| 03 Nov 2025
Deriving regional and point source nitrogen oxides emissions in China from TROPOMI using the directional derivative approach with nonlinear chemical lifetime fitting
Abstract. An appropriate representation of the NOx/NO2 ratio and NOx lifetime is essential for estimating NOx emissions from satellite NO2 observations. We introduce a satellite-based, data-driven approach that applies variable NOx/NO2 ratio and derives a nonlinear chemical lifetime using a piecewise fitting method based on the directional derivative approach (DDA). This method enables the estimation of both regional and point-source NOx emissions across China, representing the first application of a lightweight, satellite-driven method to directly capture nonlinear NOx lifetime for emission estimation over large, topographically complex region. The incorporation of a variable NOx/NO2 ratio enhances the accuracy of source divergence and emission estimates and the improved fitting scheme captures the nonlinear behavior of NOx chemistry. Anthropogenic contributions are isolated by subtracting natural sources from satellite-derived total emissions, with natural NOx identified using a seasonal criterion and further constrained by Nighttime Light (NTL) data. Estimated anthropogenic NOx emissions in China from 2019 to 2024 are 20.2 Tg, 18.5 Tg, 19.4 Tg, 18.9 Tg, 20.7 Tg and 18.8 Tg, respectively, with annual uncertainties of 27 %–30 %. These values show good agreement with both bottom-up inventories and top-down inversions, with national scale discrepancies ranging from −11.8 % to 0.8 %. The DDA captures key spatial and temporal emission patterns, including consistent decline in NOx emissions in megacities and provincial disparities linked to urbanization and economic development. The DDA estimates are consistent with previous studies on coal-fired power plant emissions, and emissions from 124 plants vary between 0.02–2.13 kg s−1 for 2019–2024, with uncertainties spanning 4 %–78 %, averaging 16 %. This satellite-based, lightweight method enables low-latency, timely long-term monitoring of NOx emissions and offers a promising alternative to bottom-up inventories and resource-intensive top-down models. The data are publicly available at https://zenodo.org/records/16787342 (Chen et al., 2025).
The authors present NOx emissions in China derived from TROPOMI column measurements and wind data. The dataset matches the scope of ESSD.
The paper is generally well written and shoul be published after dealing with the following issues:
Manuscript:
Data:
Please specify the processor version for the used TROPOMI data.
Recent processor updates have made some modifications that results in overall higher tropospheric VCDs, which would affect the discussion of the observed low bias in emissions.
Methods:
- The authors refer to previous work, in particular Ayazpour et al. It is not clear to me how far the current dataset is derived from the method described in Ayazpour, or if modifications/extensions have been made. Please explicitly specify what is new/different in this study as compared to Ayazpour.
- The lifetime and inverse scale height are fitted based on Eq. (3) for different levels of NO2 VCDs. However, the basic assumption for this is that emissions are negligible. While this is a good approximation over remote regions (most parts of WN and WS), I wonder how far this assumption can be made over EN for high NO2 VCDs. Please extend the discussion accordingly.
Results:
- Fitted lifetimes are presented in Fig. 4 and discussed in the text. Please provide and discuss the results for scale height as well. I would expect that scale height increases with distance from source regions, i.e. the assumption of constant X needs to be discussed. I would also expect that X might change seasonally due to different lifetimes.
- Table S1: Lifetimes for the test setups are far longer that for control, and also far longer than those reported in several recent studies that estimated NOx lifetimes of the order of 5 hours from (TROP)OMI NO2 patterns downwind strong sources. What is the reason for the large lifetimes in the test setups?
At the same time, fit performance seems to be often better in the test setups than in the control setup. Please comment.
Dataset:
Data is provided on zenodo in form of annual nc files.
The data is easily accessible and readable.
The following items should be clarified/improved:
- please provide some further information how this data was generated or just add a link to the ESSD paper to the attributes.
- please specify the molecular mass the given emissions are refering to (NO2?)
- the unit "tons" is misleading, as this is used differently in e.g. Europe (metric tons) and the US ("short ton").
Please switch to SI units, e.g. 1e3 kg.
- the given emissions are just given in tons. From the context, I conclude that these values are referring to "per year" (as there are annual files) and "per pixel".
This should be clarified.
Due to the link to "pixel", the results depend on the chosen grid, and can NOT be simply interpolated for comparisons to other data.
I would thus recommend to switch to emission rate densities (mass per time per area), which then could be easily interpolated and integrated to any other grid.
Then also the unit would be self-explaining.
In any case, I recommend to add an additional field "area" (lat coordinate only) which would allow for easy conversion between densities and integrated values by the user.