EISSA Methodology

‘Emission Inventory Support’ in general covers a broad range of different aspects: bottom‐up estimation of emissions, spatial and temporal allocation thereof, and spatial downscaling of existing inventories amongst others. Moreover, a suitable solution to a specific emission request strongly depends on the availability of and access to local information such as activity data (e.g. traffic volumes), emission factors, geostatistical data (e.g. population density) etc., on the scope of the request (e.g. emission reporting vs. emissions as input to air quality models), and on the scale (e.g. rough estimate of regional emissions vs. detailed local emissions).

Over the past years, VITO built up a lot of expertise in modelling emissions to air. This resulted in an integrated Air Emission Inventory Support System, which allows to assist governments, local air districts and other stakeholders in reporting emissions and controlling air quality. The integrated Air Emission Inventory Support System comprises extensive expertise and several tools. The EISSA framework acts as the heart of the integrated Air Emission Inventory Support System and is flanked by several other tools. On one hand ‘pre-processing’ tools, applied for very specific detailed emission bottom-up estimations, on the other hand ‘post-processing’ tools, in general being used for conversion of inventories towards air quality model input.

Briefly, the use of the different tools within the integrated Air Emission Inventory Support System can be summarized as follows:

  • EISSA is applied whenever emission estimation, mapping, and analyzing is required for a selection of substances, for one, several or all sectors, and making use of a bottom-up approach in case
    • computation of emissions can occur through the basic formula ‘activity’ times ‘emission factor’;
    • required activity data and emission factors are available;
    • the emission sources can be specified in a hierarchical structure of maximum five levels;
    • output at raster level (at whatever resolution) is sufficient
  • Pre-processing tools are used whenever very specific user needs require adaptations to EISSA that are conflicting with the philosophy of EISSA being generic. Examples are
    • Estimation of residential emissions, required at Tier-2 level. This assumes distinction between sectors and subsectors, fuel types and subtypes, installation types and subtypes, installation age and efficiency, … which would lead to a source structure of more than 10 hierarchical levels. Expanding the EISSA hierarchical source structure to more than 5 levels is technically possible, though was considered as too non-generic;
    • Estimation of traffic emissions for which the output is required at line segments rather than at raster level. Adding the functionality to compute line-output is not straightforward as it is not compatible with the available analysis functionalities that are based on raster formats. Expansion of the EISSA functionality is technically possible, though was considered as too non-generic;
  • Post-processing tools are in general being used to convert EISSA emissions to emissions directly suitable for use in air quality models. Examples of post-processing are
    • Complementing with emissions outside the hotspot-region, as air quality models require emissions at an extended area. This usually involves combining emission estimation for data rich and data poor areas;
    • Adding temporal profiles, allowing to estimate hourly rather than yearly emissions;
    • Formatting to specific air quality model input format, which is different for every single air quality model;

The most crucial aspect of the integrated Air Emission Inventory Support System is the  central role of EISSA, made possible by fine-tuning of the different tools so they operate closely together, and its integrating character. All tools together enable to cover the whole range of aspects involved in emission inventorying, reporting, mapping, analyzing and converting to air quality model input.

 

Bottom-up emission estimation within EISSA

The representation of the sources in EISSA is based on a basic formula stating that the gross emission (GE) of a substance is the product of an emission explanatory variable (EEV) and an emission factor (EF).

The emission explanatory variable (EVV) is the physical activity or the physical element that causes the emission. Examples are the amount of kilometers driven by a vehicle of a specific vehicle type, the amount of specific fuel, e.g. solid, liquid or gaseous fuel, used for residential heating. The emission explanatory variables are generally represented in a spatially explicit manner. They are spatial patterns showing the occurrences of the sources. To the effect, EISSA is supporting the user to generate a map featuring the pattern.

The emission factors (EF) represent the quantity of substance that is annually released per unit of the emission explanatory variable. Thus, there is an emission factor that represents the amount of emissions released per driven kilometer and another one specifying the amount of emissions per unit of  combusted fuel. In EISSA, emission factors may also have a spatial dimension.

Scientific research is required to define both the EEV and the EF per source-substance combination. EISSA supports the factual use of this scientific knowledge in policy-making by automating the necessary calculations.

Integration of dedicated tools in EISSA

Very specific user needs might require drastic adaptations to EISSA in order to allow compilation of emissions by means of a bottom-up approach and according to the desired level of detail. Examples are detailed traffic or residential emissions. As these adaptations often conflict with the philosophy of EISSA being generic, EISSA was extended with functionality enabling top-down estimation of emissions. This approach starts from emission data available at an aggregate level (national, regional, municipal) and disaggregates these emission to grid cells based on the best available geographical proxies and advanced mapping algorithms.

In the EISSA top-down approach, both the data at aggregate level (e.g. traffic emission total for Flanders, summed over all vehicle types) and the geographical proxy (e.g. the emission map as computed by the dedicated traffic emission model) are taken from the relevant dedicated model. The geographical proxy is the normalized output of the dedicated model, which is inserted in EISSA. Top-down emission estimation occurs through the basic formula stating that the gross emission (GE) of a substance is the product of the geographical proxy (P) and an emission total (E). An example is given below..

Detailed traffic emissions for Flanders are usually being computed with the dedicated road traffic emissions model, named FASTRACE. This model takes into account (over 250) different vehicle types, a user-defined amount of different road types, speed-dependent emission factors, etc.. As a result, for a specific substance, different emission maps per vehicle type on line-segments are obtained. These maps are, for instance, directly being used by air quality models that are applied to investigate traffic related policy measures. Other applications, e.g. regional modelling applications, only require less detailed traffic emissions. Instead, the traffic emissions should be complemented with emissions for all other relevant sectors. Hereto EISSA is used. Computation of traffic emissions within EISSA, requires pre-configuration of EISSA: FASTRACE output is first aggregated over all vehicle types, then converted to raster format, and finally normalized. This ‘proxy map’ is used by the system to compute gross emissions in a top-down way: this map (proportions) is multiplied with the total emissions (E), also resulting from FASTRACE.

Examples of EISSA as integrating tool

Based on our integrated approach, customized solutions can be offered for any application and for any (data‐poor or data-rich) region on Earth. Some examples are given below. For a more detailed description, see  EISSA Applications.

 

EISSA for Persistent Organic Pollutants (POPs) in Flanders

In this project an EISSA application, enabling emission estimation of persistent organic pollutants (POP’s) and the spatial allocation thereof, was developed. The tool comprises about 34 POP’s, 8 main sectors split into several subsectors and sources, and several years (1990, 1995, 2000, 2005 and 2010). All emissions are computed through a top-down approach. Proxies and emission totals are resulting from dedicated models: FASTRACE for traffic emissions, OFFREM for off-road emissions, GEOGREMIS for residential heating emissions, etc.

 

EISSA for Residential Heating Emissions in Flanders

- Under development

The  EISSA application computes and analyzes emissions stemming from residential heating (including households, agriculture and horticulture, and the tertiary sector) in Flanders. The estimation of the emissions occurs through a bottom-up approach, on a tier-2 level, in which emission factors not depend solely on the fuel type, but also on the type and age of the heating installation. 

 

EISSA applications for Residential Heating Emissions in Eastern Europe 

- Under development

In this project, an inter-regional framework for both emissions and air quality modelling will be developed. The emission modelling component will build on the existing EISS framework. EISSA will be adapted and applied to compile high resolution residential emission inventories for the different sub-regions (Małopolska and Silesia Region in Poland, the Czech Republic, and Slovakia). As for the application in Flanders, emissions will be computed bottom-up wise. These emissions will then be integrated in a trans-boundary emission data set, required for regional air quality modelling. The transboundary EISSA application will be based on a top-down approach.