TOPAS Documentation
Exposure to air pollution is the world’s largest environmental health threat, causing millions of premature deaths globally and hundreds of thousands in Europe, mainly due to fine particulate matter (PM2.5). To understand the problem, to inform the public, to fulfill reporting obligations, and to cost-effectively reduce pollution it is imperative that authorities have accurate information on its origin and effectiveness of potential measures.
TOPAS, TNO’s Operational Pollution Apportionment Service has been developed to provide key information in an efficient way, and to support and advice a variety of stakeholders, with the overall goal to reduce pollution levels and related negative health impacts.
The system provides daily information on source sectors and regions contributing to the main atmospheric components relevant for air quality policy and health. The tool is under continuous development to improve its quality and meet user needs.
Daily source attribution system
The TOPAS service delivers daily source attribution. It fully exploits the regional Chemical Transport Model (CTM) LOTOS EUROS (Manders et al., 2017), driven by ECMWFmeteorology, TNO emission information, as well as time resolved information for boundary conditions and fire emissions from CAMS. It tracks the contributions of predefined source categories (anthropogenic source sectors, natural sources and/or source regions) to air pollution levels in European cities and/or at measurement locations.
Every day, the source apportionment simulation is extended with the previous day. Evaluation of the modelled results is performed periodically, using in-situ data.
Hourly and yearly contributions
Besides the regular daily averaged source apportionment results as presented on the TOPAS website, the system also allows for more detailed analysis using the hourly outputs and can give yearly totals for a more generic overview of contributions at specific locations.
TOPAS-EU shows daily values for the past 6 weeks and yearly totals for the recent years. TOPAS-NL currently shows only daily averages. TOPAS-CH4 shows both hourly and daily contributions.
Are you interested in other time resolutions for specific TOPAS versions? Please don't hesitate to reach out!
Source attribution method
The source attribution in TOPAS is obtained by a labelling/tagging method (Kranenburg et al., 2013). After emission of pollutants, the different components are followed through the model system, while keeping track of its source sector or region. Results are valid for current atmospheric conditions. By definition, the sum of all contributions equals the total concentration, and the sum of individual contributions equals the combined contribution. In case of strong non-linearities influencing the relationship between emissions and concentrations, the contributions cannot directly be translated to a potential emission reduction impact. For these cases, we can do separate offline scenario runs or use our new TOPAS mitigation tool. The TOPAS mitigation tool is based on emission reduction model runs and provide potential impacts of emission reductions. The TOPAS mitigation tool will be available in 2026.
Labelling
Labelling of PM
The routine is implemented for primary, inert aerosol tracers as well as chemically active tracers containing a C, N (reduced and oxidized) or S atom, as these are conserved and traceable. The source attribution is valid for current atmospheric conditions as all chemical conversions occur under the same oxidant levels. For secondary aerosols consisting of two components (e.g. ammonium nitrate (NH4NO3)), the contribution is calculated by accounting half of the mass to each component source. For example for ammonium nitrate (NH4NO3) half of the mass is attributed to the ammonium source sector or location (mostly ammonia (NH3) from agricultural sources) and half of the mass to the nitrate source sector or location (a.o. nitrogen dioxide (NO2) from traffic). Both components are required for formation of the combined aerosol. The contributions are preserved when transported to another region. The chosen method leads to a larger agricultural contribution than the method where the ammonium nitrate mass is allocated to the ammonium and nitrate sources based on the respective molecular masses of these components (NH4+ is lighter than NO3-). Because the combined aerosol cannot be formed if either of the two components is unavailable, attributing equal contributions seems justified.
Methane tagging
The labelling implementation for methane follows the above method for chemically active tracers containing a carbon atom and has been developed within the Horizon Europe project Avengers. More information can be found in this report.
Ozone tagging
For the source attribution of ozone the existing labelling system as described above was extended to account for ozone production due to NOx (NOx tagging) and VOC (VOC tagging) sources similar to the TOAST method (GMD - TOAST 1.0: Tropospheric Ozone Attribution of Sources with Tagging for CESM 1.2.2).
In this extension an extra labelled family (Ox-family) is introduced with tracers in the O3 cycle (O3, NO2, NO3, N2O5, HNO3, PAN + in VOC tagging routine also HO2). To ‘activate’ the labels in the Ox-family, reactions which produce NO2 from NO and a peroxy-radical compound are taken. If a reaction takes place to convert NO into NO2 without using an O3 molecule but using an (in)organic peroxy-radical, one molecule of O3 is ‘spared’ in the photo synthetical equilibrium between NO, NO2 and O3. This can be expressed as production of O3.
In the NOx tagging, the source contributions of NO (tracked by the previously described method following conserved N) are used to update the source contribution of NO2 in the Ox-family. After this all reactions between components within the Ox family are tracked as well and labels are transferred within this group of tracers. All other processes in the model (deposition, diffusion and advection) handles the source apportionment identical to the existing method for PM and precursors.
In the VOC tagging the labels in the Ox family are not taken from the contributions of the NO involved in the chemical reactions but from the participating peroxy-radicals that are produced by VOC oxidation.
Currently TOPAS is only showing the source attribution related to the NOx sources.
Emissions
The LOTOS-EUROS model applied in TOPAS is driven by the CAMS-REG European emission inventory (Kuenen et al., 2022) produced by TNO for Europe, GRETA emissions for Germany and Dutch Emission Registration emissions for the Netherlands. Where possible temporal variable emissions are calculated based on meteorological data, such as domestic heating emissions based on heating degree days.
Emissions of biogenic volatile organic compounds, nitrogen oxides, sea salt and soil dust are modelled in LOTOS-EUROS using source parametrizations depending on meteorology, vegetation maps, salinity and soil conditions.
Emission estimates of aerosols, chemical species, and greenhouse gases from wildfires are taken on a daily basis from CAMS GFAS product.
Sector attributions
The sector source attribution product allows focus on various sectors of trade and activities in Europe as well as on natural sources and provides insight in their contributions to air pollutant concentrations. The source sector allocation product allows a differentiation between local (native) and non-local (non-native) contributions. The local contribution is defined as the mass contribution from a given sector originating from PM and precursor emissions from the grid cell in which the point of interest is located. See overview of sectors for the different sectors available.
Country attributions
The country source attribution product quantifies the trans-boundary contribution to modelled air pollution episodes and identifies the most important contributing countries. More information about the countries in TOPAS.
Flexibility
While the online TOPAS service is using a fixed set of labels as given above, note that the definition of the labels for offline use is flexible, which allows customized studies to be performed.
Tailoring the service to a particular region is also possible by including regional specific emission information for the region of interest. Local activity patterns can be used to prescribe the emission variability.
TOPAS specifications
| TOPAS-EU | TOPAS-O3 | TOPAS-NL | TOPAS-CH4 | |
|---|---|---|---|---|
|
Model version |
LOTOS-EUROS v2.3.000 | LOTOS-EUROS v2.2.003 (plus specific updates) | ||
| Spatial coverage | Europe | Netherlands | Europe | |
| Horizontal resolution | 0.2º x 0.1º (~10-15 km) | 1 km | 0.4º x 0.2º | |
| Vertical resolution | 12 layers up to ~8 km with about 7 layers in first 1000m | |||
| Time coverage | past 6 weeks and yearly averages for 2023, 2024 | past 6 weeks | past 2 weeks and yearly average past year (2024) | |
| Time resolution | daily and yearly averages from hourly data | daily averages from hourly data | hourly values and daily averages | |
| Pollutants | PM2.5, PM10 | O3 | PM2.5, PM10, NO2, SO2 | CH4 |
| Emissions | CAMS-REG v8.1 | CAMS-REG v8.1 + GRETA (Germany) + ER 2023 (Netherlands) | CAMS-REG_v5_1, wetlands emissions (LPJ-GUESS) | |
| Emission time profiles | CAMS TEMPO v5.1 + heating degree day approach for residential combustion | |||
| Fire emissions | CAMS-GFAS | |||
| Boundary conditions | IFS-COMPO excl sea salt | TOPAS-EU | CAMS Global NRT plus bias correction based on observations | |
| Meteorological drivers | 12:00 UTC D-1 (yesterday's) operational IFS forecast | |||
| Locations | cities represented by urban city core masks from urban audit, and eea observation stations | official LML and citizen science (samen meten) observation sites | ICOS observation sites at different heights | |