the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Dec 2025
China Open Biomass Burning Emissions Inventory (COBBEI) from 2020 to 2023: Multi-Satellite Fusion via FRP-Based Filtering Rules
Abstract. Pollutants released from open biomass burning (OBB) considerably impact air quality, human health, and ecosystems. Only a few studies have used geostationary satellites to monitor OBB emissions in China. Therefore, we construct the China Open Biomass Burning Emissions Inventory (COBBEI) from 2020 to 2023. This dataset included eight pollutants with a temporal resolution of 1 hour and a spatial resolution of 2 km. The COBBEI integrated multi-satellite data, including MODIS, NPP, and Fengyun-4A (FY-4A). The Fire Radiation Power (FRP) data were reconstructed to the FRP cycle, and we integrated the curves to obtain the hourly biomass burned. We also developed five filtering rules based on FRP, considering fire point frequency, radiation values, timing, and variation. These rules were applied to correct the land cover maps, and their validity was verified. The annual average emissions of CO2, CO, CH4, NOx, SO2, PM2.5, K, and LG were 46530, 2262, 132, 82, 25, 247, 11, and 12 Gg, respectively. The spatial distribution characteristics of all eight pollutants were generally consistent. Northeast China served as a major center of pollutant emissions. Different types of fires exhibited various spatial distributions. By validating the method and comparing it with other databases, it was confirmed that COBBEI reduced uncertainty in the OBB emission inventory by providing more information on fire points and effectively screening out fires that were not from OBB. The dataset could offer essential data for air quality modeling, environmental policy development, and fire emergency response strategies. The COBBEI dataset can be downloaded at https://figshare.com (last access: 2025-06-20) with the following DOI: https://doi.org/10.6084/m9.figshare.29367869.v1 (Shi and Ji, 2025).
- Preprint
(1915 KB) - Metadata XML
-
Supplement
(1117 KB) - BibTeX
- EndNote
Status: closed
- RC1: 'Comment on essd-2025-427', Anonymous Referee #1, 19 Jan 2026
-
CC1: 'Comment on essd-2025-427', Qirui Zhong, 17 Feb 2026
In this paper, the authors aim to develop biomass burning emission inventories with high temporal and spatial resolution for mainland China. By combining MODIS, VIIRS, and FY satellite products, they estimate biomass burning emissions for five biome types on an hourly basis. The new dataset developed in this work is of particular value for climate and environmental studies. However, some methodological details need to be clarified. My specific comments are as follows:
- Abstract: What does "LG" stand for? Please provide the full name.
- The pre-processing of different active fire products is based on varying criteria. How do the authors ensure that they are all at a similar confidence level?
- Lines 139–140: If a threshold of 3.5 MW is selected, will this filter out some small fires (e.g., agricultural waste burning)?
- Lines 146–154: How were the criteria (FRPpeak/FRPi+1 > 3, FRPmin/FRPmax > 0.5) selected?
- Lines 166–167: It is unclear what is meant by "single-peaked curve models" and how these models are developed. What are the predictors for FRP?
- It is unclear how data gaps (i.e., times and locations where all satellite observations are missing) are filled.
- Is there any overlap among the three satellite datasets? How do the authors handle cases where multiple data sources are available for the same grid cell?
- Figure 5: What do the different line colors represent?
Citation: https://doi.org/10.5194/essd-2025-427-CC1 -
RC2: 'Comment on essd-2025-427', Anonymous Referee #2, 17 Feb 2026
The author developed a new biomass burning emission inventory over China during 2020-2023. Although the method shows novelty and the data is valuable. However, the data shows serious weaknesses in the use for air quality modelling. I recommend the manuscript cannot be published in the current form.
- The biomass burning emission inventory developed by the authors demonstrates some methodological innovation. However, its usability is limited due to the short temporal coverage. The journal ESSD generally requires data to have broad applicability, so this is a significant drawback.
- Why does the inventory focus only on a limited number of greenhouse gases and short-lived climate pollutants? The omission of NH3, BC, and OC makes it difficult to apply this inventory in models, which severely hinders its usability.
- What exactly is the algorithm for fusing the three types of satellite data? The description provided is not clear.
- Why is the biomass burning emission inventory for North China (Figure 7) so much lower than that for Northeast China? What is the reason for this? If it's due to stricter control measures in North China, does that imply open burning is allowed in the Northeast? The authors need to carefully re-examine their inventory calculations for potential errors.
- The English throughout the manuscript should be further improved.
Citation: https://doi.org/10.5194/essd-2025-427-RC2
Status: closed
-
RC1: 'Comment on essd-2025-427', Anonymous Referee #1, 19 Jan 2026
This study constructed a high spatiotemporal resolution emission inventory for open biomass burning in China (COBBEI) covering 2020–2023. It integrated data from the FY-4A geostationary satellite with polar-orbiting satellites MODIS and VIIRS, significantly reducing interference from non-open biomass burning (OBB) fire points through FRP periodic reconstruction and multi-rule filtering methods. The dataset demonstrated distinct advantages in temporal resolution (1 h), spatial resolution (2 km), and pollutant coverage (including LG and K), offering potential applications for air quality modeling and policy formulation. Overall, the study features clear objectives, reliable data sources, a comprehensive methodological framework, and in-depth analysis, meeting the journal's fundamental requirements. However, based on my review, several issues still require further clarification and improvement, and specific comments are provided below.
- While the Introduction provides a useful overview of OBB impacts and existing BA/FRP-based inventories, the core research question addressed by this study remains implicit. So I suggest that a short paragraph be added at the end of the Introduction to articulate clearly: (i) the central scientific question (e.g., how to reliably characterize spatiotemporal patterns of OBB emissions under current management regimes), and (ii) the practical/engineering question (e.g., constructing a 2 km × 1 h national dataset for air-quality modeling). These questions should then be explicitly linked to the three stated objectives.
- The five FRP-based filtering rules represent a key methodological innovation. However, the specific threshold values are only briefly motivated. Further justification regarding these choices is necessary, and an indication of how robust the results are to these specific values would be beneficial.
- The BD–FRPpeak regressions for the 12 land-cover/region classes are central to the temporal reconstruction. I recommended that the definition of these regional classes should be briefly described to ensure clarity and reproducibility.
- The treatment of wetland fire points (land-cover = 11) requires clarification, as the manuscript currently appears to imply both exclusion and direct retention. A concise statement should be provided explaining how wetlands are handled in relation to the five filtering rules and the subsequent implications for CH₄ emissions.
- The observed bimodal pattern (midday and midnight peaks) in the diurnal cycles is a significant finding. To better highlight the added value of the hourly resolution, a discussion should be included on how neglecting this structure (e.g., relying on daily mean emissions) might bias chemical transport modeling and exposure assessments.
- A short “Limitations and Outlook” paragraph should be included to briefly acknowledge remaining uncertainties and potential future directions.
- The representativeness of the Xichang forest fire case requires justification. Specifically, it should be clarified to what extent this case reflects the diverse fire processes—in terms of type and mechanism—found across different regions of China.
- The established linear regression model between FRPpeak and BD yields R² values ranging from 0.53 to 0.95. Please explicitly identify which regions or fire types exhibit the highest fitting uncertainty and discuss the potential causes for this variance.
- Journal names in the references are inconsistent; should they be abbreviated? Please check and revise the reference format according to the journal's requirements.
Citation: https://doi.org/10.5194/essd-2025-427-RC1 -
CC1: 'Comment on essd-2025-427', Qirui Zhong, 17 Feb 2026
In this paper, the authors aim to develop biomass burning emission inventories with high temporal and spatial resolution for mainland China. By combining MODIS, VIIRS, and FY satellite products, they estimate biomass burning emissions for five biome types on an hourly basis. The new dataset developed in this work is of particular value for climate and environmental studies. However, some methodological details need to be clarified. My specific comments are as follows:
- Abstract: What does "LG" stand for? Please provide the full name.
- The pre-processing of different active fire products is based on varying criteria. How do the authors ensure that they are all at a similar confidence level?
- Lines 139–140: If a threshold of 3.5 MW is selected, will this filter out some small fires (e.g., agricultural waste burning)?
- Lines 146–154: How were the criteria (FRPpeak/FRPi+1 > 3, FRPmin/FRPmax > 0.5) selected?
- Lines 166–167: It is unclear what is meant by "single-peaked curve models" and how these models are developed. What are the predictors for FRP?
- It is unclear how data gaps (i.e., times and locations where all satellite observations are missing) are filled.
- Is there any overlap among the three satellite datasets? How do the authors handle cases where multiple data sources are available for the same grid cell?
- Figure 5: What do the different line colors represent?
Citation: https://doi.org/10.5194/essd-2025-427-CC1 -
RC2: 'Comment on essd-2025-427', Anonymous Referee #2, 17 Feb 2026
The author developed a new biomass burning emission inventory over China during 2020-2023. Although the method shows novelty and the data is valuable. However, the data shows serious weaknesses in the use for air quality modelling. I recommend the manuscript cannot be published in the current form.
- The biomass burning emission inventory developed by the authors demonstrates some methodological innovation. However, its usability is limited due to the short temporal coverage. The journal ESSD generally requires data to have broad applicability, so this is a significant drawback.
- Why does the inventory focus only on a limited number of greenhouse gases and short-lived climate pollutants? The omission of NH3, BC, and OC makes it difficult to apply this inventory in models, which severely hinders its usability.
- What exactly is the algorithm for fusing the three types of satellite data? The description provided is not clear.
- Why is the biomass burning emission inventory for North China (Figure 7) so much lower than that for Northeast China? What is the reason for this? If it's due to stricter control measures in North China, does that imply open burning is allowed in the Northeast? The authors need to carefully re-examine their inventory calculations for potential errors.
- The English throughout the manuscript should be further improved.
Citation: https://doi.org/10.5194/essd-2025-427-RC2
Data sets
COBBEI: China Open Biomass Burning Emission Inventory Jingwen Shi and Yaqin Ji https://doi.org/10.6084/m9.figshare.29367869.v1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 341 | 150 | 24 | 515 | 90 | 20 | 20 |
- HTML: 341
- PDF: 150
- XML: 24
- Total: 515
- Supplement: 90
- BibTeX: 20
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study constructed a high spatiotemporal resolution emission inventory for open biomass burning in China (COBBEI) covering 2020–2023. It integrated data from the FY-4A geostationary satellite with polar-orbiting satellites MODIS and VIIRS, significantly reducing interference from non-open biomass burning (OBB) fire points through FRP periodic reconstruction and multi-rule filtering methods. The dataset demonstrated distinct advantages in temporal resolution (1 h), spatial resolution (2 km), and pollutant coverage (including LG and K), offering potential applications for air quality modeling and policy formulation. Overall, the study features clear objectives, reliable data sources, a comprehensive methodological framework, and in-depth analysis, meeting the journal's fundamental requirements. However, based on my review, several issues still require further clarification and improvement, and specific comments are provided below.