Synergistic use of satellite and in-situ data for policy-relevant air quality information: A case study on Belgium
| dc.contributor.author | Verhoelst, T. | |
| dc.contributor.author | Compernolle, S. | |
| dc.contributor.author | Lambert, J.-C. | |
| dc.contributor.author | Vanpoucke, C. | |
| dc.contributor.author | Fierens, F. | |
| dc.date | 2025 | |
| dc.date.accessioned | 2025-11-04T13:44:15Z | |
| dc.date.available | 2025-11-04T13:44:15Z | |
| dc.identifier.uri | https://orfeo.belnet.be/handle/internal/14391 | |
| dc.description | As in most European countries, the impact of Air Quality policy in Belgium is monitored with an accurate but relatively sparse network of instruments measuring in situ the near-surface concentration of various pollutants. Assessment at unmeasured locations is based on interpolation techniques and/or numerical models. We demonstrate how today’s dedicated satellite data sets for AQ, and in particular Sentinel-5 Precursor TROPOMI NO2 , can be used in conjunction with the near-surface data to assess AQ over the entire Belgian domain in an observational approach virtually independent from assumptions on (changing) emissions. Specifically, after temporal aggregation and spatial oversampling (1x1km2), these satellite data reveal policy-relevant spatial and temporal features, and a high correlation (r - 0.9) with the near-surface measurements where these are available, in particular when adjusting the latter for local representativeness with land-cover data. This tight relation allows for a pragmatic conversion of tropospheric columns to near-surface concentrations over the entire Belgian domain using Regression Kriging, yielding uncertainties (- 12%) well below the amplitude of the larger spatio-temporal features over Belgium. This synergistic product, named LEGO-4-AQ, is evaluated against the Belgian assessment model RIO, revealing good agreement (r -m 0.85) but also interesting local deviations. Both satellite tropospheric vertical column densities and synergistic near-surface concentrations are found to have decreased by 3-10%/year over the period 2019-2024, with midday values at a 4x4km resolution now mostly below the WHO annual (all-day) exposure guideline of 10 ug/m. The three urban low-emission zones already in place in Belgium do not (yet) show stronger reductions than those observed in other (sub-)urban parts of the country. | |
| dc.language | eng | |
| dc.title | Synergistic use of satellite and in-situ data for policy-relevant air quality information: A case study on Belgium | |
| dc.type | Article | |
| dc.subject.frascati | Earth and related Environmental sciences | |
| dc.audience | Scientific | |
| dc.subject.free | Air quality | |
| dc.subject.free | Nitrogen dioxide | |
| dc.subject.free | TROPOMI | |
| dc.subject.free | Oversampling | |
| dc.subject.free | Regression Kriging | |
| dc.subject.free | Low Emission Zone (LEZ) | |
| dc.source.title | Atmospheric Environment | |
| dc.source.volume | 361 | |
| dc.source.page | A121447 | |
| Orfeo.peerreviewed | Yes | |
| dc.identifier.doi | 10.1016/j.atmosenv.2025.121447 | |
| dc.identifier.url |