Distr.

GENERAL

FCCC/SBSTA/1998/7

4 September 1998

ENGLISH ONLY

SUBSIDIARY BODY FOR SCIENTIFIC AND TECHNOLOGICAL ADVICE

Ninth session

Buenos Aires, 2-13 November 1998

Item 4 of the provisional agenda

Paragraphs Page

I. INTRODUCTION 1 - 10 3

A. Mandate 1 - 3 3

B. Scope of the note 4 - 10 3

II. ISSUES RELATED TO FLEXIBILITY 11 - 34 5

A. Methods and emission factors used by Parties 12 - 15 5

B. Impact of the choice of methods, emission factors and

activity data on emission estimates 16 - 23 6

C. Recalculation of the base year and subsequent inventories 24 - 34 8

GE.98-

Paragraph Page

III. ISSUES RELATED TO UNCERTAINTIES 35 - 56 11

A. Changes in the relative shares of different greenhouse gases

in national aggregated CO₂ equivalent emissions over time 36 - 39 11

B. Uncertainties associated with GHG emissions and

inventories over time 40 - 56 12

IV. ISSUES RELATED TO REPORTING 57 - 84 17

A. Completeness 58 - 60 17

B. Transparency 61 - 71 17

C. Relative importance of particular GHG emissions

from different source categories 72 - 76 20

D. Reporting of HFC, PFC and SF₆ emissions 77 - 81 21

E. Reporting of international bunker emissions 82 - 84 22

Tables 23 - 52

1. The Subsidiary Body for Scientific and Technological Advice (SBSTA), at its seventh and eighth sessions, requested the secretariat to prepare, for consideration at its ninth session, a number of documents on methodological issues identified by the secretariat while processing national greenhouse gas (GHG) inventories from Annex I Parties and in the course of

in-depth-reviews, taking into account the submissions by Parties (FCCC/SBSTA/1997/14,

para. 16 (c) and FCCC/SBSTA/1998/6, para. 40 (a)).

2. The SBSTA, at its eighth session, also requested the secretariat to organize a workshop with participation of experts from the roster, as well as from other relevant organizations, to develop proposals to resolve the methodological issues identified by Parties and the secretariat. The conclusion of such a workshop should be available for the tenth session of the SBSTA (FCCC/SBSTA/1998/6, para. 40 (d)).

3. Also at its eighth session, the SBSTA urged Parties to participate actively in the ongoing activities of the current programme of work on methodologies related to GHG inventories, bearing in mind their relationship with possible additions and/or amendments to the revised guidelines for the preparation of national communications by Annex I Parties, and the

longer-term methodological needs of the Kyoto Protocol, inter alia, the development of guidelines for national systems and adjustments under Article 5 of that Protocol (FCCC/SBSTA/1998/6, para. 40 (b)).

4. In response to the above-mentioned mandate, this report provides data based on the analysis of national GHG inventories and information from in-depth-reviews. Parties might also wish to consider document FCCC/SBSTA/1998/8 which identifies policy options based on this document, as well as on views submitted by Parties. The report is intended to be considered in detail during a workshop to be organized by the secretariat on 9-11 December 1998 in Bonn (see para. 2 above). It is also intended for use by the SBSTA as it considers possible additions and/or amendments to the UNFCCC guidelines for the preparation of national communications by Annex I Parties. The issues identified in this report are also relevant to the preparatory work for the first session of the Conference of the Parties (COP) serving as the meeting of the Parties to the Kyoto Protocol (COP/MOP). For example, the SBSTA may find the information helpful when it considers guidelines for national systems to account for greenhouse gas emissions by sources and removals by sinks.

5. The methodological issues identified while processing greenhouse gas inventories contained in national communications were previously analysed and presented in the compilation and synthesis documents prepared by the secretariat (A/AC.237/81; FCCC/CP/1996/12 and Add.1, and FCCC/SBI/1997/19). This report has been prepared to respond to the mandate

mentioned above, but should be considered as an integral part of the compilation and synthesis of the second national communications (FCCC/CP/1998/2).

6. This document is based on the GHG inventory data included in the national communications from 34 countries, among them all Annex I Parties which submitted their second national communications before 1 August 1998.⁽¹⁾ The report also considers GHG inventory data submitted separately from the communications of Australia, Bulgaria, Canada, the Czech Republic, Finland, France, Hungary, the Netherlands, New Zealand, Norway, Switzerland, the United Kingdom and the United States of America.

7. This document analyses how Parties followed the Intergovernmental Panel on

Climate Change (IPCC) Guidelines for national GHG inventories, and the revised guidelines for the preparation of national communications from Annex I Parties (decision 9/CP.2, annex, FCCC/CP/1996/15/Add.1), referred to below as the UNFCCC guidelines. The analysis of the inventory data was done according to the sectors, subsectors and source categories of GHG emissions corresponding to the Revised 1996 IPCC Guidelines for national inventories referred to below as the IPCC Guidelines. These sectors, subsectors and source categories are identified in italics in the narrative text.

8. Some of the analysis included in the report was carried out also taking into account possible ways in which inventory data could be used under the requirements derived from the Kyoto Protocol. For example, to assess changes in aggregated GHG emissions, the mix of gases and the assumed uncertainties in relation to the base year, the average of the five-year time period 1991-1995 was used. For the sake of brevity, this report sometimes refers to aggregated GHG emissions expressed in CO₂ equivalent as a "basket".

9. In the light of the ongoing work on methodological issues related to the estimation and reporting of emissions by sources and removals by sinks from the land-use change and forestry sector, the report does not provide information on emissions or removals from this sector. Totals, percentages and shares estimated or presented in the report exclude emissions or removals from this IPCC sector, its subsectors and categories.

10. This report's illustrative tables are placed at the end of the narrative text. Information sources are provided as well as explanatory notes for clarity as to how the information was processed. In some cases, these explanations are relevant for a full understanding of the narrative text. Some figures in the tables may differ from those submitted to the secretariat as a result of rounding and owing to the conversion of some reported estimates (for consistency and comparability) to carbon dioxide-equivalent units. In all cases the IPCC 1995 global warming potentials (GWP) with 100-year horizons were used.

11. This report provides a preliminary analysis of the extent to which Parties have chosen to use different methods, emission factors and other assumptions in preparing their inventories, based on the available information. It also provides insights into whether the current flexible approach makes a significant difference in emission estimates. A detailed consideration of these issues, as well as their implications for meeting emissions limitation or reduction objectives, by experts, Parties and the IPCC is necessary. The presentation of more complete information on national inventory data by most Parties may also be necessary for such a consideration.

12. The IPCC Guidelines, developed for a wide range of users, allow Parties great flexibility in estimating their GHG inventories. Parties may use default methods or more advanced methods, either from the IPCC Guidelines, which provide methods with different levels of complexity (tiers), or from national or other compatible methodologies, such as CORINAIR⁽³⁾. The choice of emission factors also is flexible. The IPCC Guidelines provide default emission factors, but encourage the use of national factors when more appropriate.

13. All 34 Parties whose GHG inventories were analysed by the secretariat declared that they had followed or used the IPCC Guidelines to prepare them, but the approach they used to estimate the inventories varied widely. Parties used either default methods or more advanced methods.⁽⁴⁾ Many Parties used either default emission factors, or emission factors developed on their own, or both in different source-categories. Nine Parties used CORINAIR for compiling their GHG inventory and reported them using the IPCC reporting tables. All these different approaches are compatible with the IPCC Guidelines. To illustrate the multiple approaches which Parties can use to estimate their emissions in each different sector of their GHG inventories, the following simplified scheme is provided:

EMISSION FACTORS USED	METHODS USED
	Tier 1 or other IPCC basic methods	Tier 2/3 or other IPCC advanced methods	National or other compatible methods
Defaults (IPCC or CORINAIR)
Default + national
National

14. Information provided in table 1 (page 23) clearly indicates that Parties differed in their use of tiers and/or methods to estimate their GHG emissions. Table 2 (page 24) shows that the types of emission factors used by Parties also vary widely. In many cases, because of the lack of information in the national communications, it is not possible to assess either what tier and/or what method or type of emission factors were used. The information provided in both tables indicates that the different approaches described in paragraph 12 above are possible. Most Parties did not provide complete information on what approaches they used.

15. The diversity of methods and emission factors used by Parties reflects different levels of dissagregation and data availability when preparing their national GHG inventories. The IPCC Guidelines have demonstrated their usefulness, enabling Parties to provide inventory data in most source categories and to report the results in a common reporting framework.

on emission estimates

16. To understand whether the current flexible approach in choosing methods, emission factors and assumptions could make a significant difference to a Party's aggregated carbon dioxide (CO₂) equivalent emissions, it is necessary to compare complete GHG inventories prepared using default methodologies and emission factors with those prepared using national assumptions. Since the information available to the secretariat is insufficient for completely reconstructing GHG inventories, this comparison is not possible for most Parties. However, it is possible to identify many examples, which indicate that different approaches could lead to significant differences in emission estimates for particular GHG source categories. If these emissions are large, this could also affect the annual aggregated GHG emission estimates.

17. The use of different methods and/or tiers does not always lead to significant differences in emission estimates. The comparison between the CO₂ fuel combustion emissions estimated using the reference approach (tier 1) with those estimated using detailed methods (tiers 2 and 3) presented in table 3 (page 25) demonstrates the usefulness of this self-verification procedure. Except for Bulgaria, which reported problems with fuel export statistics after 1993, the differences between the two approaches ranged from 0.3 to 4 per cent. Australia, Bulgaria, Germany and the United Kingdom provided data on comparisons for several years while Finland and the United States of America provided data for a single year.

18. For CO₂ fuel combustion emissions, the highly aggregated IPCC tier 1, which is a top-down approach, may in some cases provide more accurate estimates than the disaggregated tiers 2 and 3, which are bottom-up approaches. This is possible in the case of CO₂ fuel combustion emissions because emission factors do not vary much in the sector and because total activity is readily available from national statistics. A disaggregated bottom-up approach could provide more accurate estimates only if reliable data are available for emission factors and activity data in all fuel combustion subsectors. A comparison of the two methods allows for the identification of possible information gaps when the causes of numerical differences among methods are investigated. Parties are requested by the IPCC Guidelines to compare data obtained using both approaches and to explain possible differences. The UNFCCC guidelines request the presentation of the worksheets of the IPCC reference approach but do not request the explanation. An accurate estimation of CO₂ fuel combustion emissions is of fundamental importance, because it represents the largest share of emissions for all Annex I Parties.

19. In its second national communication, Sweden reported emissions using an improved method to calculate CO₂ emissions from road transport based on more disaggregated data. This led to a decrease of 24 per cent in the transport sector emission estimate compared to the estimate in the first national communication. Because of the importance of this sector, this change led to a 10.5 per cent reduction of the base year CO₂ equivalent emissions (see table 11, page 30) in its second national communication. This result provides an example of the importance of comparing the aggregate fuel combustion CO₂ emission estimates of different sectors using detailed technology-based approaches with the aggregated IPCC reference approach. Some Parties which recalculated their original 1990 CO₂ emissions estimates would have detected the problems earlier if they had made this comparison.

20. The data provided in tables 4, 5, and 6 (pages 26-27) demonstrate how the use of different tiers and/or methods affects estimates of sectoral emissions. Table 4 suggests that the use of tier 1, which uses default emission factors, could overestimate emissions, since the use of Czech methods results in methane (CH₄) emissions 31 per cent lower for enteric fermentation and 10 per cent lower for manure management. The information in table 5 shows that using more advanced methods (dynamic or kinetic), compared to basic methods, in the waste sector could give significantly different estimates compared to using default ones in two countries. Table 6 indicates that using the Revised 1996 IPCC Guidelines methodology to estimate

nitrous oxide (N₂O) emissions from agricultural soils could make a significant difference in relation to the 1995 version, owing to the inclusion of new sources. Values presented indicate that the range of this difference varied widely among Parties, from plus 74 per cent for Denmark to minus 45 per cent for Switzerland. The emissions grew for all cases except Switzerland as a consequence of the inclusion of new sources in the estimate. The Swiss authorities explained during the in-depth-review that their emissions decreased, despite using the new guidelines, because the emission factors used were revised.

21. The use of different emission factors could also lead to significant differences in the GHG emissions among sectors. Information provided in tables 7 and 8 (pages 28 and 27) give examples of the wide variations existing between country-specific and IPCC default emission factors. As shown in table 7, in general solid fuel emission factors vary the most, for example lignite (49 per cent), peat (17 per cent) and coke (15 per cent). Differences are generally much greater for sectors with greater uncertainties, such as N₂O from agricultural soils, as is evident from table 8, where a variation of 100 per cent is shown. Table 9 (page 29) shows how different values of emission factors could lead to different CO₂ emission estimates, using emissions from cement production as an example. In three cases the differences between the estimates are negligible, but in the other two cases the differences range from 12 to 18 per cent.

22. It should be noted that the real effect of a given emission factor on the amount of national aggregated GHG emissions depends on the relative share of the emissions estimated with that emission factor. A difference of 100 per cent between values of the emission factors in one sector, for example those in table 8, could have less of an effect on the total GHG estimates than the smaller differences shown in tables 7 and 9.

23. Differences in the gathering of activity data and in the utilization of the data result in disparate emission estimates. Quality of the activity data or the assumptions used for their application also provoke different results of emission estimates. For example, information provided by the Czech Republic indicates that they found differences between activity data from national statistics and sectoral statistics close to 10 per cent for coke, iron and cement production. Finland reported another example of how different national organizations had provided different values of activity data for consumption of hard coal, peat and black liquor, with differences of

6-7 per cent. Information in table 10 (page 29) demonstrates that the way in which such data are used (average of several years or a single year value) also influences these estimates.

24. All Parties which submitted a second national communication had recalculated their base year inventories and the subsequent inventories, except for two, as shown in table 11

(page 30).⁽⁵⁾ These two Parties have not presented a recalculated figure for their base year in their national communications, although they used updated methods/data for subsequent years. The estimation of GHG emissions for subsequent years using methods, emission factors and activity data other than those which were used for the base year inventory, affects the comparison of the target and base year figures.

25. The differences between the estimates in the first and second national communications were influenced by two factors:

(a) Replacement of the IPCC 1994 global warming potentials used in the first national communications, by the IPCC 1995 GWP in second communications. This factor influences aggregated GHG estimates in terms of CO₂ equivalent, but not the recalculations made on a gas-by-gas basis;

(b) Changes in methods, emission factors and assumptions, as well as an update of activity data and the inclusion of new sources of emissions. These changes, which are referred to in this document as "change in methods/data," are encouraged by the IPCC Guidelines aimed at improving the quality and accuracy of the inventories, and at using more appropriate data.

26. The use of different GWP values, and any associated changes in GHG emission estimates, can be avoided by fixing the GWP values in a given period of inventory date. The Conference of the Parties at its third session decided that Parties should use the IPCC 1995 GWP with a 100-year time horizon (FCCC/CP/1997/7/Add.1, decision 2/CP.3, para. 3). This decision does not imply a change of the necessity of reporting inventories on a gas-by-gas basis, as was decided by the COP at its second session (FCCC/CP/1996/15/Add.1, decision 9/CP.2, annex, para. 11). A similar provision, fixing the GWPs values, has not yet been included by the UNFCCC guidelines.

27. The COP also decided in its decision 2/CP.3 (para. 1) that Parties should use the IPCC Guidelines to estimate and report their emissions. However, owing to the characteristic of these Guidelines, this decision does not imply that Parties should use the same methods for these purposes.

28. The differences between data in the first and second national communications due to changes in methods/data are significant for many countries, as shown in table 11. This is true both for estimates made on a gas-by-gas basis and for those expressed in terms of CO₂ equivalent. The changes on a gas-by-gas basis are generally larger for CH₄ and N₂O than for CO₂, owing to the higher uncertainties inherent in the estimation of their emissions. However, the changes in CO₂ equivalent emissions depend more on the importance of the share of each GHG in the aggregated GHG emissions. Therefore, even minor changes in CO₂ emissions could cause significant changes in CO₂ equivalent emissions, because of the large share of CO₂ in the aggregated emissions. Significant changes in CH₄ and N₂O could cause a similar result.

29. For the most part, changes to the base year were affected by the desire of Parties to calculate their emission estimates more accurately. As methodologies develop, both nationally and internationally, the collection of data improves. Four elements were identified which led to changes in previous estimates submitted by Parties. They are the use of new models or methods for estimating emissions; the application of updated emission factors; the change or updating of activity data; and the change of scope due to the inclusion or exclusion of sources.

30. The application of updated emission factors that reflect actual conditions may or may not lead to the recalculation of the base year and successive years between the two submissions. The emission factors were applied retroactively when this was applicable for the time period, including the base year. However, in some cases new emission factors are applicable only from a given year onwards, when a new pattern of emissions appears as a consequence of technology change. In this case, it is not necessary to recalculate the inventories of those years prior to the change.

31. The recalculation of the base year and successive years was not limited to the inventories contained in first and second national communications. Many Parties recalculated them in each annual submission. Other recalculations were explained during the course of in-depth-reviews or by official letters submitted to the secretariat by Parties. This frequency of changes is a logical consequence of annual updates of inventories carried out in several countries. On the other hand, this high frequency of changes could complicate the assessment of their implications.

32. Inventory data would be totally comparable among countries if all countries estimated their emissions using identical methods. At the moment, there are not two Annex I Parties which use identical methods. A full standardization of inventory methods might ensure truly comparable data, facilitating the review and verification of estimates. However, such an approach ignores the different capacities of countries to prepare inventories. Standardization would affect the necessary upgrading of the inventory quality, and as a consequence, the accuracy of the estimates. A necessary balance between the required accuracy and comparability/consistency of the inventories should be achieved, taking into account the new needs of the Convention process, such as the implications for meeting emissions limitation or reduction objectives.

33. In the event that Parties consider it necessary to recalculate baseline data, this must be carried out in a consistent and transparent way. The SBSTA, at its fourth session, concluded that, to ensure comparability, Parties should recalculate the base year inventory and inventories for any subsequent years when using the 1996 IPCC Revised Guidelines (FCCC/SBSTA/1996/20, para. 30 (a)). However, the SBSTA conclusion does not indicate how to deal with the changes in the estimates introduced for reasons other than the use of the Revised Guidelines, nor does it stipulate what documentation should be provided to ensure transparency in recalculations.

34. Further guidelines for the recalculation of base year estimates, and their revision, may also be necessary. The implications of such recalculations for the comparability and consistency of the data of a given Party over time need to be considered, as well as what process, if any, could be used to review and consider revised inventories. The secretariat notes that heretofore all data have been accepted into its database. The in-depth review process has sought to clarify how GHG inventories were estimated.

35. This report provides a preliminary analysis of the changes in the relative shares of different greenhouse gases in the aggregated GHG emissions and of the changes in uncertainties over time. These changes could be relevant to assessing emission reductions that have actually occurred as well as emission limits, a methodological issue that may be relevant to the

Kyoto Protocol. A detailed consideration of these changes, and their implications, by experts, Parties, and the IPCC is necessary.

aggregated CO₂ equivalent emissions over time

36. Since GHG inventory data are now available for each year between 1990 and 1995 for almost all Annex I Parties, changes in the mix of gases can be assessed over time. The information provided in table 12 (page 32) shows how the proportion of each of the greenhouse gases (CO₂, CH₄, N₂O, hydrofluorocarbons (HFCs), perfluorocarbons (PFCs) and sulphur hexafluoride (SF₆))⁽⁶⁾ within the aggregated GHG emissions for each Party changed between the base year and the average in the period 1991-1995.⁽⁷⁾ It indicates that the mix of gases did not change much, in general, for most Parties within the period 1991 to 1995, compared to 1990. This can be illustrated by reference to CO₂ estimates, which account for the greatest proportion of all Parties' national emissions, with the exception of New Zealand. For CO₂ the mean change between the 1991-1995 average and 1990 was only 2 per cent, with a standard deviation of 3.3. The mode or most frequently reported value of this change was only 0.4 per cent. The median, which is the middle value of the range of values is 1. Six Parties exceeded the mean change. The most significant change was shown by Iceland, whose CO₂ emissions share in its aggregated GHG inventory was 9.2 per cent higher over the period 1991 to 1995 compared to 1990. In principle, a "basket" approach allows the trend of emissions to be assessed with a higher degree of confidence if the mix of gases in the "basket" does not change considerably between the base year and the years included in a given period of time.

37. The current data allow preliminary insights into whether changes in emission estimates over time are due to socio-economic driving factors or to changes in calculation procedures (table 13, page 34). The first two columns show aggregated GHG emissions in the year 1990 and over the period 1991 to 1995 inclusive, using methods applied in the second national communications. Column C shows the percentage increase or decrease between the two, which represents the real increase or decrease of these emissions in the period. This is used as a proxy for the effect of socio-economic changes.

38. Column D shows the estimate of Parties' aggregated greenhouse gas emissions between 1991 and 1995, using methods applied in the first national communications. Given that the first national communications do not, typically, provide data beyond 1993 to arrive at the estimate for the whole period, the same trend was used, year by year, as the one estimated from the data in the second national communication. The percentage difference as a result of this change in methods/data over the period 1991 to 1995 is shown in column E. These results are comparable to the results shown in the last column of table 11 for the recalculation of the base year due to changes in methods/data. It should be observed that the changes of estimates are larger than

5 per cent for 16 of these Parties.

39. Compared to the first national communication, all Parties used different methods/data for at least some and, in several cases, many source categories in compiling their inventories in the second national communications. In other words, emission estimates for the base year and for the years included in a given period of inventory data (e.g. a "commitment period" under the Kyoto Protocol) can change as a result of these changes. As Parties apply the new methods/data for all years, the effect of these changes is not obvious simply from the inventories provided in second national communications. Only some Parties provided detailed information on changes made in relation to previous inventories. The data presented in table 13 for all Parties that recalculated their baseline year were developed using the information provided by Parties in their first national communications. The information contained in the table could be used to form a preliminary assessment of the effect of these changes.

40. Information as to how Parties reported uncertainty estimates by sectors and by gases is given in table 14 (page 36). The table does not indicate uncertainty estimates for HFCs, PFCs or SF₆, as these are only provided by seven Parties. The secretariat, for purposes of consistency, uses the generic term "confidence level" in the table for other descriptions presented by Parties using different terms: "uncertainty", "emissions range", "quality of the estimate" or "error bounds". The reported confidence levels varied widely among Parties, for the same gases in the same sectors. Although information improved compared to the first national communications, it is clear that the information on uncertainties is still not comparable among Parties.

41. Thirteen Parties provided quantitative estimates about uncertainties on a gas and/or sectoral basis. The level of detail of the information about these estimates varied widely among Parties. Fifteen Parties provided the IPCC overview table, which includes a self-assessment of the completeness and quality of their inventories. This format classifies the quality of the estimates as "high"(H) , "medium" (M) and "low" (L). The use of these terms, as well as of the quantitative information mentioned above, is not really compatible among Parties. The approaches used by them differ and in many cases are not explained. Twelve Parties (AUT, DEU, EST, FRA, GRE, IRE, LAT, LIT, LUX, POL, SVN, UKR) either did not provide estimates of uncertainties or did so only partially. Five Parties (AUS, CAN, ESP, GBR, USA) provided a detailed analysis of the assumptions used to estimate uncertainties of several sectors and New Zealand only did so for energy-related CO₂ emissions.

42.  Information in national communications indicates that the ways of reducing, estimating and reporting uncertainties, and therefore also the data quality of inventories, vary among countries. The qualitative and quantitative information made available by Parties is summarized in the following box:

GHG	Confidence level	Remarks
CO2	"High" for fuel combustion and industrial processes.	"High" for energy and industrial processes means estimates which have an error range of less than 10 per cent.
CH4	"Medium" for fugitive fuel emissions.. Predominantly "medium" for fuel combustion, enteric fermentation, animal waste and waste, with some Parties reporting them as "low".	"Medium" for these categories generally means the estimate has an error range of 20 to 50 per cent.
N2O	Predominantly "high" and "medium" for industrial processes, with two Parties reporting them as "low". Predominantly "low" for fuel combustion, with some Parties reporting them as "medium". All Parties reported "low" for agricultural soils.	"Low" for these categories generally means the estimate has an error range of 50 to more than 100 per cent.
Other GHG	"Medium" and "low" for industrial process emissions "Low" for emissions from consumption	Errors were reported with a range of 50 per cent to a factor of 2. USA reported 20 per cent for industrial emissions of HFC-23.

43. The gaps and differences in the information provided by Parties about the uncertainty of GHG emissions indicate the complexity of this issue. Although Parties provided information about uncertainties associated with annual emission estimates there was no discussion about how uncertainties changed over time. The IPCC Guidelines do not provide guidance on how to estimate uncertainties over time, and neither the IPCC nor the UNFCCC guidelines request this information from Parties.

44. The need for technical and scientific work to develop compatible ways to estimate and report uncertainty of the GHG emission estimates and the overall uncertainty associated with GHG inventories is evident. In addition to the uncertainty of an emission estimate, the uncertainty of an emission reduction may also be relevant. The uncertainty of a given emission estimate may have two implicit types of errors: systematic and statistical. The former refers to errors in the methods used, and the latter to errors created by imprecise measurements or estimates. In principle, the systematic error may be cancelled out when two emission estimates are compared if the same methods are used over time, but this is not the case for statistical errors. The assessment of the extent to which systematic errors are cancelled in a "basket" approach and the degree to which statistical errors vary over time requires additional technical and scientific work.

45. To provide an insight into the influence of changing uncertainty over time, a simple method was used with the sole purpose of assessing the changes of overall uncertainty associated with national GHG inventories over time. The derived uncertainty values presented in table 15 (page 38) are not true estimates of the uncertainties associated with national GHG inventories but instead are used to illustrate the effect of changes over time.

46. To estimate the true overall uncertainty associated with a GHG inventory it would be necessary to know the inherent quantitative uncertainty for every particular GHG source category obtained from the uncertainty of the data and the associated calculations. Uncertainties for each GHG source category differ as do their share in the aggregated GHG emissions and this should be taken into account. Therefore, overall uncertainty associated with national inventories should be based on the uncertainty of each particular GHG source category emissions weighted by its relative contribution to total emissions. Insufficient data have been provided by Parties to carry out such an exercise. Even though some countries have provided relevant information it is not possible to do comparative analyses.

47. In order to approximate the overall uncertainty associated with each Party's inventory, some simplifying assumptions were made. An uncertainty factor was attributed to the totals of each main GHG emission. The uncertainty factors were approximated on the basis of the summary information presented in table 14 and summarized in paragraph 40. In aggregate, the uncertainty associated with CO₂ was +/- 7 per cent, with CH₄ +/- 35 per cent and with N₂O +/- 80 per cent. For example, this implies that, on average, a Party's CO₂ estimate may be 7 per cent higher or lower than stated. HFCs, PFCs and SF₆ were not considered, as the reporting of these emissions is inconsistent among Parties.

48. The relative shares of each GHG for a given Party presented in table 12 were multiplied by the corresponding uncertainty factors described above, equally for all Parties. The approximated overall uncertainties associated with national greenhouse gas inventories obtained by this approach are presented in the table for the base year, for every year within the period 1991-1995 and for the average of these years. With this approach the only element influencing the results is the differing shares of gases in Parties' aggregated GHG emissions. Though this can be an important source of uncertainty, it is not the only one. This approach does not take into account differences in the quality of GHG inventories between Parties or efforts made by Parties to reduce uncertainties in the period. According to this unique factor for determining uncertainty, the larger the share of CO₂ in the total, the lower the overall uncertainty of the annual inventory, given that this gas has the highest confidence level.

49. The overall approximated uncertainty in the period 1991 to 1995, inclusive, for each Party, can be compared to the approximated uncertainty associated with its inventory in 1990. Any changes will have been due to a change in the composition of the basket. It can be observed from table 15 that the change in uncertainty over time has been considerably less than its absolute level in any particular year for all Parties. Overall, the difference between uncertainties in the period 1991 to 1995 compared to 1990 ranged from 0.5 to 14 per cent. The mean difference was 3.2 per cent with a standard deviation of 4.4, a median of 2.0 and a mode of 0.8.

50. Taking into account that emission estimates throughout the analysed period are calculated using the same methodology, one way to tentatively interpret the results presented in table 15 could be as follows: the change in the approximated overall uncertainty of GHG emissions in the period 1991 to 1995, compared to the base year reflects the extent to which a systematic error remains, or is not cancelled, over time as consequence of using a "basket" approach. This percentage is reflected in column H of the table. For most Parties, as described in paragraph 47 above, this percentage is relatively small. It should be noted that the effect of the uncertainty associated with a reduction of emissions in a given period compared to the base year requires detailed analysis and is beyond the scope of this report.

51. Information as to how changes in the methods/data employed in the first and second national communications could affect, annually and over time, approximated overall uncertainties is provided in table 16 (page 40). The figures have been adjusted to take into account the change in GWP used in the first as compared to the second national communications. The differences of percentages of the approximated overall uncertainty associated with 1990 data, as reported in the second national communications, compared to that reported in the first national communications is presented in column H. The mean of the difference was 1.9 per cent. Similarly, as indicated in column I, the uncertainty associated with data for the period 1991 to 1995 inclusive differed by only 2 per cent, on average, although the standard deviation was somewhat greater at 3.6 compared to 3.2 for the 1990 data. In both cases the mode was 0.3

per cent. The median for 1990 was 1.05 and for 1991-1995 it was 0.9. For 13 Parties, these differences in uncertainty were more than 1 per cent.

52. An attempt is made to assess the impact of changed methodologies and/or data on the amount of uncertainty associated with the inventories over time. Table 16 shows the percentage change in the amount of uncertainty associated with each Party's inventory in 1990, compared to the period 1991 to 1995, inclusive. This is done for uncertainty estimates based on the first national communication (column C) and separately for uncertainty estimates based on the second national communication (column F). Any changes in the composition of the "basket" reported for any particular year will affect the uncertainty attributed to a Party's inventory. For some Parties the share of CO₂ in their total inventory increased and therefore the uncertainty was reduced over time, whereas for others the opposite is true. These changes should not be interpreted as an improvement or deterioration of the quality of the inventory of a given Party over time.

53. The difference between the percentage change over time based on first national communications and that based on second national communications is a consequence of changes in methods/data employed in both communications. Subtracting these two percentages (column F minus column C) results in a difference that is shown in column G of table 16. It can be considered as the change of uncertainty associated with the inventories over time as a consequence of changes in methods/data used. The mean of this difference is 0.77 with a standard deviation of 1.5. The mode is only 0.1 per cent. The median for this column is 0.2. For seven Parties this difference exceeded one percentage point.

54. Clear guidance on how to estimate and consider the influence of changes in methods/data over time is required. This can be illustrated by implications derived from the results of table 16 other than those described in paragraphs 51 and 52. These results could be tentatively interpreted, among other possible interpretations, in the following way:

(a) It is fair to assume that methodologies and data improved between the first and second national communications. If so, then changes in uncertainty between the communications (difference presented in column G) are a reflection of this improvement; and

(b) Variations in methods/data used can limit comparability of data over time and, perhaps, introduce an additional source of uncertainty. Although unlikely, the existing flexibility in choosing methods allows for a selection of methods that favours beneficial performance in any particular period. Furthermore, favourable adjustments to the 1990 base year could also be made as a result of methodological "improvements".

55. A more detailed analysis of the overall uncertainty associated with national inventories can be seen in table 17 (page 42). The analysis is based on the quantitative gas-by-gas sectoral uncertainty estimates provided by five Parties. It should be more accurate than the simplified approach used above. For the sake of comparison, approximated overall uncertainties associated with national inventories are also calculated using the illustrative uncertainty factors provided as an example in the reporting instructions of the IPCC Guidelines. As is to be expected, the divergence in results is large, as well as from the results of approximated uncertainty presented in table 15. It is clear that the overall uncertainty associated with an emissions inventory is highly dependent on the underlying uncertainty factors for each gas, by emissions source.

56. Table 17 suggests that the mix of gases is more important than the quality of national inventory compilation in determining the overall level of uncertainty associated with a given inventory. This is shown using New Zealand as an example, which has a well developed national inventory system yet, a higher overall level of uncertainty, despite the fact that its own uncertainty factors were used for the estimation of its GHG inventory uncertainty. Finally, it can also be observed that differences in overall uncertainties between 1990 and the period 1991 to 1995, inclusive, are considerably less than the annual level of uncertainty associated with the inventory, regardless of what uncertainty factors are used to make the estimations.

57. In general, the quality of the inventory data presented in second national communications was higher than in the first communications, but many problems remain which hamper reporting GHG inventories in a transparent, complete and consistent way. This will require an improvement and expansion of the current reporting practices and procedures for meeting emissions limitation or reduction objectives.

58. Information about how many Annex I Parties report data on different IPCC source categories is provided in table 18 (page 43). The degree of completeness in reporting varies widely. Nevertheless, all or almost all Parties reported the most significant particular GHG emissions (see table 22, page 47), such as CO₂ emissions from fuel combustion and industrial processes, CH₄ emissions from enteric fermentation and waste and N₂O emissions from agricultural soils and fuel combustion. There was little reporting of HFC, PFC and SF₆ emissions, but the number of reporting Parties increased for the year 1995. However, the percentage of reporting Parties is still low (53 for HFCs and PFCs, and 50 for SF₆).

59. It should be noted that some Parties may not have data on particular source category emissions. Many Parties did not report whether missing emissions were "not estimated," "not occurring" or "included in other sectors." Some Parties presented aggregated emission figures for broad IPCC sectors only, such as for industrial processes and fugitive emissions, and did not disaggregate their emissions on subsectors and source categories as is requested by the IPCC Guidelines. This affects the transparency and comparability of the data.

60. A comparison of the completeness of GHG emission reporting by Annex II Parties and Parties with economies in transition (EITs) is shown in table 19 (page 44). It indicates a lesser degree of completeness and dissagregation for the latter group of Parties in most subsectors. The main problems identified for EITs relate to the reporting of HFC, PFC and SF₆ emissions

(see paras. 77-81) and bunkers (see paras. 82- 84), emissions from industrial processes and the dissagregation of the emissions by subsectors in the fuel combustion sector. This may have implications for the Kyoto Protocol.

61. In order to ensure transparency, Parties were requested to provide sufficient information for reconstructing inventories from national activity data, emission factors and other assumptions. Table 20 (page 45) indicates how Parties followed key elements of this request.

62. Only 18 Parties provided the IPCC standard data tables, while Ireland presented these tables for 1993 only. The UNFCCC guidelines request that Parties provide these tables. The IPCC standard data tables do not provide the level of detail necessary to reconstruct inventories in all sectors. For this reason the Revised 1996 IPCC Guidelines request countries to submit the worksheets or equivalent information for all sectors instead of the IPCC standard data tables. The UNFCCC guidelines also request, in addition to the IPCC standard data tables, the provision of the worksheets for fuel combustion, agricultural soils and the land-use change and forestry sector.⁽⁸⁾ However, only Bulgaria, Germany, Hungary, Slovakia and the United Kingdom submitted the worksheets, and in all cases only for the fuel combustion sector.

63. There are many reasons why the worksheets used to estimate GHG emissions are not presented. These include the different level of dissagregation of the national inventories, the use of different methods to prepare inventories which do not require worksheets, the large quantity of worksheets, the preparation of worksheets in national languages, and the undefined nature of the terms "worksheet" and "equivalent information", among other possible factors. The level of additional information provided by Parties varies considerably and in general the information was not sufficient for reconstructing the inventories.

64. The usefulness of the worksheets using the IPCC reference approach for CO₂ fuel combustion emissions was discussed in paragraphs 17-19. The importance of these worksheets for verification and self-verification purposes is directly related to the importance of CO₂ fuel combustion in the aggregated GHG emissions. They constituted 50 per cent or more of the aggregated CO₂ equivalent emissions for all Annex I Parties, except New Zealand. For 18 Annex I Parties this share exceeded 75 per cent (see table 22, page 47). It should be noted that in order to carry out a precise comparison between top-down and bottom-up methods it is necessary to have compatible estimations of the CO₂ non-energy use, or feedstocks. In addition, how they were allocated in the different IPCC sectors and source categories must be considered. Feedstocks could constitute close to 10 per cent of the total CO₂ fuel combustion emissions of many Parties. Only 10 out of 34 Parties provided a description of how feedstocks were considered in their inventories.

65. Although the standard data tables of the 1995 version of the IPCC Guidelines do not provide the level of detail necessary to enable the reconstruction of an inventory, they provide information on the aggregated activity data and emission factors used. They serve the purpose of comparing among Parties and checking transparency, and completeness. Undoubtedly, the 18 Parties which provided these tables presented a more transparent inventory than those which did not. Only 15 Parties provided the overview table requested by the IPCC Guidelines, which is part of the reporting format of both the 1995 and the 1996 IPCC Guidelines. This table is very useful in an evaluation of the completeness of an inventory, because it provides the information on the gaps referred to in paragraph 59.

66. All reporting Parties except Canada and Ireland reported nitrogen oxides (NO_x), carbon monoxide (CO) and non-methane volatile organic compounds (NMVOC), known as precursors, as requested by IPCC and UNFCCC guidelines. Sixteen Parties optionally reported sulphur dioxide (SO₂) emissions, which is encouraged by the revised versions of both guidelines. These gases are important radiative constituents in the atmosphere, influencing the radiative forcing or more simply global warming rates. The calculation of specific GWPs for precursors is not currently possible because of the inadequate characterization of many of the atmospheric processes involved. It should be noted that Annex I Parties except Australia, Japan and

New Zealand are Parties to the Convention on Long-range Transboundary Air Pollution, which has adopted five protocols. Parties to the Convention and its protocols systematically report the emissions of these gases and have programmes for their reduction.

67. Twenty-seven Parties reported their emissions in terms of CO₂ equivalents. All of these Parties used the IPCC 1995 GWP 100-year horizon, except Lithuania which used an old IPCC GWP. The reporting of GHG emissions in terms of CO₂ equivalent is not required by the IPCC Guidelines and is optional under the UNFCCC guidelines. The latter do not provide guidance on whether the GHG emissions expressed in CO₂ equivalent should be provided at a sectoral or at national level, or on the format (e.g. tables) to be used to present the data. The reporting of CO₂ equivalent emissions should not affect the reporting on a gas-by-gas basis using mass units which is essential for purposes of transparency, as requested by the UNFCCC guidelines.

68. Most Parties did not specifically indicate what IPCC tier or own method they used to estimate their emissions. Nine Parties (AUT, BEL⁽⁹⁾, CHE, ESP, FRA, ITA, IRE, LUX, POR) used CORINAIR for compiling their GHG inventory. This is a bottom-up method which is built on a detailed source category split. The quality of the CORINAIR to IPCC conversion seems better in the second national communications than in the first ones, which reflects the progress achieved in the harmonization of these two methodologies. However, the information provided by the Parties using CORINAIR to back up their inventory data differed very much among them and, in general, was not sufficient for reconstructing inventories. This situation is common for most reporting Parties, not only for those that used CORINAIR.

69. In general the documentation on emission factors provided by Parties is poor

(see table 21, page 46). Only a few Parties presented a detailed explanation of why they chose country-specific emission factors to estimate their GHG emissions. Many Parties only provided comments in this regard or referred the reader to several sources, which in many cases are in their national languages. Many Parties did not provide any information.

70. The information in table 21 also indicates that many Parties did not provide any numerical values for the emission factors. Many of them only provided aggregated emission factors at the IPCC sectors level, mainly through the IPCC standard data tables. Few Parties presented more disaggregated information, identifying values of specific emission factors which were applied in key sectors. The level of documentation varies widely among Parties. In general, the information presented underlines the fact that the instructions provided in the guidelines for documenting the emission factors used by Parties are vague and need to be improved.

71. Supporting documents with detailed data on activity statistics, emission factors and methods used could potentially lead to more consistent and transparent inventories. The UNFCCC guidelines request this background documentation without a clear definition of the specific information to be provided. This definition should be improved in the future. For example, the GHG inventory from Spain used 29,738 individual activity data and 2,555 emission factors, 1,225 of them for the estimation of emissions of the main GHGs. The French GHG inventory is built on a detailed source split of more than 500 items that could lead to a larger quantity of emission factors than the previous example. The emission registration system used by the Netherlands to compile its GHG inventory, which includes many point sources, aggregated over 400 compounds in hundreds of subcategories. Therefore, the provision of all basic background data is virtually impossible. Clear guidance as to what background information should be provided under the Convention and Kyoto Protocol is needed.

72. The aggregated GHG emissions of a given Party is the sum of particular GHG emissions from source categories in the different sectors and subsectors of the IPCC Guidelines. Emissions are estimated on a gas-by-gas basis, because each GHG in each source category requires specific emission factors. In practice, the particular GHG emission estimates from source categories act as individual "blocks" in "building" the inventory. The relative importance of each of these particular GHG emissions can be estimated by considering the percentage share of these emissions expressed in CO₂ equivalent related to the aggregated GHG inventory, expressed in the same units. These data are provided in table 22 (page 47) for the largest, referred to here as the "top," particular GHG emissions for most Parties.

73. This table demonstrates that the relative importance of each particular GHG from source categories varies among Parties, reflecting the differentiated structure of their economies. The top five particular GHG emissions for each Party are shown in bold.

74. The information in table 23 (page 49) indicates the percentage shares of the "top" five, seven, or 10 particular GHG emissions from different source categories of a Party's aggregated GHG emissions inventory. It can be observed that for 90 per cent of the reporting Parties, 30 out of 34, more than 90 per cent of their aggregated GHG emissions is attributable to their five top particular GHG emissions, more than 94 per cent is attributable to their seven top GHG particular emissions and more than 96 per cent of their emissions is accounted for by their 10 top particular GHG emissions. It should be noted that the relative importance of particular GHG emissions from different source categories in the inventory differs by country.

75. The current IPCC sectors and subsectors have at least 60 particular GHG emissions from source categories for which emission factors and activity data are required to estimate emissions. The UNFCCC guidelines request the inclusion of sufficient information to back up the data presented. Given that a limited number of particular GHG emissions from different source categories in each country are responsible for the bulk of emissions, as shown in table 23, emphasis could be placed on giving detailed information on only these emissions for each Party.

76. The extent to which different sources appear in the top five, seven and 10 particular GHG emissions from different source categories of the Parties is provided in table 24 (page 50). The table provides evidence that for most Parties, CO₂ from fuel combustion and industrial processes, CH₄ from enteric fermentation and solid waste and N₂O from agricultural soils and fuel combustion (mainly transport) are among the most common top particular GHG sources.

77. Reporting emissions of HFCs, PFCs and SF₆ on a mandatory basis is a new requirement of the UNFCCC guidelines. Only 21 Parties provided data but not all of them reported the three types of gases. The quality of the data provided by the 21 Parties varies widely, and in general data are incomplete. Of the 13 Parties which did not report these emissions at all, some indicated that emissions of one or more of these gases are negligible. Many of them, mainly those that are undergoing the process of transition to a market economy, reported that they do not have systems in place to estimate these emissions. Only some indicated that they will provide these data in future inventories. Methods for estimating emissions of these gases were included for the first time in the Revised 1996 IPCC Guidelines.

78. The differences between GWP values for different HFC and PFC species are large. Five Parties (BEL, CHE, FIN, ICE, NZL) only reported aggregated figures for HFCs and for PFCs. The UNFCCC guidelines do not provide clear guidance on reporting emissions of these substances in a disaggregated way, that is by the different types of chemicals. Another inconsistency in the reporting mentioned by some Parties is that the HFC, PFC and SF₆ species contained in imported or exported equipment are not reported in the inventories of these emissions.

79. Information on the reporting of actual and potential emissions for these gases is provided in table 25 (page 51). Seven out of 21 Parties which reported HFC emissions used an actual approach. Among them, three (Canada, Denmark and the United Kingdom) also reported estimates using a potential approach. Eight Parties used a potential approach and it is not clear what approach was used by the other five Parties. The reporting of PFCs and SF₆ contained similar problems. Decision 2/CP.3 affirms that actual emissions of these gases should be used for the reporting of these emissions. The SBSTA, at its fourth session, encouraged Parties to report both actual and potential emissions (FCCC/SBSTA/1996/20, para. 31).

80. The differences between the values of emission factors based on different rates of release to the atmosphere are also large. The lower portion of table 25 provides the ratio of potential to actual emissions for those six Parties which estimated both of these for HFCs or SF₆. This ratio is based on the estimated leakage rates of these GHGs in different applications. The reported differences are large, for example 16:1 in Canada, 6:1 in Denmark and 2:1 in the Netherlands. Therefore, estimates for actual and potential emissions can also vary to a significant degree, depending on the mix of different types of HFCs and PFCs. These large differences have implications for estimating the amount of the aggregated GHG emissions of a given Party.

81. The draft of the 1998 United States inventory points out the usefulness of examining atmospheric concentrations in this regard. The actual increase in the atmospheric concentration of SF₆ shows a higher accumulation than could be expected with the leakage rate of one per cent annually claimed by most manufacturers of circuit-breakers and gas-insulated substations. It was assumed that roughly three-quarters of SF₆ production is used in electric equipment. This example suggests the feasibility of using the measurement of the atmospheric concentrations of these gases as a tool to identify real ratios between potential and actual emissions. Also, the example indicates the usefulness of reporting emissions using both approaches, because each of them provides useful complementary information. However, even this approach may not suffice for policy-making related to specific sources and more information may be needed.

82. Bunker emissions were reported separately from fuel combustion emissions by only 27 Parties. The Czech Republic, Estonia, Latvia, Lithuania, Slovakia, Slovenia and Ukraine did not provide this information. The information is requested by the IPCC and the UNFCCC guidelines. Some of these Parties explained that either these emissions are not significant to their aggregated emissions, or their national statistics are currently not able to make this differentiation. The emission share of the bunker emissions to the total GHG emissions of reporting Parties range from 0.1 (United States) to 19 per cent (Netherlands), as shown in

table 26 (page 52).

83. Fourteen out of the 27 Parties which reported these emissions separated them into marine and aviation bunkers in their national communications or supporting materials. This separation is requested by the UNFCCC guidelines and COP decision 2/CP.3. The method used to estimate bunkers is also not consistent among Parties. For example, Switzerland only estimated bunker emissions which are emitted from the national territory, whereas Canada and Portugal estimated emissions based on fuel sales to air and marine vessels of foreign registration. Iceland and Denmark estimated emissions on the basis of fuel sold, but not specifying to whom. The UNFCCC guidelines include the same wording for reporting fuel sold without requesting the specification as to whom the fuel was to be sold to. Finland developed its own model to estimate aviation emissions, including international aviation emissions. Most Parties did not specify what methods they used. The impact of different methods on the bunker emission estimate has not been assessed.⁽¹⁰⁾

84. Eleven Parties reported bunker emissions of CO₂ only. For 13 out of 15 Parties which also reported CH₄ and N₂O bunker emissions the share of the CO₂ emissions in the aggregated GHG bunker emissions is higher than 98 per cent. For Finland it is 89 per cent, but the reasons for this lower figure were not explained. The share was not estimated for Canada as this Party reported only aggregated bunker emissions in terms of CO₂ equivalent.

Table 1. Tiers and methods used by Parties for estimating GHG emissions

IPCC categories	Tier 1	Tier 2	Tier 3	Use of different tiers	No information provided
	Number of reporting Partiesa)
ENERGY:
Stationary combustion	7	15		10
Mobile combustion	3	9	11	7	2
Fugitive solid fuel emissions	12	4	3	1	4
Fugitive oil and gas emissions	13	4	5	1	5
AGRICULTURE:
Enteric fermentation	15	5		7	6
Manure management	15	6		5	5
	Basic IPCC 1995	Basic IPCC 1996	Country specific	Use of different methods	No information provided
Agricultural soils	12	7	7	2	6
Rice cultivation	8	---	2		2
Field burning of residues	10	---			3
WASTE:	Default	Kinetic	Country specific	Use of different methods	No information provided
Solid waste disposal	16	6	8		3
Waste water	14	---	6	2	3
Waste incineration	4	---	5	1	2

^a) Number of Parties reporting refers specifically to CO₂ emissions from stationary and mobile combustion and waste incineration, to CH₄ emissions from enteric fermentation, manure management, rice cultivation, field burning of residues, solid waste disposal and waste water and to N₂O emissions from agricultural soils. Industrial processes emissions are not considered in the table because the methods and tiers used to estimate most of these emissions are quite similar. Reporting of HFC, PFC and SF₆ emissions is covered in table 25.

Comments

Table 1 indicates the methods chosen by countries within the reporting framework of the IPCC Guidelines. The secretariat has compiled this table based on the information submitted by Parties in their second national communications and supporting materials. In those cases where a clear statement of the method or tier used by Parties was made, this was used as the basis for the allocation. In other cases the allocation was inferred from the relevant narrative text or inventory tables. Some Parties did not provide sufficient information for the allocation of methods used in the IPCC reporting framework. The differences between the number of reporting Parties in the different IPCC source categories is a consequence of the varying degree of completeness in the reporting (see table 18).

Table 2. Types of emission factors used by Parties

IPCC categories	Types of emission factors used
	Defaults (IPCC or CORINAIR)	Country- specific (CS)	Mix of defaults/CS	No information provided
	Number of reporting Partiesa)
ENERGY:
Stationary combustion	9	11	5	7
Mobile combustion	10	11	4	7
Fugitive solid fuel emissions	7	7	3	7
Fugitive oil and gas emissions	9	9	2	8
INDUSTRIAL PROCESSES:
Mineral production	5	7	4	18
Chemical industry	8	5	4	17
Metal production	7	6	5	16
AGRICULTURE:
Enteric fermentation	12	8	7	6
Manure management	12	8	5	6
Agricultural soils	13	9	3	9
Rice cultivation	5			7
Field burning of residues	4			9
WASTE:
Solid waste disposal	11	13	1	8
Waste water	9	7	3	6
Waste incineration	4	5	1	2

^a) Number of Parties reporting refers specifically to CO₂ emissions from stationary and mobile combustion and waste incineration, to CH₄ emissions from enteric fermentation, manure management, rice cultivation, field burning of residues, solid waste disposal and waste water and to N₂O emissions from agricultural soils. For CO₂ emissions from mineral production the secretariat included those Parties that specifically reported these emissions (23) and those that reported total CO₂ emissions from the industrial processes sector. This was done because some Parties only provide total CO₂ emissions from the industrial processes sector. For the other industrial subsectors emissions of the three main GHGs were considered.

Comments

The table indicates what types of emission factors were used by Parties to compile their GHG inventories. The secretariat has compiled this table based on the information submitted by Parties in their second national communications and supporting materials. In those cases where a clear statement of the type of emission factors used by Parties was provided, this was used as the basis for the allocation. In other cases the allocation was inferred from the relevant narrative text or inventory tables. To the extent possible, the values of reported emission factors were compared with the default ones, if the source of the emission factors was not indicated. Some Parties did not provide sufficient information to allocate the emission factors according to the classification provided in the table. The differences between the number of reporting Parties in the different IPCC source categories are a consequence of the different degrees of completeness in the reporting (see table 18).

Table 3. Comparison of CO₂ emissions from fuel combustion using the reference approach (tier 1)

and detailed methods (tiers 2 and 3) (Gg)

	Germany					United Kingdom					USA	Finland
	1990	1991	1992	1993	1994	1990	1991	1992	1993	1994	1990	1994
Reference approach	983 528	948 549	904 325	896 608	--	576 457	586 922	575 049	553 964	554 260	1 424 000	58 281
Detailed technology- based approach	986 640	950 625	901 383	893 100	--	563 908	570 847	555 019	540 454	533 180	1 436 000	58 331
Difference (%)	0.3	0.2	-0.3	-0.4	--	-2.2	-2.8	-3.6	-2.5	-4.0	0.8	0.8
	Bulgaria						Australia
	1990	1991	1992	1993	1994	1995	1990	1991	1992	1993	1994	1995
Reference approach	78 716	61 286	56 570	61 772	58 041	61 098	262 473	264 642	267 408	271 171	276 771	287 349
Detailed technology- based approach	76 484	60 626	55 416	57 650	54 322	56 255	262 569	270 269	273 031	276 545	280 180	287 144
Difference (%)	-2.9	-1	-2	-6.7	-6.4	-7.9	0.04	2.1	2.1	2.0	1.2	-0.07

Comments

This table is based on the information submitted in national communications and supporting materials.

Table 4. Comparison of CH₄ emissions from enteric fermentation and manure management estimated using

tier 1 and tier 2 methods in the Czech Republic

Livestock type	Number of animals	Emission factors enteric fermentation		Emissions from enteric fermentation		Emission factors manure management		Emissions from manure management
	(thousands)	(kg/head/year)		(Gg/year)		(kg/head/year)		(Gg/year)
		Tier 1	Tier 2	Tier 1	Tier 2	Tier 1	Tier 2	Tier 1	Tier 2
Dairy cattle	1 195	94.00	67.16	112.33	80.26	6.00	3.29	7.17	3.93
Non-dairy cattle	2 165	48.00	subdivided	103.92	53.00	14.00	1.50	30031	3.25
Sheep	430	8.00	7.87	3.44	3.38	0.19	0.33	0.08	0.14
Goats	42	5.00	4.59	0.21	0.19	0.12	0.21	0.01	0.01
Horses	25	18.00	47.22	0.45	1.18	1.39	3.63	0.03	0.09
Swine	4 569	1.50	3.41	6.85	15.58	3.00	8.80	13.71	40.21
Poultry	33 278	--	0.07	0.00	2.33	0.08	0.02	2.63	0.67
Total				227.2	155.92			53.94	48.29
				Difference	71.28			Difference	5.64
				%	-31.4			%	-10.5

Comments

The data used for comparison in the table were taken from: "Greenhouse Gas Inventories: National Reporting Processes and Implementation Review Mechanisms in the EU. Case Study on the Czech Republic". M. Tichy and W. Katscher, pp. 38-39. Forschungszentrum Jülich GmbH (KFA), Programme Group Technology Assessment (TFF). KFA series 1996. The data presented were reorganized.

Table 5. Comparison of CH₄ emissions from solid waste disposal using basic and

more advanced methods in the Czech Republic and Finland

Czech Republic	CH4 emissions (Gg)	Finland	CH4 emissions (Gg)
Basic methodology	149	Mass balance	105
Kinetic methodology	125	Dynamic model	71
Difference (%)	16		48

Comments

The information presented in the table was taken from Finland's National Greenhouse Gas Inventory, 1992-1994, supporting material prepared by the Ministry of Environment of Finland, 19.06.1996 and from "Greenhouse Gas Inventories: National Reporting Processes and Implementation Review Mechanisms in the EU: Case Study on the Czech Republic". M. Tichy and W. Katscher, p. 41 Forschungszentrum Jülich GmbH (KFA), Programme Group Technology Assessment (TFF).

KFA series 1996.

Table 6. Comparison of N₂O emissions from agricultural soils estimated

using the Revised 1996 and 1995 IPCC Guidelines

N2O emissions, 1990 (Gg)
	Revised 1996 IPCC Guidelines	1995 IPCC Guidelines	Difference (%)
Denmark	33.0	8.5	74.2
Finland	6.0	4.0	33.3
New Zealand	44.9	37.0	17.5
Switzerland	9.2	13.4	-45.7
United States	196.0	183.2	6.5

Comments

The information presented in the table was taken from the first and the second national communications submitted by Parties reporting the use of the Revised 1996 IPCC Guidelines for the estimation of N₂O emissions from agricultural soils.

Table 8. Comparison between IPCC default and country-specific emission factors for N₂O emissions from agricultural soils, New Zealand

	IPCC default value	New Zealand value	Difference (%)
Leaching	0.30	0.15	100
Pasture emissions	0.02	0.01	100

Comments

The information presented in the table was taken from the second national communication of

New Zealand.

Table 7. Comparison between IPCC default and country-specific emission factors for CO₂ emissions from the

fuel combustion sector (t CO₂/TJ)

	Crude oil	Anthracite	Hard coal (coking coal)	Bituminous coal	Sub-bituminous coal (brown coal)	Lignite	Peat	Wood (solid biomass)	Coke (petroleum coke)	Car gasoline	Aviation gasoline	Jet kerosene	Gas/ diesel oil	Residual fuel oil	LPG	Natural gas	Coke oven gas	Blast furnace gas	Refinery gas
Austria			93-97	97	101 108				92-100	74	74	71	74	77-79	64	55
Belgium	72.6		92.7						107	69		71	73	77	63	56	46	280	73
Denmark			95					102	102	73	72?	72	74	78	65	57			57
Finland			91.3?				104	109		73	71	72	73 74	77	66	56
France	73?		94.1	93	96	100		92		72.4		74	74.8	74	64	57	47	268	56
Germany	73			95		95	106		108	69	72	72	74	77	63	56	44
Greece			92.7			73.2-122		97.2	92.7-99.8	68.6			73.3	76.6	62.4	55.8			66.1
Ireland			86				104 115			70		71	73	76	64	55
Sweden			91				97.1 107.1			73	72	73	75	75.3 76.2	78		60	103
United Kingdom	73.3	91		90.3	99.6			104.7	106.4	67.3		68.8	69.8	74.1	67.9	55.6
Default IPCC 1996	73.3	98.3	94.6	94.6	96.1	101.2	106	109.6	108.2	69.3	71.9	71.5	74.1	77.4	63.1	56.1	47.7	242	66.7
Maximum difference (%)	0	-4	3	3	12	21	9	-1	0	7	3	3	1	2	8	2	26
Minimum difference (%)	-1	-4	-9	-5	0	-28	-8	-16	-15	-3	-1	-4	-6	-4	-1	-2	-8

Comments

The country-specific emission factors presented in the table were taken from national communications or supporting materials submitted by Parties. Only a group of Parties which presented disaggregated emission factors in the fuel combustion sector were chosen. Other Parties which also provided disaggregated emission factors were not included. Default emission factors were taken from the IPCC Guidelines.

Table 9. Comparison of CO₂ emissions from cement production estimated using IPCC default and

country-specific emission factors (1995)

	Quantity of cement produced (kt)	Country-specific emission factor (t CO2/ t cement produced)	CO2 emitted (Gg)	IPCC default emission factor (t CO2/ t cement produced)	CO2 emitted (Gg)	Difference in emissions estimates (Gg)	Difference in emissions estimates (%)
Australiaa)	6 116	0.51	3 168	0.50	3 101	67	2
Canada	10 722	0.5	5 361	0.49	5 345	16	0.3
Iceland	81	0.44	50	0.49	41	-4.6	-11.7
New Zealand	989	0.51	503	0.49	492	10	2.2
Norway	1 683	0.5	842	0.49	839	2.6	0.3
Switzerland	420	0.59	248	0.49	210	384	18.4

^a) Data for Australia are from 1990. This Party reported emissions based on clinker instead of cement production. Therefore, the emission factors are different from those used by other Parties.

Comments

Activity data and country-specific emission factors presented in the table were taken from standard data tables provided in the national communications of these Parties. Emission estimates made with these data were also taken from the communications. Default emission factors were taken from the IPCC Guidelines. The secretariat estimated emissions using default emission factors and the activity data provided by Parties. Percentage differences were estimated assuming that national estimates were 100 per cent.

Table 10. Change in emission estimates as a consequence of different assumptions

used to choose activity data, United States

	N2O emissions from agricultural soils (Gg)
	1990	1991	1992	1993	1994
N2O (1 year)a)	184.7	188.3	190.9	190.0	210.9
N2O (3 years)b)	183.2	188.0	189.7	197.3	200.4
Difference (%)	0.8	0.2	0.6	3.7	5.0

^a) Activity data correspond to each single year.

^b) Activity data correspond to the average of the three previous years.

Comments

Information presented in table 10 was taken from the Inventory of US Greenhouse Gas Emissions and Sinks: 1990-1994,

submitted by this Party as supporting material.

Table 11. Differences in GHG emission estimates for the year 1990 due to changes in methods/data used (percentage change)^a)b)

	A	B	C	D	E	F
				All GHG emission estimates (CO2 equivalent)
Partyc)	CO2 emission estimates	CH4 emission estimates	N2O emission estimates	Total change	Effect of the use of different GWPs	Effect of changes in methods/ datad)
Australia	-5.5	-17.1	+31.4	-12.1	-4.4	-7.7
Austria	+4.5	-2.6	+183.2	+3.4	-2.7	+6.1
Bulgariae)	-0.7	+0.8	+31.6	-2.6	-4.1	+1.5
Canada	<1	+3.6	-9.9	-1.9	-2.1	+0.2
Czech Republic	<1	-5.8	+7.5	-2.2	-1.8	-0.4
Denmark	<1	+3.4	+230.1	+9.7	-2.5	+12.0
Estonia	--	-67.4	-4.2	-2.4	-2.8	+0.4
Finland	<1	-2.4	-18.2	-3.8	-1.0	+2.8
France	+3.2	+4.2	+2.8	+0.8	-2.4	+3.2
Germany	--	--	+7.1	-1.4	-1.8	+0.4
Greece	+3	+29.2	+26.0	+4.6	-1.7	+6.3
Iceland	-1.1	-39.1	-30.0	-12.2	-2.0	-10.2
Ireland	--	+1.9	-30.6	-10.8	-5.2	-5.6
Italy	<1	-40.3	+36.7	-5.5	-1.8	-3.7
Japan	-2.6	+14.0	+90.8	-1.3	-0.6	-0.7
Latvia	+7.8	+17.2	+838.8	+29.0	-2.4	+31.4
Luxembourg	+16.3	+0.5	+14.5	+14.7	-0.7	+15.4
Netherlands	--	+4.1	-0.6	-1.6	-2.1	+0.5
New Zealand	--	-14.1	+177.9	-4.5	-7.82	+3.32
Norway	<1	+49.0	--	+3.9	-3.3	+7.2
Polandf)	+14.9	-48.2	-4.1	-13.8	-2.0	-11.8
Portugal	+11.8	+257.7	+33.3	+34.1	-4.2	+38.3
Russian Federation	<1	-1.9	+151.9	-2.6	-3.1	+0.5
Slovakia	+3	+17.9	-21.9	0.8	-2.1	+2.9
Spain	<1	+1.4	+0.3	-2.8	-2.8	+0.0
Sweden	-9.5	-1.5	-39.5	-12.2	-1.9	-10.3
Switzerland	--	-26.7	-26.3	-7.6	-1.8	-5.8
United Kingdom	+1.2	-1.5	+10.8	-1.1	-2.3	+1.2
United States	<1	+9.5	+3.3	-0.6	-1.9	+1.3

^a) Percentage deviation of the inventory of the second national communication relative to the inventory submitted in the first national communication. Negative values denote that the latest submitted inventory has a lower figure. All figures rounded.

^b) Where no value is shown this indicates that the difference is zero between the emissions in the first and second national communications.

^c) Hungary was not included because it did not provide a recalculated figure for the base year (average of

1987-1989) or for the year 1990 in its second national communication.

d) This change represents the effects of changes in methods/data. The effect of the use of different GWPs in the first and second national communication is excluded here.

^e) The data for Bulgaria are for 1990 because the base year was not recalculated in the second national communication despite the fact that the years from 1990 and on were recalculated.

^f) Data for Poland represent the changes in their base year (1988).

Comments

Table 11 indicates the effects of changes in methods/data and change of GWP used in the first and second national communications on the aggregated GHG emission estimates for the base year (column F). The total changes of these emissions (column D), including the effect of the use of different GWPs (column E), are also shown. Changes in the estimates on a gas-by-gas basis are also presented (columns A, B, C).

(1) Differences in CO₂ emissions were estimated from the percentage change in estimates between the first and second national communications. The formula used is the following:

Change = 100 (1- (CO₂ 2^nd NC/CO₂ 1^st NC)). The results are presented in column A.

Differences in CH₄ and N₂O emissions were estimated in the same way. The results are shown in columns B and C, respectively.

For the estimation of the CO₂ equivalent emission differences, the estimates of CH₄ and N₂O were multiplied by the corresponding GWP and aggregated with the CO₂ estimates. HFC, PFC and SF₆ emissions were not included for purposes of consistency, as the reporting of these gases is inconsistent among Parties.

(2) The procedure used for obtaining the total changes between CO₂ equivalent emissions of the first and the second national communications, as well as the effect of change in methods/data used, was the following:

(a) The value of aggregated GHG emissions from the second national communications obtained using 1995 GWPs was divided by the value of aggregated GHG emissions from the first national communications obtained using the 1994 GWPs to estimate the total changes of estimations between the two communications. To express the value in per cent the following formula was used:

100*[(CO₂ eq. 2^nd NC using GWP 95/CO₂ eq. 1^st NC using GWP 94)-1]. These results are presented in column D.

(b) The effect of using different GWPs was estimated by dividing the aggregated GHG emissions from the 2^nd national communication using the 1995 GWPs by the aggregated GHG emissions of the 2^nd NC using the 1994 GWPs. To express the value in per cent the following formula was used:

100*[(CO₂ eq. 2^nd NC using GWP 95/CO₂ eq. 2^nd NC using GWP 94)-1]. These results are presented in column E.

(c) Finally, the effects of the change in methods/data were estimated by the algebraic subtraction of the effect of the use of different GWPs (E) from the total change (D). As the previous operation was a subtraction of percentages, these results are also expressed in per cent.

Table 12. Changes in the percentage share of emissions of particular GHGs in aggregated national

CO₂ equivalent emissions (average for 1991-1995 compared to 1990)

Party	CO2			CH4			N2O			Other GHG
	1990	1991-95	Diff. (%)h)i)	1990	1991-95	Diff.(%)h)i)	1990	1991-95	Diff.(%)h)i)	1990	1991-95	Diff.(%)h)i)
Australia	66.5	68.0	2.3	26.3	25.8	-1.9	6.0	6.1	1.8	1.2	0.1	-91.8
Austria	79.5	79.2	-0.4	15.9	15.6	-1.9	4.6	5.0	7.4	--	0.2
Belgium	83.2	83.4	0.2	9.5	9.3	-2.2	6.8	6.8	-1.0	0.4	0.5	28.2
Bulgariaa)	68.4	68.4	0.0	24.15	25.0	3.4	7.5	6.7	-10.6	--	--
Canada	81.9	81.0	-1.1	11.9	12.4	4.0	4.7	5.1	8.1	1.6	1.4	-9.7
Czech Republic	86.1	85.8	-0.3	9.7	9.9	2.0	4.2	4.3	3.0	0.0	--
Denmark	72.8	75.6	3.8	12.3	11.3	-8.8	14.7	12.9	--11.8	0.2	0.1	-60.0
Estonia	92.8	91.9	-1.0	5.4	6.3	14.3	1.8	1.7	-1.8	--	--
Finlandb)	83.4	83.8	0.5	8	7.8	-2.6	8.6	8.4	-3.3	--	0.1
France	74.8	76.4	2.1	12.5	12.1	-3.3	11.1	10.5	-5.5	1.5	1.0	-34.2
Germany	83.6	83.4	-0.2	9.8	9.5	-3.2	5.8	6.2	6.8	0.7	0.9	21.6
Greece	85.2	85.6	0.5	9.4	9.3	-1.1	5.4	5.1	-5.5	--	--
Hungarya)c)	82.3	76.1	-7.5	13.7	20.8	34.1	3.9	3.1	-22.0	--	--
Iceland	74.3	81.1	9.2	10.2	10.5	2.9	4.5	4.3	-3.7	11.0	4.1	-62.8
Ireland	54	56.8	5.2	30	29.4	-2.0	16.0	13.8	-13.5	--	--
Italyd)	81.1	80.2	-1.1	9.2	10.1	8.9	9.6	9.4	-1.4	0.2	0.3	87.5
Japane)	89.8	89.9	0.1	2.6	2.0	-30.0	2.6	2.1	-21.3	4.9	6.1	24.2
Latvia	69.4	64.2	-7.5	11	11.4	3.5	19.6	24.4	24.7	--	--
Luxembourgd)	94.8	94.0	-0.8	3.6	4.1	12.2	1.5	1.9	29.6	--	--
Netherlands	77.8	79.2	1.8	10.8	10.2	-5.9	7.4	7.9	7.2	4.1	2.7	-33.7
New Zealand	33	35.7	8.2	46.4	44.8	-3.6	19.2	18.7	-2.5	1.5	0.8	-46.3
Norway	65.8	68.4	4.0	16.8	18.0	6.7	8.6	8.3	-4.0	8.8	5.3	-39.6
Polanda)f)	84.5	84.7	0.2	11.7	11.8	0.8	3.5	3.5	0.3	--	--
Portugalg)	68.9	70.2	1.9	24.8	23.8	-4.2	6.3	6.0	-5.1	--	--
Russian Fed.	78	77.2	-1.0	18.3	19.2	4.7	2.3	1.8	-20.0	1.4	1.8	31.4
Slovakia	82.2	82.5	0.4	11.8	12.3	4.1	5.3	4.5	-16.0	0.7	0.8	17.6
Spaing)	75.1	75.4	0.4	15.2	15.5	1.9	9.7	9.1	-6.1	--	--
Sweden	83.4	84.5	1.3	10.2	9.8	-4.1	4.3	4.2	-1.3	2.0	1.6	-21.6
Switzerland	83.9	83.4	-0.6	9.5	9.4	-1.1	6.6	6.8	2.5	--	0.4
United Kingdom	80	80.9	1.1	12.8	12.4	-3.2	5.1	4.4	-13.7	2.1	2.3	9.5
United States	85.5	85.2	-0.4	10.7	10.7	0.0	2.3	2.3	2.5	1.5	1.8	18.4

^a) Bulgaria, Hungary and Poland have base years other than 1990. Values here represent 1990.

^b) Finland did not report emissions for 1991. The average represents 1992-1995 emissions.

^c) Hungary did not recalculate its inventory of the year 1990. Therefore, emission estimates for 1991-1995 were calculated using different methods.

^d) Italy and Luxembourg did not report emissions for 1991-1993. The averages represent 1994-1995 emissions.

^e) Japan did not report CH₄ or N₂O emissions for 1995. The separate and aggregated averages for all gases represent 1991-1994 emissions.

^f) Poland did not report emission data for 1991, 1993 or 1995. The average represents 1992 and 1994 data.

^g) Portugal and Spain did not report emissions for 1995. Averages represent 1991-1994 emissions.

^h) The formula used for CO₂, CH₄, N₂O and Other GHG to arrive at the values for the difference is: Difference = ((GHG_1991-95-GHG₁₉₉₀)*100)/GHG₁₉₉₀

ⁱ⁾ A negative difference indicates a decrease from 1990 to 1991-1995 in the share of that particular gas, in the mix.

Comments:

This table was prepared by calculating the share of the emissions of each GHG in the aggregated GHG emissions for 1990 and 1991-1995. The values reported by Parties for emissions of each GHG in the year 1990 and in the years between 1991 and 1995 were expressed in terms of CO₂ equivalent, multiplying the emissions estimated by the corresponding 1995 IPCC GWP. The share of each GHG expressed in per cent is calculated in relation to the aggregated GHG emissions of each Party.

The first column for each gas indicates the share for 1990 and the second column shows the average share of the gases for 1991-1995, unless otherwise indicated. The difference is the percentage change from 1990 to 1991-1995. Negative values signify that the relative importance of a particular gas decreased in 1991-1995 in relation to 1990. This could mean a decrease in the emissions of that gas, but also an increase in the emissions of any of the other gases.

Table 13. Changes in aggregated national GHG emissions, expressed in CO2 equivalent, due to socio-economic driving factors and to changes in methods/data (average for 1991-1995 compared to 1990)a)
	Aggregated GHG emissions 1990, using the 2nd NCs (Gg) (A)	Average GHG emissions 1991-1995, using the methods from the 2nd NCsb) (Gg) (B)	Percentage increase or decrease of average aggregated GHG emissions 1991-1995, related to the emissions of the year 1990, reflecting only changes in socio-economic driving factorsc) C=100*((B/A)-1)	Average GHG emissions 1991-1995, using the methods/data from the 1st NCs, assuming the same rate of change reported in column C (Gg) (D)	Percentage of increase or decrease of the average aggregated emissions 1991-1995 due to changes in methods/data in the 2nd NCs.c) E=100*(1-(B/D))
Australiad)	405 553	416 211	2.6	450 228	7.6
Austriad)	77 814	77 507	-0.4	72 846	-6.4
Belgiumd)	138 943	141 933	2.2	116 872	-21.4
Bulgariae)	123 432	91 074	-26.2	87 608	-4.0
Canada	557 860	577 110	3.5	576 319	-0.1
Czech Republic	192 130	159 584	-16.9	160 318	0.5
Denmarkd)	71 658	79 991	11.6	71 180	-12.4
Estonia	40 719	28 338	-30.4	31 542	10.2
Finland	64 546	65 834	2.0	67 329	2.2
Franced)	498 067	501 725	0.7	485 670	-3.3
Germany	1 203 537	1 097 797	-8.8	1 093 556	-0.4
Greeced)	99 232	102 079	2.9	96 234	-6.1
Hungarye)f)	101 634	80 897	-20.4	68 972	-17.3
Icelandd)	2 571	2 630	2.3	2 906	9.5
Irelandd)	56 861	57 506	1.1	61 234	6.1
Italy	532 048	528 168	-0.7	544 157	2.9
Japan	1 190 250	1 242 550	4.4	1 253 912	0.9
Latvia	35 669	23 074	-35.3	17 505	-31.8
Luxembourg	13 448	11 448	-14.9	10 244	-11.8
Netherlands	206 602	216 440	4.8	215 626	-0.4
New Zealandd)	76 034	75 646	-0.5	72 113	-4.9
Norwayd)	49 266	49 776	1.0	46 732	-6.5
Polande)	459 048	438 970	-4.4	565 524	22.4
Portugald)	68 442	72 270	5.6	52 954	-36.5
Russian Fed.	2 998 767	2 111 366	-29.6	2 100 614	-0.5
Slovakia	72 496	57 732	-20.4	56 162	-2.8
Spain	301 431	305 190	1.2	305 363	0.1
Swedend)	65 101	66 155	1.6	74 057	10.7
Switzerlandd)	53 749	53 381	-0.7	56 489	5.5
United Kingdom	714 691	679 819	-4.9	671 301	-1.3
United States	5 713 320	5 845 742	2.3	5 782 546	-1.1

^a) Parties which have not submitted more than one communication are not included in this table.

^b) When Parties did not provide emissions data for 1990 through 1995, only those years reported were used. Finland's average represents 1992-1995 emissions. Italy and Luxembourg averages represent 1994-1995 emissions. Japan did not report CH₄ or N₂O emissions for 1995. The separate and aggregated averages for all gases represent 1991-1994 emissions. Poland's average represents 1992 and 1994 data. Portugal and Spain averages represent 1991-1994 emissions.

^c) Negative values in columns C and E indicate a decrease in the emissions between 1990 and 1991-1995.

^d) For Australia, Austria, Belgium, Denmark, France, Greece, Iceland, Ireland, New Zealand, Norway, Portugal, Sweden and Switzerland the percentage increase or decrease of the average aggregated emissions 1991-1995 due to changes in methods/data in the second national communication exceeded the percentage increase or decrease of emissions due to change in socio-economic driving factors in the period.

^e) Bulgaria, Hungary and Poland have base years other than 1990. Values here represent 1990.

^f) Hungary did not recalculate its inventory of the year 1990. Therefore, emission estimates for 1991-1995 were calculated using different methods. The situation does not allow changes reflected in columns C and E to be comparable with changes for other Parties.

Comments:

Column A shows the aggregated GHG emissions for the year 1990, taken from the second national communications and using the GWPs from 1995 (CO₂ equivalent = 21 times CH₄ emissions, 310 times N₂O emissions). The table does not include HFCs, PFCs and SF₆. Column B shows the average aggregated GHG emissions for the years 1991 through 1995, except when other years are used, as indicated in the footnotes. The figures are from the second national communications, and the 1995 GWPs are used. Column C shows the percentage change from 1990 to the average of years 1991-1995. Because GWP and other methods are the same for all six years, the change in emissions represented in this difference is due to changes in socioeconomic driving factors. A negative value therefore indicates that there was a decrease in the aggregated emissions between 1990 and 1991-1995.

Column D shows the estimate of Parties' aggregated GHG emissions between 1991 and 1995, using methods applied in the first national communications. The values were estimated using the same trend of emissions related to 1990, year by year, as estimated from the data of the second national communications (reflected in column C) but using the base year estimates provided in the first national communications. 1995 IPCC GWPs were used, therefore changes in base year estimates between first national communications and second national communications are a consequence of changes in methods/data and are not influenced by the use of different GWPs. Column E indicates the effect of a change in methods between the first and second national communications, in per cent, and was calculated using the formula E=100*(1-(B/D)). A negative value signifies that the change in methods has caused the emissions to be reported as lower than they would be had the Parties used the methods from the first national communications.

Table 14. Confidence levels^a) (qualitative^b) or quantitative (+/- per cent)) of GHG emission estimates in the

main IPCC source categories

Gas and source

AUS

BEL

BUL

CAN

CHE

CZE

DNK

FIN

GBR

HUN

ICE

ITA

JAP

NLD

CO2

2

4c)

H

8-10

1-2

H/M

H 5

H

H 2

Fuel combustion

<5

H

3

H

H/M

H

H

4-5

H

10

10

H

Industrial processes

<10

H

15

H

H

H

H

10

10

H

Changes in forest

25

H

H

M

60

60

M

Other LUCF

>80

H

100

100

CH4

30

30

M

40

*d)

M/L

M 25

M/L

M 25

Fuel combustion

>20

M

40

M

20-30

L

M

M

M

Fugitive: oil & gas

>20

M

30

M

20-30

M

M

3-4

60

60

M

: coal mining

>20

M

40

40-50

5-7

60

60

Enteric fermentation

20-80

H

30-50

M

20-30

M

M

M

25

25

M

Animal waste

20-80

M

50

M

20-30

M

L

L

20

20

L

Waste

>50

M

30

M

M

M

L

100

100

M

N2O

50

40

M/L

80-100

*d)

M

L 50

L

L 50

Fuel combustion

>20

M

50-60

M

M

L

L

L

Inorganic chemicals

M

30

M

M

L

50

50

L

Organic chemicals

M

15

50

50

Agricultural soils

20-80

M

60-100

L

M

L

L

200

200

L