Source Apportionment

Source contributions per concentration bin in the Netherlands, divided per sector.
One of the main targets in air quality modeling is to know the origin of air pollution. Ideally, information about the contribution of individual emission sources is known at a certain location, for example locations within a large urban or industrialized area. The different sources can be for example road traffic versus industry, emissions from abroad versus emissions from inside the country of interest and many others. This information of the origin of air pollution can be used for policy measures to create an effective reduction of air pollution.
Until now the individual contributions of sources to pollutant concentrations are mostly calculated by so-called scenario runs. In such a scenario run, the emissions of only one (or a few) sources are turned off in the model. The difference between the original run and the scenario run is used to determine the impact of the emission source. This means that for each emission source of interest, the model must be run separately. Another issue in the approach of scenario runs is that for chemical active tracers, the non-linear chemical regime is influenced. This implies that the difference between the basis run and the scenario run is not equal to only the contribution of the emission source(s), which is (are) switched off.
Within the chemistry transport model LOTOS-EUROS, a labeling method is developed to make a better source apportionment study. In this labeling method, the emissions of several sources of interest can be labeled. During all the model processes, the labels are tracked, such that the resulting concentrations can be coupled with the originating emissions. For the non-linear processes, the labeling method is more accurate to make a coupling of concentrations with originating emissions.
The labeling method is used in several projects. For example, an application of the labeling method is used for the BOP campaign; in this campaign a study of the origin of modeled PM concentrations in the Netherlands is done. In the top figure, one of the intermediate results is shown; the concentrations nitrate aerosol (left) and total pm10 (right) which is from road transport. Note that the city hotspots are visible in this picture. In the bottom figure the contributions of different sources to PM10 in the Netherlands are given with respect to the PM10 concentrations. In the left panel one can clearly notice that the Dutch sources contribute only 20-25% to the total concentration, while the foreign sources contribute about 45-50%. The contribution of the boundary conditions can be allocated to windblown dust from the Sahara and from long-lived tracers which comes from outside the European domain. In the right panel one can see that the contribution of agriculture and road transport grows if the concentrations are higher. On the other hand the contribution of natural sources is large when the concentrations are low.
Example of modelled nitrate aerosol from Dutch road transport (left), and modelled PM10 from Dutch road transport (right) in the Netherlands
Source contributions per concentration bin in the Netherlands. Sources divided in Dutch and foreign (left) and divided per sector (right).
Kranenburg, R., Hendriks, C., Schaap, M., and Segers, A.
Source apportionment using LOTOS-EUROS: module description and evaluation.
Geosci. Model Dev., 6, 721-733, 2013, doi:10.5194/gmd-6-721-2013