AgriCarbon-EO: v1.0.1: Large Scale and High Resolution Simulation of Carbon Fluxes by Assimilation of Sentinel-2 and Landsat-8 Reflectances using a Bayesian approach

Wijmer, Taeken; Al Bitar, Ahmad; Arnaud, Ludovic; Fieuzal, Rémy; Ceschia, Eric

doi:https://doi.org/10.5194/egusphere-2023-48

Preprints

https://doi.org/10.5194/egusphere-2023-48

Preprints

22 Feb 2023

| 22 Feb 2023

AgriCarbon-EO: v1.0.1: Large Scale and High Resolution Simulation of Carbon Fluxes by Assimilation of Sentinel-2 and Landsat-8 Reflectances using a Bayesian approach

Taeken Wijmer, Ahmad Al Bitar, Ludovic Arnaud, Rémy Fieuzal, and Eric Ceschia

Abstract. Soil carbon storage is a well identified climate change mitigation solution. The extensive in-situ monitoring of the soil carbon storage in cropland for agricultural policy and offset carbon markets is prohibitive, especially at intra-field scale. For this reason, comprehensive Monitoring, Reporting and Verification (MRV) of soil carbon and its explanatory variables at large scale needs to rely on remote sensing and modelling tools that provide the spatio-temporal dynamics of the carbon budget and it’s components at high resolution with associated uncertainties. In this paper, we present AgriCarbon-EO v1.0: an end-to-end processing chain that enables the estimation of carbon budget components of major crops and cover crops at intra-field resolution (10 m) and large scale (over 110×110 km) by assimilating remote sensing data in physically-based radiative transfert and agronomic models. The data assimilation in AgriCarbon-EO is based on a novel Bayesian approach that combines Normalised Importance Sampling (NIS) and Look-Up Table (LUT) generation. This approach propagates the 10 m uncertainties across the processing chain from the reflectances to the output variables. The chain considers as input a land cover map, multi-spectral reflectance maps from the Sentinel-2 and Landsat-8 satellites, and daily weather forcing. The PROSAIL radiative transfer model is inversed in a first step to obtain Green Leaf Area Index (GLAI). The GLAI time series are then assimilated into the SAFYE-CO2 crop model taking into consideration their uncertainty. The chain is applied over winter wheat in the south-west of France during the cropping seasons 2017 and 2019. We compare the results against the net ecosystem exchange measured at the FR-AUR ICOS site (RMSE = 1.69 - 2.4 gC m⁻² , R² = 0.88 - 0.88), biomass (RMSE = 250 g m⁻² , R² = 0.9), and combine harvester yield maps. We quantify the difference between pixel and field and pixel scale simulations of biomass (bias = -47 g m⁻² , -39 % variability), and the impact of the number of remote sensing acquisitions on the outputs (-66 % of mean uncertainty of biomass). Finally, we conduct a coherency analysis at regional scale to test the consistency of the observed patterns with soil texture, altitude and exposition variability. Results show higher biomass for higher clay soils and earlier emergence and senescence for south western exposition.

Received: 13 Jan 2023 – Discussion started: 22 Feb 2023

Download & links

Taeken Wijmer et al.

Status: final response (author comments only)

RC1:
'Comment on egusphere-2023-48', Anonymous Referee #1, 06 Mar 2023

Dear authors,

thank you very much for your submission. I think that your research field is of high relevance and that your data assimilation approach is of high interest to the community. However, I have severe doubts, if the current quality of the presentation will enable the community to really benefit from your study. Consequently, I recommend rather substantial changes, which more correspond to "reject and resubmit" (an option that is not available for GMD) than to "major revisions".

Please, see my detailed comments in the attached PDF.

Yours sincerely

Citation: https://doi.org/10.5194/egusphere-2023-48-RC1
- AC1: 'Comment on egusphere-2023-48', Taeken Wijmer, 30 Mar 2023
  
  Dear Referee 1,
  We thank you for your detailed review of the paper and for your positive comment on the assimilation approach which is a central part of the paper. In-depth answers to comments are provided in the attached PDF.
  We provide in the attached pdf clarifications and detailed answers to the comments. Taking into account the Copernicus GMD publication procedure we will provide a final reply and an edited manuscript at the end of the open discussion stage.
  Sincerely.
  
  Citation: https://doi.org/10.5194/egusphere-2023-48-AC1
- AC3: 'Comment on egusphere-2023-48', Ahmad Al Bitar, 04 Jul 2023
  
  Dear Referee 1,
  We would like to add the following elements before finalizing the discussion. We hope that they bring some clarifications to the remaining questions and give an overview of the modifications we envision.
  We thank you for the discussion phase. We are confident it will enhance our manuscript.
  Sincerely.
  
  Citation: https://doi.org/10.5194/egusphere-2023-48-AC3
AC1: 'Comment on egusphere-2023-48', Taeken Wijmer, 30 Mar 2023

Dear Referee 1,
We thank you for your detailed review of the paper and for your positive comment on the assimilation approach which is a central part of the paper. In-depth answers to comments are provided in the attached PDF.
We provide in the attached pdf clarifications and detailed answers to the comments. Taking into account the Copernicus GMD publication procedure we will provide a final reply and an edited manuscript at the end of the open discussion stage.
Sincerely.

Citation: https://doi.org/10.5194/egusphere-2023-48-AC1
CEC1:
'Comment on egusphere-2023-48', Juan Antonio Añel, 06 Apr 2023

Dear authors,
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy": https://www.geoscientific-model-development.net/policies/code_and_data_policy.html
We can not accept embargoes such as registration or previous contact with the authors to get access to the code. It must be published, permanently stored and fully public at the same moment of submitting your work to the journal. In fact, your manuscript should have never been published in Discussions with such a problem.
Therefore, unless you publish the code of AgriCarbon-EO v1.0.1 and SAFYE-CO2 v2.0.5 in one of the suitable repositories following our policy, we will have to reject your manuscript. In this way, you must reply to this comment with the link to the repositories for the code used in your manuscript, with their DOIs. Both the reply and the repository should be available as soon as possible without waiting until the end of the Discussions stage.
Please, note that you must include a license with your code. If you do not include a license, the code continues to be your property and can not be used by others, despite any statement on being free to use. Therefore, when uploading the model's code to the repository, you could want to choose a free software/open-source (FLOSS) license. We recommend the GPLv3.
In case you address this issue, and the Topical Editor of your manuscript considers that it deserves a new round of revisions or publication, you must include in a potentially revised version of your manuscript the modified 'Code and Data Availability' section, with the DOIs of the code.
Juan A. Añel
Geoesci. Model Dev. Executive Editor

Citation: https://doi.org/10.5194/egusphere-2023-48-CEC1
- CC1:
  'Reply on CEC1', Ahmad Al Bitar, 07 Apr 2023
  
  Dear Prof. Juan A. Añel,
  We greatly appreciate the commitment of Geoesci. Model Dev. in ensuring that all papers are following the highest publishing standards. The posted comment comes as a surprise because we have spent a considerable amount of time clarifying the code availability issue and providing clarifications for our first reviewer. In fact, we had an exchange with the Topical Editor during the submission phase prior to the publishing of the discussion paper, that addressed mainly the code availability. We then ensured that our article fulfilled the requirement specified in the “code and data availability” of Geoesci. Model Dev.: “https://www.geoscientific-model-development.net/policies/code_and_data_policy.html” (accessed on 06/04/2023). Typically, we provided anonymous access to the code for the reviewers and the editor via “zenodo”.
  Note that, code availability is independent of us as the governing public institutions (Université Toulouse 3, CNES, CNRS, IRD, and INRAE) of our laboratory CESBIO who are our employers, decided to use a proprietary license for the Agricarbon-EO and SAFYE-CO2 codes. The management of the license is delegated to Toulouse Technology Transfer (TTT) which manages the intellectual properties of the tools. The repository does include a license issued by the UNIVERSITE TOULOUSE III - PAUL SABATIER, SIRET No. 193 113 842 00010. The license allows the sharing of the code for research and teaching purposes for free after a formulated demand. We also understand that Geoesci. Model Dev. encourages authors “to deposit software, algorithms, and model code in FAIR-aligned repositories/archives whenever possible.”
  https://www.geoscientific-model-development.net/submission.html#assets
  Our situation is covered by item 2 of the section “Core principles” of “code and data policy“ which states :
  “2. Where the authors cannot, for reasons beyond their control, publicly archive part or all of the code and data associated with a paper, they must clearly state the restrictions. They must also provide confidential access to the code and data for the editor and reviewers in order to enable peer review. The arrangements for this access must not compromise the anonymity of the reviewers.”
  We insist on the fact that we were authorized to share the codes with the reviewers and editor in this process but not to make them publicly available. This is why we deposited the release v1.0.1 of the code and documentation in “zenodo” with restricted access. The repository has been created for the reviewing process and can be downloaded anonymously using a provided link during the submission process. This ensures no contact with the reviewers. This procedure has actually been also used lately by authors of other papers in Geoesci. Model Dev.
  The topical Editor was also very keen on ensuring that we provided the correct referencing, DOI versions for the tools, and enhancements for the paper. Following the exchanges, we provided the Topical Editor with the modifications, they were validated on 22/02/2023 and the manuscript was sent out for review.
  We do not see why our paper should not have passed the discussion stage while several papers have been published recently in Geoesci. Model Dev. with a code made available for reviewers over Zenodo in restricted access in compliance with section 2 of core principles.
  We greatly appreciate it if the Executive Editors could investigate the case of our submission closely considering the elements that we explained in this comment.
  Sincerely,
  The authors
  
  Citation: https://doi.org/10.5194/egusphere-2023-48-CC1
  - CEC2: 'Reply on CC1', Juan Antonio Añel, 09 Apr 2023
    
    Dear authors,
    First, many thanks for your quick reply to this comment. Also, I apologise; I had not seen that you replied about this during the pre-Discussions interactions with the Topical Editor. You addressed this in the pdf file of the response letter to the Topical Editor on the 9th of February. However, the information you provide about the limitations is missing in the "Code and Data Availability" section of your submitted manuscript, which is what we primarily check and what readers can access. It would have been good if you included the explanations there.
    However, I have to say that your explanation is not enough to comply with our policy. You say that you are employed by several institutions and that "these institutions" had ordered you not to release the code. We can not accept this unless you provide clear evidence of it. The part of the policy that you mention that enables authors not to publish the code is thought for cases where this is forbidden, for example, by law or by copyright inherited and decided at the governmental level. We can hardly accept that, for example, the director of an institute (or several together) chooses to and order not to release a model without anything that forbids them to do it. If we accepted such a thing, anyone could decide not to publish the code of a model simply arguing something similar. Also, I find the situation that you mention pretty strange. We receive submissions from many French institutions (including some of the ones you say), and they can release their code. For example, what prohibits the model you have developed from being released under the CECILE license?
    Therefore, you must provide evidence of what forbids you or your institution from releasing the code. I want to highlight here that those that develop a code retain the intellectual property, independently of the exploitation rights. If you have developed a module that works with a model, you could release it independently of the fact that the remainder of the model is not released.
    Is there an order, law, regulation,... a document with instructions about this to you as employees of those institutions? To summarise, we need more evidence and a more complete explanation of what precludes the open release of the code.
    Finally, I would like to point out that the Zenodo repository for PROSAIL does not contain a license. If you do not include a license, the code continues to be the property of their developers and can not be used by others. Consequently, if the code can not be executed (in this case, for legal reasons), the replicability of your work can not be verified. Therefore, we would ask you to contact the developer of PROSAIL and ask to add a license to the code in Zenodo. We recommend the GPLv3. You only need to include the file 'https://www.gnu.org/licenses/gpl-3.0.txt' as LICENSE.txt with your code. Also, in Zenodo, it is possible can choose other options: GPLv2, Apache License, MIT License, etc.
    Best regards,
    Juan A. Añel
    Geosci. Model Dev. Executive Editor
    
    Citation: https://doi.org/10.5194/egusphere-2023-48-CEC2
    
    AC4: 'Reply on CEC2', Ahmad Al Bitar, 04 Jul 2023
    
    Dear Prof. Juan A. Añel,
    As we finalize the discussion phase, we thank you for the exchanges concerning the code availability and GMD Code and Data Policy.
    Also for accepting the continuation of the discussion process after the needed documents were provided to the Editors.
    Sincerely.
    
    Citation: https://doi.org/10.5194/egusphere-2023-48-AC4
RC2:
'Comment on egusphere-2023-48', Anonymous Referee #2, 21 Jun 2023
General comments
This preprint describes a newly developed method of estimating the soil carbon budget of major crops at high spatial resolution. It addresses a “hot” and very relevant topic in climate research, specifically the field of estimating soil carbon fluxes. The method embodies a novel combination of both high spatial resolution (the intra-field scale) with large spatial coverage (up to 100x100 km) of carbon flux-related variables and their variability. It assimilates high-resolution remote sensing data into an agronomic model and accounts for uncertainties across the processing chain. The paper deals with the evaluating the model accuracy at multiple scales, assessing the impact of spatial and temporal scale in the remote sensing-based input data, and integration of multiple models/dataset for a specific crop type and study site (winter wheat in SW France).
Although the topic of quantifying soil carbon fluxes holds high significance, the way the proposed methodology is presented in the manuscript offers only little knowledge gain and shows only limited potential to be used for further studies of the soil carbon budget.
First, although the paper starts with the need for better quantifications of soil organic carbon, it is too hung up on general carbon flux, data/model integration and comparison to make interesting statements about the variability and accuracy of soil carbon estimation. To me, rather than addressing the estimation of soil carbon fluxes, it seems that the proposed approach is a crop biomass model based on correlations from remote sensing-derived GLAI.
Second, it remains unclear why exactly the study area was chosen and to what extent it has greater relevance for other potential applications in different regions of the world with different crop types, different climatic and soil characteristics.
Furthermore, it is not clear which possible users the presented processing chain has and which questions - apart from comparing/evaluating data and model parameters - it can answer.
Lastly, the paper touches on too many aspects – the discussion of scale (spatial and temporal), inter-comparison of different data/models, different crop-related parameters –so that it is rather difficult to read. The focus on relevant questions within the carbon modelling community and clear answers to those got lost.
Overall, the paper holds certain potential for wide scientific interest, but needs to be substantially revised for better informative value on soil carbon fluxes and scalability of the method for other regions. I recommend performing major revisions before considering for publication in GMD.
I suggest to address the following major points:
the many spelling and grammatical errors in the manuscript (I can’t list all of these)

missed topic: the discrepancy between the introduction of the topic of soil carbon fluxes and the presentation of the approach to a crop model that rather maps biomass and NPP instead of soil carbon.

Clearer designation and focussed answering of relevant research questions/objectives

Justification for the selection of the study area in terms of what we can learn from it for other regions

Specific comments
Lines73ff (objectives)

Please be more specific about the objectives here. What are specifically “the outputs” you aim to check for accuracy and coherence? Instead of throwing in buzz words like “multi-scale validation”, name the research questions precisely, e.g. what is the impact of (a) spatial resolution and (b) temporal resolution of the input data? “verifying the coherency of the outputs through intra-field as well as regional analysis” sounds spongy too me. What exactly is meant by “coherency”, what is “intra-field and regional analysis”?
Line 112/land cover map

So, only the border extents of the shape file are used here to download the remote sensing and weather forcing data? Please describe how the land cover information is used, if at all, and how the land cover categories should look like.
Figure 4

What do the different colours mean? Why not use different colours for each dataset groups: input, validation and regional datasets?
Line 571f: “So high resolution allows more accurate estimates of the mean DAM values at field scale which enables more accurate field scale estimates of SOC changes by soil models in the perspective of monitoring SOC stock changes”

SOC was not really analysed and assessed here, so how would you know the effect of high-spatial resolution input data on soil organic carbon modelling? I think that this conclusion is simply too far-fetched here.
Line 665: “an intra-field scale quantification of the DAM and of the biomass that returns to the soil is needed for accurate monitoring of soil organic carbon stock changes”

Again, this conclusion seems far-fetched here. It is not clear to me how you assessed soil organic carbon with the proposed modelling chain, as it was never computed or derived from the output variables.

Technical corrections (selection, please perform comprehensive spelling correction!)
Line 8: misspelling – “radiative transfert”

Line 10/11: misspelling - “a land cover maps”

Line 14: misspelling - “agaisnt"

Line 103: misspelling – “corpping”

Line 111: misspelling - “Landcover”

Line 165: misspelling – “In Contrast”

Line 268 (and every other occurrence): misspelling “lykelyhoods”

Line 304: misspelling- “dry Biomass”

Line 322: “in the field 27” What is 27?

Line 325: misspelling – “is applied over a for winter wheat”

Table 3: “Quantify spatial and variability” What variability?

Line 353: “Dry Aboveground biomass, DAM measurements” Please present abbreviations in a uniform manner.

Line 453: misspelling - “respiration”

etc.
Citation: https://doi.org/10.5194/egusphere-2023-48-RC2
- AC2: 'Comment on egusphere-2023-48', Ahmad Al Bitar, 04 Jul 2023
  
  Dear Referee 2,
  We thank you for the review of the paper and the constructive comments. We provide in the attached pdf file clarifications and suggestions for modifications to the manuscript.
  Sincerely.
  
  Citation: https://doi.org/10.5194/egusphere-2023-48-AC2
AC2: 'Comment on egusphere-2023-48', Ahmad Al Bitar, 04 Jul 2023

Dear Referee 2,
We thank you for the review of the paper and the constructive comments. We provide in the attached pdf file clarifications and suggestions for modifications to the manuscript.
Sincerely.

Citation: https://doi.org/10.5194/egusphere-2023-48-AC2
AC3: 'Comment on egusphere-2023-48', Ahmad Al Bitar, 04 Jul 2023

Dear Referee 1,
We would like to add the following elements before finalizing the discussion. We hope that they bring some clarifications to the remaining questions and give an overview of the modifications we envision.
We thank you for the discussion phase. We are confident it will enhance our manuscript.
Sincerely.

Citation: https://doi.org/10.5194/egusphere-2023-48-AC3

Taeken Wijmer et al.

Data sets

AgriCarbon-EO Winter wheat Net Ecosystem Exchange and Biomass over South-west France at 10 m resolution Taeken Wijmer, Ahmad Al Bitar, and Eric Ceschia https://doi.org/10.5281/zenodo.7534280

Taeken Wijmer et al.

Viewed

Total article views: 1,057 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
730	301	26	1,057	6	7

HTML: 730
PDF: 301
XML: 26
Total: 1,057
BibTeX: 6
EndNote: 7

Views and downloads (calculated since 22 Feb 2023)

Month	HTML	PDF	XML	Total
Feb 2023	113	36	4	153
Mar 2023	147	63	6	216
Apr 2023	157	42	6	205
May 2023	49	14	0	63
Jun 2023	72	19	2	93
Jul 2023	92	69	7	168
Aug 2023	49	26	0	75
Sep 2023	51	32	1	84

Cumulative views and downloads (calculated since 22 Feb 2023)

Month	HTML	PDF	XML	Total
Feb 2023	113	36	4	153
Mar 2023	147	63	6	216
Apr 2023	157	42	6	205
May 2023	49	14	0	63
Jun 2023	72	19	2	93
Jul 2023	92	69	7	168
Aug 2023	49	26	0	75
Sep 2023	51	32	1	84

Viewed (geographical distribution)

Total article views: 1,010 (including HTML, PDF, and XML) Thereof 1,010 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 01 Oct 2023

Short summary

Quantification of Carbon fluxes of crops is an essential brick for the construction of a Monitoring, Reporting and Verification approach. We developed an end-to-end platform (AgriCarbon-EO) that assimilates through an efficient Bayesian approach, high resolution (10 m) optical remote sensing data into radiative transfer and crop modelling at regional scale (100 x 100 km). Large-scale estimates of carbon flux are validated against in-situ flux towers, yield maps, and analysed at regional scale.


Total:	0
HTML:	0
PDF:	0
XML:	0