the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Oct 2024
Two different mechanisms cause severe hot droughts in the western Amazon
Abstract. In the last few decades, the Amazon basin has experienced several periods of simultaneous extreme drought and heat. It is crucial to gain a better causal understanding of these compound events, which can severely damage the rain forest ecosystem. Here, we study the role of atmospheric mass, moisture and heat transport to the western Amazon, an area where pristine tropical rain forests can still be found. To investigate how anomalous the atmospheric transport during extreme events is compared to the climatological mean, we use a Lagrangian reanalysis dataset created with the particle dispersion model FLEXPART driven with meteorological input data from the ERA5 reanalysis. For the period 1979–2021, air masses over the western Amazon are selected every three hours and traced 10 days backward in time. We compared the overall transport climatology for the end of the dry season (August–September) with the transport for 21 extreme compound events that were selected for their extremity in terms of high air temperature and low soil water content. We find that extreme events over the western Amazon that happen during El Niño events have very different causes from those that take place under La Niña or neutral ENSO (El Niño Southern Oscillation) conditions. For the five extreme events that occurred under El Niño conditions, around 50 % of the air, as compared to 30 % for the climatology, is located over the tropical Atlantic Ocean 10 days prior to arrival. Air that is already anomalously dry and warm is transported from this region towards the western Amazon, highlighting the role of long-range transport for these extremes. The other 16 extreme events occurred under La Niña or neutral ENSO conditions and show similar transport patterns to the overall transport climatology until 3 days prior to arrival. During the last 3 days, however, air is preferentially transported over the southern Amazon, where already dry soil conditions cause a malfunctioning of moisture recycling, resulting in the propagation of extreme conditions to the western Amazon. Air arriving in the western Amazon during these events is travelling over areas heavily affected by deforestation. Therefore, we expect that landuse changes over the southern Amazon have a stronger impact on compound drought and heat extremes in the western Amazon under La Niña or neutral ENSO conditions than during El Niño events.
- Preprint
(37978 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2024-2801', Anonymous Referee #1, 05 Nov 2024
In the present manuscript, the authors analyzed the source and path of air masses related to compound hot and dry events over the western Amazon. The topic is very relevant, given the recent droughts in the region. The proposed analysis is robust and provides valuable insights into the impacts of different circulations on the development of heat waves over the region. However, the categorization of the events into ENSO phases may not be the best option here. There is no clear difference in the characteristics of the events during El Nino compared to those during non-El Nino years. This weakens the conclusions of the manuscript. Thus, I recommend the rejection of the manuscript but strongly encourage its re-submission after restructuring the analysis. See more details below.
Major Comments
- My main issue with the current manuscript is the categorization of the drought events into ENSO phases. The selected events are clearly different from the climatology (outside the one standard deviation around the climatological mean), but there is no clear distinction between E+ and E0- events. In most of the graphics (Figs 3 to 5 and A1 to A3), the red lines indicating E+ events are within the envelope of the lines related to E0-. This is also evident when comparing the values of the variables listed in Table 1 and Table A1. On the other hand, there is a large variability in the E0- events, resulting in the median (dashed blue lines) being quite close to the climatology in some cases. In fact, some of the most interesting events are classified as E0-, as described in different parts of the manuscript. This makes me wonder if using ENSO events to classify the drought events is the best option here. I agree that there is a clear difference in the amount of air coming from southern Atlantic Ocean (SAT) and tropical Atlantic Ocean (TAT), as seen in Table A1 and Figures 2c, A3c and A4. Thus, maybe a better option is to categorize the events according to the source of the air masses and then verify if there is any influence of ENSO or other modes of variability on the events. Particularly, the western Amazon responds to anomalies in the northern tropical Atlantic Ocean and the ITCZ behaviour (as mentioned in the introduction but not explored in the manuscript), which, in turn, respond to the Southern Annular mode and the South Atlantic Dipole, for example. Thus, separating the events using a source criterium may lead to clearer results.
- In the conclusions, the authors mentioned the importance of the SST anomalies over SAT as a source of warm and dry air to the region. I agree that the Atlantic ocean plays an important role here (as mentioned in the previous point). However, the relationship between the SST anomalies and the characteristics of the air masses (temperature and humidity) is not clear. Most (but not all) E0- cases are related to hot and dry air masses from the Southern Atlantic, with warm SST anomalies. However, if the South Atlantic is warmer than normal, I would expect more moisture in the air, not less. This relationship should be better explained. Also, what is the source of the SST data? This is not mentioned anywhere in the manuscript.
- The information from the soil moisture could also be used to better understand the events. The dip in soil moisture anomalies in Fig 5e is quite relevant and shows that, over WAM, the soil moisture seems to be responding to the atmosphere. However, this is not true for all events, with some maintaining very low soil moisture throughout the 20-day period. Here, it is important to also the into account the land-atmosphere coupling and the characteristics of the previous months, not just 20 days, because of the large memory of the soil moisture. Other points related to the soil moisture data:
- How reliable is the soil moisture data in ERA5? This region has very few measurements so it would be interesting to have some reference about it.
- How different from climatology are the anomalies in Fig 5d and e? It would be helpful to add a measurement of standard deviation or significance to the graphic.
Minor comments
- Figure 1a and Eastern Amazon region (EAM). This region is in Northeastern Brazil, which is not part of the Amazon. This region is a semi-arid region, characterised by prolonged droughts. Thus, naming it Eastern Amazon is incorrect and misleading.
- Page 8, lines 144-146: “Even the air arriving in the lowest 1 km of the WAMt originates from mid-tropospheric levels (mostly over the TAT region), likely air that has previously been lifted by convection over the eastern part of the TAT region or Africa.” Is there any reference that supports the statement in the second part of the sentence?
- Page 9, lines 197-199: “Overall, there seems to exist a stronger surface-atmosphere coupling over the continent during the last few days of transport, and a tendency for some cases for this to occur over the SAM region.”. Which results corroborate this statement?
- Figure 6: the blue region labelled +SSTA over the Southern Atlantic Ocean is located south of 30S, while the results in Fig 1b and A4 indicate that the air masses into WAMt are originating over tropical Atlantic (north of 20S in the majority of the cases). Together with the blue arrow pointing into the region, this figure can be misinterpreted as indicating some influence of extratropical air masses into the region.
- Table A1: what is the whole Amazon region AMZ in the 5th column? Is it the combination of SAM, WAM and EAM?
Citation: https://doi.org/10.5194/egusphere-2024-2801-RC1 -
AC1: 'Reply on RC1', Katharina Baier, 25 Feb 2025
We thank the reviewer for the constructive criticism. Addressing these comments certainly helped to further improve our paper.
Major Comments
1.We understand the concern of the reviewer and see the points that are mentioned here. However, we decided to classify the extreme events using ENSO events, as we know from previous studies (Baier et al., 2022) that ENSO has a strong impact on the Amazon region. Categorizing the events was also implemented to streamline the analysis, as we aimed to avoid individually examining each event. The focus of our study is to look at a climatological scale and not individual events. We added this statement to the end of the Introduction. We also do not agree that there is little difference between E+ and E0- events. For instance, the transport from the Tropical Atlantic is clearly different (Fig. 2c, Fig. 3a). The fact that temperatures (Fig. 3c) are relatively similar for the events is to be expected, given that we exclusively study strong temperature anomalies. Still, we show that the mechanisms by which these high temperatures are produced are different for E+ and E0- events.
As we understand the concerns, we suspect that the title may have been somewhat misleading (as mentioned in the second review). Therefore, we have revised it in the hope that this will clarify the focus of the study. The new title reads:
The causal role of atmospheric transport for severe hot droughts in the western Amazon
2.When examining the tracked particles, we observe that they travel at slightly higher altitudes than those in the climatology (Fig. 3b), which is the reason why they contain less moisture. We expect that due to higher sea surface temperatures over the Atlantic Ocean, the particles can reach higher altitudes and therefore become drier. At the same time, the available latent energy from the warm ocean is converted to sensible heat, thus being causally responsible for the high temperatures during the heat waves, due to adiabatic warming once the air masses descend again before arrival.
We used SSTs from ERA5. We added this to the Methods.
3.The reviewer is right that soil moisture has a long memory. It certainly is also an important causal factor. In this paper, however, we are mainly interested in the contribution of atmospheric transport anomalies to causing drought events. Nevertheless, in the last paragraph of section 3.4 we point out that the first E0- WAM extreme heat and drought events in a season are usually related to transport from an already dry SAM region. This indeed suggests that low soil moisture in the SAM region leads to a progression of drought towards the WAM region. For this, definitely longer-lasting soil moisture anomalies are responsible.
To clarify this point, we added the following statement to the paper:
Since our primary focus is on the contribution of atmospheric anomalies, we believe that a 20-day period of soil moisture anomalies is sufficient for our purpose. However, it is important to note that soil moisture has a long memory, and conditions from previous months might also play a role. This may explain why the first event of a given year tends to have a stronger impact, with subsequent events being more influenced by pre-existing conditions.
- Ground-based soil moisture data are indeed very sparse in this region. However, ERA5 is the first ECMWF reanalysis to assimilate satellite observations of soil moisture (available since 1991) (see Hersbach et al., 2020). ECMWF executes stringent quality controls, so we think these data are trustworthy at least to the extent required for our paper. However, it is beyond the scope of our paper to validate the ERA5 soil moisture data.
- We agree that this point is indeed missing and would be quite helpful. Therefore, we added the values of the climatology for WAM and SAM (including the standard deviation) to the text, as there is already a lot of information in the graph, (now Fig. 6d,e) .
Minor comments
1.Thanks for pointing this out. While our box partly contains the Amazon delta, we agree that the major part of the box is outside the Amazon Basin. We have therefore renamed it North-East South-America (NES), as it is used in Iturbide et al. (2020)
2.We added the following reference and the following statement:
The connection between the latter region and the Amazon is already known from previous studies (e.g. Ben-ami et al, 2010).
Ben-Ami, Y., Koren, I., Rudich, Y., Artaxo, P., Martin, S. T., and Andreae, M. O.: Transport of North African dust from the Bodélé depression to the Amazon Basin: a case study, Atmos. Chem. Phys., 10, 7533–7544, https://doi.org/10.5194/acp-10-7533-2010, 2010.Although they had a different focus, they still showed via trajectories that the Amazon and Africa are directly linked via atmospheric transport.
3.We rephrased this sentence to the following:
For some events, especially those occurring over the SAM region, there seems to exist a stronger surface-atmosphere coupling over the continent during the last few days of transport, as more particles are located below 1km (Fig. 5b).
4.Thanks for pointing this out. In an attempt to more clearly distinguish the origin of the three arrows, we have indeed put this arrow too far south. Therefore, we changed the blue arrow in the figure (now Fig. 7). Now it should be consistent with the quantitative analysis.
- Ground-based soil moisture data are indeed very sparse in this region. However, ERA5 is the first ECMWF reanalysis to assimilate satellite observations of soil moisture (available since 1991) (see Hersbach et al., 2020). ECMWF executes stringent quality controls, so we think these data are trustworthy at least to the extent required for our paper. However, it is beyond the scope of our paper to validate the ERA5 soil moisture data.
-
RC2: 'Comment on egusphere-2024-2801 - to editor', murilo rev lemes, 11 Nov 2024
Caro Editor
Primeiramente, gostaria de agradecer pelo envio do artigo. O artigo apresenta interesse científico significativo, especialmente para o estudo do transporte de umidade e calor para a Amazônia e também regionalmente. Acredito que meus comentários são específicos, visando principalmente melhorar a discussão e a qualidade do material. Todas as observações estão no arquivo .zip que enviei. Portanto, recomendo sua publicação após abordar os comentários.
Obrigado!
Melhor,
-
AC2: 'Reply on RC2', Katharina Baier, 25 Feb 2025
We thank the reviewer for the positive assessment of our paper and the encouraging comments. Addressing the specific comments has helped improving the paper.
Main Comments
1.Indeed, SST anomalies exert an important control on the studied extreme events. In Table 1, the average sea surface temperature anomalies over the Southern Atlantic Ocean are reported for each individual extreme event. However, we realize that this may not be informative enough. We have therefore added maps of SST anomalies for the E+ and E0- events (Figure 5).
2.It is important to note that our study focuses exclusively on the boreal summer, whereas Ruv Lemes et al. (2020) primarily concentrated on the austral summer. To make a direct, qualitative comparison with their findings, we would need to trace the particles forward in time from the Amazon region. However, for El Nino events, with SST anomalies mainly occuring over the central Pacific Ocean, referred to Central El Nino events, they found a reduction in moisture transport from the Amazon to Southeastern Brazil during austral summer (e.g. 2014/15). This might be directly caused by pre-existing dry conditions over the Amazon during the end of the boreal summer, as discussed in our paper.
Although the impact of droughts in the Amazon on other regions in South America is beyond the focus of our study, we would still like to briefly mention this important topic and have added the following statement to the conclusion:
This may also indicate that drought in the Amazon further leads to reduced moisture transport to southern South America, causing dry conditions there. However, this might only be the case for the western Amazon and for the identified extreme events.
In addition, we address the issue of moisture transport in more detail in the following comments.
Minor Comments
1.Thanks for the recommendation. We agree and have changed the title to:
The causal role of atmospheric transport for severe hot droughts in the western Amazon
2.As reference we considered the study by Costa et al. (2022), who looked at heat waves happening under extreme dryness, highlighting the end of the dry season (around August and September), for such events.
As the statement “end of the dry season” might be a bit confusing, we changed it in the paper to the following:We focused on the dry season and compared the overall transport climatology for August and September with the transport for 21 [...]
3.Changed
4.We included the following statement:
However, over the last few years the Amazon experienced increased carbon emissions, associated with deforestation and climate change (Gatti et al., 2021; Gatti et al, 2023). Moreover, the western Amazon became a carbon source during 2019 and 2020.
5.We added the following statement to the second paragraph of the introduction:
“In addition, the transport of moisture from the Amazon has been shown to be important to other regions in South America, such as the South-East of Brazil, with an amplified contribution to its rainfall during El Nino events (Ruv Lemes, 2020). Replacing the Amazonian rainforest with savannah, has been shown to cause reduced rainfall throughout Brazil (Bottino et al., 2024), emphasizing the importance of the rainforest for the continental water cycle.”
Furthermore, we added the following statement to the conclusion:
“The future of the Amazon under climate change and deforestation is uncertain. ENSO events have been shown to intensify through global warming (Cai et al., 2014), possibly causing lasting effects on the Western Amazon, and circulation patterns could be altered, further affecting the moisture flow toward and from the Amazon (Ruv Lemes, 2023). Since heat and moisture transport are likely affected by these changes, further studies will be needed.”
Cai, W., Borlace, S., Lengaigne, M., van Rensch, P., Collins, M., Vecchi, G., . . . Jin, F.-F. (2014, February). Increasing frequency of extreme El Niño events due to greenhouse warming. Nature Climate Change, 4(2), 111–116.
Bottino, Marcus & Nobre, Paulo & Giarolla, Emanuel & Junior, Manoel & Capistrano, Vinicius & Malagutti, Marta & Tamaoki, Jonas & Oliveira, Beatriz & Nobre, Carlos. (2024). Amazon savannization and climate change are projected to increase dry season length and temperature extremes over Brazil. Scientific Reports. 14. 10.1038/s41598-024-55176-5.
Ruv Lemes, M. D. C., de Oliveira, G. S., Fisch, G., Tedeschi, R. G., & da Silva, J. P. R. (2020). Analysis of moisture transport from Amazonia to Southeastern Brazil during the austral summer. Revista Brasileira de Geografia Física, 13(06), 2650-2670.
Ruv Lemes, M., Sampaio, G., Fisch, G., Alves, L. M., Maksic, J., Guatura, M., & Shimizu, M. (2023). Impacts of atmospheric CO2 increase and Amazon deforestation on the regional climate: A water budget modelling study. International Journal of Climatology, 43(3), 1497-1513.
6.We agree that this paper is worth mentioning. However, we added a reference in the section where we discuss studies that focus on South-America and/or Amazon by using Lagrangian methods, as we find it fits better there.
7.We added the following statement to the third paragraph of the introduction:
A large portion of the Amazon Basin’s precipitation depends on local evapotranspiration (Marengo, 2006), stressing the importance of land-surface-atmosphere feedbacks in the Amazonian hydroclimate, where the recycling of precipitation is in part dependent on solar heating, which further governs the water and energy surface budgets (Nobre, Carlos A., et al. See below). Droughts could even be amplified by remote land-use, as rainfall has been linked to past air passage over forests (Spracklen et al., 2012). Differences in land-use could also potentially affect circulation, for example through differences in surface roughness and albedo (Ruiz-Vasequez et al., 2020), but can also impact the availability of atmospheric moisture, through a myriad of different mechanisms (Flores and Staal, 2022), either way influencing the transport of energy and moisture.
8.We added a statement to the Methods, to clarify our decision:
We considered the period from June to September (JJAS), when dry and warm conditions predominate, due to the northward shift of the ITCZ. However, we found extreme events only during August and September (AS), and therefore we focus on this period only and compare our results to the corresponding AS climatology.
9.We agree that this acronym might disrupt the flow, therefore we removed it from the text.
10.We added a short statement, that our result is consistent with previous studies:
“, which is consistent with previous studies (Drumond et al., 2014; Leyba et al., 2023).”
However, our study focuses more on the transport changes associated with extreme events, whereas previous studies have looked more at the climatological moisture transport towards the Amazon basin, also using a Lagrangian method.
Citation: https://doi.org/10.5194/egusphere-2024-2801-AC2
-
AC2: 'Reply on RC2', Katharina Baier, 25 Feb 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 214 | 57 | 137 | 408 | 16 | 15 |
- HTML: 214
- PDF: 57
- XML: 137
- Total: 408
- BibTeX: 16
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 114 | 28 |
| France | 2 | 22 | 5 |
| Austria | 3 | 19 | 4 |
| Germany | 4 | 19 | 4 |
| Brazil | 5 | 18 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 114