgmentinformed by previous work in the literature (Schopp et al2005 Gregg et al 2008 Eyring et al 2010) comparisonswith previous versions of this work and changes over time inEPA inventories (see Supplement S13 S14) These sourcessuggest that overall uncertainty is smallest where emissionsare directly measured such as in coal-fired power plants andis relatively larger for emissions from petroleum products(except for countries with well-enforced and comprehensivesulfur standards) and process emissions

In recent decades sulfur emissions in most high-incomecountries have come under increasingly stringent control

regimes In earlier years information on sulfur emissionswas less complete and we therefore assume that uncertain-ties are larger at these times For similar reasons we also as-sume that emissions are more uncertain in countries withoutcomprehensive control regimes or where such regimes haveonly been implemented recently In addition information onactivity levels are also more uncertain in the past and in de-veloping countries generally Because common assumptionsand data sources are used for large portions of the world weassume that uncertainties with each source category are per-fectly correlated within 14 world regions

This procedure assumes uncertainties are symmetric Thisis likely not strictly true since for example sulfur removal(for petroleum and metal smelting) is bounded above sulfurretained in ash is bounded below and some emissions drivershave potential biases in one direction ndash for example underre-porting of consumption (Logan 2001) It is not clear how-ever if a more nuanced calculation is warranted given thenumber of assumptions that would need to be made

The uncertainty estimate calculated as described above re-sults in uncertainty bounds on annual global total SO2 emis-sions that are relatively small 6ndash10 over the 20th centuryThis low value is due to cancellation between source cate-gories and regions This uncertainty level would appear tobe unrealistically low given that a number of previous globalsulfur dioxide emissions estimates do not fall within this es-timated uncertainty bound (Smith et al 2001Om et al1996 van Aardenne et al 2001 Lefohn et al 1999 Spiro etal 1992 Stern 2006) The reason is that additional essen-tially correlated uncertainties are present that add to the un-certainty value estimated above Examples include reportingor other biases in global data sets for energy sulfur removaland other driver data methodological assumptions and theuse of common default assumptions for sources where lit-tle data exists Comparing the present inventory with thatof Smith et al (2004) for example indicates that the differ-ences between these two estimates involve several method-ological and data changes that impacted emissions estimatesover multiple world regions (see Supplement S13)



To include the impact of such systemic effects we add tothe uncertainty estimate for each source category an addi-tional uncertainty amounting to 5 of total emissions (halfthis value for countries with well-specified inventories) withthe additional uncertainty combined again in quadrature be-tween source categories This latter assumption is made sincethere is little overlap in assumptions between sulfur contentsemissions controls or driver data between the broad cate-gories used in this calculation The uncertainty estimates areagain authorrsquos judgments but are sufficient to increase theoverall uncertainty estimate enough to encompass a largerfraction of the existing global estimates (S13)

The global value of the systemic uncertainty component isless than 3 since 1960 due to statistical cancellation acrosssectors increasing to 4 by 1920 as emissions from coalcombustion become more dominant The addition of thiscorrelated component to uncertainty has a large impact on thefinal uncertainty value The uncertainty range is increased bya factor of 13 to 15 depending on the year Even with thiscomponent however the global uncertainty in sulfur diox-ide emissions over the 20th century is still only 8ndash14 Theglobal emissions estimate of Smith et al (2004) now largelyfalls within this expanded uncertainty range While this com-ponent has a large relative impact on the global uncertaintythe impact of this assumption on regional uncertainties issomewhat smaller For example the largest uncertainty inrecent years is in emissions from China The addition of thiscorrelated component increases the magnitude of the total es-timated uncertainty for China emissions by a factor of 12(from 25 of total emissions to 29 of total emissions)

The combined uncertainty bounds for global emissions areshown in Fig 3 The uncertainty bounds are not wide enoughon a global level to change the overall character of emissionsover time with global emissions peaking in the 1970s witha significant decrease over the 1990rsquos and likely increasingslightly in recent years In 2000 the estimated uncertainty inglobal emissions isplusmn9 600 Gg SO2 While a number of pre-vious estimates lie within the uncertainty bounds estimatedhere there are some significant differences In particular anumber of estimates are larger than even the upper uncer-tainty bound in 1990 (see Supplement S15) We return tothis point in the discussion

The estimated uncertainty bounds for the United Statesand China are also shown in Fig 3 The uncertainty rangefor the United States is smaller even in earlier years thanthe estimate for China In recent years a large portion of thisdifference is due to the assumed lower uncertainty by source(Table 1) Also contributing to a lower overall uncertaintyestimate is that emissions from the United States throughoutthe 20th century are from a wider variety of sources as com-pared to China where emissions are dominated by coal con-sumption in all periods This results in a larger statistical can-celation of uncertainties between sectors in the United StatesExamples of such effects are discussed for European emis-sions estimates in Schopp et al (2005) Offsetting this some-

0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year

Global SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions

Fig 3 Sulfur dioxide emissions from fuel combustion and processemissions with central value (solid line) and upper and lower un-certainty bounds (dotted lines)(a) Global(b) China and(c) USAChina and USA graphs exclude shipping emissions

what is a similar relative impact of systematic uncertaintiesUncertainty in emissions from China is large enough that sig-nificant differences in trends over recent decades are possibleif trends in uncertain parameters also change over time

The regional uncertainty values before combination toglobal values as described above are shown in Fig 4 Chinais the largest single contributor to emissions uncertainty



0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year

Regional (before addition) and Global Uncertainty

India

Korea

Eastern Europe

South amp East Asia

Central amp South America

Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty

Fig 4 Regional uncertainty in sulfur dioxide emissions includingboth uncorrelated and correlated components before addition (seetext) The global uncertainty after addition is also shown (bold line)

since about 1980 with an estimated uncertainty in 2000 ofplusmn6 200 Gg SO2 a value comparable to our estimated emis-sions from India or the Middle East The second largestsource of uncertainty since the mid 1990s is emissions frominternational shipping The largest contributors to uncer-tainty have changed over the years with countries of the For-mer Soviet Union and Western Europe dominant during the1960s and 1970s and the United States dominant during theearly to mid-20th century

Sulfur emissions are less uncertain than emissions of mostother air pollutants because emissions depend largely on sul-fur contents rather than combustion conditions There is alarge contrast with emissions of another important aerosolblack carbon Bond et al (2004) estimated global fossil blackcarbon emissions in 1996 as 30 TgC with an uncertaintyrange of 20ndash74 TgC or +150 andminus30 with the up-per range an order of magnitude larger than the uncertaintyin sulfur dioxide emissions estimated here

Note that uncertainty in the spatial allocation was not as-sessed but will likely be much higher than the uncertainty innational or sectoral emissions totals due to the simple down-scaling methods used here

4 Results and discussion

41 Emission trends

Sulfur dioxide emissions since 1850 have shown substantialsectoral and regional shifts that reflect economic and techno-logical trends over this period Global SO2 emissions from1850ndash2005 by source including emissions from forest andgrassland fires and by end-use sector are shown in Fig 5Table 2 presents a summary of total emissions by region fordecadal years with global emissions from open burning alsoshown Uncertainty bounds for the same regions are shownin Table 3

In 1850 global sulfur dioxide emissions over land ar-eas were split roughly evenly between emissions from openburning and anthropogenic industrial activities Over thenext 50 years this changed dramatically as anthropogenic sul-fur dioxide emissions increased by an order of magnitudedriven by increased use of coal In the early 20th century thesteady growth of emissions was slowed by a global depres-sion and the second world war followed by the post war eco-nomic expansion resulting in an unprecedented absolute rateof emissions growth averaging 3400 Ggdecade from 1950to 1970 Global emissions peaked in the 1970s and have de-clined overall since 1990 with an increase between 2002 and2005 largely due to strong growth of emissions in China

The fraction of emissions from coal and petroleum haveremained remarkably constant since 1980 with around 50of emissions from coal and 30 from petroleum productsWhile terrestrial petroleum emissions have declined emis-sions from international shipping have increased This is asector where high sulfur bunker fuel could until recently beused without significant restriction in most of the world

Trends in emissions from coal combustion include a steadydecline since the 1970s in Europe and North America com-bined with large changes in coal combustion in the states ofthe Former Soviet Union after 1990 and a large increase incoal combustion in China in recent years Emissions frommetal smelting have decreased since their peak around 1970due to increased sulfur removal in part due to the widespreadintroduction of flash smelter technologies

From an end-use perspective (Fig 5) emissions from en-ergy conversion largely from electricity generation havesteadily increased over the 20th century to now comprisenearly 50 of SO2 emissions Global emissions from do-mestic and industrial combustion peaked around 1970 asthese sectors switched to electricity and natural gas Whileenergy (conversion) now comprises the largest emission sec-tor the remaining emissions are distributed over a rangeof end-uses notably smelting industrial combustion oceanshipping domestic combustion and other process emissionsSurface transportation contributes a relatively small portionof total sulfur dioxide emissions The largest portion of emis-sions from petroleum combustion arise from use of heavyfuel oils including residual fuel which are not often usedin surface transportation other than for domestic shippingThe only end-use sectors that have increased emissions be-tween 1970 and 2005 are the energy sector and internationalshipping Total process emissions (smelting fuel process-ing and extraction pulp and paper etc) appear to haveexceeded emissions from industrial combustion since 1990These trends point to the central role of energy-sector emis-sions in policies to reduce sulfur emissions overall but alsothe increasing importance that other sectors such as interna-tional shipping can play once energy sector emissions comeunder emission control regimes and begin to decrease (Miolaet al 2010 Eyring et al 2010)



Table 2 Total emissions estimates by decade (for the indicated year) and region (Gg SO2) with global emissions shown from sectors notincluded in the country-level estimates (open biomass burning from forest grassland and agricultural waste burning on fields see text)

Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005

USA amp Canada 311 707 1380 2701 5443 9345 16424 20436 21193 21475 24513 25849 34980 27809 24066 17054 15131Mexico 1 2 2 2 17 52 218 254 331 404 658 766 1003 2003 2729 2991 2145Central America 7 1 1 1 1 1 9 17 115 84 187 444 760 816 778 867 708South America 33 56 68 76 75 79 278 408 757 1101 1346 2740 3648 5328 5314 4719 4432Western Europe 1292 1929 2754 3972 5295 7213 9323 9035 11649 12519 13974 20599 29329 26759 18206 7998 6242Central Europe 34 75 219 518 903 1323 1756 1448 2280 3334 3942 7235 11462 13589 12316 5704 4832Russia 40 39 54 85 125 278 389 120 517 1897 2565 5744 7926 10143 10632 6352 5975Ukraine 7 9 15 33 55 140 192 76 260 907 1710 3249 4320 5505 4921 1548 1470Other Former Soviet Union 6 6 7 10 19 39 39 21 125 525 695 2864 3684 3777 4116 2516 3338China 162 147 138 140 142 148 357 458 581 1089 1070 8261 7327 11981 17194 21393 32673Japan 25 26 27 54 119 261 533 1081 1347 1641 1099 2188 5337 1317 975 885 834Other South amp East Asia 25 27 29 32 35 49 70 107 213 325 241 780 1796 3654 5489 6330 6111India 35 39 43 48 64 102 161 221 281 338 425 702 1114 1674 3302 5363 6275Australia amp New Zealand 9 24 32 53 61 149 264 213 339 466 486 927 1553 1692 1648 2438 2606South Africa 1 2 3 11 20 29 170 271 252 399 588 833 1491 1860 2283 2392 2477Africa 11 12 14 16 18 20 26 50 220 823 1087 2070 3168 3784 3350 3322 2690Middle East 3 3 4 4 5 5 5 11 20 136 324 572 1179 2490 3436 5218 5490International Shipping 61 120 208 365 587 892 1450 1927 2274 2233 3247 4680 6469 6607 7041 9779 12078Global Total (Comb + Process) 2063 3224 4996 8120 12983 20126 31665 36155 42756 49694 58158 90502 126544 130788 127795 106869 115507

Additional Emissions (Global)Forest amp Grassland Burning 2447 2447 2447 2447 2447 2447 2439 2244 2035 1958 1859 1920 2198 2649 3357 3836 3836Agricultural Waste Burning 83 88 94 100 106 110 115 120 126 134 141 153 155 180 211 205 205

Table 3 Estimated emissions uncertainty as 5ndash95 confidence interval for the indicated year and region (Gg SO2)

Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005

USA amp Canada 82 191 384 748 1310 1922 3250 3931 3734 3404 3658 2998 3212 2664 2357 1650 1440Mexico 1 1 1 1 6 17 71 77 102 124 210 257 370 857 1219 1301 822Central America 3 0 0 0 0 0 3 6 55 38 59 123 184 187 184 172 127South America 11 20 24 29 24 20 53 96 199 310 329 587 774 1181 1166 836 764Western Europe 272 385 536 661 761 981 1275 1359 1514 1699 1577 2172 2771 2097 1197 654 521Central Europe 7 15 42 110 189 267 341 277 348 512 571 995 1418 1600 1532 559 604Russia 14 14 17 24 34 82 116 39 165 703 1044 1917 2337 2633 2443 1382 1280Ukraine 3 3 5 13 20 56 84 29 115 449 884 1501 1705 1581 1106 427 394Other Former Soviet Union 2 2 2 2 3 7 7 4 26 137 192 701 843 770 825 590 738China 95 87 81 82 83 86 113 139 170 318 310 2542 2094 3367 5139 6229 9579Japan 11 12 12 13 20 45 94 205 249 299 185 356 1178 229 186 160 142Other South amp East Asia 6 6 7 7 8 9 11 15 33 45 34 94 265 522 689 831 713India 21 21 22 23 25 31 43 58 75 94 108 173 263 399 791 1334 1496Australia amp New Zealand 3 7 9 12 13 32 57 41 70 100 92 183 315 245 144 226 238South Africa 0 1 1 4 7 7 47 73 76 111 167 230 378 485 587 669 779Africa 3 3 3 4 4 5 5 9 52 254 273 562 739 758 582 555 423Middle East 1 1 1 2 2 2 2 3 5 47 82 123 245 525 701 954 1010International Shipping 26 51 88 155 249 379 615 817 965 789 918 1324 1830 1869 1991 2766 3416Global Total 334 493 755 1166 1781 2551 4079 4808 4893 5009 5515 7519 8906 9284 9975 9857 12827

Regional emissions trends are shown in Fig 6 As seenin previous work (Smith et al 2001 2004 Stern 2006) wefind that there has also been a shift in the regional distributionof sulfur emissions with an increasing proportion of emis-sions from Asia Until the middle of the 20th century emis-sions were dominated by Europe and North America Sincethat time emissions from other regions Asia in particularhave increased with approximately 40 of global emissionsoriginating from Asia by 2005 Emissions from other regions(Africa Middle-East Central and South America) have alsoincreased but to a lesser extent Emissions from China haveincreased to 28 of the estimated global total in 2005

For most of the historical period sulfur dioxide emissionsincreased in rough proportion to activity levels This be-

gan to change after 1970 due to concern over the impactsof these emissions on regional scales Precursors to thischange are evident in smelting and petroleum refining sec-tors The amount of sulfur removed from crude oil has beensteadily increasing since 1958 (Fig 2) and a similar patternexists for metal smelting where around half the sulfur con-tained in metal ores was removed at the smelter by 1980Removal of sulfur from coal combustion by means of post-combustion scrubbers began somewhat later but is now amajor driver of sulfur emission reductions Shifts in coalsupply to favor lower sulfur coal have also been a factor al-though it is difficult to estimate the magnitude of this effectglobally In the United States where detailed informationon coal sulfur content and emissions are available the shift



0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel

Grassland and Forest Fires

Other Process

Metal Smelting

International Shipping

Petroleum Combustion

Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector

WST amp Ag Waste Burn

Grass amp Forest Fires

Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb

Fig 5 Global sulfur dioxide emissions by(a) source and(b) end-use sector Emissions by source are the primary inventory result from thiswork Emissions were then mapped to sector Included are emissions from forest and grassland fires from van der Werf et al (2006) Schultzet al (2008) and Mieville et al (2010) as used in the RCP historical inventory exercise (Lamarque et al 2010)

013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13

Global Anthropogenic SO2 Emissions13

North America13Europe13FSU13East Asia13South-East Asia amp AustNZ13India13South and Central America13Middle East13Africa13International Shipping13

Fig 6 Global sulfur dioxide emissions by region (North Amer-ica = USA + Canada East Asia = Japan China+ and South Korea)

to low-sulfur coal was a major driver of emissions reduc-tions until recent years Flue-gas desulfurization has playedan increasing role in recent years when the coal sulfur con-tent actually increased but emissions continued to decrease(see Supplement Fig S-2)

The impact of these changes can be seen in Fig 7 whichshows the aggregate emissions factor (emissions over fuelcombusted) by region Note that the split between coal andpetroleum emissions is approximate in some regions whichmeans that there is uncertainty in these emissions factorsThe aggregate emissions factor shown in Fig 7 includes theimpact of changes in sulfur content sectoral shifts and emis-sions controls

The change in coal emissions factor varies greatly by re-gion with some regions showing little change while emis-sions factors in many regions including the Japan EuropeUnited States Canada and South Korea have decreased sub-stantially since 1970 As a result the global average emis-sions factor for combusted coal has decreased by 2005 to60 of the 1970 value Shifts to lower sulfur coal and flue-gas desulfurization have contributed to lower relative emis-sions from coal combustion although emissions data aloneis not sufficient to quantify these effects individually

The emissions factor for petroleum has decreased in allworld regions The global average emissions factor forpetroleum products in 2005 is about half of the 1970 valueFrom a top-down perspective this is due to an increase inthe fraction of sulfur removed from crude oil at oil refiner-ies From a bottom-up perspective the decrease is due tolimits on the sulfur content of end-use fuels and a reductionin the fraction of residual oil consumed Countries such asMexico and South Korea had particularly high percentagesof residual oil in their consumption mix until around 1990and a decrease in this fraction since then has contributed tothe decline in the aggregate petroleum emissions factor

42 Comparison with other estimates

The methodology used here whereby regional inventoriesare used to calibrate a bottom-up emissions estimate wasused in order to produce an estimate that uses what wasjudged to be the best data from various regions Use ofsuch inventory information automatically takes into accountemissions control efforts providing the inventory data usedaccurately takes these factors into account Uncertainty



013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13

Aggregate Coal Emissions Factor13

USA13 Canada13Western Europe13 Japan13Aus amp NZ13 FSU13China+13 South amp East Asia13Eastern Europe13 Korea13Average13

0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year

Aggregate Petr Emissions Factor

USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average

Fig 7 Aggregate emissions factors for(a) coal and(b) petroleumcombustion by region

assumptions were coupled with the use of inventory datawith emissions estimates assumed to be more uncertain inyears before inventory data were available

The annual global emissions estimate in this work issimilar before 1970 to the previous estimate using similarmethodologies (Smith et al 2004) with an average absolutedifference of 4 from 1900ndash1970 After 1970 the currentglobal emissions estimate is consistently lower than Smith etal (2004) and the difference increases to 13 over 1995ndash2000 The current estimate is lower for the Former SovietUnion China and South and East Asia but larger for inter-national shipping The largest difference is lower emissionsfrom coal with lower emissions also from petroleum (asidefrom shipping) The reasons for the lower emissions varyby region and include different treatment of coking coal as-sumptions for coal emissions in China different calibration

0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year

Global Anthropogenic Sulfur Emissions

Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)

Benkovitz et al (1996)

Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41

Fig 8 Global sulfur emissions from this and previous studiesCofala et al 2007 GEIA (Benkovitz et al 1996) Lefohn et al1999Om et al 1996 Smith et al 2001 2004 Spiro et al 1992SRES (Nakicenovic and Swart 2001) Stern 2006 EDGAR 20(Olivier et al 1996) EDGAR 32 (Olivier and Berdowski 2001)and EDGAR 41 (JRCPBL 2010) Open biomass burning emis-sions where available were not included in the totals shown Co-fala et al (2007) estimates adjusted to include international ship-ping as estimated here The 2000 points from SRES and Smith etal (2001) are shown as open symbols as these were projected emis-sions estimates made before year 2000 data were available

data for the Former Soviet Union and inclusion of additionalemissions control measures in Asia and lower sulfur stan-dards in most regions Other differences include updated in-formation on sulfur removals from refineries and smeltersand improved methodologies See the Supplement (S13) forfurther discussion and analysis

The current estimate is compared to other estimates ofglobal sulfur dioxide emissions in Fig 8 The current esti-mate is somewhat below many recent estimates particularlyin the 1970s and 1980s The estimate here for 1990 1995and 2000 agrees very closely with that of Cofala et al (2007)in part due to the use of similar assumptions for China andcalibration to country inventory data for many OECD coun-tries The uncertainty bounds estimated in section 4 encom-pass many of the other estimates (see also S15) although theSpiro et al (1992) and EDGAR 32 (Olivier and Berdowski2001) estimates lie above the uncertainty range as does theestimated year 2000 emissions data point used in the SRESemissions scenarios

The recent estimates from the EDGAR 41 (JRCPBL2010) inventory are on a global basis somewhat higher thanthe estimates here (see Supplement S16) While the globalemissions value is similar in 1970 and 2005 there are largeregional differences in all years The EDGAR inventory usesa consistent approach for all regions and gases but does notcalibrate to country-level inventories Some differences aretherefore expected since not all region- and country-specific



data can be included There are particularly large differencesin the two estimates in the Former Soviet Union an areawhere comprehensive emissions information is still relativelysparse Large relative differences are also seen for Japanthe FSU South amp East Asia South Korea India Argentinaand international shipping The reasons for these differencesshould be examined once methodological details from theEDGAR estimate are available The global totals from theEDGAR-HTAP inventory for 2000ndash2005 which substituteregional inventory values for the EDGAR 41 values whereavailable are within 1ndash2 of the values reported here

It is also useful to compare 2005 emissions with previousestimates from global scenario projections The 2005 emis-sions estimated here as 115 510plusmn 12 830 Gg largely over-lap with the range of future SO2 emissions scenarios fromSmith et al (2005) This range is significantly lower thanthe SRES scenario projections for 2005 but does overlapwith many of the scenarios in the post-SRES literature (vanVuuren et al 2008) The recent RCP emissions scenarios(Moss et al 2010) closely match the range estimated heresince they were calibrated to an earlier version of this inven-tory (see Supplement S17)

Emissions in Asia are of particular interest due to theirrecent increase and it is useful to compare emissions herewith previous estimates As shown in the Supplement evenrecent estimates for SO2 emissions in Asia (Ohara et al2007 Streets et al 2003 Zhang et al 2009 and Klimont etal 2009) sometimes differ substantially (S4) Focusing onChina and India the two countries with the largest emissionsthese recent estimates all lie within the uncertainty range es-timated here with the REAS (Ohara et al 2007) estimatefor 2000 consistently larger than the other estimates quotedhere but within the estimated uncertainty range The esti-mate here for China is within a few percent of the estimate inLu et al (2010) for 2000ndash2005 The current estimate for In-dia is close to the estimate from Garg et al (2006) in 1985 butdiverges to be 30 higher than the Garg et al value by 2005but similar to other recent inventories (see Supplement S4)

The current estimate for China is 20ndash30 lower than ear-lier estimates based on the RAINS Asia project (Arndt etal 1997 Streets et al 2000 Streets and Waldhoff 2000Klimont et al 2001) for 1895ndash1990 This difference iswithin the estimated uncertainty range While the current es-timate for India for 2000 and 2005 is similar to a number ofother regional estimates for these years (although higher thanGarg et al 2006) the current estimate is 40 lower than thecirca 1987 estimates from Arndt et al (1997) and Streets etal (2000) Note that such historical differences are not lim-ited to developing countries as indicated above (Sect 21)for past emissions from petroleum in the United States

It is not clear if such historical differences are due to vari-ation in energy consumption data parameter assumptionssuch as sulfur content or ash retention fraction or othermethodological differences such as treatment of shifts inpetroleum product consumption and sulfur contents Anal-

ysis of these differences is beyond the scope of the currentproject Given that the current estimate is similar to recentliterature for later years there may have been changes in as-sumptions and methodology in recent estimates that wouldimply changes in these earlier emissions estimates as wellBecause emissions in China are dominated by coal consump-tion any uncertainty in coal emission parameters will trans-late directly into uncertainty in total emissions A historicalanalysis of coal production and consumption in China par-ticularly data on coal production by source in order to trackchanging sulfur contents would be especially useful in betterdetermining historical emissions

43 Uncertainty

The uncertainty methodology used here uses a relatively sim-ple procedure whereby confidence intervals based largely onthe authorsrsquo judgment but also informed by analysis of in-ventory differences are applied to broad emissions sectorsand regions The relatively small resulting global uncertaintythat results indicates that a more complex global uncertaintyanalysis may not be warranted Regional uncertainty can befar higher than global uncertainty however and more de-tailed analysis of high-emitting regions and the countries ofthe Former Soviet Union in particular may be useful to bet-ter bound current and past environmental impacts of sulfurdioxide emissions

Because the simple methodology used here does not in-corporate correlations between uncertainty assumptions indifferent regions and sectors (parameters that can be diffi-cult to estimate in any event) a systemic uncertainty com-ponent was added Systematic errors and biases are difficultto quantify but are particularly important for emissions suchas sulfur dioxide where most input values are only weaklycorrelated between regions which results in relatively smallglobal uncertainties as random errors tend to statistically can-cel across regions

When comparing data sets it should be noted that mostof these estimates rely on similar if not identical data setsfor historical fossil fuel use and for historical emissions fromEurope and the United States Errors or biases in these datasuch as the apparent underestimate of SO2 emissions frompetroleum products in the United States prior to 1980 arelikely to be common to most of these estimates

As shown by an analysis of inventory values from theUSA (sectS14) however significant changes can occur overtime in national inventory values Substantial changes havealso occurred for inventory estimates for Europe (eg seeKonovalov et al 2008 for NOx emissions) Analysis of thesources of such changes would be valuable for both improv-ing both inventory methodologies and uncertainty estimates



5 Future work

Future estimates of historical sulfur dioxide emissions couldbe improved through improved regional data sets for fos-sil fuel consumption fuel properties and industrial processdrivers (including metal smelting amounts ore propertiessulfuric acid production and pulp and paper production)Limited information was available for sulfur dioxide emis-sions resulting from petroleum and natural gas extraction op-erations and the use of sour gas even though these sourcescould be regionally significant Many historical data seriesare only readily available at the country level Improvedspatial estimates of historical emissions would require sub-national consumption and activity information particularlyfor geographically large countries such as the United StatesRussia and China

Differences among estimates of both current and pastemissions point to the need for further research to identifycurrent and historical fuel sulfur contents and the characteris-tics of fuel-emissions technology combinations (such as sul-fur retention in ash) While the bottom-up estimation meth-ods used in part in this work are appropriate for countrieswith few emissions controls direct emissions monitoring oflarge sources will become even more important for accurateemission estimates as the use of sulfur control technologiesbecomes more widespread

Improved and updated emissions estimates for countrieswithout national inventories will depend in part on accuraterepresentations of changing sulfur content standards Thesectors and specific fuel types subject to standards and thelevel of enforcement will all impact the resulting emissionsThe increase in international shipping emissions should slowand ultimately decline once recent agreements on sulfur stan-dards for bunker fuels are implemented1 Regionally any ex-pansion of Sulfur Emission Control Areas where lower sul-fur fuel limits will apply will further reduce shipping emis-sions

The use of satellite observations (Krotkov et al 2008 Liet al 2010 Lu et al 2010 Gottwald et al 2010) to ver-ify trends shows promise although estimation of absoluteemission values through use of satellite data is more diffi-cult Figure 9 shows the SO2 column concentration aboveEastern China as estimated by Gottwald et al (2010) usingsatellite data The increase in estimated SO2 column from2000 to 2005 is about 15 which is the same change as intotal emissions over China from this period The compari-son for the earlier period is less clear The satellite estimateindicates a slight increase in concentrations from 19961997to 20002001 while the inventory estimate is nearly flat tak-ing the average over the indicated years There is of course

1 Agreements conducted under the International Convention forthe Prevention of Pollution From Ships commonly refered to asMARPOL Marine Environment Protection Committee (MEPC) ndash57th session 31 Marchndash4 April 2008

Fig 9 Monthly (lines) and annual average (bars) estimated sul-fur dioxide column over Eastern China as estimated from satellitemeasurements (Gottwaldov et al 2010)

substantial uncertainty in the emissions estimate and a dif-ferent trend over this period would be within the uncertaintyrange In addition a part of this difference could be due toshifts in SO2 source distribution over this time period Con-centrated emissions from large sources which are also loftedhigher into the atmosphere are more likely to be detected bysatellite instruments From 1995 to 2000 we estimate thatthe fraction of total emissions from power plants in Chinaincreased from 46 to 60 which is in the right directionto help explain the difference in trends Meteorological andatmospheric chemistry effects on SO2 transport and lifetimewill also affect the relationship between SO2 column mea-surements and emissions Additional uncertainty arises fromchanges in aerosol loading leading to changes in satellite sen-sitivity See Gottwald et al (2010) and the papers cited abovefor further discussion of these effects It is not yet clear howthese uncertainties in satellite estimates compare to the un-certainty in the inventory-based emissions estimate as quan-tified here

Historical estimates of sulfur dioxide emissions are neces-sary for estimating past trends in acid deposition the associ-ated impacts and past climate forcing The current estimaterepresents a consistent global data set with annual resolutionthat can be used for historical modeling studies The annualemissions data described in this paper are available from thecorresponding author The 05 gridded emissions data re-leased for the RCP project are available at the RCP web site2

Supplementary material related to thisarticle is available online athttpwwwatmos-chem-physnet1111012011acp-11-1101-2011-supplementpdf

AcknowledgementsWork at PNNL on this project was supportedby the Office of Science (BER) US Department of EnergyZ Klimont and GAINS work was supported by ACCENT andEUCAARI RJA was sponsored by US Department of EnergyOffice of Science Biological and Environmental Research (BER)

2httpwwwiiasaacatweb-appstntRcpDb



programs and performed at Oak Ridge National Laboratory(ORNL) under US Department of Energy contract DE-AC05-00OR22725 The authors would like to thank Elaine Chapman fora very useful internal review Andreas Richter for Figure 9 andhelpful comments and the many colleagues and organizations thatshared data used in this project We also thank Mayda NathanTerri Weber and Andrew Mizrahi for assistance with proofreadingand formatting The authors would like to thank the two anonymousreferees for their detailed comments which were of great value inimproving the paper

Edited by R Harley

References

Andreae M O and Merlet P Emissions of trace gases andaerosols from biomass burning Global Biogeochem Cy 15955ndash966doi1010292000GB001382 2001

Andres R J Fielding D J Marland G Boden T A and Ku-mar N Carbon dioxide emissions from fossil-fuel use 1751ndash1950 Tellus 51 759ndash765 1999

Arndt R L Carmichael G R Streets D G and BhattiN Sulfur dioxide emissions and sectorial contributions tosulfur deposition in Asia Atmos Environ 31 1553ndash2310doi101016S1352-2310(96)00236-1 1997

Benkovitz C M Scholtz M T Pacyna J Tarrason L DignonE C Voldner P A Spiro Logan J A and Graedel T AGlobal gridded inventories of anthropogenic emissions of sul-phur and nitrogen J Geophys Res 101(D22) 29239ndash29253doi10102996JD00126 1996

Bond T C Streets D G Yarber K F Nelson S M Woo J-H and Klimont Z A technology-based global inventory ofblack and organic carbon emissions from combustion J Geo-phys Res 109 D14203doi1010292003JD003697 2004

Bond T C Bhardwaj E Dong R Jogani R Jung S Ro-den C Streets D G and Trautmann N M Historicalemissions of black and organic carbon aerosol from energy-related combustion Global Biogeochem Cy 21 GB2018doi1010292006GB002840 2007

Center for International Earth Science Information Network(CIESIN) Columbia University and Centro Internacional deAgricultura Tropical (CIAT) Gridded Population of the WorldVersion 3 (GPWv3) Population Density Grids PalisadesNY Socioeconomic Data and Applications Center (SEDAC)Columbia University available athttpsedacciesincolumbiaedugpw 2005

Cofala J Amann M Klimont Z Kupiainen K and Hoglund-Isaksson L Scenarios of Global Anthropogenic Emissions ofAir Pollutants and Methane Until 2030 Atmos Environ 384769ndash4778doi101016jatmosenv200707010 2007

Corbett J J and Kohler H W Updated emissionsfrom ocean shipping J Geophys Res 108(D20) 4650doi1010292003JD003751 2003

Corbett J J and Kohler H W Considering alternative input pa-rameters in an activity-based ship fuel consumption and emis-sions model Reply to comment by Oslashyvind Endresen et al onldquoUpdated emissions from ocean shippingrdquo J Geophys Res109 D23303 doi1010292004JD005030 2004 2004

Corbett J J Fischbeck P S and Pandis S N Global nitrogenand sulfur inventories for oceangoing ships J Geophys Res104(D3) 3457ndash3470doi1010291998JD100040 1999

Eggleston S Buendia L Miwa K Ngara T and Tanabe K(Eds) 2006 IPCC Guidelines for National Greenhouse Gas In-ventories Intergovernmental Panel on Climate Change IGESJapan 2006

Endresen Oslash Soslashrgard E Behrens HL Brett PO and Isak-sen ISA A historical reconstruction of shipsrsquo fuel con-sumption and emissions J Geophys Res 112 D12301doi1010292006JD007630 2007

Energy Information Administration (EIA 2008) International En-ergy Annual 2006 Washington DC available atwwweiadoegov) 2004

Energy Information Administration (EIA 2009) Annual EnergyReview 2008 Washington DC DOEEIA-0384 2009

Environment Canada 1985ndash2006 Historical SOx emissions forCanada (version 2 8 April 2008) available athttpwwwecgcca 2008

European Environment Agency (EEA 2002) Anthropogenic emis-sions of sulphur (1980ndash2010) in the ECE region (September2000)

Eyring V Kohler H W van Aardenne J and Lauer A Emis-sions from international shipping 1 The last 50 years J Geo-phys Res 110 D17305doi1010292004JD005619 2005

Eyring V Isaksen I S A Berntsen T Collins W J CorbettJ J Endresen O Grainger R G Moldanova J SchlagerH and Stevenson D S Transport impacts on atmosphere andclimate Shipping Atmos Environ 44 4735ndash4771 2010

Fernandes S D Trautmann N M Streets D G Roden C Aand Bond T C Global biofuel use 1850ndash2000 Global Bio-geochem Cy 21 GB2019doi1010292006GB002836 2007

Fletcher M E ldquoFrom coal to oil in British shippingrdquo in WilliamsDavid M The World of Shipping by Aldershot Hants Eng-land Ashgate Brookfield VT 1997

Forster P Ramaswamy V Artaxo P Berntsen T Betts R Fa-hey D W Haywood J Lean J Lowe D C Myhre GNganga J Prinn R Raga G Schulz M and van DorlandR Changes in Atmospheric Constituents and in Radiative Forc-ing In Climate Change 2007 The Physical Science Basis Con-tribution of Working Group I to the Fourth Assessment Reportof the Intergovernmental Panel on Climate Change edited bySolomon S Qin D Manning M Chen Z Marquis M Av-eryt K B M Tignor M and Miller H L Cambridge Uni-versity Press Cambridge United Kingdom and New York NYUSA 129ndash234 2007

Fujita S Sanseiu kenkyu 100 nen no rekishi to sono hensen JRes Environ 29 82ndash88 1993

Garg A Shukla P R and Kapshe M The sectoral trendsof multigas emissions inventory of India Atmos Environ 404608ndash4620 2006

Goldewijk K K Three Centuries of Global Population GrowthA Spatial Referenced Population (Density) Database for 1700ndash2000 Population and Environment 26doi101007s11111-005-3346-7 2005

Gottwald M and Bovensmann H (Eds) SCIAMACHY Explor-ing the Changing Earthrsquos Atmosphere 1st edition SpringerISBN 978-90-481-9895-5 2011

Gregg J S Andres R J and Marland G China the emis-



sions pattern of the world leader in CO2 emissions from fossilfuel consumption and cement production Geophys Res Lett35 L08806doi1010292007GL032887 2008Gschwandtner G Gschwandtner K Eldridge K Mann Cand Mobley D Historic emissions of sulfur and nitrogen oxidesin the United States from 1900 to 1980 Journal of Air PollutionControl Association 36 139ndash149 1986

International Energy AgencyOrganization for Economic Coopera-tion and Development (IEAOECD) Energy Statistics and Bal-ances International Energy Agency Paris 2006

International Maritime Organization (IMO) Prevention of Air Pol-lution From Ships Sulphur monitoring for 2007 (Marine Envi-ronment Protection Committee 57th session MEPC 57424)2007

European Commission Joint Research Centre (JRC)NetherlandsEnvironmental Assessment Agency (PBL) Emission Databasefor Global Atmospheric Research (EDGAR) release version 41available athttpedgarjrceceuropaeu 2010

Klimont Z Cofala J Schopp W Amann M Streets D GIchikawa Y and Fujita S Projections of SO2 NOx NH3 andVOC Emissions in East Asia up to 2030 J Water Air Soil Poll130 193ndash198 2001

Klimont Z Cofala J Xing J Wei W Zhang C Wang SKejun J Bhandari P Mathur R Purohit P Rafaj P Cham-bers A Amann M and Hao J Projections of SO2 NOx andcarbonaceous aerosols emissions in Asia Tellus B 61 602ndash6172009

Konovalov I B Beekmann M Burrows J P and Richter ASatellite measurement based estimates of decadal changes in Eu-ropean nitrogen oxides emissions Atmos Chem Phys 8 2623ndash2641doi105194acp-8-2623-2008 2008

Krotkov N A McClure B Dickerson R R Carn S A Li CBhartia P K Yang K Krueger A J Li Z Levelt P F ChenH Wang P and Lu D Validation of SO2 retrievals from theOzone Monitoring Instrument over NE China J Geophys Res113 D16S40doi1010292007JD008818 2008

Lamarque J-F Bond T C Eyring V Granier C Heil AKlimont Z Lee D Liousse C Mieville A Owen BSchultz M G Shindell D Smith S J Stehfest E van Aar-denne J Cooper O R Kainuma M Mahowald N Mc-Connell J R Naik V Riahi K and van Vuuren D P His-torical (1850-2000) gridded anthropogenic and biomass burningemissions of reactive gases and aerosols methodology and ap-plication Atmos Chem Phys 10 7017ndash7039doi105194acp-10-7017-2010 2010

Lefohn A S Husar J D and Husar R B Estimating histori-cal anthropogenic global sulfur emission patterns for the period1850ndash1990 Atmos Environ 33 3435ndash3444 1999

Li C Zhang Q N Krotkov D G Streets K He S-C Tsay and Gleason J F Recent large reduction insulfur dioxide emissions from Chinese power plants ob-served by the Ozone Monitoring Instrument GRL 37 L08807doi1010292010GL042594 2010

Logan J Diverging energy and economic growth in China wherehas all the coal gone Pacific and Asian Journal of Energy 111ndash13 2001

Lu Z Streets D G Zhang Q Wang S Carmichael G RCheng Y F Wei C Chin M Diehl T and Tan Q Sulfurdioxide emissions in China and sulfur trends in East Asia since

2000 Atmos Chem Phys 10 6311ndash6331doi105194acp-10-6311-2010 2010

Mieville A Granier C Liousse C B Guillaume Mouil-lot F Lamarque J-F Gregoire J-M and Petron GEmissions of gases and particles from biomass burning usingsatellite data and an historical reconstruction Atmos Environ101016jatmosenv201001011 2010

Miola A Ciuffo B Giovine E and Marra M The State of theArt on Methodologies Technologies and Policy Options Publi-cations Office of the European Union Luxembourg 2010

Moss R H Edmonds J A Hibbard K Carter T Emori SKainuma M Kram T Manning M Meehl J Mitchell JNakicenovic N Riahi K Rose S Smith S J Stouffer RThomson A M van Vuuren D Weyant J and WilbanksT Representative Concentration Pathways A New Approachto Scenario Development for the IPCC Fifth Assessment ReportNature 463 747ndash756 2010

Mylona S Sulphur dioxide emissions in Europe 1880ndash1991 andtheir effect on sulphur concentrations and depositions Tellus B48 662ndash689 1996

Nakicenovic N and Swart R (Eds) Emissions Scenarios 2000Special Report of the Intergovernmental Panel on ClimateChange Cambridge UK Cambridge University Press 2001

National Institute of Environmental Research-Korea (NIER) avail-able at httpairemissniergokrDownloaded 09-11-20082008

Ohara T Akimoto H Kurokawa J Horii N Yamaji KYan X and Hayasaka T An Asian emission inventory ofanthropogenic emission sources for the period 1980-2020 At-mos Chem Phys 7 4419ndash4444doi105194acp-7-4419-20072007

Olivier J G J and Berdowski J J M Global emissions sourcesand sinks in The Climate System edited by Berdowski JGuicherit R and Heij B J 33ndash78 A A Balkema Pub-lishersSwets amp Zeitlinger Publishers Lisse The NetherlandsISBN 90 5809 255 0 2001

Olivier J G J Bouwman A F van der Maas C W MBerdowski J J M Veldt C Bloos J P J Visschedijk AJ H Zandveld P Y J and Haverlag J L Description ofEDGAR Version 20 A set of global emission inventories ofgreenhouse gases and ozone-depleting substances for all anthro-pogenic and most natural sources on a per country basis and on1times 1 degree grid National Institute of Public Health and the En-vironment (RIVM) report no 771060 002TNO-MEP report noR96119 1996

Olivier J G J van Aardenne J A Dentener F Pagliari VGanzeveld L N and J A H W Peters Recent trends in globalgreenhouse gas emissions regional trends 1970ndash2000 and spa-tial distribution of key sources in 2000 Environ Sci 2(2ndash3)81ndash99doi10108015693430500400345 2005

Om G Hansson U and Rodhe H Historical worldwide emis-sions of anthropogenic sulfur 1860ndash1985 Department of Mete-orology Stockholm University Report CM-91 1996

Schopp W Klimont Z Suutari R and Cofala J Uncertaintyanalysis of emission estimates in the RAINS integrated assess-ment model Environmental Science and Policy 8 601ndash6132005

Schultz M G Heil A Hoelzemann J J Spessa AThonicke K Goldammer J Held A C Pereira J M



and van het Bolscher M Global Wildland Fire Emissionsfrom 1960 to 2000 Global Biogeochem Cy 22 GB2002doi1010292007GB003031 2008

Smith S J Pitcher H and Wigley T M L Global and Re-gional Anthropogenic Sulfur Dioxide Emissions Global PlanetChange 29(1ndash2) 99ndash119 2001

Smith S J Andres R Conception E and Lurz J Sulfur Diox-ide Emissions 1850ndash2000 (JGCRI Report PNNL-14537) 2004

Smith S J Pitcher H and Wigley TML Future Sulfur DioxideEmissions Clim Change 73(3) 267ndash318 2005

Spiro P A Jacob D J and Logan J A Global inventory ofsulfur emissions with a 1 times1 resolution J Geophys Res 976023ndash6036 1992

Stern DI Reversal of the trend in global anthropogenic sulfuremissions Global Environ Change 16 207ndash220 2006

Streets D G and Waldhoff S T Present and future emissions ofair pollutants in China SO2 NOx and CO Atmos Environ 34363ndash374 2000

Streets D G Tsai N Y Akimoto H and Oka K Sulfur dioxideemissions in Asia in the period 1985ndash1997 Atmos Environ 344413ndash4424 2000

Streets D Bond T C Carmichael G R Fernandes S D FuQ He D Klimont Z Nelson S M Tsai N Y Wang MQ Woo J-H and Yarber K F An inventory of gaseous andprimary aerosol emissions in Asia in the year 2000 J GeophysRes 108 8809doi1010292002JD003093 2003

US Environmental Protection Agency (USEPA 1996) NationalAir Pollutant Emission Trends 1900ndash1995 EPA-454R-96-007Washington DC 1996

UK National Atmospheric Emissions Inventory available athttpwwwnaeiorguk 2009

UNFCCC National Inventory Submissions of Annex I Parties tothe UNFCCC available athttpunfcccint 2008

United Nations The United Nations Energy Statistics Database(UN 1996) (United Nations Statistics Division New York June1996

United Nations Environmental Program (UNEP) Sulphur Levels inDiesel Fuel Matrix tables various regions Partnership for CleanFuel and Vehicles downloaded 1092008 2008

United Nations Framework Convention on Climate Change UN-FCCC 2008 Country submissions Downloaded 122009 avail-able athttpunfcccint 2009

van Aardenne J A Dentener F J Olivier J G J Klein Gold-ewijk C G M and Lelieveld J A 1times 1 degree resolutiondataset of historical anthropogenic trace gas emissions for theperiod 1890ndash1990 Global Biogeochem Cy 15 909ndash928 2001

van der Werf G R Randerson J T Giglio L Collatz G JKasibhatla P S and Arellano Jr A F Interannual variabil-ity in global biomass burning emissions from 1997 to 2004 At-mos Chem Phys 6 3423ndash3441doi105194acp-6-3423-20062006

van Vuuren D Smith S J and Riahi K Downscaling Driversof Global Environmental Scenarios Wiley Interdisciplinary Re-views Climate Change 1(3) 393ndash404doi101002wcc502010

van Vuuren D P Meinshausen M Plattner G-K Joos FStrassmann K M Smith S J Wigley T M L Raper S CB Riahi K de la Chesnaye F den Elzen M Fujino J JiangK Nakicenovic N Paltsev S and Reilly J M Temperatureincrease of 21st century mitigation scenarios P Natl Acad Sci105 15258ndash15262 2008

Vestreng V Myhre G Fagerli H Reis S and TarrasonL Twenty-five years of continuous sulphur dioxide emis-sion reduction in Europe Atmos Chem Phys 7 3663ndash3681doi105194acp-7-3663-2007 2007

Xu Y Williams R H and Socolow R H Chinarsquos rapid deploy-ment of SO2 scrubbers Energy Environ Sci 2 459ndash4652009

Yevich R and Logan J An assessment of biofuel use and burn-ing of agricultural waste in the developing world Global Bio-geochem Cy 17 1095ndash1034 2003

Zhang Q Streets D G Carmichael G R He K B Huo HKannari A Klimont Z Park I S Reddy S Fu J S ChenD Duan L Lei Y Wang L T and Yao Z L Asian emis-sions in 2006 for the NASA INTEX-B mission Atmos ChemPhys 9 5131ndash5153doi105194acp-9-5131-2009 2009



2 Methodology

Sulfur emissions from combustion and metal smelting can inprinciple be estimated using a bottom-up mass balance ap-proach where emissions are equal to the sulfur content of thefuel (or ore) minus the amount of sulfur removed or retainedin bottom ash or in products Data limitations howevermake the bottom-up approach uncertain since sulfur contentsvary and information on sulfur removals is not always re-ported Countries with air pollution policies in place gener-ally require detailed reporting of emissions including directmeasurement of emissions for many large sources Thesedata are likely to be more accurate than bottom-up estimatestherefore we constrain our calculations to match country-level inventories where these data are available and judgedto be reliable This method produces an emissions estimatethat is consistent with the available country inventory datacontains complete coverage of all relevant emissions sources(assuming the country inventory data are complete) and isconsistent across years

Emissions were estimated annually by country for the fol-lowing sources coal combustion petroleum combustionnatural gas processing and combustion petroleum process-ing biomass combustion shipping bunker fuels metal smelt-ing pulp and paper processing other industrial processesand agricultural waste burning (AWB) The approach is sum-marized in Fig 1 and further detailed in the Supplement(S2 S3) The first step was to develop a detailed inventoryestimate by sector and country for three key years 19902000 and 2005 These years were chosen due to data avail-ability and because these years span a time period of sig-nificant change in emissions controls Initial estimates forother years were constructed by interpolating emissions fac-tors with final estimates by source and country found af-ter calibrating to country-level emissions inventories wherethose were available (S3) This methodology accounts forchanging patterns of fuel consumption and emission controlsin a context of limited detail for earlier years

Emissions by end-use sector for decadal years were esti-mated for a set of standard reporting sectors (energy indus-try transportation domestic AWB) before downscaling to a05 spatial grid Descriptions of each element in this calcu-lation are given below

21 Fossil fuel combustion

Emissions from coal and petroleum combustion were es-timated starting with the regional emissions factors fromSmith et al (2001) A composite fossil fuel consumptiontime series was constructed (see Supplement S2) using datafrom IEAOECD (2006) UN energy statistics (1996) andAndres et al (1999) Country-level emissions estimates forEurope North America Japan Australia and New Zealandwere compiled from UNFCCC (2009) for 1990ndash2005 En-vironment Canada (2008) the EEA (2002) Fujita (1993)

(1)13Emissions by

fuel and sector13

1990 2000 and 200513

(3)13Default

emissions by country and

fuel13

1850-200513

Fuel consumption

and other drivers13

1850-200513

(2)13Estimates for

smelting (mass balance) and international

shipping13emissions13

1850-200513

Emissions inventory estimates13

where available13

(4)13Final emissions

by fuel amp process13

1850-200513

Sectoral split percentages13

1850-200513

(5)13Emissions by

fuel and sector13

1850 1860 hellip 1990 2000

200513

Fig 1 Calculation summary The key steps in the calculation are(1) development of an inventory by sector and fuel for three keyyears(2) development of detailed estimates for smelting and inter-national shipping(3) calculation of a default set of emissions byinterpolating emissions factors from the key years(4) calculationof final annual emissions values by fuel that match inventory valuesand(5) estimate sectoral emissions

Mylona (1996) Gschwandtner et al (1986) UK NationalAtmospheric Emissions Inventory (2009) USEPA (1996)and Vestreng et al (2007) as detailed in the Supplement(S4 S5) Fossil-fuel emissions factors were scaled suchthat total emissions matched inventory values for those yearswhere inventory values are available While total emissionsfor these countries are therefore constrained by inventoryvalues emissions by fuel are not as well constrained by theavailable data

The ratio of default to inventory emissions in 1975 rangedfrom 03 to 2 for countries in Europe for example givingan indication of the differing rates at which emissions factorschange in different countries These changes can be due todifferent fuel sulfur contents and emission control policiesA similar range was found when comparing 1975 to 1960values (see Supplement S3)

The emission estimates for other Asian countries devel-oped here were compared to a number of regional estimates(Ohara et al 2007 Streets et al 2003 Zhang et al 2009NIER 2008 Klimont et al 2009) and the initial estimateshere were adjusted where necessary to better match thesestudies (see Supplement S4) These estimates differ fromeach other in some countries and sectors highlighting a needfor improved inventories in this rapidly changing regionSpecific assumptions for China and the Former Soviet Unionare described in the Supplement (S6 S7) including the op-eration of SO2 scrubbers in China (Xu et al 2009)

Petroleum emissions are particularly difficult to estimatein the absence of detailed country-level inventory data For








013

1000013

2000013

3000013

4000013

5000013

6000013

195013 196013 197013 198013 199013 200013

Sulfu

r as S

O2 (

Gg)13

Year13














23 Metal amelting




24 Other emissions


















3 Uncertainty









uncertainty=

radicsumr

sums


s

)2

+

sumr

radicsums


)2 (1)















0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References

























































































013

1000013

2000013

3000013

4000013

5000013

6000013

195013 196013 197013 198013 199013 200013

Sulfu

r as S

O2 (

Gg)13

Year13














23 Metal amelting




24 Other emissions


















3 Uncertainty









uncertainty=

radicsumr

sums


s

)2

+

sumr

radicsums


)2 (1)















0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References
























































































23 Metal amelting




24 Other emissions


















3 Uncertainty









uncertainty=

radicsumr

sums


s

)2

+

sumr

radicsums


)2 (1)















0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References
































































































3 Uncertainty









uncertainty=

radicsumr

sums


s

)2

+

sumr

radicsums


)2 (1)















0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References
























































































uncertainty=

radicsumr

sums


s

)2

+

sumr

radicsums


)2 (1)















0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References
























































































0

20000

40000

60000

80000

100000

120000

140000

160000

1850 1875 1900 1925 1950 1975 2000

Gg S

O2

Year


0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg

SO

2

Year

China SO2 Emissions

0

5000

10000

15000

20000

25000

30000

35000

40000

1900 1925 1950 1975 2000

Gg S

O2

Year

USA SO2 Emissions






0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References




















































































0

5000

10000

15000

20000

25000

30000

1850 1875 1900 1925 1950 1975 2000

Unce

rtai

nty

(G

g S

O2y

r)

Year


India

Korea

Eastern Europe

South amp East Asia


Africa

Middle East

China+

FSU

Aus amp NZ

Japan

Western Europe

Canada

USA

Intl Shipping

Global Uncertianty






41 Emission trends









Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References





















































































Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005




Region 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2005







0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References




















































































0

25000

50000

75000

100000

125000

150000

1850

1875

1900

1925

1950

1975

2000

Em

issi

ons

(Gg S

O2)

Year


by primary source

Waste

Ag Waste Burning

Biomass fuel


Other Process

Metal Smelting



Coal Combustion

1850

1875

1900

1925

1950

1975

2000

Year


by end-use sector



Domestic

Shipping

Transp

Other Process

Smelting

Industry-Comb

Energy-Comb


013

1000013

2000013

3000013

4000013

5000013

185013 190013 195013 200013

Emiss

ions

(Gg

SO2)13

Year13












013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References




















































































013

513

1013

1513

2013

2513

197013 197513 198013 198513 199013 199513 200013 200513

Emiss

ions

Fac

tor (

tSk

t)13

Year13



0

5

10

15

20

25

1970 1980 1990 2000

Em

issi

ons

Fac

tor

(tS

kt)

Year


USA

Canada

Western Europe

Japan

Aus amp NZ

FSU

China+

Middle East

Africa


South amp East Asia

Eastern Europe

Korea

India

Argentina

Brazil

Mexico

Average




0

20000

40000

60000

80000

100000

120000

140000

160000

1900 1925 1950 1975 2000

Em

issi

on

s (G

g S

O2)

Year


Smith etal (2004)

Smith et al (2001)

Oumlm et al (1996)


Spiro et al (1992)

EDGAR 20

Lefohn et al (1999)

Edgar 32

SRES

Stern (2006)

Cofala et al (2007)

Current

EDGAR 41













43 Uncertainty







5 Future work















Edited by R Harley

References


























































































43 Uncertainty







5 Future work















Edited by R Harley

References




















































































5 Future work















Edited by R Harley

References





















































































Edited by R Harley

References


































































































































































Documents

Anthropogenic sulfur dioxide emissions: 1850–2005