Annex 8 Data Quality

Because there is some level of uncertainty in all data, compiling inventories always entails uncertainty. There is uncertainty associated with data on both activity rates (e.g. reliability of data sources and data collection procedure) and emission factors (e.g. quality of measurement data). This is not a problem, since the purpose of the inventories is to assess the national situation by identifying sources and estimating releases of particular pollutants, setting priorities, developing an action plan to minimize these releases, and evaluating the progress achieved by assessing the trends observed over time. If the same methodology is consistently being applied when reassessing the country’s situation regarding PCDD/PCDF emission, then this purpose can be met.

The following sources of uncertainty may be considered when developing a national PCDD/PCDF inventory:

Default emission factors: Two questions need to be answered when assessing emission factor quality:

• How reliable are the data that were used to derive the emission factor? and
• How well does the emission factor represent the emission source under different national circumstances, i.e. can it be appropriately used as a global average emission factor for a specific source activity?

Evaluation of the quality of emission factors included or to be included in the Toolkit is the mandate of the Toolkit experts, who should ascertain that only scientifically-sound data are included into the Toolkit. Data quality ratings are assigned by the Toolkit experts to all emission factors published in the Toolkit.

Classifying sources: There is uncertainty in classifying sources and, consequently, selecting appropriate emission factors from the Toolkit while developing the national inventory.

Activity rates: There is uncertainty in generating activity data for inventory compilation.

The overall level of uncertainty in the complete inventory is a combination of the above three elements.

The most complete guidance document currently available is the IPCC Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories. However, this document was developed with a different purpose, and is not suitable for developing PCDD/PCDF emission inventories under the Stockholm Convention.

For the purpose of emission inventories under the Stockholm Convention, a simple approach using qualitative data quality rating is recommended. An emission factor's rating is a general indication of the reliability, or robustness, of that factor. This rating is assigned based on the estimated reliability of the experiments used to develop the factor and on both the amount and the representative characteristics of those data. This approach can be used to assess the underlying confidence of the author of the inventory in the data used to generate emissions estimates. It is most suitable for estimation approaches that rely on emission factors, as well as estimates of activity rates. In all cases, higher quality ratings would be given to more direct approaches based on measurement data.

The following criteria are used to assign quality ranks to the Toolkit’s emission factors:

• Data/information used to derive emission factors were evaluated by a formal process of peer review. Publications/reports, reviewed by the Toolkit expert group will equally be considered as peer-reviewed.
• Data range: High variability of available data may trigger the use of median emission factors that do not fully reflect certain operational circumstances. A wide range of the data used to derive an emission factor would therefore reduce the confidence in applying the respective emission factor to a concrete situation.
• Geographical coverage: Experimental data with low variability derived at many locations over the world would increase the confidence in using the respective emission factor in different national circumstances.
• Need for extrapolation: The need for extrapolation/expert judgment in order to fill data gaps reduces the confidence in the emission factors derived by making certain assumptions e.g. based on information from similar classes.
• Process stability: High stability of the process generating PCDD/PCDF generally increases the confidence in the experimental results used to derive emission factors. It should be noted that high variability of emission data may be obtained even with a stable process. High confidence levels are to be assigned to emission factors derived from data with well-described variability.

Taking into account the criteria described above, each emission factor is assigned a data quality rating according to the following definitions:

Qualifier / Level of confidence	Criteria
High	Peer review Low data range Broad geographical coverage Assumptions and/or expert judgment are not required High stability of the process
Medium	Any combination of high and low criteria
Low	No peer review Wide data range Limited geographical coverage Extrapolation is needed e.g. EF derived from similar class Low stability of the process

The studies used to derive the Toolkit’s emission factors are made accessible for the users whenever possible, to ensure that emission factors are applied with the specified level of confidence only when they match a certain situation. Where expert judgment is used to derive emission factors, clear information on the extrapolation process is provided, along with justification of the assumptions made.

Ratings can be similarly assigned to the activity or production data, taking into account the reliability of the data source, data collection process, number of data points (e.g. for questionnaires) etc. According to these criteria, activity data taken from national statistics, or derived from questionnaires with a high return rate may be used with high confidence. Activity data based on assumptions about certain sources (especially with respect to technologies in place), where specific information could not be collected, will be assigned a low level of confidence.

Due to the iterative revision process of the Toolkit, the emission factors are periodically verified, updated and supplemented via targeted projects and research. Furthermore, as the focus of the Toolkit is exclusively on unintentional POPs, it can be regarded as a comprehensive compilation of emission factors for POPs releases to air, water, soil, products and residues.