the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Jul 2023
Spatially explicit re-harmonized terrestrial carbon densities for calibrating Integrated human-Earth System Models
Abstract. Soil and vegetation carbon densities play a critical role in global and regional human-Earth system models and MultiSector Dynamics Models. These densities affect variables such as land use change emissions and also influence land use change pathways under climate forcing scenarios where terrestrial carbon is assigned a carbon price. Recently, more spatially explicit, fine resolution data have become available for both soil and vegetation carbon. However, for models to effectively use these data the fine resolution data need to be reharmonized to the initial land use and land cover conditions represented by these models. Without such reharmonization the carbon values may be inaccurate for particular land types and places where the source data and the model disagree on the land use/cover type. Here we present reharmonized soil and vegetation carbon densities both at the 5-arcmin resolution grid cell level and also aggregated to 235 water sheds for 4 land use types and 15 land cover types. These data are particularly useful as initial land carbon conditions for global Multisectoral Dynamic Models (MSD). Moreover, these data include six different statistical states calculated using distinct resampling methods for each of the land use and land cover types. These statistical states are used to define a range of possible carbon values for each land classification, and any state can be used for defining initial conditions of soil and vegetation carbon in MSD models. Users can also estimate any percentile of the carbon distribution defined by these six summary states. We make use of these statistical states to calculate spatially distinct uncertainties in the carbon densities by land type. We have implemented these data in a state-of-the-art multi sector dynamics model, namely the Global Change Analysis Model (GCAM), and show that these new data improve several land use responses in the model, especially when terrestrial carbon is assigned a carbon price. The statistical states in our data are validated against similar estimates in the literature both at a grid cell level and at a regional level.
- Preprint
(2780 KB) - Metadata XML
-
Supplement
(498 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2023-251', Anonymous Referee #1, 03 Sep 2023
Narayan’s study aims to produce a carbon density dataset including aboveground biomass, belowground biomass and soil organic carbon for human-earth system models. The authors reharmonized several existing datasets to produce biomass and soil carbon datasets for GCAM. The topic is important, but this study lacks the innovation, and the manuscript looks more like a technical report or product manual. The manuscript still need much more work to highlight the importance of this idea, and introduce the method clearly. Besides, the manuscript was written very poorly, and the authors wrote very casually. I did not enjoy reading at all. Page 1, Line 38-39: what does this sentence mean? Actually, process-based ecosystem models can simulate biomass and soil organic carbon, and do not use directly the existing datasets into the models. Page 2, Line 10-18: I really do not understand what the authors mean here. The carbon density datasets with fine resolution basically were generated by using jointly field surveys and satellite-based observations, and were weakly dependent on the land cover type. In addition, HYDE also integrate the satellite-based land cover product in it during the recent decades. Page 2, Line 19-28: what kind challenges for resolution mismatch? As the datasets of biomass and soil carbon have fine resolution, it is easy to be aggerated to coarse resolution or geopolitical units which was used by the models. Why is this mismatch a challenge? Page 2, Line 29-39: it seems that authors have introduced the main points why they conducted this study. However, I did not feel importance of this study yet. And this paragraph is also really mixed with many information which should be put into the method section. Page 2, Line 40: I totally lost and did not understand the purpose of the authors for this paragraph. Page 3, Line 8-23: it will be better to put part of this paragraph forward, i.e., why GCAM need a prescribed biomass and soil carbon dataset. And the rest of this paragraph concerning how to make it should be moved to result section. The authors used about 3-4 paragraphs to briefly introduce the results in the introduction section, which is not really appropriate. Because, it is hard to understand what you mean here before seeing the method details. 2.1 section is too simply method. Generating datasets with 5 arcmin from 300 m resolution, and only used several common resample methods, which did not indicate any ecological meanings. How about considering the heterogeneity of vegetation to upscaling to 5 arcmin based on satellite-based observations? Table 2: this information is like common knowledge, and it is necessary to show in this table? 2.2.2 section, the method used here is also simple method, and the manuscript lacks the scientific innovation and more likes a technical report or dataset manual. Table 4 and other tables and figures: the authors seem not to draw them carefully, and they look like semi-finished products. The authors casually drew them. Besides the equation and other main text also have a lot of incorrected 2.3 section: if you did not use the above methods, and simply to aggregate the values in the pixels to the regions, what kind differences would be? Page 12, line1-5: the serial number is not correct, and suggest use brackets. Fig. 2 and other figures: the authors should put the top text line to figure captions. Page 14, line 7: should you use 3.1 instead of 3A for serial number? Table 5 and other tables, the table captions should put the top of the table which is different to the figures. Page 25, line 10-23: this part should be put in the method section. Page 27, line 7-18: this section looks important for this manuscript, but the authors failed to provide evidences for this model setting. Page 33, line 13-21: this paragraph is to introduce results, not discussion. Same to other four paragraphs. The discussion is very shallow, and basically is repeat with results.
Citation: https://doi.org/10.5194/essd-2023-251-RC1 -
RC2: 'Comment on essd-2023-251', Anonymous Referee #2, 07 Oct 2023
A globally reliable soil carbon database is of utmost importance for scientific research. This study is timely and fulfills a fundamental data requirement for earth science studies, particularly within the context of decarbonization. However, the innovation and improvement of this study should be justified. Here are a few suggestions to further benefit the readers:
- In the abstract, it is important to provide a summarized process of the re-harmonization and also specify the validation results of the statistical states mentioned by the authors. This will help readers gain a better understanding of the study's methodology and performance in GCAM.
- In the introduction, when introducing the new dataset implemented in GCAMs, it would be beneficial for the authors to provide an explanation of what GCAMs are and compare the accuracy of outputs using current soil density datasets. This will help establish the significance and objectives of this dataset. The detailed processing of the dataset can be discussed in a separate section, rather than summarizing it in the introduction
- It would be helpful for readers if the variables used in the study are clearly defined. For example, explaining that LT_MASK_300m is a binary variable representing the state of a specific land type (where 1 signifies that a particular land type is present and 0 signifies otherwise).
- It should be noted that there are actually 396 aggregations, not 366.
- In section 2.2.2, it would be beneficial to provide more information about how neighboring cells were selected and specify which nearest neighbor algorithm was used for this purpose.
- More explanations and validations regarding the distribution of carbon in Figures 4 and 5 would be helpful for readers.
- Figures 6c and 7c could include histograms for different vegetations to provide more comprehensive information. Additionally, on page 18 lines 23-28, although potential reasons were mentioned, it would be valuable to explore potential pathways for improvement as well.
- While this study implemented data in GCAM and compared it to Houghton's results, it would be beneficial to provide suggestions for primary choices of resampling methods that could be used to reduce workload when applying these data to different models in future implementations.
- On page 34 line 13, further justification is needed for the mentioned reduction of 30%. The rationale or reasoning behind this reduction should be explained to provide a clearer understanding of why this value was chosen.
Citation: https://doi.org/10.5194/essd-2023-251-RC2
Status: closed
-
RC1: 'Comment on essd-2023-251', Anonymous Referee #1, 03 Sep 2023
Narayan’s study aims to produce a carbon density dataset including aboveground biomass, belowground biomass and soil organic carbon for human-earth system models. The authors reharmonized several existing datasets to produce biomass and soil carbon datasets for GCAM. The topic is important, but this study lacks the innovation, and the manuscript looks more like a technical report or product manual. The manuscript still need much more work to highlight the importance of this idea, and introduce the method clearly. Besides, the manuscript was written very poorly, and the authors wrote very casually. I did not enjoy reading at all. Page 1, Line 38-39: what does this sentence mean? Actually, process-based ecosystem models can simulate biomass and soil organic carbon, and do not use directly the existing datasets into the models. Page 2, Line 10-18: I really do not understand what the authors mean here. The carbon density datasets with fine resolution basically were generated by using jointly field surveys and satellite-based observations, and were weakly dependent on the land cover type. In addition, HYDE also integrate the satellite-based land cover product in it during the recent decades. Page 2, Line 19-28: what kind challenges for resolution mismatch? As the datasets of biomass and soil carbon have fine resolution, it is easy to be aggerated to coarse resolution or geopolitical units which was used by the models. Why is this mismatch a challenge? Page 2, Line 29-39: it seems that authors have introduced the main points why they conducted this study. However, I did not feel importance of this study yet. And this paragraph is also really mixed with many information which should be put into the method section. Page 2, Line 40: I totally lost and did not understand the purpose of the authors for this paragraph. Page 3, Line 8-23: it will be better to put part of this paragraph forward, i.e., why GCAM need a prescribed biomass and soil carbon dataset. And the rest of this paragraph concerning how to make it should be moved to result section. The authors used about 3-4 paragraphs to briefly introduce the results in the introduction section, which is not really appropriate. Because, it is hard to understand what you mean here before seeing the method details. 2.1 section is too simply method. Generating datasets with 5 arcmin from 300 m resolution, and only used several common resample methods, which did not indicate any ecological meanings. How about considering the heterogeneity of vegetation to upscaling to 5 arcmin based on satellite-based observations? Table 2: this information is like common knowledge, and it is necessary to show in this table? 2.2.2 section, the method used here is also simple method, and the manuscript lacks the scientific innovation and more likes a technical report or dataset manual. Table 4 and other tables and figures: the authors seem not to draw them carefully, and they look like semi-finished products. The authors casually drew them. Besides the equation and other main text also have a lot of incorrected 2.3 section: if you did not use the above methods, and simply to aggregate the values in the pixels to the regions, what kind differences would be? Page 12, line1-5: the serial number is not correct, and suggest use brackets. Fig. 2 and other figures: the authors should put the top text line to figure captions. Page 14, line 7: should you use 3.1 instead of 3A for serial number? Table 5 and other tables, the table captions should put the top of the table which is different to the figures. Page 25, line 10-23: this part should be put in the method section. Page 27, line 7-18: this section looks important for this manuscript, but the authors failed to provide evidences for this model setting. Page 33, line 13-21: this paragraph is to introduce results, not discussion. Same to other four paragraphs. The discussion is very shallow, and basically is repeat with results.
Citation: https://doi.org/10.5194/essd-2023-251-RC1 -
RC2: 'Comment on essd-2023-251', Anonymous Referee #2, 07 Oct 2023
A globally reliable soil carbon database is of utmost importance for scientific research. This study is timely and fulfills a fundamental data requirement for earth science studies, particularly within the context of decarbonization. However, the innovation and improvement of this study should be justified. Here are a few suggestions to further benefit the readers:
- In the abstract, it is important to provide a summarized process of the re-harmonization and also specify the validation results of the statistical states mentioned by the authors. This will help readers gain a better understanding of the study's methodology and performance in GCAM.
- In the introduction, when introducing the new dataset implemented in GCAMs, it would be beneficial for the authors to provide an explanation of what GCAMs are and compare the accuracy of outputs using current soil density datasets. This will help establish the significance and objectives of this dataset. The detailed processing of the dataset can be discussed in a separate section, rather than summarizing it in the introduction
- It would be helpful for readers if the variables used in the study are clearly defined. For example, explaining that LT_MASK_300m is a binary variable representing the state of a specific land type (where 1 signifies that a particular land type is present and 0 signifies otherwise).
- It should be noted that there are actually 396 aggregations, not 366.
- In section 2.2.2, it would be beneficial to provide more information about how neighboring cells were selected and specify which nearest neighbor algorithm was used for this purpose.
- More explanations and validations regarding the distribution of carbon in Figures 4 and 5 would be helpful for readers.
- Figures 6c and 7c could include histograms for different vegetations to provide more comprehensive information. Additionally, on page 18 lines 23-28, although potential reasons were mentioned, it would be valuable to explore potential pathways for improvement as well.
- While this study implemented data in GCAM and compared it to Houghton's results, it would be beneficial to provide suggestions for primary choices of resampling methods that could be used to reduce workload when applying these data to different models in future implementations.
- On page 34 line 13, further justification is needed for the mentioned reduction of 30%. The rationale or reasoning behind this reduction should be explained to provide a clearer understanding of why this value was chosen.
Citation: https://doi.org/10.5194/essd-2023-251-RC2
Data sets
Spatially explicit re-harmonized terrestrial carbon densities for calibrating Integrated Multisectoral Models Kanishka B. Narayan; Alan Di Vittorio; Evan Margiotta; Seth A. Spawn; Holly Gibbs https://zenodo.org/record/7884615
Model code and software
moirai land data system Alan Di Vittorio; Evan Margiotta; Kanishka B. Narayan; Chris Vernon https://github.com/JGCRI/moirai/tree/master/ancillary/carbon_harmonization
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 510 | 170 | 46 | 726 | 72 | 48 | 57 |
- HTML: 510
- PDF: 170
- XML: 46
- Total: 726
- Supplement: 72
- BibTeX: 48
- EndNote: 57
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
