A reanalysis of the BOP dataset : Source apportionment and mineral dust
Een heranalyse van de BOP dataset
05 July 2013, PDF |
76 pages |
Mooibroek D, van der Swaluw E, Hoogerbrugge R
RIVM Report 680356001
In order to determine the effect of policy measures on particulate matter concentrations, it is necessary to get accurate knowledge on the behavior of particulate matter. This report describes results from scientific research based on measurements and calculations of particulate matter. RIVM, TNO and ECN have performed detailed measurements on particulate matter at six locations in the Netherlands. These measurements were performed as part of the first Netherlands Research Program on Particulate Matter (BOP), which aimed at obtaining knowledge on the composition and distribution of particulate matter. The measurements made a distinction between the total amount of particulate matter (PM10) and its finer fraction (PM2.5). This BOP dataset also contains information on the chemical composition of PM. This composition provides important information about the sources of particulate matter, such as industry and traffic. However, some questions about the origin of particulate matter remained unanswered after the first BOP program. Because this lack of knowledge affects the understanding of the effectiveness of policy measures, RIVM has performed a reanalysis of the BOP dataset in the second BOP program, which focused on two questions.
Explanation for inconsistent composition of PM10 and PM2.5 per source Using a source apportionment technique (Positive Matrix Factorization), the PM10 and PM2.5 datasets from BOP, were separately analyzed. The obtained results were similar for the main components of particulate matter. However, for elemental carbon (soot), the result was inconsistent. This inconsistency could result from the fact that the analysis was performed on each dataset separately. To check this, the analysis was repeated, this time on the combined dataset of PM10 and PM2.5. The inconsistency was however not resolved. This means that the source of soot, using the source apportionment analysis, is still not fully understood. Further analysis of the existing data set will probably not answer this question. The fact that the result is not changed by the different PMF, means that the analysis, which is also promoted by the EU, is more robust than expected.
Differences between mineral dust contribution to PM in the Netherlands and Flanders The second question concerned the contribution of mineral dust to PM10 concentrations. According to previous calculations, mineral dust concentrations in Flanders were twice as large as compared to concentrations in the Netherlands. Here it is shown that this difference can largely be explained by the used calculation method. It is shown that in the Flemish calculation some chemical components are assessed differently than in the Netherlands, hence the difference. Therefore, in this case the differences are not attributable to the measurement data. Hence, the results are therefore sufficient for reliable estimates of source contributions and the possible effect of policy measures.