Satellite-Based Estimation of Carbon Dioxide Budget in Tropical Peatland Ecosystems
Round 1
Reviewer 1 Report
This manuscript presents the carbon dioxide budget in tropical peatland ecosystems estimated using remote sensing data and data from local flux towers in parts of islands of Indonesia and Malaysia. Attention was given to the drought conditions of the ecosystems using a modified Keetch-Byram-Drought Index which incorporates ground water level data.
Overall, it is a relevant paper that covers an interesting topic. The paper is fairly well written, but the text could be improved as indicated below. The study is based on sufficient large datasets, and the figures are nicely presenting the research results. The figure captions; however, should contain more information so readers can interpret the figure without digging into the main text.
Furthered, I do not understand why the authors do not use directly the entire MODIS LST time series for the KDBI computations, instead of using partly MODIS and partly MTSAT LST.
There is one big issue. The paper deals with carbon budgets of peatlands, but the MODIS GPP data sets were calibrated on the flux tower in a forest ecosystem. So the derived CO2 originates from the footprint of forests. Also the ecosystem respiration is based on forest based flux tower data. How can this discrepancy be reconciled?
Some other questions/comments are given below.
L13: “forest conditions”: this is really vague. Be more specific;
L14: “… THE groundwater table …”;
L19: insert space between CO2 and released; “… into THE atmosphere…”;
L22: which datasets? Be more specific!
L23: “… in THE study area …”;
L27-29: What are the authors trying to tell us? Please clarify; What do you mean with “delineations”;
L40: Explain what you mean with “discharge”; Water discharge? Should it not be plant water availability? So not too wet, not too dry? Clarify.
L41-43: This is not a proper phrased sentence; Carbon offset???
L45: I prefer the use of “shallow groundwater table”;
L54: “… due to the Mega Rice Project”; the use of “from” is not OK;
L61: remove “detected”;
L65: “In that study, spatial distributions of GWT ….”:
L78: remove “,” and “;”; End with “.”;
L84-87: this sentence is not phrased well;
Figure 1. Change the gray background in white as was done in the other maps. Include lore text: Spatial distribution of tropical peatlands in ….accordingly to …..(include reference). The location of the flux tower is marked in red….
Section 2.2.1. is on KDBI, but no single word is mentioned on KDBI. Perhaps start with the reference that KDBI is based on temperature and precipitation and refer to section 2.3.1.
L120: blending? What do the authors mean? Is that proper terminology?
L138: insert space between 50 and km²;
L153: Should be “… anthropogenic induced and wild fires.
L158: VISIT: which spatial and temporal resolution of the used datasets?
L161-162: not properly phrased. Rewrite. Include the fits for the different biomes in an appendix section.
Figure 2. No reference in the main body of the text to Figure 2. This flowchart needs an extensive caption in order to understand the connection. So add a large description;
L180, Eq 1: Q is not explained in the text. It is the initial KBDI value and should be noted as Q0;
L187, Eq 2: the multiplication sign is wrong; What is the effect of substituting air temperature with LST, and P with Pannual?
L188, Eq 3: How strong is this relation? Add R² into the text;
L191: “Pann is the annual precipitation…”; Is Pann used per year, or is a long-term average used over 17 years?
L192: What us meant with inundation timing? Is this flooding? Is it the time when the ground water table reaches the surface?
L195-196: Please rephrase; What do the authors mean with “denote”?
L200: “… was used as supplemental data for …”;
L200-201: What do the authors mean? Please rephrase;
L206: “reverse”? This is wrong. Soil respiration occurs also during the day. And soil respiration is a complete different process compared to photosynthesis. Not a reverse! What is more, ecosystem respiration is the combination of heterotrophic and autotrophic respiration (Verstraeten et al., 2006, On temperature and water limitation of net ecosystem productivity: Implementation in the C-Fix modeling Ecological Modelling). I am not sure for your flux sites it they derive ecosystem or heterotrophic respiration.
L212: “… were calculated ASSUMING that all …”
Figure 3: Peat forests? Should it not be Forest on peat land? Please include in the caption that data derived from flux towers were used.
L223: product produced? Should be something like “MODIS-based GPP time series underestimate the GPP derived from flux towers ….”; Note that GPP is not measured by flux towers, NBP is measured.
L226: “… around 0.29 based on the least square regression between MODIS GPP and flux tower derived GPP. ”; Then directly refer to Figure 4.
L229: “The BEST fitting …”;
L229-236: Rephrase! This is not proper English;
L241: showed, showed; bracket?
L250: Intercepted more plots? I do not understand this. This is improper wording;
L255: “… modified THE MODIS GPP …”;
Figure 4. GPP data DERIVED from the DF flux tower ….;
L271: “… of THE validation ….”;
L272: CO2 UPTAKE by vegetation, not absorption!
L285-286: “… between KBDI-based and on-site measured GWT ….”;
L294: Are the T and P measured at the flux site?
L301: “for”;
L322: Conspicuous? This is a very vague terminology! Which spatial dependencies? Be specific!
L327: remove “conditions”; Why should the ecosystem respiration be high if the forest GPP is high? I agree that the autotropic respiration (roots) are positively linked to GPP, but the heterotrophic respiration is not necessarily linked to vegetation carbon uptake. So please rephrase.
L344: “Annual average GPP range from 2100 to …”;
L349: “… GPP compared with flux tower based GPP …”; Some uncertainties: which ones? Be specific.
L351: Essential datasets? I do not get it. Please rephrase.
L356-357: Some differences: which ones? Be specific.
L358: discrepancies in … could …. : which ones? Be specific.
Figure 9: Annual averaged modified GPP …. Base on ….: Give more information.
L375: should be 0.80;
L378-379: “would”? Improper word;
L385-387: Rephrase.
Figure 11: The annual averaged fire emissions ….. Why are they no data in West Papua New Guinea? Please include and compare the GFED map for this region.
L400: “The NBP of tropical peatlands were calculated including biomass burning (Eq. 9).”;
L401-402: “Using the Re for UF the monthly average …., respectively. ”;
L403: “... the NBP values were … on average …);
L407-408: Do you mean homogenous distribution? Please rephrase. Use visual, instead of maps.
L409: remove of;
L410: “The dynamics of NBP seems to be dominated by the temporal behavior of the fire emissions.”
L421: “producing”? Improper wording; Perhaps affecting?
L418-421 and Figure 12: The El Nino index seems to be shifted to the right compared to the fire emissions. The fire emissions peak first, prior to the peak of the El Nino index. How do you explain that? What is more, how do you explain the low respiration values in some years (i.e. end 2009, beginning of 2011, 2016, late 2018)? Please include the standard deviations of RE, GPP, FE, NBP. Elaborate in the figure caption on what you present? Is the data based on spatial averages? Which region? Please help the reader to understand your figure.
L435-436: “enabling”? Rephrase.
L444: “evaluated the ecosystem of …”. This is vague. You are evaluating the ecosystem carbon balance.
L448: “… artificial disturbance data”? This is awkward terminology. Use fires.
L449: “compensative”? Improper wording.
L458: “… peatlands are a net emitter of carbon to the atmosphere.”
L464: artificial? Use natural and anthropogenic induced emissions.
L585: should be “van der Werf”;
Author Response
Dear. Reviewer 1,
Thank you for your comments and suggestions. We tried to apply your comments appropriately in the manuscript for improving it. Especially, we carefully revised the descriptions to make it specific for better understanding. Not only the grammar of English, but also the structure of paragraph were modified. Regarding the methodology, the terminologies were revised, and the description of equations were also edited according to your comments. More discussions were conducted adding the limitations of this study. Please see the attachment file for the response to your comments.
Thank you.
Best regards,
Haemi Park, Wataru Takeuchi, and Kazuhito Ichii
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This study describes a novel and brief method for reveal the sinks and sources of carbon (C) in tropical peatland ecosystems by NBP derived from satellite-based data. Essentially, the method involves modifying or improving the estimations of GWT, FE, GPP and Re using satellite observations combined with the flux tower measurements. Lastly the manuscript revealed the CO2 budget of target peatland areas using NBP derived from these estimated variables. It demonstrated the interesting results of the sinks and sources of carbon for tropical peatland ecosystems. Thereby, I would like to recommend this paper for publication after the minor revision suggested below.
Line 89-94: The describe of this studying objectives, which listed the calculation of GWT, FE, GPP and Re, mostly like research content of this studying. If this studying objective should be to better understand the net carbon budget in the tropical peatland ecosystems via the satellite-based estimation of those variables accurately?
Line 103-104: The legend description of Figure 1 should be put in the illustrate of Figure 1 (Line 106).
Line 105: Figure 1 only showed one point of Flux tower in red circle. But the studying area include three sites described in Line 125-126, why?
Line 108: title of section doesn’t match with the content described below.
Line 172: Detailed the illustrate of flowchart, which should be the flowchart deriving NBP by satellite-based observations.
Line 419-420: This result, “the higher emissions were found in El Nino years consistently”, is interesting, which is consistent with the results from the other studies below based on the satellite observation of atmospheric CO2 as well.
[1] Liu, J.; Bowman, K.W.; Schimel, D.S.; Parazoo, N.C.; Jiang, Z.; Lee, M.; Bloom, A.A.; Wunch, D.; Frankenberg, C.; Sun, Y., et al. Contrasting carbon cycle responses of the tropical continents to the 2015-2016 El Nino. Science 2017, 358.
[2] Guerlet, S.; Basu, S.; Butz, A.; Krol, M.; Hahne, P.; Houweling, S.; Hasekamp, O.P.; Aben, I. Reduced carbon uptake during the 2010 Northern Hemisphere summer from GOSAT. Geophysical Research Letters 2013, 40, 2378-2383
[3] He, Z.; Lei, L.; Welp, L.; Zeng, Z.-C.; Bie, N.; Yang, S.; Liu, L. Detection of Spatiotemporal Extreme Changes in Atmospheric CO2 Concentration Based on Satellite Observations. Remote Sensing 2018, 10, 839
However there are two questions for the results shown in Figure 12, (1) the time appearing the higher NBP is slightly ahead of Nino.3’s time; (2) NBP did not show the higher value corresponding to the event of El Nino in 2010. Please demonstrate the reasons.
Line 433-436: The manuscript demonstrated the reliability of modified satellite data for the result of NBP derived from this studying, that is much greater than the other studies described in Line 428-430. Please adding the reasonability of GWT, FE, GPP and Re data in detail. It can be seen from Figure 12 that the FE seem mostly to respond to the effects of El Nino, which likely demonstrate the effects of FE. I wonder if this studying derived more accurate FE than the other studies.
Author Response
Dear. Reviewer 2,
Thank you for your comments and suggestions. We tried to apply your comments appropriately in the manuscript for improving it. Especially, we appreciated for the suggestion regarding the other studies as more references. Accordingly, we added the suggested reference papers, and revised our manuscript by discussing ENSO cycle and the carbon balance in tropical peatlands. Please see the attachment file for the response to your comments. Thank you.
Best regards,
Haemi Park, Wataru Takeuchi, and Kazuhito Ichii
Author Response File: Author Response.pdf
Reviewer 3 Report
The main objective of this study is to estimate the CO2 (carbon dioxide) budget of tropical peatlands in Indonesia and Malaysia (6 N–11 S, 95 – 141 E). By using satellite data and calibrated with in-situ data, this study summarized the natural net carbon budget including ecosystem respiration (Re) and GPP-C assimilations with fire emission (FE) to calculate the net carbon budget. Keetch–Byram Drought Index was applied to estimate the groundwater table represent the number of Re, also net biome production concept was used to calculate the carbon dioxide budget. Moreover, this study might have the potential to be published. However, some comments below need to be considered
Need a slight correction for typos Regarding the visualization of the study area (map), I suggest that the authors could provide the labels or names of locations considering that Malaysia and Indonesia are different countries also Indonesia has 5 major islands (Sumatra, Java, Kalimantan, Sulawesi, Papua). It will be easy to understand, knowing that not all islands in Indonesia have the potential of peatlands (focused on the Sumatra, Kalimantan, and Papua Island). Knowing that this study uses multi-resolution imageries such as GSMaP (10 km), MTSAT (4 km) + MOD11A1 (1 km), MOD17A2. I suggest that the authors could explain how the integration of satellite data was conducted, given that the difference in resolution affects the accuracy of the data used. In Figure 2, mentioned that the spatial resolution of MODIS GPP (MOD17A2) is 1 km - 8 days, while in the description part (lines 139-140) it is explained that the product used has 500m spatial resolution, between the two which one is true?
Moreover, previous study have shown that MODIS GPP (especially product with a 1 x1 km2 resolution) have some limitations even though this product widely used to study the global carbon in relation to terrestrial ecosystems. The limitation lies in the sub-grid scale process, where forests are located close to mountainous areas, and show heterogeneity in vegetation types. In this connection, has this study considered the complexity of ecosystems? can calibration with Flux Tower GPP handle MODIS GPP limitations? can Flux Tower observations located in Palangkaraya, Kalimantan be able to calibrate the MODIS GPP data for the whole of the study area from Sumatra to Papua also to cover Borneo-Malaysia?
Note: not only land cover data (MCD12Q1), integration of MODIS GPP with land use data is recommended to increase the robustness of the dataset.
Figure 7, 9, 11, found boundary errors for Kalimantan-Indonesia and Borneo-Malaysia (line 412) what does "5 N–5 S, 90-150 W" refer to? Are the geographical coordinates of the study area? I think 90-150 W should be 90 - 150 E
Author Response
Dear. Reviewer 3,
Thank you for your comments and suggestions. We tried to apply your comments appropriately in the manuscript for improving it. Especially, we appreciated for the comments regarding study area in detail. Accordingly, we revised all figures with specific boundaries in the study area. Please see the attachment file for the response to your comments. Thank you.
Best regards,
Haemi Park, Wataru Takeuchi, and Kazuhito Ichii
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Overall it is a relevant manuscript that has been corrected well. However, extensive editing of English language and style is really required. Some suggestions:
L33: Should be “drainages and fires”;
L89-91: Should be “Finally for balancing the CO2 emissions from the surface, the MODIS GPP product (MOD17A2) was used [34].”;
L95: Should be “… via the more accurate estimated carbon balance components using satellite observations. “;
L114: Should be “is used” and “marked as red circles”;
L119: Should be “… were used in the computation of KBDI.”
L190: Should be “As CO2 sources emitting to the atmosphere, ….”
L190: Should be “by balancing Re, ….”;
L204: Should be “Q0 is the initial moisture deficiency at time zero.”;
L246: Should be “…during nighttime, and …”; either you use “night time” or “nighttime”; the same with daytime;
L247: Remove “with”;
L267: twice “use” in one sentence; An alternative could be “were applied”;
L373-375: Should be “Pixels with values larger than 1000 g C/m2/year were not confirmed in the Re maps ….”;
L382: Should be “from THE satellite based …”;
L401: Should be “… (GPP) from observations (FLUX tower derived GPP), …”;
L431: “… was calculated …”;
L442-443: Should be “…The net CO2 balance is positive indicating that tropical peatlands are net sources of CO2.”;
L443: Should be “….THE dominant factor …”;
L471-472: Should be “For future studies, with more flux towers representing peatlands and associated biomes the GPP calibration would be improved by considering LUE, and plant functional types using satellite-based land cover maps [60] ”;
L474: Should be “In addition, we compare our results with another study, ….”;
L478: remove “from whole study area”;
L492: Should be “However, fires that destroy potential carbon uptake sources (vegetation) and disturb the hydrological cycle … have most impact in changing the carbon balance of peatlands [62]. The FE in THE Papua region …with GFED 4.1s since the visual interpretation of the map was difficult.”;
L500: remove “are”;
L501: “finishing”? Improper wording.
L505: twice show in one sentence;
L522-523: effect and affecting in one sentence;
L524: remove “in”;
L526-544: reconsider this paragraph; it is not proper English
L566-567: Should be “A MODIS-based fire emission (FE) dataset was used as source of anthropogenic fire emissions. “;
Table A1, also include the correlation coefficients in the table.
Figure A1 is hard to read. Please use white background.
Author Response
Dear. Reviewer 1,
We really appreciate for your contribution to edit our manuscript. Authors tried to apply all of your comments appropriately to the current version of manuscript.
Please see the attachment.
Best regards,
Haemi Park, Wataru Takeuchi, and Kazuhito Ichii
Author Response File: Author Response.pdf
Reviewer 3 Report
i think the current version is ready to be published
Author Response
Dear. Reviewer 3,
We really appreciate for your kind response to our revision. In current manuscript, we revised some minor problems. Accordingly, we attached the current version of modified manuscript.
Thank you for your cooperation.
Best regards,
Haemi Park, Wataru Takeuchi, and Kazuhito Ichii
