the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Dec 2025
Seasonal and interannual variability of atmospheric ammonia over Guatemala driven by land use, biomass burning, and meteorological circulation
Abstract. Ammonia (NH3) is a key atmospheric precursor of fine particulate matter and a marker of agricultural and biomass burning emissions. In Central America, NH3 variability remains largely unquantified. This study presents the first integrated spatiotemporal assessment of atmospheric NH3 over Guatemala (2015–2023) using multi-satellite observations (IASI A/B/C), combined with MODIS fire data, Sentinel-2 land cover, ERA5 meteorology, and CAMS reanalysis. Annual median NH3 columns remained relatively stable, reflecting persistent agricultural sources dominated by fertilizer use and livestock. Significant anomalies occurred in 2016, 2020, and 2023, with 2020 showing the highest annual and monthly NH3 levels. Seasonal peaks in April–May coincided with the regional fire season, followed by a sharp decline after rainfall onset. Hotspots were consistently detected in northern (Petén–Quiché) and southern (Escuintla) agricultural regions. The most extreme episode in April 2020 recorded 957 active fires over ~1,486 km2, largely within the Maya Biosphere Reserve. Elevated temperature (+ 0.3 °C above the 2015–2023 mean) and high precipitation (+ 17 % above average) favored NH3 accumulation despite reduced anthropogenic activity during the COVID-19 lockdown. These results indicate that Guatemala’s NH3 variability is shaped by a stable agricultural baseline with superimposed fire-driven peaks, modulated by climatic anomalies. Continuous satellite monitoring is essential to improve emission inventories and support strategies to reduce biomass burning impacts across Central America.
- Preprint
(37760 KB) - Metadata XML
-
Supplement
(9365 KB) - BibTeX
- EndNote
Status: open (until 11 Feb 2026)
- RC1: 'Comment on egusphere-2025-5668', Anonymous Referee #1, 08 Jan 2026 reply
-
RC2: 'Comment on egusphere-2025-5668', Anonymous Referee #2, 02 Feb 2026
reply
General comments:
This paper presents the first study on the seasonal and interannual trends of NH3 over Guatemala using IASI A, B, and C total columns. By combining satellite observations with land use and fire data, the authors conclude that the NH3 trends over Guatemala between 2015 and 2023 were mainly driven by agricultural emissions and episodic fire events.
A number of studies have used IASI observations to look at NH3 trends and variabilities at country and regional scales, but most focused on major agricultural regions such as North America, Europe, and Asia, where elevated NH3 plumes are easily detectable from space because the column concentrations are one to two orders of magnitude higher than the instrument’s detection limit. This study adds value to the existing body of literature because NH3 concentrations in Guatemala are generally near the detection limit of IASI, shedding light on the technical limitations of satellite measurements. This is a potentially interesting point that remains under-explored in the current paper.
The paper also needs some reorganization. For example, in the first paragraph of Methodology, the authors are already discussing results. This section should start with a description of the datasets and mathematical/statistical methods used, which are currently placed later in the section. My other concerns and questions can be found below.
Major comments:
L41-49: This paragraph heavily focuses on vehicle emissions as an underestimated source of NH3, so I was expecting to see some analysis on vehicle emissions, but there were none except one sentence mentioning that traffic activity decreased in April 2020. If you are not going to delve into traffic emissions, I recommend shortening this paragraph to one sentence just saying that vehicle emissions remain an underestimated source in NH3 emission inventories and merging it with the previous paragraph.
L98: The authors seem to use the words “emission” and “concentration” interchangeably throughout the paper, including here where it says the authors will first detail the spatiotemporal variability of NH3 emissions. Another example is Figures 2 and 3, which captions both say “NH3 emissions (molecules/cm^2)”, but these are the units of column concentrations (or more precisely, total columns), not emissions. “Emission” and “concentration” have different scientific meanings and should not be mixed, as deriving emissions from column concentrations requires more than satellite observations and typically involves a modeling approach.
L172: This detection limit (4–6×10^15 molecules/cm^2) was for an earlier version of IASI product. Is there an updated number for version 4? As mentioned earlier, this becomes relevant because the reported annual mean NH3 concentrations over Guatemala (4.12–6.67×10^15 molecules/cm^2) are very close to the detection limit of IASI. Can we realistically draw conclusions on the interannual trends when the mean concentrations are so low? Would other satellite instruments that have higher sensitivities to surface NH3 be better options?
Figure 2: How much do the measurements vary between the three IASI instruments? Can you show some statistics of NH3 columns by instrument? Without knowing this, this figure feels like comparing apples to oranges because not all years are covered by the same IASI instruments.
It would also be interesting to see if the number of IASI pixels are significantly different 1) between the years, and 2) between the low concentration grids and high concentration grids. This helps to understand whether the low concentrations are affected by sampling bias. Can you also describe how the IASI algorithm handles spatial averaging/weighting in the L3 data?
L407: “…which points to a general inverse relationship where higher NH3 concentrations tend to coincide with lower PBLH values, and vice versa. This pattern suggests that a shallower boundary layer, which restricts the vertical dispersion, leads to an accumulation of NH3 near the surface.” This explanation would hold true for surface NH3 concentrations. However, IASI measures vertically integrated columns which are less sensitive to fluctuations in BLH. Figure 7 also does not quite reflect an inverse relationship between IASI and BLH. In fact, the Pearson correlation matrix in Figure S3 even shows NH3 and BLH are moderately positively correlated.
L565: Sometimes I find it difficult to associate the argument with the results supporting it. For example, here it says the high number of fires in April 2020 was likely driven by prolonged dry conditions. This initially confused me because the following sentence says the year 2020 recorded the highest total precipitation. I had to look through the paper again to see that April 2020 indeed had one of the lowest monthly precipitations according to Figure 7, but this was never explicitly called out.
Finally, a lot of interesting results presented in Figure S3 were never discussed in the main text. In particular, it was mentioned in section 2.1 that the interrelationships between NH3 and other air pollutants (NO2 and SO2) will be explored, but they were not. Also, what are the sources of the NO2 and SO2 data (satellite observations or surface measurements)?
Minor comments:
L81: When you say “previous research has often overlooked the integrated role of multiple environmental drivers…” without giving examples, it is unclear to me which studies you are referring to, since you also mention there is a significant lack of studies on Central America. Can you add some references?
L174: The authors may also want to double check the existing citations. For example, the 2012 paper cited here seems unrelated to IASI.
L195: This equation does not look right. It is missing an equal sign, and the square root sign should extend further. Please double check this equation and also Equation (2).
L240: This should be Equation (3).
L314: “3.2 Land cover classification driven a k-means clustering of NH3 emissions” This subtitle needs some grammar editing.
L341: Here it says Cluster 1 covers the smallest area compared to Cluster 2 and Cluster 3. However, the paragraph above says Cluster 3 has the smallest territorial extent. I do not understand the difference.
L355: Why does it suddenly say ammonium (NH4+) here?
L453: ppmv does not need to be in uppercase. Also, instead of writing 1.55 × 10^–3 ppmv, it is more natural to say 1.55 ppbv.
Citation: https://doi.org/10.5194/egusphere-2025-5668-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 211 | 123 | 20 | 354 | 49 | 57 | 64 |
- HTML: 211
- PDF: 123
- XML: 20
- Total: 354
- Supplement: 49
- BibTeX: 57
- EndNote: 64
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The topic addressed in this manuscript is relevant, and the study's objective is important. However, I have major concerns regarding conceptual inconsistencies and the clarity of the methodology and discussion. Because of the above, I consider that the manuscript cannot be accepted for publication.
MAJOR GENERAL COMMENTS
A major issue throughout the manuscript is the interchangeable use of the terms NH₃ “emissions”, “concentrations”, and “total column”, which are not equivalent. This conceptual confusion seriously undermines the soundness of the data and the interpretation of the results.
Additionally, the document contains internal inconsistencies that raise concerns about its scientific rigor, making it challenging to evaluate its conclusions.
The manuscript is written in a very repetitive style, where similar ideas are repeated both within paragraphs and across sections where they do not belong. Furthermore, the manuscript does not follow the standard structure typically found in scientific writing. These patterns remind me of other manuscripts written using extensive AI assistance. While I do not believe the use of AI is necessarily a reason for rejection, the authors should clearly describe the extent to which AI was used for data retrieval, analysis, interpretation, and writing.
Overall, the manuscript would greatly benefit from a revision that addresses the issues outlined above. Some specific examples of these issues are detailed below.
SPECIFIC MAJOR COMMENTS
Section 2.1 (Data Analysis). This section is an example of not following the common manuscript structure, as it is a summarized narrative of the methodology, mixed with results and conclusions. This type of section is not usually found. Furthermore, results and conclusions are not supposed to be included in a Methodology section.
Section 2.3 (ESRI Land Cover Class Map). “The… LUC dataset… derived from Sentinel-2 satellite imagery…” and “The LULC … dataset, based on data from the European Space Agency’s Sentinel-2 mission”. The above sentences are repetitive and appear in the same paragraph: This is only one example of repetitive writing found throughout the manuscript; there are several more.
Figure 1. Regions mentioned in the text, such as Petén and Quiché, should be indicated on the map to help readers follow the discussion more easily.”
L150 “Atmospheric total column measurements of NH3 were obtained…”. L184 “To identify spatial patterns in atmospheric NH3 emissions…”. Both NH3 observations (total column and emissions) are mentioned in section 2.4, but only the retrieval of the NH3 total column is described. Regarding NH3 emissions, either the authors are using this term as equivalent total column, or they are omitting how they obtained the total column. In the rest of the document, “total column”, “emissions”, and “concentration” are used interchangeably.
L216 “Correlation analyses between these variables and atmospheric NH3 concentrations were performed to better understand the meteorological drivers influencing NH3 variability across Guatemala.” The “Pearson correlation analysis” is also mentioned in L109; however, the manuscript presents no quantitative correlation analysis. There is only a qualitative comparison between the meteorological conditions and the NH3 data described in the following sentence: L407 “… which points to a general inverse relationship where higher NH3 concentrations tend to coincide with lower PBLH values, and vice versa.“
The following paragraph is internally inconsistent (L273 - L280): “The concatenated data from IASI A/B/C data of spatial distribution of annual mean of NH₃ emissions…” “A consistent spatial pattern of elevated NH₃ concentrations was observed…”
L273 “The concatenated data from IASI A/B/C data of spatial distribution of annual mean of NH3 emissions across Guatemala from 2015 to 2023 exhibited a consistent pattern of elevated emissions in the northern and southern regions. A consistent spatial pattern of elevated NH3 concentrations was observed, with discernible hotspots in the northern, north-central, and southern regions of the country. Conversely, the central midlands and highlands consistently displayed lower emission rates throughout the period." Please define terms such as “northern and southern regions”, “north-central and southern regions“, and “north-central and southern regions“; otherwise, it is difficult to corroborate the description in the text. For example, I do not see the “consistently elevated NH3 concentrations in the northern regions” that is mentioned. Also, “NH3 concentrations” are mentioned, but Figure 2 declares to show only “NH3 emissions”. In addition, the methodology section never mentions how the concentrations are determined. It seems, then, that both terms are being incorrectly used interchangeably.
Figure 5. What are the boxplots showing? FRP or fire counts? Units in the vertical axis are missing.
L419 “a monthly means detected by IASI A (2.32 × 1016 molecules/cm2), B (2.48 × 1016 molecules/cm2) and C (2.43 × 1016 molecules/cm2). “ Why are the three values shown? Everywhere else in the manuscript, only one value (I guess it is the average of the three) is shown.
Figure 9. Units on the color scale are inconsistent: ppmv (unit of concentration) is not equivalent to molecules/cm2 (units of total column).
L466: “April 2020 recorded the highest number of active fires across Guatemala throughout the entire study period (957), representing…” The reported number of fires in April 2020 (957) does not match Figure 10a, which shows only 85, approximately.
L474: The term “climatological peak” is illogical, as April 2020 represents an exceptional event.
L475: ”This case study demonstrates that April 2020 was… an expanded spatial footprint compared to other years.” There is no data demonstrating how April 2020 compares to other years.
Figure 11. All the discussion in this section is about April of 2020 as a case study, so why is April combined with May data in this figure?
Section 4 (Discussion) The Discussion section mainly repeats the Results and provides no additional interpretation or new insights. A substantially shorter version of this section could be included as the summary required by ACP in the Conclusions section.
OTHER COMMENTS
L23. The statement “Continuous satellite monitoring is essential to … support strategies to reduce biomass burning” seems unrealistic and is not supported by the analysis shown in the manuscript.
L87 and L141 (maybe more lines). “Environmental Systems Research Institute (ESRI)” The ESRI acronym was defined on line 87, so it does not need to be defined again on line 141. In fact, once the acronym is defined, the full name should not be used again. There are several other cases similar to this in which the acronyms are not used correctly (i.e. next comment).
L249 “Moderate Resolution Imaging Spectroradiometer (MODIS)” The acronym MODIS is defined in L249, but it is used several times before this line.
L261 “The respective boxplots reveal a generally stable NH3 concentration throughout the study period.” L269 “… indicating substantial variability and peak emissions.” Both sentences in the same paragraph are contradictory.
L278 - L281. The information was already stated in the previous paragraph.