the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Apr 2024
Synchrony of African rainforest solar induced chlorophyll fluorescence and environmental factors
Abstract. Global atmospheric carbon dioxide concentrations are largely driven by terrestrial photosynthesis, of which tropical forests account for one third. Relative to other tropical regions, less is known about the seasonality of African tropical forest productivity and its synchrony with environmental factors due to a lack of in situ carbon flux data. To help fill this knowledge gap, we use spaceborne solar-induced chlorophyll fluorescence (SIF), vegetation indices, and climate data to investigate the seasonality and synchrony of photosynthesis in Africa’s tropical forest ecoregions. We find West African SIF to increase during the dry season and peak prior to precipitation, as has been observed in the Amazon. In Central Africa, we find a continental-scale bimodal seasonality in SIF, the minimum of which is synchronous with precipitation, but its maximum is likely less related to environmental drivers.
- Preprint
(2577 KB) - Metadata XML
-
Supplement
(1567 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2023-3022', Anonymous Referee #1, 07 Jun 2024
Main comments
The authors aim to link seasonality, minima and maxima of SIF and vegetation indices to environmental factors (precipitation, PAR and VPD). This is I think very valuable, but the manuscript and data analysis should be improved considerably. Firstly the abstract and the title only refers to SIF, while in the text also EVI and NDVI was mentioned. However, EVI is discussed and NDVI not or much less. In the M&M the pro’s and con’s of EVI, NDVI, NIRv and LSWI vegetation indices are mentioned, but in the Results it is not fully justified why EVI mosty is used and not the other three.
Discussion of the data is generally vague and often speculative. As a consequence, the relation (i.e. synchrony) with other environmental drivers (VPD, PAR) cannot be claimed/justified. This results a bit in overselling the paper.
It is also surprising that the discussion section is focused one comparison with the Amazon, which I thought was not the aim of the manuscript.
Finally, I think also more detailed and rigorous data analyses can be done, e.g. by, zero order and partial correlations. As such, as example, the response of precipitation controlled for PAR and/or VPD can be examined. I think this will add much more information to the discussion. This will allow a more rigorous and better structured discussion. The discussion is now very descriptive, mostly vague and sometimes speculative discussion.
In addition, most of the ecoregions studied have 1700 mm or less of rainfall, this should also be discussed, as this is at low end for an evergreen tropical forest.
Specific comments
Line 12: I think also fossil fuel emissions drive CO3 concentrations.
Line 22: the northern hemispheric biosphere is the key driver for intra-annual variation in CO2 concentrations.
Line 24: Carbon store? Write “carbon stock”.
Line 25, 26: It is rather the regional water cycle for each tropical basin. Please add some more recent references here.
Line 34: I think the main conclusion is rather constant carbon gain for African intact tropical forest, which diverting from the Amazon.
Line 40: is this really the case? I think you need to down turn this statement a bit as we don’t have the complete evidence for this.
Line 41-43: What is “Congolian”? Write Congo basin.
Line 51: What about mortality? Was this not observed?
Line 53: What you mean with “coastal forests”?
Line 56: SIF is I think an indirect observation and hence still a proxy for productivity. Can you elaborate here.
Line 59: Leaf physiology: can you be more specific.
Line 66: VPD and temperature are linked; please make clear in the text
Line 82 can you reformulate this hypothesis as not al forest in Africa you test are “moist”, see Table S1.
Line 107: is 735-758 correct, i.e. it is not 743-758 nm?
Line 111: in the Results you also present soil moisture data. It is not clear how this is calculated from the Materials and Methods section.
Line 127: you indicate 4 vegetation indices and there pro’ s and con’s. But in the Results you focus on EVI and a little bit on NDVI. Please explain/justify better your choice.
Line 147: why this selection of African forest types and not the one proposed by Réjou-Méchain et al. 2021, Nature ?
Line 152-154 : Where do we see this ? Some of the text here is also redundant.
Line 155: I though the classification in 8 11 regions was done a priori and not because of the seasonality.
Line 157: Elaborate Beter how were seasonality can be seen, i.e. by better referring to figure 1. But the legend and caption in figure 1 is not clear and needs to be improved.
Line 170 etc. In this section moments of the year are given. However, I think this should be done more precise be giving months of the year and not expressions like “mid-year”, ect.
Figure 2: The caption is not complete as also VPD is shown here. Justify why only “TROPOMI SIF” is shown (also valid for Figure 3).
Line 194 etc. In this section a mostly qualitative description is given of synchrony with precipitation, PAR and VPD. I think more efforts can be done to make these relationships more quantitative. Hence my suggestion on top for zero order and partial correlations,… And I am sure other techniques exists.
Figure 4 and related text: I do not understand why a difference in correlations is proposed here; why VPD + Precip is given, etc.
Figure 5: can you explain why these periods are chosen?
Figure 6: the caption of this figure needs to be improved to read the figure independently.
Line 269: It is surprising to see that the discussion is now mostly geared towards a comparison with the Amazon. Is this realy the scope of the paper? And than suddenly the discussion is on PAR and VPD, while this was not emphasised in the introduction of the paper or the results section.
Line 305, but where are the physiology and phenology data? You refer here to the vegetation indices? This is not clear.
Citation: https://doi.org/10.5194/egusphere-2023-3022-RC1 - AC1: 'Reply on RC1', Russell Doughty, 09 Nov 2024
-
RC2: 'Comment on egusphere-2023-3022', Anonymous Referee #2, 26 Aug 2024
The paper explores the seasonality of African tropical forest productivity and its synchrony with environmental factors using satellite SIF vegetation indices, and climate reanalysis. While the topic is important and interesting, the paper focuses much on empirical correlation relationships without a solid justification and explanation of the mechanism behind the observed statistical relationship. Therefore, the data analysis and discussion should be improved considerably.
Major comments:
(1) The authors grouped the 11 ecoregions into three categories based on their precipitation regimes, using various thresholds such as high monthly variability (standard deviation ≥ 120 mm), relatively high mean annual rainfall (> 2400 mm), low monthly variability (standard deviation ≤ 100 mm), and relatively lower mean annual rainfall (< 2200 mm). However, the selection of these threshold values appears arbitrary, with no references provided to justify their choice.
(2) For the second hypothesis, the authors sought to test whether SIF would be more strongly linked to precipitation in less moist African forests and whether SIF and VPD would be positively correlated in moist forests. However, this hypothesis was not addressed in the Results and Discussion. There is a lack of explanation from the perspective of plant physiology regarding how these two distinct mechanisms—soil moisture deficit and atmospheric dryness—operate in African forests.
(3) The Introduction should be reworked to emphasize the importance and current understanding of African forest. Now there is little description of the background of African forest. I would like to see (i) what is the climate background (ii) what is dominant vegetation species and (iii) what is the current understanding of trend, inter-annual variations and seasonality (most important since this is the topic of the study).
(4) In the Introduction, the author mentions the use of satellite SIF to explore the relationship between photosynthesis and environmental factors. However, since satellite vegetation indices are also used later in the study, it would be beneficial to introduce these indices earlier and provide justification for the use of multiple satellite vegetation indicators.
(5) The differing patterns of SIF in comparison to EVI and NIRv in response to precipitation in Fig. 4 are interesting. I would expect the authors explore this more thoroughly, providing a deeper explanation rather than only describing the empirical relationship. Does this suggest a decoupling between photosynthesis and canopy structure? Additionally, I would suggest the authors test whether the sampling times for SIF and MODIS vegetation indices are approximately the same. The authors mention that SIF retrieval is less sensitive to cloud cover, whereas MODIS vegetation indices are more affected by clouds.
Minor comments:
Abstract
Line 16: the authors did not investigate actual photosynthesis, instead, they investigate SIF, which is an indicator of photosynthesis.
Line 17: maybe delete the Amazon here since the aim of the study is not to compare Africa tropical forest and Amazon.
Introduction
Line 25: carbon store -> carbon storage?
Line 31-33: add references
Line 35: What is the meaning of “carbon gains”? Carbon fluxes or storage?
Paragraph Line 41: Please rephrase this paragraph. The first sentence mentions little work has been done to study the seasonality of West African tropical forests. However, the other part does not mention anything related to the seasonality. It seems the authors tried to introduce several knowledge gaps but without very clear and organized presentation.
Line 65: Maybe add Wang et al (2023) in the introduction, which explores how different climate factors influence tropical forests using SIF.
Wang, Y., Liu, J., Wennberg, P.O., He, L., Bonal, D., Köhler, P., Frankenberg, C., Sitch, S. and Friedlingstein, P., 2023. Elucidating climatic drivers of photosynthesis by tropical forests. Global Change Biology, 29(17), pp.4811-4825.
Line 67-68: not a complete sentence
Line 75: define ecoregions
Methods
2.1 OCO-2 and OCO-3 SIF and 2.2 TROPOMI SIF and 2.5 MODIS Surface Reflectance and Vegetation Indices: Not sure how the authors gridded these data to match climate reanalysis
2.4 ERA5 Reanalysis: I am not sure whether the author used CHIRPS Precipitation instead of ERA5 precipitation.
2.7 Ecoregions: Please expand the description. It would be good to provide a map of Ecoregions of the study area. How many types of Africa’s tropical forest types do you have? How are they defined?
Results
Line 212, 215 and 216 “sites” -> “ecoregions”?
Fig. 4: What don't you show the scattering plot at the pixel level?
3.3 Robustness of the SIF signal: I think this part should go into discussion or Supplementary. Also there is no unit in Fig. 7.
Discussion
Line 271: why suddenly discuss Amazon vs. African forests? I think the paper does not intend to compare Amazon and African forests since the authors did not mention any results of Amazon before.
Line 305: Not sure whether the authors have defined physiology and phenology/
The authors may add a paragraph to describe the uncertainty or limitation of this study.
Citation: https://doi.org/10.5194/egusphere-2023-3022-RC2 - AC2: 'Reply on RC2', Russell Doughty, 09 Nov 2024
Status: closed
-
RC1: 'Comment on egusphere-2023-3022', Anonymous Referee #1, 07 Jun 2024
Main comments
The authors aim to link seasonality, minima and maxima of SIF and vegetation indices to environmental factors (precipitation, PAR and VPD). This is I think very valuable, but the manuscript and data analysis should be improved considerably. Firstly the abstract and the title only refers to SIF, while in the text also EVI and NDVI was mentioned. However, EVI is discussed and NDVI not or much less. In the M&M the pro’s and con’s of EVI, NDVI, NIRv and LSWI vegetation indices are mentioned, but in the Results it is not fully justified why EVI mosty is used and not the other three.
Discussion of the data is generally vague and often speculative. As a consequence, the relation (i.e. synchrony) with other environmental drivers (VPD, PAR) cannot be claimed/justified. This results a bit in overselling the paper.
It is also surprising that the discussion section is focused one comparison with the Amazon, which I thought was not the aim of the manuscript.
Finally, I think also more detailed and rigorous data analyses can be done, e.g. by, zero order and partial correlations. As such, as example, the response of precipitation controlled for PAR and/or VPD can be examined. I think this will add much more information to the discussion. This will allow a more rigorous and better structured discussion. The discussion is now very descriptive, mostly vague and sometimes speculative discussion.
In addition, most of the ecoregions studied have 1700 mm or less of rainfall, this should also be discussed, as this is at low end for an evergreen tropical forest.
Specific comments
Line 12: I think also fossil fuel emissions drive CO3 concentrations.
Line 22: the northern hemispheric biosphere is the key driver for intra-annual variation in CO2 concentrations.
Line 24: Carbon store? Write “carbon stock”.
Line 25, 26: It is rather the regional water cycle for each tropical basin. Please add some more recent references here.
Line 34: I think the main conclusion is rather constant carbon gain for African intact tropical forest, which diverting from the Amazon.
Line 40: is this really the case? I think you need to down turn this statement a bit as we don’t have the complete evidence for this.
Line 41-43: What is “Congolian”? Write Congo basin.
Line 51: What about mortality? Was this not observed?
Line 53: What you mean with “coastal forests”?
Line 56: SIF is I think an indirect observation and hence still a proxy for productivity. Can you elaborate here.
Line 59: Leaf physiology: can you be more specific.
Line 66: VPD and temperature are linked; please make clear in the text
Line 82 can you reformulate this hypothesis as not al forest in Africa you test are “moist”, see Table S1.
Line 107: is 735-758 correct, i.e. it is not 743-758 nm?
Line 111: in the Results you also present soil moisture data. It is not clear how this is calculated from the Materials and Methods section.
Line 127: you indicate 4 vegetation indices and there pro’ s and con’s. But in the Results you focus on EVI and a little bit on NDVI. Please explain/justify better your choice.
Line 147: why this selection of African forest types and not the one proposed by Réjou-Méchain et al. 2021, Nature ?
Line 152-154 : Where do we see this ? Some of the text here is also redundant.
Line 155: I though the classification in 8 11 regions was done a priori and not because of the seasonality.
Line 157: Elaborate Beter how were seasonality can be seen, i.e. by better referring to figure 1. But the legend and caption in figure 1 is not clear and needs to be improved.
Line 170 etc. In this section moments of the year are given. However, I think this should be done more precise be giving months of the year and not expressions like “mid-year”, ect.
Figure 2: The caption is not complete as also VPD is shown here. Justify why only “TROPOMI SIF” is shown (also valid for Figure 3).
Line 194 etc. In this section a mostly qualitative description is given of synchrony with precipitation, PAR and VPD. I think more efforts can be done to make these relationships more quantitative. Hence my suggestion on top for zero order and partial correlations,… And I am sure other techniques exists.
Figure 4 and related text: I do not understand why a difference in correlations is proposed here; why VPD + Precip is given, etc.
Figure 5: can you explain why these periods are chosen?
Figure 6: the caption of this figure needs to be improved to read the figure independently.
Line 269: It is surprising to see that the discussion is now mostly geared towards a comparison with the Amazon. Is this realy the scope of the paper? And than suddenly the discussion is on PAR and VPD, while this was not emphasised in the introduction of the paper or the results section.
Line 305, but where are the physiology and phenology data? You refer here to the vegetation indices? This is not clear.
Citation: https://doi.org/10.5194/egusphere-2023-3022-RC1 - AC1: 'Reply on RC1', Russell Doughty, 09 Nov 2024
-
RC2: 'Comment on egusphere-2023-3022', Anonymous Referee #2, 26 Aug 2024
The paper explores the seasonality of African tropical forest productivity and its synchrony with environmental factors using satellite SIF vegetation indices, and climate reanalysis. While the topic is important and interesting, the paper focuses much on empirical correlation relationships without a solid justification and explanation of the mechanism behind the observed statistical relationship. Therefore, the data analysis and discussion should be improved considerably.
Major comments:
(1) The authors grouped the 11 ecoregions into three categories based on their precipitation regimes, using various thresholds such as high monthly variability (standard deviation ≥ 120 mm), relatively high mean annual rainfall (> 2400 mm), low monthly variability (standard deviation ≤ 100 mm), and relatively lower mean annual rainfall (< 2200 mm). However, the selection of these threshold values appears arbitrary, with no references provided to justify their choice.
(2) For the second hypothesis, the authors sought to test whether SIF would be more strongly linked to precipitation in less moist African forests and whether SIF and VPD would be positively correlated in moist forests. However, this hypothesis was not addressed in the Results and Discussion. There is a lack of explanation from the perspective of plant physiology regarding how these two distinct mechanisms—soil moisture deficit and atmospheric dryness—operate in African forests.
(3) The Introduction should be reworked to emphasize the importance and current understanding of African forest. Now there is little description of the background of African forest. I would like to see (i) what is the climate background (ii) what is dominant vegetation species and (iii) what is the current understanding of trend, inter-annual variations and seasonality (most important since this is the topic of the study).
(4) In the Introduction, the author mentions the use of satellite SIF to explore the relationship between photosynthesis and environmental factors. However, since satellite vegetation indices are also used later in the study, it would be beneficial to introduce these indices earlier and provide justification for the use of multiple satellite vegetation indicators.
(5) The differing patterns of SIF in comparison to EVI and NIRv in response to precipitation in Fig. 4 are interesting. I would expect the authors explore this more thoroughly, providing a deeper explanation rather than only describing the empirical relationship. Does this suggest a decoupling between photosynthesis and canopy structure? Additionally, I would suggest the authors test whether the sampling times for SIF and MODIS vegetation indices are approximately the same. The authors mention that SIF retrieval is less sensitive to cloud cover, whereas MODIS vegetation indices are more affected by clouds.
Minor comments:
Abstract
Line 16: the authors did not investigate actual photosynthesis, instead, they investigate SIF, which is an indicator of photosynthesis.
Line 17: maybe delete the Amazon here since the aim of the study is not to compare Africa tropical forest and Amazon.
Introduction
Line 25: carbon store -> carbon storage?
Line 31-33: add references
Line 35: What is the meaning of “carbon gains”? Carbon fluxes or storage?
Paragraph Line 41: Please rephrase this paragraph. The first sentence mentions little work has been done to study the seasonality of West African tropical forests. However, the other part does not mention anything related to the seasonality. It seems the authors tried to introduce several knowledge gaps but without very clear and organized presentation.
Line 65: Maybe add Wang et al (2023) in the introduction, which explores how different climate factors influence tropical forests using SIF.
Wang, Y., Liu, J., Wennberg, P.O., He, L., Bonal, D., Köhler, P., Frankenberg, C., Sitch, S. and Friedlingstein, P., 2023. Elucidating climatic drivers of photosynthesis by tropical forests. Global Change Biology, 29(17), pp.4811-4825.
Line 67-68: not a complete sentence
Line 75: define ecoregions
Methods
2.1 OCO-2 and OCO-3 SIF and 2.2 TROPOMI SIF and 2.5 MODIS Surface Reflectance and Vegetation Indices: Not sure how the authors gridded these data to match climate reanalysis
2.4 ERA5 Reanalysis: I am not sure whether the author used CHIRPS Precipitation instead of ERA5 precipitation.
2.7 Ecoregions: Please expand the description. It would be good to provide a map of Ecoregions of the study area. How many types of Africa’s tropical forest types do you have? How are they defined?
Results
Line 212, 215 and 216 “sites” -> “ecoregions”?
Fig. 4: What don't you show the scattering plot at the pixel level?
3.3 Robustness of the SIF signal: I think this part should go into discussion or Supplementary. Also there is no unit in Fig. 7.
Discussion
Line 271: why suddenly discuss Amazon vs. African forests? I think the paper does not intend to compare Amazon and African forests since the authors did not mention any results of Amazon before.
Line 305: Not sure whether the authors have defined physiology and phenology/
The authors may add a paragraph to describe the uncertainty or limitation of this study.
Citation: https://doi.org/10.5194/egusphere-2023-3022-RC2 - AC2: 'Reply on RC2', Russell Doughty, 09 Nov 2024
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 365 | 178 | 42 | 585 | 61 | 40 | 44 |
- HTML: 365
- PDF: 178
- XML: 42
- Total: 585
- Supplement: 61
- BibTeX: 40
- EndNote: 44
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 357 | 58 |
| China | 2 | 58 | 9 |
| Germany | 3 | 24 | 3 |
| France | 4 | 24 | 3 |
| Netherlands | 5 | 16 | 2 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 357