the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Dec 2025
A detailed comparison of the Dutch emission inventory with satellite-derived NOx emissions
Abstract. Nitrogen oxides are one of the most important air pollutants with a large impact on human health. Their emissions are monitored by national emission inventories that are the basis for emission related policies. Because of their large impact on policies these emission data should ideally be verified against independent data, such as emission estimates derived from atmospheric observations. However, this is not yet a widely established practice. Here, we present a detailed comparison of NOx emissions from the Dutch national emission inventory with completely independent emission data derived with the DECSO algorithm from satellite observations by TROPOMI on board of sentinel 5-P. This is enabled by the introduction of a new high-resolution DECSO version DECSO-HR 6.5. We find good agreement in overall emission levels, the spatial emission pattern, the 5-year emission trend, and regional emissions, with deviations in the yearly variation of emissions and at large point sources. Our results demonstrate the robustness of the national inventory and the satellite-derived emissions. This approach might serve as a use-case for the adoption of similar methods in other countries.
- Preprint
(1797 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-6036', Anonymous Referee #1, 17 Jan 2026
- AC1: 'Reply on RC2', Hannes Witt, 19 Mar 2026
-
RC2: 'Comment on egusphere-2025-6036', Anonymous Referee #2, 19 Jan 2026
I found the paper interesting and well written. However, below I list some points that I would like to see being addressed, before publication:
- section 2.5 deals with uncertainties of the different dataset used. DECSO has uncertainty at 8%, the Dutch inventory at 19%...could you please elaborate more on how you came out with such numbers? 8% seems to be very low, and I would like to better understand from where this value comes from
- I think a graphical scheme of how DECSO works would help the readers (and also me). DECSO is well-known, but a graphical scheme would help the reader to grasp the main features of the proposed approach
- the authors present the Dutch inventory (section 2.2) and CAMS-REG-ANT (section 2.4). As far as I know, CAMS-REG-ANT is based on the officially reported (Dutch) data but with different gridding...but not 100% sure about this. Could you please better elaborate on the difference between these 2 inventories?
- in the paper the DECSO, then the Dutch inventory and also CAMS-REG-ANT are briefly described and used for further analysis. But which is the final conclusion of the authors? One should use the Dutch inventory and simply use DECSO for validation? or use DECSO directly, and for example use the Dutch inventory to split in sectors? Please elaborate on this
- could you please explain if the presented approach can also be used for other pollutants, also i.e. in view of future satellite sensors that will be available in the future?
Citation: https://doi.org/10.5194/egusphere-2025-6036-RC2 - AC1: 'Reply on RC2', Hannes Witt, 19 Mar 2026
Status: closed
-
RC1: 'Comment on egusphere-2025-6036', Anonymous Referee #1, 17 Jan 2026
The authors present a detailed comparison of NOx emissions derived from the Dutch national inventory and DECSO emissions. While the work appears scientifically sound, the manuscript would benefit from significant improvements in writing.
General comments:
- The authors state, “we therefore present a detailed comparison of the Dutch national emission inventory and emission data generated by DECSO with the goal to verify the national emission inventory that forms the basis for emission policies. To that end, a high resolution version of DECSO-HR was developed and a new version DECSO 6.5 is introduced.” While the introduction provides some context, it is unclear why DECSO-HR is specifically required to compare the Dutch inventory with DECSO. Providing additional background to clarify this point would strengthen the introduction. Furthermore, the final paragraph of the introduction contains a summary of the paper's major findings, including statements like, "The national inventory and DECSO-HR 6.5 showed good agreement of spatial emission patterns with a general agreement in total annual and monthly national emissions. On the other hand, the respective emission trends showed deviations in the years 2020 and 2021 and there were local deviations at locations with high emission intensities." Summarizing results at this stage is premature; such content is better suited for the conclusion section.
- The authors assert that validating bottom-up emission inventories with satellite-derived emissions is not a widely established practice, framing this as a key motivation for the study. However, I disagree with this claim. For example, the U.S. National Emissions Inventory (NEI) has been extensively validated using emissions inversion techniques. The main novelty of this work appears to lie in the development and application of DECSO-HR compared to the original DECSO. As such, the motivation would be better framed around the introduction of DECSO-HR and using Dutch emissions as a case study. The result session should also include a subsection to discuss the improvements/changes introduced in DECSO-HR, along with a more thorough examination of the uncertainties associated with these innovations.
- Section 3.2. The discrepancies in trends observed between bottom-up emissions and DECSO inventory from 2021 to 2022 are intriguing. Although the authors list several plausible contributors to this discrepancy on the bottom-up side, I wonder whether the fixed lifetime of NOx assumed in DECSO-HR is a significant factor. It may also partly explain larger differences for high-emission-density grid cells discussed in Section 3.3. A sensitivity analysis exploring the uncertainty of the fixed NOx lifetime is strongly recommended to provide clarity.
- Section 3.3. The uncertainties associated with wind information are not accounted for in the analysis. Incorrect wind data could potentially explain the mismatched emission locations, considering the spatial resolution of wind data is coarser than TROPOMI. Performing an additional sensitivity analysis would be valuable in addressing this concern.
Specific comments:
- line 93. The statement, "Note that the a-priori emissions are not from an external inventory but based on the emissions of the previous day," is confusing. I assume that the emissions inventory for the "first" day relies on an external inventory. I don’t quite get how the a-priori emissions are not this external inventory. Perhaps rephrasing this sentence for clarity would help avoid confusion.
- Figure 2 and 4. I recommend using an alternative color to highlight Schiphol airport's runways, as the current color choice makes it difficult to identify them within the plots. Improving the visual distinction would enhance clarity and accessibility for readers.
Citation: https://doi.org/10.5194/egusphere-2025-6036-RC1 - AC1: 'Reply on RC2', Hannes Witt, 19 Mar 2026
-
RC2: 'Comment on egusphere-2025-6036', Anonymous Referee #2, 19 Jan 2026
I found the paper interesting and well written. However, below I list some points that I would like to see being addressed, before publication:
- section 2.5 deals with uncertainties of the different dataset used. DECSO has uncertainty at 8%, the Dutch inventory at 19%...could you please elaborate more on how you came out with such numbers? 8% seems to be very low, and I would like to better understand from where this value comes from
- I think a graphical scheme of how DECSO works would help the readers (and also me). DECSO is well-known, but a graphical scheme would help the reader to grasp the main features of the proposed approach
- the authors present the Dutch inventory (section 2.2) and CAMS-REG-ANT (section 2.4). As far as I know, CAMS-REG-ANT is based on the officially reported (Dutch) data but with different gridding...but not 100% sure about this. Could you please better elaborate on the difference between these 2 inventories?
- in the paper the DECSO, then the Dutch inventory and also CAMS-REG-ANT are briefly described and used for further analysis. But which is the final conclusion of the authors? One should use the Dutch inventory and simply use DECSO for validation? or use DECSO directly, and for example use the Dutch inventory to split in sectors? Please elaborate on this
- could you please explain if the presented approach can also be used for other pollutants, also i.e. in view of future satellite sensors that will be available in the future?
Citation: https://doi.org/10.5194/egusphere-2025-6036-RC2 - AC1: 'Reply on RC2', Hannes Witt, 19 Mar 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 456 | 267 | 30 | 753 | 43 | 46 |
- HTML: 456
- PDF: 267
- XML: 30
- Total: 753
- BibTeX: 43
- EndNote: 46
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors present a detailed comparison of NOx emissions derived from the Dutch national inventory and DECSO emissions. While the work appears scientifically sound, the manuscript would benefit from significant improvements in writing.
General comments:
Specific comments: