the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Nov 2025
Hydrological drivers of hydrogen cyanide wildfire emissions from Indonesian peat fires during the 2015, 2019, and 2023 El Niño events
Abstract. Indonesian peatlands store vast amounts of carbon that are highly vulnerable to fire during El Niño-driven droughts. When ignited, peats release large quantities of greenhouse gases and other species with significant environmental impacts, including hydrogen cyanide (HCN), a sensitive tracer of smouldering combustion. In this work, we use satellite retrievals from the Infrared Atmospheric Sounding Interferometer (IASI), TOMCAT atmospheric model simulations, hydrological information, and fire activity observations to evaluate the factors driving trace gas emissions during the 2015, 2019, and 2023 El Niño events.
The 2015 El Niño produced large burdens of HCN and CO unprecedented in the satellite observational era, driven by exceptionally low soil moisture, depressed groundwater levels, and deep burn depths. In contrast, the 2019 and 2023 events exhibited markedly weaker emissions despite similar Oceanic Niño Index (ONI) anomalies, reflecting more favourable hydrological conditions. Comparisons of the satellite trace gas observations with simulations of the TOMCAT model show that burned-area-based inventories such as GFED substantially overestimate emissions from peat fires in 2015, while a new peat-specific database, FINNpeatSM, better represents fire season timing and burn depth by incorporating soil moisture constraints. From satellite-derived HCN:CO enhancement ratios, we provide new emission factors for HCN that offer benchmarks for new emission inventories.
Our results show that peat fire intensity and emissions are driven not only by El Niño strength but also by local hydrological conditions such as soil water content and precipitation. Integrating hydrological indicators with satellite observations of atmospheric composition is therefore critical for improving fire emission inventories.
- Preprint
(7719 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5109', Anonymous Referee #1, 12 Dec 2025
-
RC2: 'Comment on egusphere-2025-5109', Anonymous Referee #2, 02 Mar 2026
Review Bruno et al., 2025
Bruno et al. show in their manuscript unusual strong HCN emissions due to peatland fire emissions during El-Nino periods. They compare IASI HCN to TOMCAT simulations and suggest a scaled HCN emission for September 2015. Further, they compare the extreme year 2015 with other El-Nino years, and connect HCN emissions with soil moisture and calculate enhancement ratios for HCN.
The manuscript is well within the scope of ACP, however it needs major improvements. In particular the objective and the structure of this manuscript us unclear to me, and statements in the conclusions are not well supported. In addition, the new IASI HCN data product used in this manuscript is currently unpublished and undocumented, which makes it difficult for me to evaluate this manuscript.
General comments:
- The reference "Moore et al., in preparation" is, according to the ACP submission guidelines (https://www.atmospheric-chemistry-and-physics.net/submission.html) allowed to use in the review stage of the submission, but referees need to have access to this unpublished work. This is not the case at the moment for me as referee, so the ACP guidelines are not met here, even though Copernicus staff pointed out that issue (see MS overview). Please provide the Moore et al. reference as asset. In particular, I need more information about the statements made about retrieval sensitivities in different atmospheric layers, which are essential to understand major findings of the current study.
- The structure of the manuscript lacks consistency. In some sections (e.g. 3.1.1), the paragraphs of this section are barely connected to each other. In section 3, many sub-sections start again with details of the data product, which should have been introduced in section 2. Further, figures are sometimes only marginally introduced and barely discussed. Also, the order of appearance in the text is not the same as the numbering of the figures and the reader needs to jump from one figure to another quite often.
- The objective of the paper is not clear to me. The first part reads as a model study showing that emissions should be improved for 2015 to better match IASI observations. Then enhancement ratios are calculated, and further years are investigated. In the end, again only 2015 is investigated regarding details of the fire processes and plume dynamics. Most of these activities are treated rather standalone, I am missing the connecting element in the manuscript.
- The conclusions contain statements that are not covered by the findings discussed earlier in the manuscript. Please prepare all statements in the main part of the manuscript based on the presented data.
- Given the rather small scales that can occur if fires are active and plumes are transported, I am wondering if the TOMCAT T42 horizontal resolution is good enough for this task. Also it is not clear to me why the older and coarser ERA-Interim reanalysis was chosen for nudging instead of the supreme ERA5 data set. What about vertical transport in the reanalysis, in particular in the tropics. Further, vertical transport of biomass burning plumes due to pyroconvection may occur - how was that handeled by the model?
- The manuscript contains 17 figures of different quality. The authors should carefully think about which figures are really necessary to support their findings and move all other figures to a supplement. Figures used in the manuscript should have similar font sizes (same as font sizes of the text) and line widths. Figure 15 could be a nice example for the figure style (in my opinion).
Specific comments:
- Abstract: Please define all used abbreviations and acronyms (e.g. CO, TOMCAT, GFED, FINNpeatSM).
- L76: "high spectral resolution and sampling": This is relative regarding the background of the reader. Some may have ground-based FTS systems in mind, others radiometers. Please briefly mention the typical IASI spectral resolution/sampling and also the spectral range employed by IASI.
- L95: "For HCN, the a priori profile was constructed from INTEX-B aircraft information ...": I do not understand how the retrieval is using aircraft measurements prior 2009 for the retrievals. I guess the data was further processed before using it in the retrievals. Again, it is difficult to understand the quality of the data used in this study without the knowledge of Moore et al.
- L120: Why is the information about the different latency products relevant for this section? For a study published years after the events discussed, I would always assume that final data is used instead of near-real-time data. In my opinion it is sufficient to mention that IMERG Final Run data was used together with the reference describing more details of different data products.
- L140: In this chapter, SWI is introduced (but the acronym is not defined here, only later in Figure 4). For a typical reader of ACP, it would be good to connect SWI to GWL, which are occasionally used together in the article.
- L155: So the fire emissions are resolved much better than the meteorology driving the model (GRED: 0.25° vs ERA-I 0.75°)?
- Figure 1: The color bar used is made for highlighting positive and negative differences around the center of the color bar. Here, however, absolute columns are shown and for columns around 1.5e16 molecules/cm2, it is difficult to discriminate this concentration from missing values. I suggest to change to a uniform color bar like viridis or turbo.
- Figure 1 right panel: Please explain why the regridded monthly mean data reveals these abrupt changes in total column at +- 20° latitude. Is this an instrument, data processing or plotting artifact?
- Figure 2: Please explain abbreviations like "T_HCN" in the figure caption.
- L210/Figure 3: Please mark these longitudes of the extent of Indonesia in the Plots in Figure 3.
- L215: At which altitudes is IASI sensitivity greatest? And I guess the last sentence of this paragraph refers again to Figure 2 as well as the first sentence of the following paragraph?
- L227: In my opinion, this and the following paragraphs should be moved into Section 2.1.2
- L246/Figure 5: The text says Fig. 5 shows all HCN emissions by the model, the figure caption says it only shows GFED emissions. What is really shown in Fig. 5? Further, it would be interesting to see which emission source is responsible for the suggested overestimation of HCN emissions in September. So it could be interesting to show the combined emissions together with GFED emissions only.
- Figure 5 caption: "(molecules cm−2 s+−1)" The "+-1" looks strange here. Please use a more common way to indicate uncertainties (in case this was meant by the "+-1")
- Figure 6: Is there a good reason why Figure 1 shows data from 1 November 2015, while Figure 6 shows data from 2 November 2015? Further, I think a better comparison would be to grid the model data on the measurement geolocations (and apply the averaging kernel) instead of regridding the measurement data on the regular model horizontal grid. The comment about the color bar for Figure 1 also applies for this Figure.
- Equation (2): What is meant by "smoke" here? Is this really only for smoke conditions (i.e. with large aerosol concentrations)? In that case, IASI would not be able to measure these air masses and would be not suitable for this kind of study. Maybe the authors should instead refer to "polluted air" or "plume"?
- L283: What does a low correlation factor indicate here?
- Table 1: I am not sure if I understand this table correctly: The first four columns are cited from the authors which are mentioned in column 1? And the columns with EF_HCN are derived using findings from this study and equation (3)? In this case the "author" column is kind of misleading for the EF_HCN columns. Further, I am not sure why there are EF_HCN results for so many different correlation coefficients presented. Again: What does the correlation coefficient tell me here?
- L319: Since MLS is mentioned here: Since MLS measured HCN and CO vertically resolved in 2015 - wouldn't it be useful to include this data into this study?
- L321: "Before mid-October, we likely underestimate the HCN amounts as plumes are closer to the surface.": Does that mean that the original emissions (blue curve in Fig. 5) could be correct and only the poor sensitivity of IASI underestimates the true amounts of HCN here?
- L325: Please give a reference for the de-facto standard.
- L327: Figures 9 and 11 are mentioned earlier than Figure 8. Please adjust the order of figures.
- L343: Why is the HCN burden introduced? How was the burden calculated? What does the burden give as additional information compared to the columns shown before? Figure 10 again shows the CO columns in the same style as the HCN burdens in Figure 8, so there is a chance for misinterpretation.
- Figure 11: What is the advantage of showing monthly fire radiative power in a cumulative way?
- L422: Why is a cross-correlation analysis performed? How was that analysis performed? Why does this analysis help the reader to understand the temporal dynamics of plume transport better? The whole paragraph should be motivated and explained better. From the paragraph and Figure 17, I can hardly tell what "Lag" is shown here on the x-axis.
- L427: I think the peak is at -10 days instead of 10?
- L428: Earlier in the manuscript, it was stated that IASI has different maximum sensitivity altitudes for HCN and CO, here both species are mentioned together.
- L430: Again, it should be -10 instead of +10.
- L452: The HCN EFs provided by this study heavily rely on previous work, which provided the CO EFs. Is there a recommended range for HCN EF? This recommendation would make the current manuscript very useful for readers looking for HCN EFs, instead of only providing the table with a set of different EFs.
- L455: I would argue that this work only showed that GFED emissions are overestimating HCN emissions in September 2015. The rather general formulation in this paragraph is not supported by the findings of this manuscript.
- L460: I think I missed the part of the manuscript, where the FINNpeatSM emissions were used in TOMCAT to show that these match better the observations than the GFED emissions with scaling applied in September 2015?
- L480: I am missing here a platform for the VNP14IMG data (e.g. a website or if available a DOI). Further, are the IASI, IMERG, SMAP, SWI, GFED, FINNpeatSM data available? Please add these information.
- L482: The TOMCAT data on Zenodo only contains HCN loss and gain values (HCN1-12), but (as long as I understand the file format) no HCN concentrations or columns, as they were used in the manuscript. Please also provide these data or improve the file format (e.g. using CF standards), so I can find these data entries in the files.Citation: https://doi.org/10.5194/egusphere-2025-5109-RC2
Data sets
Hydrological drivers of hydrogen cyanide wildfire emissions from Indonesian peat fires during the 2015, 2019, and 2023 El Niño events - TOMCAT data part 1 Antonio G. Bruno et al. https://doi.org/10.5281/zenodo.17194012
Hydrological drivers of hydrogen cyanide wildfire emissions from Indonesian peat fires during the 2015, 2019, and 2023 El Niño events - TOMCAT data part 2 Antonio G. Bruno et al. https://doi.org/10.5281/zenodo.17194065
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 390 | 182 | 31 | 603 | 31 | 31 |
- HTML: 390
- PDF: 182
- XML: 31
- Total: 603
- BibTeX: 31
- EndNote: 31
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Bruno et al. analyse the conditions that drive Indonesian peat fire emissions. They identify the importance of hydrological conditions like soil water content and precipitation in addition to El Niño. While the results are interesting and advance our understanding of Indonesian peatland fires, the study misses key aspects required for publication in ACP. At its core, the study relies on newly derived CO and HCN retrievals based on IASI observations. The authors fail to provide a satisfactory explanation as to why alternative retrievals, in addition to the methodologies that are already in place, are required. No retrieval data are provided, the retrieval description is insufficient, and the retrievals have not been evaluated. The only reference provided is to another paper by the authors which is 'in preparation'. I recommend rejecting the manuscript for the time being, and reconsidering it once the new IASI retrievals for CO and HCN have been published and evaluated, and the retrieval method has undergone proper peer review. Before reconsidering the manuscript after this process has been finished, the authors need to address the following comments.
Further major comments:
I have major concerns about the TOMCAT model setup used, as it seems inadequate for the analysis performed:
(1) The authors themselves highlight the importance of circulation patterns and vertical transport (e.g. Figure 3), which are based on ERA5 data. However, their model setup relies on outdated ERA-Interim reanalysis data for meteorological forcing, which lacks improved vertical resolution (Hersbach et al., 2020). I strongly suggest switching to ERA5 reanalysis data as the meteorological driver. This would increase the temporal resolution of the meteorological input from 6 hours to hourly, greatly improving the representation of transport processes.
(2) During the 2015 peatland fires, the Asian monsoon anticyclone was ongoing. The model resolution used in this study (T42, 2.8° by 2.8°) is too coarse to adequately resolve these transport patterns. Please increase the model resolution as frequently employed in global model studies.
(3) The authors state that the TOMCAT model uses monthly averaged emissions, which are then resampled in time. This approach may be valid for emission sources that are not highly time-variable, but it is inappropriate for fire emissions due to their high time-variability. How is this interpolation performed for peatland emissions, and how representative is this approach? Why not rely on datasets that include daily wildfire emissions based on daily fire activity (e.g. GFEDv5; van der Werf et al., 2025)? What influence does your approach have on the overestimation of HCN emissions in September 2015?
The outline and style of the current version need improvement. The order in which the figures are presented does not reflect their usage in the main body of the text. Section 2 provides a simple listing of all the datasets used in this study. The manuscript would benefit greatly if, at the beginning of Section 2, an overview was provided illustrating which datasets will be used and how they will be analysed. For example, the use of the precipitation and soil moisture data is unclear, as is its relevance to the study which only becomes evident later on.
Minor comments:
- Overall, not all abbreviations are properly introduced. In particular, the abstract introduces some abbreviations, while others are assumed to be familiar.
- Lines 28–30: Please provide references for these statements.
- Line 51: HCN has already been introduced on line 29.
- Lines 51–53: Provide reasons and references explaining why these are poorly suited.
- Line 95: How does the TOMCAT model compare to other atmospheric chemistry models for the troposphere? What uncertainty in the retrieval would be introduced if another model would be used?
- Lines 93–97: How representative are the derived profiles?
- Line 109: This is unclear. Does 'fixed' refer to a fixed yearly emission profile, or are the emissions averaged over a year and then fixed?
- Lines 111–113: Provide details on the HCN chemistry used, as well as on other physical loss processes.
- Lines 157–159: Please provide a source for this statement.
- Line 197: The plume of HCN is clearly visible in the 'left' panel, not the 'right' panel, of Figure 1.
- Lines 197–198: This statement reads more like a report than a scientific paper.
- Figure 1: How is the monthly average calculated? How is it ensured that the monthly average at a given location is not skewed towards a single observation, or a few observations, if all other observations within the month have been invalid due to for example cloud cover?
- Figure 2: Why are there two gaps between November and December? Was no output created for these periods by TOMCAT?
- Line 212: Please introduce the ERA5 data in Section 2.
- Lines 227–237: Should this not be included in Section 2.1.2?
- Line 231: Is the daily sample from the model the same as the IASI overpass time? Does the model provide averaged or instantaneous data?
- Line 241: Replace 'observed' with 'simulated'.
- Line 253: This 75% reduction comes out of the blue. How was this number obtained? What could be the reason for such a significant overestimation of emissions?
- Figure 3: Please ensure that the x-axis and y-axis labels are shared across all subplots. Remove the colour bars from (a), (c) and (e).
- Figure 3: Comment on how well ERA5 represents vertical velocities at the equator.
- Figure 4: Add labels (a) and (b).
- Figure 5: Correct the unit in the caption.
- Line 270: Please provide a reference for this statement.
References:
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.-N.: The ERA5 global reanalysis, Quarterly Journal of the Royal Meteorological Society, 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020.
van der Werf, G. R., Randerson, J. T., van Wees, D., Chen, Y., Giglio, L., Hall, J., Roland, V., Mu, M., Binte Shahid, S., Barsanti, K. C., Yokelson, R., and Morton, D. C.: Landscape fire emissions from the 5th version of the Global Fire Emissions Database (GFED5), Sci Data, 12, 1870, https://doi.org/10.1038/s41597-025-06127-w, 2025.