the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Oct 2023
Diagnosing uncertainties in global biomass burning emission inventories and their impact on modeled air pollutants
Abstract. Biomass burning (BB) emission inventories are often used to understand the interactions of aerosols with weather and climate. However, large uncertainties exist among current BB inventories, so the choice of inventories can greatly affect model results. To quantify the differences among BB emission inventories and reveal their reasons, we compared carbon monoxide (CO) and organic carbon (OC) emissions from seven major BB regions globally from 2013 to 2016. The current inventories are based on two basic approaches: (1) bottom-up approach, which establishes inventories based on observed surface data, and (2) top-down approach, which based on the release rate of radiative energy from vegetation burning. In this study, we selected mainstream bottom-up inventories Fire INventory from NCAR 1.5 (FINN1.5) and Global Fire Emissions Database version 4s (GFED4s), and the top-down inventories Quick Fire Emissions Dataset 2.5 (QFED2.5) and VIIRS-based Fire Emission Inventory version 0 (VFEI0). We find that the total global CO emissions fluctuate between 252 and 336 Tg and the regional bias is even larger, which can be up to six times. Dry matter is responsible for most of the regional variation in CO emissions (50–80 %), with emission factors accounting for the remaining 20–50 %. Uncertainties in dry matter often come from biases in the calculation of bottom fuel consumption and burned area, which are closely related to vegetation classification methods and fire detection products. In the tropics, peatlands contribute more fuel loads and higher emission factors than grasslands. At high latitudes, as cloud fraction increases, the bias between burned area (or fire radiative power) increases by 20 %. In addition, due to the corrected emission factors in QFED2.5, global BB OC emissions have higher variability, fluctuating between 14.9 and 42.9 Tg.
Finally, we applied the four sets of BB emission inventories to the Community Atmosphere Model version 6 (CAM6) and compared the model results with observations. Our results suggest that the simulations based on the GFED4s agree best with the MOPITT-retrieved CO. We also compared the simulation results with satellite or ground-based measurments, such as Moderate Resolution Imaging Spectroradiometer (MODIS) AOD and AErosol RObotic NETwork (AERONET) AOD. Our results reveal that there is no global optimal choice for the BB inventories, but we give certain inventory recommendations based on different study areas and spatiotemporal scales. This study has implications for reducing the uncertainties in emissions or improving BB emission inventories in further studies.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(5862 KB)
-
Supplement
(2214 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5862 KB) - Metadata XML
-
Supplement
(2214 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1822', Anonymous Referee #1, 16 Nov 2023
This study diagnoses uncertainties in global biomass burning emission inventories and discusses the causes of large biases. In this study, the authors compared gas- and particulate-phase emissions from four biomass burning emission inventories established by bottom-up and top-down approaches. The authors quantified the contribution of different factors to the uncertainty in biomass burning emissions and proposed that dry matter was the main cause of regional bias in CO emission estimates. Vegetation classification methods and fire detection products led to the uncertainties in bottom fuel consumption and burned area calculations, resulted in biases in dry matter. They reported that the variability of particulate-phase emission was even higher than that of gas-phase emission. In addition, they compared the simulated results with satellite measurements, and given certain inventory recommendations based on different study areas and spatiotemporal scales. This study is well written and well organized, and could support improvements in biomass burning emission inventories in further studies. I recommend accepting it after minor revisions.
- The second paragraph beginning with “Recent studies” can be combined with the third paragraph beginning with “Previous studies”.
- Section 2. The description of Biomass Burning emission inventories and Quantitative statistical methods can be shortening. Some details can be moved to supplementary information.
- Line 545-561. There is a little bit of confusion about this paragraph. The authors said, “The total QFED FRP is 1.5 times higher than VFEI0, but DM in QFED2.5 inventory is 30% lower than VFEI0”, and also said: “Therefore, although the two top-down emission inventories use similar algorithms, significant bias occurs under high cloud fraction conditions, with QFED2.5 estimating DM much higher than VFEI0”. So, does the low DM in the QFED2.5 mainly occur under low or medium cloud fraction? Could the authors give some specific values?
- Figure 2 showed that FINN1.5 estimated much larger CO emissions than other emission inventories in EQAS. The authors also selected the EQAS as one of the important biomass burning regions based on the fact that “(1) regional BB CO emissions above 20Tg/yr, (2) BB CO emissions account for more than 70% of the total”. However, table 3 shows similar CO column averages across the four BB emission inventories. Could the authors also explain why? Additionally, it is important to show simulated near-surface/different vertical layer CO concentrations.
- Line 660-665. Is there any reference to support the conclusion that the overestimation of SOA in South Hemispheric South America is due to biogenic sources?
Citation: https://doi.org/10.5194/egusphere-2023-1822-RC1 - AC1: 'Reply on RC1', Sijia Lou, 02 Feb 2024
-
RC2: 'Comment on egusphere-2023-1822', Anonymous Referee #2, 27 Dec 2023
This study mainly focused on investigating the differences among widely adopted emission inventories of biomass burning and revealing the main reasons. It is a necessary work for improving the emission inventories in the future. The manuscript can be accepted after the following questions addressed.
- The abstract should be shorted and refined.
- Line 41: we give certain inventory recommendations based on different study areas and spatiotemporal scales. What are your certain recommendations for the global emission inventory? Please briefly describe in abstract, which is important to help readers find the key points of the paper.
- The introduction should be re-written again. The authors listed so many literatures, while they do not better summarize them. At the end of each paragraph, the main contents or research shortages should be given. Line 123-150 is not necessary to give these equations in the introduction. Line 181-190, wordy sentences.
- The calculation of DM (Line 411-419), FC (Line 508-510), EF (567-569) may not be included in part 3. This is the introduction of methods, instead of the analysis of emission inventories.
- Line 550-553: How empirical factor affect the amount of DM? It is also an important factor with uncertainty.
- The uncertainty of emission inventories was impacted by a combination of those factors (EF, DM). Monte Carlo simulations were usually performed in articles to evaluate the estimation uncertainty quantitatively for pollutant emissions. So those combined uncertainty results of the emission inventories in different regions can be compared, which may be associated to the regional applicability of BB emission inventories.
Conclusion：wordy sentences. Line 715-720, we all know these, and they are not your conclusions or new findings.
From Figure.1 to Figure. 7, all the datasets are from the literatures or former studies. These should not be the main contents of this study. The modeling works, especially for the comparison between the modeling results, the bias and the reasons should be emphasized.
Table 1, all the emission factors adopted in the four emission inventories can not reflect the real emission situation. Can the author optimize the emission factors with your modeling works? I think it will contribute more to science, but not just compare the values.
Citation: https://doi.org/10.5194/egusphere-2023-1822-RC2 - AC2: 'Reply on RC2', Sijia Lou, 02 Feb 2024
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1822', Anonymous Referee #1, 16 Nov 2023
This study diagnoses uncertainties in global biomass burning emission inventories and discusses the causes of large biases. In this study, the authors compared gas- and particulate-phase emissions from four biomass burning emission inventories established by bottom-up and top-down approaches. The authors quantified the contribution of different factors to the uncertainty in biomass burning emissions and proposed that dry matter was the main cause of regional bias in CO emission estimates. Vegetation classification methods and fire detection products led to the uncertainties in bottom fuel consumption and burned area calculations, resulted in biases in dry matter. They reported that the variability of particulate-phase emission was even higher than that of gas-phase emission. In addition, they compared the simulated results with satellite measurements, and given certain inventory recommendations based on different study areas and spatiotemporal scales. This study is well written and well organized, and could support improvements in biomass burning emission inventories in further studies. I recommend accepting it after minor revisions.
- The second paragraph beginning with “Recent studies” can be combined with the third paragraph beginning with “Previous studies”.
- Section 2. The description of Biomass Burning emission inventories and Quantitative statistical methods can be shortening. Some details can be moved to supplementary information.
- Line 545-561. There is a little bit of confusion about this paragraph. The authors said, “The total QFED FRP is 1.5 times higher than VFEI0, but DM in QFED2.5 inventory is 30% lower than VFEI0”, and also said: “Therefore, although the two top-down emission inventories use similar algorithms, significant bias occurs under high cloud fraction conditions, with QFED2.5 estimating DM much higher than VFEI0”. So, does the low DM in the QFED2.5 mainly occur under low or medium cloud fraction? Could the authors give some specific values?
- Figure 2 showed that FINN1.5 estimated much larger CO emissions than other emission inventories in EQAS. The authors also selected the EQAS as one of the important biomass burning regions based on the fact that “(1) regional BB CO emissions above 20Tg/yr, (2) BB CO emissions account for more than 70% of the total”. However, table 3 shows similar CO column averages across the four BB emission inventories. Could the authors also explain why? Additionally, it is important to show simulated near-surface/different vertical layer CO concentrations.
- Line 660-665. Is there any reference to support the conclusion that the overestimation of SOA in South Hemispheric South America is due to biogenic sources?
Citation: https://doi.org/10.5194/egusphere-2023-1822-RC1 - AC1: 'Reply on RC1', Sijia Lou, 02 Feb 2024
-
RC2: 'Comment on egusphere-2023-1822', Anonymous Referee #2, 27 Dec 2023
This study mainly focused on investigating the differences among widely adopted emission inventories of biomass burning and revealing the main reasons. It is a necessary work for improving the emission inventories in the future. The manuscript can be accepted after the following questions addressed.
- The abstract should be shorted and refined.
- Line 41: we give certain inventory recommendations based on different study areas and spatiotemporal scales. What are your certain recommendations for the global emission inventory? Please briefly describe in abstract, which is important to help readers find the key points of the paper.
- The introduction should be re-written again. The authors listed so many literatures, while they do not better summarize them. At the end of each paragraph, the main contents or research shortages should be given. Line 123-150 is not necessary to give these equations in the introduction. Line 181-190, wordy sentences.
- The calculation of DM (Line 411-419), FC (Line 508-510), EF (567-569) may not be included in part 3. This is the introduction of methods, instead of the analysis of emission inventories.
- Line 550-553: How empirical factor affect the amount of DM? It is also an important factor with uncertainty.
- The uncertainty of emission inventories was impacted by a combination of those factors (EF, DM). Monte Carlo simulations were usually performed in articles to evaluate the estimation uncertainty quantitatively for pollutant emissions. So those combined uncertainty results of the emission inventories in different regions can be compared, which may be associated to the regional applicability of BB emission inventories.
Conclusion：wordy sentences. Line 715-720, we all know these, and they are not your conclusions or new findings.
From Figure.1 to Figure. 7, all the datasets are from the literatures or former studies. These should not be the main contents of this study. The modeling works, especially for the comparison between the modeling results, the bias and the reasons should be emphasized.
Table 1, all the emission factors adopted in the four emission inventories can not reflect the real emission situation. Can the author optimize the emission factors with your modeling works? I think it will contribute more to science, but not just compare the values.
Citation: https://doi.org/10.5194/egusphere-2023-1822-RC2 - AC2: 'Reply on RC2', Sijia Lou, 02 Feb 2024
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 420 | 150 | 21 | 591 | 36 | 16 | 17 |
- HTML: 420
- PDF: 150
- XML: 21
- Total: 591
- Supplement: 36
- BibTeX: 16
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Wenxuan Hua
Sijia Lou
Xin Huang
Ke Ding
Zilin Wang
Aijun Ding
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5862 KB) - Metadata XML
-
Supplement
(2214 KB) - BibTeX
- EndNote
- Final revised paper