Statistics Iceland

Business Statistics

Borgartún 21A,

105 Reykjavík

Iceland

Air emissions accounts (AEA) record flows of gaseous and particulate materials emitted into the atmosphere as a result of economic activity.

AEA is a subset of environmental-economic accounts. They offer a detailed breakdown for 64 emitting economic activities (NACE), plus households, as defined in the national accounts of EU countries. They are aligned with economic statistics and GDP. These features make them suitable for integrated environmental-economic analyses and modeling – for example of 'carbon footprints' and climate-change modeling scenarios.

National Statistical Institutes (NSI) submit AEA to Eurostat through a mandatory annual data collection. The data collection includes an electronic questionnaire and this quality report.

The AEA dataset has the following dimensions:

1)           Air pollutant: Emissions to air of the following gaseous and particulate substances are collected (greenhouse gases, air pollutants):

Carbon dioxide without emissions from biomass (CO2),

Carbon dioxide from biomass (Biomass CO2)*,

Nitrous oxide (N2O), Methane (CH4),

Perfluorocarbons (PFCs),

Hydrofluorocarbons (HFCs),

Sulphur hexafluoride (SF6) including nitrogen trifluoride (NF3),

Nitrogen oxides (NOx),

Non-methane volatile organic compounds, (NMVOC),

Carbon monoxide (CO),

Particulate matter < 10μm (PM10),

Particulate matter < 2,5μm (PM2,5),

Sulphur dioxide (SO2),

Ammonia (NH3)

2)           Geopolitical entity: EU Member States, EFTA Countries, Candidate Countries etc.

3)           Economic activities: include 64 production activities (classified by NACE rev.2 A*64), and households’ consumption (3 sub-classes).

4)           Time: reference year for which air emissions are reported

5)           Unit: tonnes and thousand tonnes

The data refer to national economies as defined in the system of national accounts. Greenhouse gases and air pollutants emitted by resident units representing the national economy are covered.

Conceptually AEA belong to the international system of environmental economic accounting (SEEA-Central Framework). Furthermore, AEA is one of several physical modules of Eurostat's programme on European environmental economic accounts. It is covered by Regulation (EU) No.691/2011 on European environmental economic accounts.

AEA are closely related to concepts and definitions of national accounts. Most notably, they follow the residence principle, i.e. they record emissions related to resident unit's activities, regardless where those occur geographically.

Further methodological guidelines are provided in various publications by Eurostat (see Eurostat website > Environment > Methodology, heading: 'Air emissions accounts').

Data refer to emissions by resident economic units in the sense of SEEA CF 2012 and National Accounts (ESA), including households.

The national economy is as defined in SEEA CF 2012 and National Accounts (ESA), i.e. all economic activities undertaken by resident units.

The reference area is the economic territory as defined in SEEA CF 2012 and National Accounts (ESA). A unit is said to be a resident unit of a country when it has a centre of economic interest in the economic territory of that country, that is, when it engages for an extended period (1 year or more) in economic activities in that territory.

By following this residence principle, the Air Emission Accounts record emissions from resident units' activities, regardless where they occur. This is the main conceptual difference to emission inventories for greenhouse gases (UNFCCC) and air pollutants (CLRTAP).

Data is available for each year from 1995 until 2019. Data for 2020 is a preliminary reporting based on financial activity data

Not applicable because AEA are not reported as indices.

The unit of measure is tonnes or thousand tonnes.

F-gases (HFC, PFC, SF₆ and NF₃) are reported in tonnes of CO₂ equivalents.

SO_X are reported in tonnes of SO₂ equivalents, and NO_X are reported in tonnes of NO₂ equivalents.

The data refer to calendar years.

Air emissions accounts (AEA) are legally covered by Regulation (EU) 691/2011on European Environmental Economic Accounts.

Not applicable at national level.

The working group at Statistics Iceland has agreed that no data in the collection are to be subject to confidentiality

The working group within Statistics Iceland reviewed the policy for the publication of data from economic sectors that have very few registered industries.

It was concluded that the policy of "more than ten companies per sector" should not be followed in the environmental accounts, since doing so would be confusing to the Icelandic population. Underlying here is that the largest contributors to GHG emissions from Iceland (viz. metal industry and air transport) are few in numbers. 

Does not apply

Data from the AEA account should be disseminated to the public as close to the publication of other air emissions accounts, e.g. the NIR report. Emissions of greenhouse gasses should be published in its fullest detail as well as in a simplified view of the data to aid the public in drawing informed conclusions from the data.

The NIR report is generally published in March.

• The initial publication of the AEA should be completed in April
• Publication of the AEA including estimation for the last year should be completed in June
• Reviewed data should be published in December(including corrections requested by Eurostat)

Does not apply

Data regarding GHG emissions (CO2, CH4, N2O, HFC, PFC, and CO2-eq) are published in full NACE*64 detail, as well as in a summary view on a publicly available data repository.

Publication of the data is announced in a press brief that should highlight key changes from the previous year. 

Data is disseminated annually.

News releases are online

No ad-hoc publications are done of the AEA

The data is available on a px-web format

no data available

Some of the microdata around the AEA are available, although the data is not immediately associated with the AEA. The microdata includes

• Data regarding the energy consumption of the vehicle fleet per NACE*64 sector
• Data regarding the number of live animals, estimates of live weight and energy consumption per month
• Data regarding the material flow of industrial raw material and produced goods.

none that applies

none this year

Description of the methodology and compilation instruction is available in Icelandic. The documentation is a part of the R markdown package run in order to collect, clean and validate the data.

Standard documentation according to the format requirements by Statistics Iceland is on the Statistics Iceland website

The metadata is reviewed every year

Time series submitted each year are compared with the data submitted the previous year. 

The year-to-year change for each submitted value is monitored. These values can change due to:

1. Changes in computations, or assumed emission coefficient for the activity
2. Changes in reported activity or quantity of material used in the activity
3. Changes in the classification of the NACE category of the activity
4. Changes in the reported values in the NIR or CLRTAP reports (completed by the Icelandic Environment Agency)

Significant changes in the values are reviewed by the working group.

Sections of the AEA report was sent out for review and quality checking by industry affiliated focus group. In 2021 these groups were

• Fisheries Iceland, which is an umbrella organization that speaks for fishing fleet operators and fish producers
• Federation of Icelandic Industries, which is an umbrella organization that speaks for manufacturing and other industrial sectors in Iceland

The AEA report was presented to the Environment Agency of Iceland and the Icelandic Energy Authority, which are major data providers.

Data is reviewed in its entirety between years and any changes greater than 0,5% in any value from the previous year reviewed. 

Emission values from industry sectors are compared to financial profit/loss statements and employment statistics. The assumption is that indicators, such as emission per employee and emission per million kr turnover should be somewhat constant from year to year. Unlikely changes are reviewed further.

Energy authority data (Orkustofnun)

• The data received from Orkustofnun 2020 suffer in quality due to a lack of detail, specifically in terms of fuel distribution to marine transport. Data in this category was not separated by the national origin of the effective operator of the boat.
• Data received from Orkustofnun 2021 have been revised. Fuel sold to non-Icelandic marine transport vessels is now available from 2016
• Data on fuel distributed to road transport reported by Orkustofnun were compared with the road-tax collected by the Icelandic authorities that is attached to fuel sales. This showed a 1-3% difference in the values for some years. Orkustofnun has corrected values retroactively, but some difference remains due to the distribution of "colored" fuel (fuel for off-road vehicles, and thus not taxed with road taxes). 

NIR/CLRTAP data from the Icelandic Environment Agency

• Data in the NIR dataset from 2020 differs considerably from the data submitted in 2019. Changes in emission values could be as high as 15% of the value reported previously. Reasons for this are unclear but may be due to a review in computation methods at the Environment Agency of Iceland.
• Data in the NIR dataset from 2021 differs considerably from the data submitted in 2020. The changes largely undo the difference between the submissions of 2020 and 2019 (see the previous comment). The reason for the change is unclear.

Results from the AEA are widely discussed in the public sphere. Industrial groups such as the Icelandic Fisheries republish results from the AEA in their annual reports and the tourism industry mentions any favorable trends in the results. Part of this results in the deliberate recruiting of an industry group to review and comment on the AEA data processing and conclusions prior to final reporting.

Statistics Iceland prepared presentations of the AEA to ministries in order to educate users.

Experimental statistics were developed in 2020 where emissions values are reported down to monthly timescale for the time period of 2016 to the current year. The lag-time in the reporting is 2-5 months (reported quarterly). This data received much interest during the Covid-19 shut-downs and the subsequent re-opening of the economy in 2021

Users are primarily concerned about the timeliness of the AEA data. Data from 2018 are largely irrelevant in 2020, but this timeliness is limited by the timeliness of the NIR report. Statistics Iceland decided therefore to extend the dataseries to 2019 using financial records and material flow data (MFA) and energy flow data (PEFA).

Users are also interested in getting more detailed information for a per-quarter or per-month view of the data.

User satisfaction is in general high. 

Common confusions and criticism include:

• Users are often looking for emission reporting that can be connected to the value or production chain of specific products (carbon footprint idea). The AEA is not suited for this purpose since products and services often cross between country boundaries and NACE services.
• Users are often looking to connect the AEA with specific UNFCCC goals or treatises that Iceland has signed. Although the agreements are usually connected with the NIR-reported emissions, users find the reporting there obtuse and are looking to identify the industry contribution to emissions in the framework of the UNFCCC reporting, especially considering road transportation and emissions from waste.
  • The emissions from road transport are available in the microdata of the AEA but has not been published thus far.
  • Emissions from waste management are not identifiable in the microdata but should be extractable from the WASGEN module
  • Emissions associated with energy consumption can be obtained from the microdata associated with the PEFA module
• Some industrial groups are more interested (e.g. the tourism sector) are interested in connecting emissions within NACE sectors (and sub-sectors) that are specifically due to services to non-nationals. Such micro-analysis is being discussed in forums and meetings, but is unlikely to result in the publication of new datasets in the near future.

not applicable

Not applicable; To ensure comparability, this will be calculated and provided by EUROSTAT in the European quality report using a standardised method.

No absolute value exists for the exact emission of greenhouse gasses from Iceland. Accuracy can therefore only be estimated from the complexity and number of approximations used in processing

See Annex 1 - 2021

Not applicable because data are not based on a sample survey.

Not applicable because data are not based on a sample survey.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

Not applicable.

The compilation of the AEA data can be conducted immediately following the reception of the NIR data and IEA fuel distribution data. This event replaces estimated data that is pre-prepared. Data review and critique are then completed the following month. 

Estimates of emission from the economy per NACE*64 sectors are done every month from economic activities, import/export reports, employment records, and industrial reporting. This monthly emission reporting uses reported annual emission values to train a model that predicts the correct annual value within 0,5% of the correct value. This model is then used to predict values where official AEA data is missing.

It is understood that emission calculation based on imported goods has a built-in lag time and calculations based on exported goods have a built-in lead time. An attempt is made to correct or distribute the emission values if import/export or other financial transaction records are sparse. 

Computation of official AEA data happens immediately upon insertion of the source data (data from the NIR and IEA energy statistics) into the database.

The lag time between completion of the data and the publication is therefore due to review procedures and correction filing.

Not applicable.

Not applicable.

Thus far the data has been ready for publication well in advance of the desired publication date

Thus far the data has been ready for publication well in advance of the desired publication date

2022 is the fifth year that Statistics Iceland publishes the AEA. Most of the effort in developing data structure has been spent on ensuring the coherence of the reported data, although the final product is primarily based on the NIR/IEA reported values.

AEA is compiled according to harmonized guidelines provided by Eurostat and hence comparable across European countries reporting AEA to Eurostat.

Not applicable.

See Annex 2.

Not applicable; To ensure comparability, this will be calculated and provided by EUROSTAT in the European quality report using a standardised method.

none

Not applicable, because AEA data are annual.

none

not explored

Estimated cost in manhours

• Compilation of AEA data report from NIR/IEA data: 15 minutes
• Review (meeting with focus groups and presentation): 2 weeks
• Code review, code improvement to dev-stack: 2 weeks
• Code review, code improvement to production-stack: 2 days
• Dissemination, publication, news release: 2 weeks

The data production stack is constantly under revision. Each revision is done on a development stack prior to being factored into the production stack.

Data is revised if a substantial improvement can be achieved in the accuracy of reported emission values or NACE*64 distribution of emission values that reaches at least 10 years back. 

Accuracy is evaluated by reviewing three key emission indicators

1. emission of CO2 per million kr turnover
2. emission of CO2 per full employee
3. emission of CO2 per exported produced goods, or per tonne of fish/meat/vegetable produced
4. Year-to-year fluctuations in reported values of CO2, CH4 and N2O. Values that change by more than 20% are reviewed

A single industry or material segment is identified each year to be reviewed and revised.

In 2019 data from the national vehicle registration database, including make/model/driven distance by owner/operator was integrated into the data processing. This improved significantly the estimated emission from road transport per NACE*64.

In 2020 a new model for energy consumed per kg of fish produced was factored into the processing. This should improve the consistency of the data regarding fuel consumption and emission of refrigerants from fishing and fish production

In 2021 discrepancies between fluorine gas distribution and consumption were of concern. A program surveying the import and consumption pattern (and technological status) was launched in February, but the survey was unsatisfactory. Emissions from waste sites were also reviewed and some data revision was performed. 

In 2022 we successfully conducted a study with the main consumers and distributors of Fluorine-containing gasses. This study showed that our current approach to where emissions were originating from was incorrect and a new distribution model was built. A new data source was also introduced into this process. This has changed significantly the information on what NACE sector is responsible for the emissions from previous years' submissions. but the data is greatly improved by this effort.

An example of reported emission values for HFC for the reference year 2018 is collected in an Annex document ("Annex - HFC_2018_reported_by_different_years.csv"). This document shows the submitted values by submission years where a major change appears in the submission 2022. Furthermore, the total adjusted emissions are identical to the reported CLRTAP emissions.

Not applicable; To ensure comparability, this will be calculated and provided by EUROSTAT in the European quality report using a standardised method.

All microdata and source data is stored on a central database. 

Data processing is done using in steps using R-markdown workbooks, that return quality and success report for each step in the process. 

The three main data sources are

1. NIR data submitted by the Icelandic Environment Agency to UNFCCC
2. Energy statistics data submitted by the Icelandic Energy Authority to the IEA
3. Statistics Iceland database on physical trade
4. Statistics Iceland database on tax reporting by companies and individuals
5. Statistics Iceland database on financial trade (credit card transactions) in Iceland and for nationals abroad
6. Vehicle registration and inspections records including owner/operator reporting
7. Import, distribution, and waste management records for Fluorinated gases

Data used to check consistency are:

1. Statistics Iceland database for employment records
2. Statistics Iceland database for tourism
3. Statistics Iceland database on physical trade
4. Statistics Iceland database for industrial production report (these are somewhat unreliable)
5. Statistics Iceland database for VAT reporting by companies to the Tax Authority
6. Statistics Iceland database over company annual reports
7. Icelandic Environment Agency database on annual environmental reporting
8. Energy Authority database on licensed waste sites, mines, and offshore constructions
9. Statistics Iceland database on agriculture production and live animals count
10. Fishing Authority database on in/out the weight of fish through harbors
11. Icelandic Harbor Authority reports on cargo, passenger, and crew records for fishing, shipping, and recreational boats

Data from the in-house database is updated using trigger functions that monitor the line-by-line change of data

Data from external API endpoints are updated by PUSH notifications or by REST protocols

External providers submit data through the Statistics Iceland web portal that ensures that data stays confidential (ISO 27001 certified).

A survey on the use of fluorinated gasses was done in February 2022. The result participation was satisfactory and a new statistical model could be constructed

Data processing returns a set of validation documents where time-series consistencies are checked and year-to-year changes are highlighted for each processed data entry.

Consistency is primarily checked on values that are 1% or more of the national total emission. Suspect entries are flagged based on consistency criteria or based on alternative calculation routine methods where applicable. If suspect values can be reasoned they are manually passed. If inconsistent data has been provided the provider is asked to revise the data before a manual adjustment is entered.

A number of single values reported in the NIR or IEA data need to be "combed" into different NACE*64 categories and can not be calculated ab-initio for the AEA. This means that a "comb" is developed based on financial transaction records, employment data, or other business statistics available. Such combs include:

• Distribution of emission values of refrigerant between the manufacturing sector, the transportation sector, and the energy production sector. 
• Distribution of coal import/coal consumption between C24, C20, food production industries (grilling/smoking) and households
• Distribution of fuel to boats outside fishing vessels

Not applicable.

Economic activity is measured by several key indicators

1. Number of employees
2. The annual income from the sales of goods and sales of services
3. Number of known operation centers (detected by electricity purchases records)
4. Export quantity/value of manufactured goods
5. Registry of vehicles and inspection records
6. Property registry

Each NACE number gets a scaling factor for each indicator in order to obtain a measure of the activity per year. the scale factors consider fuel purchases (by PEFA fuel type), key non-energy related material consumption (such as F-gases, lubricants, coals, wood etc.). This allows for a consistent dispersion of the values reported in the NIR. The dispersed values are then tested against "first principles" values obtained by direct surveys.

Emissions from road transport are calculated from data we obtain from the vehicle registry.

This data contains:

1. Registry of ownership
2. Registry of operator
3. Inspection of vehicle conditions, including annual distance report
4. Type of vehicle, including energy source

The records take status points for two months of the year (February and August). Estimates are then produced for vehicles that have not come in for inspection during the year and for the energy needed for the vehicle. This information is used to estimate the total fuel consumed by road vehicles for the entire fleet in Iceland. The estimated fuel consumption is then compared with the reported volume sold (from fuel-tax records, and reported values from the Icelandic Energy Authority) and the model is iterated until the difference is within acceptable levels (3% of the true value). Thus, this information gives estimated fuel consumption per industrial sector (assuming that each vehicle only changes ownership max twice per year), as well as an estimate of the CO2, N2O, CH4 and NMVOC. The emission values are compared to values reported in the NIR, but we expect our information to be more complete than this reporting.

Fuel consumption of heavy machinery and non-road-going vehicles is calculated from reported ownership of said equipment and the reported fuel distribution by the Icelandic Energy Authority.

Information on fuel import and export is collected by the Toll authority (tax authority) and used here.  Purchases of foreign fishing vessels are reported as an export in this data source as well as the purchases of Icelandic operators abroad. Sales of Jet fuel is estimated from arrival and departure information collected by ISAVIA along with survey data of fuel purchase receipts from fuel distribution operator at Keflavík.

Consumption by foreigners operating rental vehicles on the territory is estimated from credit card transaction records from car rental companies as well as purchases of fuel at gas pumps (using credit cards or other electronic payments).

no specific method for seasonal adjustment is used in compiling the AEA data

Not applicable.