the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Jul 2023
Terrestrial Ecosystem Model in R (TEMIR) version 1.0: Simulating ecophysiological responses of vegetation to atmospheric chemical and meteorological changes
Abstract. The newly developed offline land ecosystem model Terrestrial Ecosystem Model in R (TEMIR) version 1.0 is described here. This version of the model simulates plant ecophysiological (e.g., photosynthetic, stomatal) responses to varying meteorological conditions and concentrations of CO2 and ground-level ozone (O3) based on prescribed meteorological and atmospheric chemical inputs from various sources. Driven by the same meteorological data used in the GEOS-Chem chemical transport model, this allows asynchronously coupled experiments with GEOS-Chem simulations with unique coherency for investigating biosphere-atmosphere chemical interactions. TEMIR agrees well with FLUXNET site-level gross primary productivity (GPP) in terms of both the diurnal and monthly cycles (correlation coefficients R2 > 0.85 and R2 > 0.8, respectively) for most plant functional types (PFTs). Grass and shrub PFTs have larger biases due to generic model representations. The model performs best when driven by local site-level meteorology rather than reanalyzed gridded meteorology. Simulation using gridded meteorology agrees well for annual GPP in seasonality and spatial distribution with a global average of 134 Pg C yr–1. Application of Monin-Obukhov similarity theory to infer canopy conditions from gridded meteorology does not improve model performance, predicting a uniform increase of +21 % for global GPP. Present-day O3 concentrations simulated by GEOS-Chem and an O3 damage scheme at high sensitivity show a 2 % reduction in global GPP with prominent reductions by up to 15 % in eastern China and the eastern US. Regional correlations are generally unchanged when O3 is present, and biases are reduced especially for regions with high O3 damages. An increase in atmospheric CO2 concentration by 20 ppmv from year-2000 to year-2010 level modestly decreases O3 damage due to reduced stomatal uptake, consistent with ecophysiological understanding. Our work showcases the utility of this version of TEMIR for evaluating biogeophysical responses of vegetation to changes in atmospheric composition and meteorological conditions.
- Preprint
(5663 KB) - Metadata XML
-
Supplement
(9367 KB) - BibTeX
- EndNote
Status: closed
-
CEC1: 'Comment on egusphere-2023-1287', Juan Antonio Añel, 03 Aug 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.htmlActually, it contains several flaws regarding our policy, and your manuscript should not have been accepted in Discussions, given this lack of compliance. Therefore, the current situation with your manuscript is irregular. In this way, if you do not fix the issues listed next, we will have to reject your manuscript for publication in our journal.
First, as the handling topical editor pointed out earlier, your code repository does not include a license. If you do not include a license, the code continues to be your property, and nobody can use it. Therefore, it is not possible to test or replicate your work. In this way, you must add a license to your repository. You could want to choose a free software/open-source (FLOSS) license. 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, you can choose other options that Zenodo provides: GPLv2, Apache License, MIT License, etc.
Second, the instructions to execute your code include links to repositories that we do not accept. Again, please, check our policy. Some of them even request a password that it is said needs to be obtained by contacting by email. Our policy is crystal clear that all the assets used to produce the submitted work must be published openly and publicly at submission time and that the need to request access is not acceptable. Therefore, all the input data necessary for your work and the output data generated during your experiments must be published in a repository that we accept according to our policy.
I note here, too, that your manuscript does not even contain a "Data availability" section.
Therefore, please, add a license to your code, publish your data in one of the appropriate repositories, and reply to this comment with the relevant information (link and DOI) as soon as possible, as it should be available before the Discussions stage.
Also, you must include in a potentially reviewed version of your manuscript the new 'Data Availability' section and the DOI of the repository.
Juan A. AñelGeosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/egusphere-2023-1287-CEC1 -
AC1: 'Reply on CEC1', Amos Tai, 04 Aug 2023
We profoundly apologize for the inconvenience caused by overlooking the data/code licensing and repositing requirement, and hereby provide the necessary corrections. The recommended license (GNU General Public License) has been added as "LICENSE.txt" as part of the code, which is now updated and reposited here (DOI: 10.5281/zenodo.8215332):
https://doi.org/10.5281/zenodo.8215332
The input dataset necessary for running TEMIR v1.0 has also been uploaded, licensed and provided via the public repository here (DOI: 10.5281/zenodo.8215152):
https://doi.org/10.5281/zenodo.8215152
A "Data availability" section with the above updated links and DOIs has also been added to the manuscript, and is attached here in this reply. We thank the editor again for the necessary suggestions to ensure compliance of our manuscript with GMD's rigorous requirements.
-
AC1: 'Reply on CEC1', Amos Tai, 04 Aug 2023
-
RC1: 'Comment on egusphere-2023-1287', Anonymous Referee #1, 21 Sep 2023
The study presents an easy to use ecosystem model, which is a step to open up ecosystem modelling for a wider community.
With a background in ecosystem modelling, I find this really appealing. Although I can not fully understand/buy into what you are talking about in the introduction regarding the use of R by ecologists. Sure, we in the ecosystem modelling community are using lower level computer language, but we are also using R, Matlab or Python for our analysis. I suggest that you re-write that section.
Generally, I like your model description, but there are two things that I beleive needs a bit more attention. First of all, PFT:s (Plant Functional Types), are you really simulating PFT:s? From what I have read, it is more like you are simulating different land tiles with monocultures of PFT:s planted on them. There is a mix, in the text, between talking about a PFT as plant and PFT as a land-cover/vegetation type. For instance in section 2.2.1. For instance, bare ground can not be a PFT.
Depending on your answer to the question above, if you are truly simulating different PFT:s in a tile, then competition needs to be presented. If not, then clearly state that you are not simulating competition between different strategies, and that the spatial extent of these strategies is fixed. Something that I was missing from the discussion as well. How is that affecting your results, that you have a fixed fraction of strategies?
I would claim that for instance having different sensitiveness to ozone would affect the composition if [O3] will change in the future.Some specific corrections:
line 10, page 2; there is a missing "is" on that line.Eq. 14 is missing something
Citation: https://doi.org/10.5194/egusphere-2023-1287-RC1 -
RC2: 'Comment on egusphere-2023-1287', Anonymous Referee #2, 30 Sep 2023
Review comments for egusphere-2023-1287
Title: Terrestrial Ecosystem Model in R (TEMIR) version 1.0: Simulating ecophysiological responses of vegetation to atmospheric chemical and meteorological changes
This study developed a terrestrial ecosystem model using the R programming language, which is beneficial for more ecologists to use ecosystem model. I like the idea of developing an open and free ecosystem model using the user-friendly programming language R, it will help expand the ecosystem modeling community. The TEMIR is capable of modeling plant ecophysiological responses to varying meteorological conditions and CO2 concentrations and ground-level ozone (O3) based on prescribed meteorological and atmospheric chemical inputs. The authors validated the model performance at both site and global scales. In general, this manuscript is well organized, but still has several drawbacks. The reviewer has the following concerns and suggestions for the authors to consider:
(1) Is the spatial distribution and fractional coverage of PFTs fixed during the simulation period? Does it mean that the model doesn’t consider land use change effect? If so, it is necessary to discuss the effect of lacking LULCC change on GPP estimation.
(2) Please elaborate what MODIS product did you use to derive PFTs distribution maps. And how did you sperate C3 and C4 crops, since MODIS products don’t have such information.
(3) MERRA-2 has a native resolution of 0.5° x 0.625°, why didn’t you conduct simulations at this relative high resolution?
(4) Please elaborate how is the dark respiration (Rd) rate calculated in the model.
(5) The TEMIR contains a two-layer soil model. Does this soil model have carbon cycling and water transport processes?
(6) The canopy decay coefficient for nitrogen Kn is 0.30 in the model. How was this number determined? Shouldn’t it vary with leaf nitrogen content or nitrogen availability? It seems that the current version of TEMIR doesn’t have an explicit nitrogen cycle. If so, it is necessary to discuss how does lack of considering nitrogen cycle affect GPP estimation.
(7) The model doesn’t perform well at several crop sites. Could you add one paragraph discussing the potential measures to improve model performance for crops?
(8) For the site-level simulation, why did you fix CO2 concentration at the level of 390 ppmv, rather than using the actual CO2 concentration during the study period?
(9) For the site CH-Cha, can you explain why the model performance is good using reanalysis meteorological input data?
(10) You explained that the underestimation of GPP at the FR-Fon site is most likely due to inaccurate parameterization overcompensating for the uncertainties of satellite derived LAI. Have you tried to quantify how uncertainties in LAI affect GPP estimation. Maybe you can compare the difference using tow simulations: one use MODIS LAI as input, one use observed LAI as input.
(11) I suggest the author add a paragraph describing what meteorological variables are needed to drive the TEMIR.
(12) Please add one table listing the key parameters for different PFTs.
Citation: https://doi.org/10.5194/egusphere-2023-1287-RC2 -
AC2: 'Comment on egusphere-2023-1287', Amos Tai, 09 Jan 2024
We thank the two reviewers for their insightful and helpful comments, which have helped greatly improve the manuscript, and the editor for handling this manuscript. Our point-by-point response is now attached here as a Supplement. For the editor's easier reference, the revised manuscript is linked below:
https://gocuhk-my.sharepoint.com/:b:/g/personal/amostai_cuhk_edu_hk/EfEfLBgcsZ9GkNWMhzDTkRwB-EPwzwANA1-RFA1OMKkaRQ?e=dZs3Vj
Status: closed
-
CEC1: 'Comment on egusphere-2023-1287', Juan Antonio Añel, 03 Aug 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.htmlActually, it contains several flaws regarding our policy, and your manuscript should not have been accepted in Discussions, given this lack of compliance. Therefore, the current situation with your manuscript is irregular. In this way, if you do not fix the issues listed next, we will have to reject your manuscript for publication in our journal.
First, as the handling topical editor pointed out earlier, your code repository does not include a license. If you do not include a license, the code continues to be your property, and nobody can use it. Therefore, it is not possible to test or replicate your work. In this way, you must add a license to your repository. You could want to choose a free software/open-source (FLOSS) license. 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, you can choose other options that Zenodo provides: GPLv2, Apache License, MIT License, etc.
Second, the instructions to execute your code include links to repositories that we do not accept. Again, please, check our policy. Some of them even request a password that it is said needs to be obtained by contacting by email. Our policy is crystal clear that all the assets used to produce the submitted work must be published openly and publicly at submission time and that the need to request access is not acceptable. Therefore, all the input data necessary for your work and the output data generated during your experiments must be published in a repository that we accept according to our policy.
I note here, too, that your manuscript does not even contain a "Data availability" section.
Therefore, please, add a license to your code, publish your data in one of the appropriate repositories, and reply to this comment with the relevant information (link and DOI) as soon as possible, as it should be available before the Discussions stage.
Also, you must include in a potentially reviewed version of your manuscript the new 'Data Availability' section and the DOI of the repository.
Juan A. AñelGeosci. Model Dev. Executive EditorCitation: https://doi.org/10.5194/egusphere-2023-1287-CEC1 -
AC1: 'Reply on CEC1', Amos Tai, 04 Aug 2023
We profoundly apologize for the inconvenience caused by overlooking the data/code licensing and repositing requirement, and hereby provide the necessary corrections. The recommended license (GNU General Public License) has been added as "LICENSE.txt" as part of the code, which is now updated and reposited here (DOI: 10.5281/zenodo.8215332):
https://doi.org/10.5281/zenodo.8215332
The input dataset necessary for running TEMIR v1.0 has also been uploaded, licensed and provided via the public repository here (DOI: 10.5281/zenodo.8215152):
https://doi.org/10.5281/zenodo.8215152
A "Data availability" section with the above updated links and DOIs has also been added to the manuscript, and is attached here in this reply. We thank the editor again for the necessary suggestions to ensure compliance of our manuscript with GMD's rigorous requirements.
-
AC1: 'Reply on CEC1', Amos Tai, 04 Aug 2023
-
RC1: 'Comment on egusphere-2023-1287', Anonymous Referee #1, 21 Sep 2023
The study presents an easy to use ecosystem model, which is a step to open up ecosystem modelling for a wider community.
With a background in ecosystem modelling, I find this really appealing. Although I can not fully understand/buy into what you are talking about in the introduction regarding the use of R by ecologists. Sure, we in the ecosystem modelling community are using lower level computer language, but we are also using R, Matlab or Python for our analysis. I suggest that you re-write that section.
Generally, I like your model description, but there are two things that I beleive needs a bit more attention. First of all, PFT:s (Plant Functional Types), are you really simulating PFT:s? From what I have read, it is more like you are simulating different land tiles with monocultures of PFT:s planted on them. There is a mix, in the text, between talking about a PFT as plant and PFT as a land-cover/vegetation type. For instance in section 2.2.1. For instance, bare ground can not be a PFT.
Depending on your answer to the question above, if you are truly simulating different PFT:s in a tile, then competition needs to be presented. If not, then clearly state that you are not simulating competition between different strategies, and that the spatial extent of these strategies is fixed. Something that I was missing from the discussion as well. How is that affecting your results, that you have a fixed fraction of strategies?
I would claim that for instance having different sensitiveness to ozone would affect the composition if [O3] will change in the future.Some specific corrections:
line 10, page 2; there is a missing "is" on that line.Eq. 14 is missing something
Citation: https://doi.org/10.5194/egusphere-2023-1287-RC1 -
RC2: 'Comment on egusphere-2023-1287', Anonymous Referee #2, 30 Sep 2023
Review comments for egusphere-2023-1287
Title: Terrestrial Ecosystem Model in R (TEMIR) version 1.0: Simulating ecophysiological responses of vegetation to atmospheric chemical and meteorological changes
This study developed a terrestrial ecosystem model using the R programming language, which is beneficial for more ecologists to use ecosystem model. I like the idea of developing an open and free ecosystem model using the user-friendly programming language R, it will help expand the ecosystem modeling community. The TEMIR is capable of modeling plant ecophysiological responses to varying meteorological conditions and CO2 concentrations and ground-level ozone (O3) based on prescribed meteorological and atmospheric chemical inputs. The authors validated the model performance at both site and global scales. In general, this manuscript is well organized, but still has several drawbacks. The reviewer has the following concerns and suggestions for the authors to consider:
(1) Is the spatial distribution and fractional coverage of PFTs fixed during the simulation period? Does it mean that the model doesn’t consider land use change effect? If so, it is necessary to discuss the effect of lacking LULCC change on GPP estimation.
(2) Please elaborate what MODIS product did you use to derive PFTs distribution maps. And how did you sperate C3 and C4 crops, since MODIS products don’t have such information.
(3) MERRA-2 has a native resolution of 0.5° x 0.625°, why didn’t you conduct simulations at this relative high resolution?
(4) Please elaborate how is the dark respiration (Rd) rate calculated in the model.
(5) The TEMIR contains a two-layer soil model. Does this soil model have carbon cycling and water transport processes?
(6) The canopy decay coefficient for nitrogen Kn is 0.30 in the model. How was this number determined? Shouldn’t it vary with leaf nitrogen content or nitrogen availability? It seems that the current version of TEMIR doesn’t have an explicit nitrogen cycle. If so, it is necessary to discuss how does lack of considering nitrogen cycle affect GPP estimation.
(7) The model doesn’t perform well at several crop sites. Could you add one paragraph discussing the potential measures to improve model performance for crops?
(8) For the site-level simulation, why did you fix CO2 concentration at the level of 390 ppmv, rather than using the actual CO2 concentration during the study period?
(9) For the site CH-Cha, can you explain why the model performance is good using reanalysis meteorological input data?
(10) You explained that the underestimation of GPP at the FR-Fon site is most likely due to inaccurate parameterization overcompensating for the uncertainties of satellite derived LAI. Have you tried to quantify how uncertainties in LAI affect GPP estimation. Maybe you can compare the difference using tow simulations: one use MODIS LAI as input, one use observed LAI as input.
(11) I suggest the author add a paragraph describing what meteorological variables are needed to drive the TEMIR.
(12) Please add one table listing the key parameters for different PFTs.
Citation: https://doi.org/10.5194/egusphere-2023-1287-RC2 -
AC2: 'Comment on egusphere-2023-1287', Amos Tai, 09 Jan 2024
We thank the two reviewers for their insightful and helpful comments, which have helped greatly improve the manuscript, and the editor for handling this manuscript. Our point-by-point response is now attached here as a Supplement. For the editor's easier reference, the revised manuscript is linked below:
https://gocuhk-my.sharepoint.com/:b:/g/personal/amostai_cuhk_edu_hk/EfEfLBgcsZ9GkNWMhzDTkRwB-EPwzwANA1-RFA1OMKkaRQ?e=dZs3Vj
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 429 | 238 | 42 | 709 | 60 | 24 | 26 |
- HTML: 429
- PDF: 238
- XML: 42
- Total: 709
- Supplement: 60
- BibTeX: 24
- EndNote: 26
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
