the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Mar 2024
Environmental drivers and remote sensing proxies of post-fire thaw depth in Eastern Siberian larch forests
Abstract. Boreal fire regimes are intensifying because of climate change and the northern parts of boreal forests are underlain by permafrost. Boreal fires combust vegetation and organic soils, which insulate permafrost, and as such deepen the seasonally thawed active layer and can lead to further carbon emissions to the atmosphere. Current understanding of the environmental drivers of post-fire thaw depth is limited but of critical importance. In addition, mapping thaw depth over fire scars may enable a better understanding of the spatial variability in post-fire responses of permafrost soils. We assessed the environmental drivers of post-fire thaw depth using field data from a fire scar in a larch-dominated forest in the continuous permafrost zone in Eastern Siberia. Particularly, summer thaw depth was deeper in burned (mean = 127.3 cm, standard deviation (sd) = 27.7 cm) than in unburned (98.1 cm, sd = 26.9 cm) landscapes one year after the fire, yet the effect of fire was modulated by landscape and vegetation characteristics. We found deeper thaw in well-drained landscape positions, in open larch forest often intermixed with Scots pine, and in high severity burns. The environmental drivers, site moisture, forest type and density, and fire severity explained 73.4 % of the measured thaw depth variability at the study sites. In addition, we evaluated the relationships between field-measured thaw depth and several remote sensing proxies. Albedo, the differenced Normalized Burn Ratio (dNBR), land surface temperature (LST), and pre-fire Normalized Difference Vegetation Index (NDVI) derived from Landsat 8 imagery together explained 66.3 % of the variability in field-measured thaw depth. Based on these remote sensing proxies and multiple linear regression analysis, we estimated thaw depth over the entire fire scar, and found that LST displayed particularly strong correlations with post-fire thaw depth (r = 0.65, p < 0.01). Our study reveals some of the governing processes of post-fire thaw depth development and shows the capability of Landsat imagery to estimate thaw depth at a landscape scale.
- Preprint
(3010 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-469', Anonymous Referee #1, 16 Apr 2024
Review of ‘Environmental drivers and remote sensing proxies of post-fire thaw depth in Eastern Siberian larch forests’ by Diaz et al. (2024)
This paper investigated post-fire thaw depth and its driving factors in East Siberian employing both field measurement and remote sensing proxies. Their findings indicate that: 1) fire exacerbates thaw depth, particularly when compared to unburned regions; 2) a combination of site moisture, forest composition, and fire severity accounts for 73.4% of thaw depth variability based on field investigations, while remote sensing proxies such as albedo, differenced Normalized Burn Ratio, land surface temperature, and NDVI contribute to explaining 66.3% of the variability.
The research explored topographical, vegetative, and burning effects on post-fire thaw depth in permafrost soil and mapped thaw depth with remote sensing data. Although the framework looks promising, some clarifications and elucidations are necessary to bolster more convincing findings. There are some limitations that I believe require further investigation.
My major comments are:
- The manual selection process of driving factors for the MLR model appears insufficiently rigorous. Despite burn depth showing the highest correlation with thaw depth, it's omitted from the regression model. Do you have any thoughts on this selection?
- Could you provide the significance for all the correlation matrices in Figures A1 and B1? Burn depth exhibits a positive correlation to thaw depth (0.53) and soil moisture (0.04), respectively, however, the thaw depth has a negative correlation (-0.53) to site moisture. This raises questions.
- Regarding the application of multi-linear regression, are you utilizing the original data or standardized data? Expanding on this in section 2.4 Statistical Approach would enhance clarity.
- Are the environmental factors and remote sensing proxies of thaw depth consistent between burned and unburned plots if you explore the data separately? How does the correlation coefficient fluctuate between burned and unburned regions?
- According to the MLR model, site moisture seems to play a more significant role in driving variations in thaw depth than fire severity. However, thaw depth in burned areas typically surpasses that of unburned areas on average. How do you consider the relative contributing importance of site moisture and burning severity? What are your thoughts on the potential driving mechanism of thaw depth by a comprehensive interpretation of the statistical model in this study? Furthermore, given the potential contribution of site moisture to thaw depth, why wasn't soil moisture remote sensing data considered?
Other small comments:
- Line 118: The reference for Johnstone et al. (2008) is missing.
- Line 234: Does the larch tree play a certain function in inducing boreal fires? What is the reason for retaining the larch proportion?
- Lines 280 – 282: why do plots with fewer larch trees thaw deeper? You may expand some discussion here on larch proportion.
- Figures A1 & B1: please add significance to the correlation matrix.
- Figures A2 & 3: For unburned areas where burn depth and GeoCBI are 0, it’s worth showing what drives the thaw depth when there is no fire rather than explaining everything by one statistic model.
- Line 405: please double-check all the references.
- Figure 4 (b) & (d): The fire scar wasn’t fully covered.
- Figure 8 (a) & (b): The fire scar wasn’t fully covered.
Citation: https://doi.org/10.5194/egusphere-2024-469-RC1 -
AC1: 'Reply on RC1', Lucas Ribeiro Diaz, 28 May 2024
Dear Editor and Reviewers,
Thank you for considering our manuscript for publication in ESD. We appreciate your time and your valuable comments.
Please see the attached file for responses to the comments of reviewer #1.
Sincerely,
Lucas R. Diaz & Sander Veraverbeke
-
RC2: 'Comment on egusphere-2024-469', Anonymous Referee #2, 22 May 2024
In this manuscript, the authors investigated post-fire thaw depth within one fire event in the Republic of Sakha. They used a combination of field data collected one year post-burn and compare this field data with multiple remote sensing indices derived from Landsat optical and thermal data. The environmental characteristics assessed included a variety of vegetation, fire severity, and thaw depth characteristics. The remote sensing techniques included several pre- and post-fire indices, including land surface temperature. Through their field work, the authors found deeper thaw in burned areas and well-drained areas. The authors found that the remote sensing characteristics assessed explained 66.3% of the variability in the field-measured thaw depth. Additionally, it was found that land surface temperature correlated highly with post-fire thaw depth (42.9% of the variability explained).
This was a well-written manuscript which clearly described the research planned and conducted, both in the field, and with the remote sensing techniques. The use of Landsat thermal data to assess thaw depth was a new application, and it was surprising that the correlation was so high, especially considering that the resolution of the data was 100m. The discussion section mentioned some of the concerns with these new techniques and adequately addressed them, including the resolution of the Landsat thermal data, the small sample size of the field dataset, and the timing of the collection of the field data (mid-summer, as opposed to end of summer when active layer thickness could be collected). The authors also provided a worthwhile discussion of future research including the use of more advanced machine learning techniques, collecting additional field data, and incorporating radar data into such an analysis in the future.
Comments:
- Line 22 and 232 – Was the thaw depth significantly deeper in burned than unburned plots? The mean and standard deviation are provided, but the significance level is not. Please provide it if possible.
- Section 2.3 – Consider a table to show the indices used and the formulas, as a way for readers to have a quick overview. Perhaps this could go in the Appendix.
- Line 160-162 – The pre-fire imagery is from 2 years prior to the fire, and 2 scenes needed to be mosaicked together to cover the entire fire event – Could this have affected any of the results? Consider adding a clarifying statement in either the methods or discussion section.
- Line 289 – The case studies of the 2 burned/unburned plot pairs undoubtedly helped in separating the impact of fire on thaw from topographic and vegetation influences, but this is still a very small sample size, and should be treated as such. Perhaps soften the language here from “enabled”, to show that this small sample size would not fully address all situations in separating influences on thaw depth.
- Figure A2 – Figure 2 provides a description of the meaning of the triangle and the bounds of the box plot, but it is not repeated in Figure A2. Consider adding it again here as the Appendix is separate from the main manuscript and readers could be confused.
Citation: https://doi.org/10.5194/egusphere-2024-469-RC2 -
AC2: 'Reply on RC2', Lucas Ribeiro Diaz, 28 May 2024
Dear Editor and Reviewers,
Thank you for considering our manuscript for publication in ESD. We appreciate your time and your valuable comments.
Please see the attached file for responses to the comments of reviewer #2.
Sincerely,
Lucas R. Diaz & Sander Veraverbeke
Status: closed
-
RC1: 'Comment on egusphere-2024-469', Anonymous Referee #1, 16 Apr 2024
Review of ‘Environmental drivers and remote sensing proxies of post-fire thaw depth in Eastern Siberian larch forests’ by Diaz et al. (2024)
This paper investigated post-fire thaw depth and its driving factors in East Siberian employing both field measurement and remote sensing proxies. Their findings indicate that: 1) fire exacerbates thaw depth, particularly when compared to unburned regions; 2) a combination of site moisture, forest composition, and fire severity accounts for 73.4% of thaw depth variability based on field investigations, while remote sensing proxies such as albedo, differenced Normalized Burn Ratio, land surface temperature, and NDVI contribute to explaining 66.3% of the variability.
The research explored topographical, vegetative, and burning effects on post-fire thaw depth in permafrost soil and mapped thaw depth with remote sensing data. Although the framework looks promising, some clarifications and elucidations are necessary to bolster more convincing findings. There are some limitations that I believe require further investigation.
My major comments are:
- The manual selection process of driving factors for the MLR model appears insufficiently rigorous. Despite burn depth showing the highest correlation with thaw depth, it's omitted from the regression model. Do you have any thoughts on this selection?
- Could you provide the significance for all the correlation matrices in Figures A1 and B1? Burn depth exhibits a positive correlation to thaw depth (0.53) and soil moisture (0.04), respectively, however, the thaw depth has a negative correlation (-0.53) to site moisture. This raises questions.
- Regarding the application of multi-linear regression, are you utilizing the original data or standardized data? Expanding on this in section 2.4 Statistical Approach would enhance clarity.
- Are the environmental factors and remote sensing proxies of thaw depth consistent between burned and unburned plots if you explore the data separately? How does the correlation coefficient fluctuate between burned and unburned regions?
- According to the MLR model, site moisture seems to play a more significant role in driving variations in thaw depth than fire severity. However, thaw depth in burned areas typically surpasses that of unburned areas on average. How do you consider the relative contributing importance of site moisture and burning severity? What are your thoughts on the potential driving mechanism of thaw depth by a comprehensive interpretation of the statistical model in this study? Furthermore, given the potential contribution of site moisture to thaw depth, why wasn't soil moisture remote sensing data considered?
Other small comments:
- Line 118: The reference for Johnstone et al. (2008) is missing.
- Line 234: Does the larch tree play a certain function in inducing boreal fires? What is the reason for retaining the larch proportion?
- Lines 280 – 282: why do plots with fewer larch trees thaw deeper? You may expand some discussion here on larch proportion.
- Figures A1 & B1: please add significance to the correlation matrix.
- Figures A2 & 3: For unburned areas where burn depth and GeoCBI are 0, it’s worth showing what drives the thaw depth when there is no fire rather than explaining everything by one statistic model.
- Line 405: please double-check all the references.
- Figure 4 (b) & (d): The fire scar wasn’t fully covered.
- Figure 8 (a) & (b): The fire scar wasn’t fully covered.
Citation: https://doi.org/10.5194/egusphere-2024-469-RC1 -
AC1: 'Reply on RC1', Lucas Ribeiro Diaz, 28 May 2024
Dear Editor and Reviewers,
Thank you for considering our manuscript for publication in ESD. We appreciate your time and your valuable comments.
Please see the attached file for responses to the comments of reviewer #1.
Sincerely,
Lucas R. Diaz & Sander Veraverbeke
-
RC2: 'Comment on egusphere-2024-469', Anonymous Referee #2, 22 May 2024
In this manuscript, the authors investigated post-fire thaw depth within one fire event in the Republic of Sakha. They used a combination of field data collected one year post-burn and compare this field data with multiple remote sensing indices derived from Landsat optical and thermal data. The environmental characteristics assessed included a variety of vegetation, fire severity, and thaw depth characteristics. The remote sensing techniques included several pre- and post-fire indices, including land surface temperature. Through their field work, the authors found deeper thaw in burned areas and well-drained areas. The authors found that the remote sensing characteristics assessed explained 66.3% of the variability in the field-measured thaw depth. Additionally, it was found that land surface temperature correlated highly with post-fire thaw depth (42.9% of the variability explained).
This was a well-written manuscript which clearly described the research planned and conducted, both in the field, and with the remote sensing techniques. The use of Landsat thermal data to assess thaw depth was a new application, and it was surprising that the correlation was so high, especially considering that the resolution of the data was 100m. The discussion section mentioned some of the concerns with these new techniques and adequately addressed them, including the resolution of the Landsat thermal data, the small sample size of the field dataset, and the timing of the collection of the field data (mid-summer, as opposed to end of summer when active layer thickness could be collected). The authors also provided a worthwhile discussion of future research including the use of more advanced machine learning techniques, collecting additional field data, and incorporating radar data into such an analysis in the future.
Comments:
- Line 22 and 232 – Was the thaw depth significantly deeper in burned than unburned plots? The mean and standard deviation are provided, but the significance level is not. Please provide it if possible.
- Section 2.3 – Consider a table to show the indices used and the formulas, as a way for readers to have a quick overview. Perhaps this could go in the Appendix.
- Line 160-162 – The pre-fire imagery is from 2 years prior to the fire, and 2 scenes needed to be mosaicked together to cover the entire fire event – Could this have affected any of the results? Consider adding a clarifying statement in either the methods or discussion section.
- Line 289 – The case studies of the 2 burned/unburned plot pairs undoubtedly helped in separating the impact of fire on thaw from topographic and vegetation influences, but this is still a very small sample size, and should be treated as such. Perhaps soften the language here from “enabled”, to show that this small sample size would not fully address all situations in separating influences on thaw depth.
- Figure A2 – Figure 2 provides a description of the meaning of the triangle and the bounds of the box plot, but it is not repeated in Figure A2. Consider adding it again here as the Appendix is separate from the main manuscript and readers could be confused.
Citation: https://doi.org/10.5194/egusphere-2024-469-RC2 -
AC2: 'Reply on RC2', Lucas Ribeiro Diaz, 28 May 2024
Dear Editor and Reviewers,
Thank you for considering our manuscript for publication in ESD. We appreciate your time and your valuable comments.
Please see the attached file for responses to the comments of reviewer #2.
Sincerely,
Lucas R. Diaz & Sander Veraverbeke
Data sets
Burned and Unburned Boreal Larch Forest Site Data, Northeast Siberia C. J. F. Delcourt et al. https://doi.org/10.5281/zenodo.10840088
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 473 | 196 | 38 | 707 | 22 | 24 |
- HTML: 473
- PDF: 196
- XML: 38
- Total: 707
- BibTeX: 22
- EndNote: 24
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1