the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Highly-resolved satellite remote sensing based land-use change inventory yields weaker surface albedo-induced global cooling
Abstract. Land-use change (LUC) is ranked as the second anthropogenic source of climate change after fossil fuel burning and yields negative albedo-induced radiative forcing (ARF). This cooling effect has been assessed using low spatiotemporally resolved LUC datasets derived from historical statistical data with large uncertainties. Herein, we implement a satellite remote sensing derived highly resolved LUC dataset into a compact earth system model and reassess the global and regional surface ARF by LUC from 1983 to 2010 relative to 1750. We find that the magnitude of negative ARF obtained from the present study is lower by 20 % than that estimated by the Intergovernmental Panel on Climate Change, implying a weaker cooling effect. The result reveals that the global LUC-induced surface albedo change may not significantly slow down global warming as was previously anticipated. Sub-Saharan Africa made the largest net contribution to the magnitude of global ARF (39.2 %), due to substantial land use conversions, typically the conversion from forest to other vegetation lands, which accompany with higher surface albedos. The most remarkable land cover changes occurred in East and Southeast Asia, which dominated the changes in global ARF in recent decades. Based on major land cover types in these two regions, we infer that vegetation lands exert a most vital effect on global ARF variation.
- Preprint
(2812 KB) - Metadata XML
-
Supplement
(4868 KB) - BibTeX
- EndNote
Status: open (until 17 Dec 2024)
-
RC1: 'Comment on egusphere-2024-1497', Anonymous Referee #2, 23 Nov 2024
reply
The manuscript egusphere-2024-1497 introduces a satellite-derived historical land cover product to a climate model, recalculates the radiative forcing (RF) of land use change (LUC) from 1983 to 2010, and demonstrates that satellite-derived results show weaker LUC RF compared to the original model's coarse-resolution LUC input. This study is well designed, and the results sufficiently support the conclusions. I have some questions regarding the interpretation of the results, and I believe that addressing these concerns will strengthen the manuscript and facilitate its publication in ACP.
Major Comments
- Inter-Annual Variability of Satellite Land Cover Product
My main concern is the excessive annual variability in the satellite-derived land cover product, particularly when it is claimed to represent land use change. Typically, land use change reflects human activities. However, Figure 1 shows significant fluctuations in the global average of LUC-derived RF between the late 1980s and early 1990s, with increases and decreases that nearly double the overall magnitude observed since the industrial era. Similar abrupt changes are noted in South Asia and Russia in the late 1990s (Figure 2). These fluctuations seem unrealistic and undermine the reliability of the input satellite data. I recommend exploring additional satellite datasets, if available, and comparing the results for inter-validation. - Land Cover and Land Use Classification
How do the authors reconcile the differences between the satellite-derived land cover classifications and the land use classifications in the original model input (LUH1)? Land cover and land use are distinct concepts, and their categories differ. For example, LUH1 includes "pasture" as a category, while GLASS-GLC uses "grassland," which are not equivalent. Clarification on the mapping or harmonization process is needed. - Sensitivity Analysis Methodology
The sensitivity analysis is a critical foundation for this study. Is the method employed here commonly used for quantifying LUC radiative forcing? If not, how does it compare with approaches used in previous studies? Providing context and justification for this methodology is essential.
ÂOther Comments
- Abstract: Clarify the apparent contradiction between "Sub-Saharan Africa made the largest net contribution" and "East and Southeast Asia dominated the changes in global ARF."
- Lines 55–57: Elaborate on the distinction between the well-investigated "LUC on climate balance" and the research gaps in "LUC-induced climate forcing."
- Lines 74–75: Suggest investigating multiple satellite products, rather than relying on a single dataset.
- Section 2.1: Provide an introduction to how OSCAR converts land use types into albedo values and their subsequent effects on radiative forcing and climate.
- Line 188: Explain the rationale for using a 20% threshold in the analysis.
- Figure 1(b): Indicate the time periods covered by other studies for better comparability.
- Figure 2: Justify the chosen regional separations and clarify whether latitude weighting was applied to calculate the regional means.
- Lines 347–349: This statement is unnecessary and could be removed to streamline the manuscript.
- Supplementary Table S2: Explain why the values for "Rest of East Asia" are notably larger than those for other regions.
Citation: https://doi.org/10.5194/egusphere-2024-1497-RC1 - Inter-Annual Variability of Satellite Land Cover Product
-
RC2: 'Comment on egusphere-2024-1497', Anonymous Referee #3, 27 Nov 2024
reply
This manuscript, which estimates albedo-induced radiative forcing (ARF) using satellite-derived land-use change (LUC) data at fine spatial resolution, has the potential to significantly impact our understanding of LUC effects. The authors' report of a lower ARF estimation using fine spatial resolution data than published values, suggesting a weaker cooling effect of LUC, is a promising finding. The manuscript is well written and interesting, the motivation to the work is strong, the methodology is well described, and the figures are engaging and effective.
Â
I have two main concerns:
- As the authors have mentioned, LUH2 is more recent and at a finer spatial resolution than LUH1. Despite this, why was LUH2 data not used instead of LUH1 for comparison to GLASS-GLC?
- Temporal variation in LUC appears large for all regions, resulting in large fluctuations in ARF (Figure 1 and 2). Such large variations in LUC should be justified or studies reporting similar fluctuations should be cited.
Â
Additional comment:
The authors have provided websites for downloading the GLASS-GLC data and OSCAR code but have not shared a repository to access the outputs of OSCAR model generated and analyzed in this study. I would encourage them to share a link to their model simulations.
Citation: https://doi.org/10.5194/egusphere-2024-1497-RC2 -
RC3: 'Comment on egusphere-2024-1497', Anonymous Referee #1, 28 Nov 2024
reply
Â
Land use change has been demonstrated to show large impacts on regional or even global climate change. This study quantified the LUC-induced albedo change and its radiative forcing based on high-resolution remote sensing-derived LUC dataset. Thank the authors carefully resolved my comments in the previous round of review. Please see below for my further comments.
Major concerns:
- Line 102-103: The authors stated that they assigned a 5% uncertainty in OSCAR modeled ARF based on LUC data uncertainty. However, what is the LUC data uncertainty? How did the authors derive this 5% threshold? Please provide more details.
- Line 152: The authors mentioned that they conducted extensive sensitivity experiments by reducing each LU transition area by 20% within five major LU types. However, why did the authors select this 20% threshold? Please clearly clarify it.
- Line 142: The authors mentioned that they neglected the LUC-induced surface roughness change. Please discuss the potential uncertainty from this.
- The latest LUH2 dataset is available. There are some improvements in LUH2 compared to LUH1. Please use the latest version of LUH2 rather than the out-of-date LUH1.
- Although the GLASS-LUC has a higher spatial resolution, the authors upscaled them to national and regional levels. Please clearly clarify this point.
- However, GLASS-LUC also include uncertainties, and is not necessarily more accurate than LUH2 data. I suggest the authors include more remote sensing datasets e.g., MODIS data to increase the robustness of the results.
- The authors mentioned that OSCAR does not estimate surface albedos itself. Instead, it collected surface albedos in different countries and regions from literature and other climate models. However, as I know, surface albedo shows large spatial variation even for the same land types. Please provide a direct evaluation of the OSCAR surface albedo using the available remote sensing data, e.g., MODIS. Without such evaluation, the results from this study can be unreliable.
Â
Minor concerns:
- L136-137: Please provide the citation.
- Data availability: Please also share the model output in the study.
Â
Below are my comments from the previous round of review.
Land use change has a large impact on global climate change. This study quantified the LUC-induced albedo change and its radiative forcing based on remote sensing-derived LUC dataset. The study is interesting and the results and conclusions are meaningful. However, some issues are needed to be carefully revised: 1) The authors set a lot of thresholds when calculating RFs and carrying out the sensitivity analysis, without accounting the corresponding reasons. 2) The urban change is not accounted for, which can induce large uncertainty. Please see below for my specific comments.
Major concerns:
- Line 15-18: Land use change has complex impacts on climate change. Whether it is cooling or warming effect depends on the specific conversions from one land use to another land use. Land use change can emit GHG, change surface albedo and ET, and further affect climate. However, which factor dominates depends on the specific conditions.
- L103: The authors assigned a 5% uncertainty in modeled ARF induced by LUC uncertainty. However, why the authors set this value is unclear. How did the authors use this in the model?
- L131: Some studies (e.g., Ouyang et al., 2022) have shown that urbanization has an albedo-induced warming effect. However, this study neglected the urban change, which may induce large uncertainties.
Ouyang, Z., Sciusco, P., Jiao, T. et al. Albedo changes caused by future urbanization contribute to global warming. Nat Commun 13, 3800 (2022). https://doi.org/10.1038/s41467-022-31558-z
- L139: In the sensitivity experiments, why did the authors set this threshold of 20%?
- GLASS LC data cover 1982-2015. Why did the authors just analyze the data from 1983-2010?
- Section 2.1: Please clarify how OSCAR model uses the land use data, considering it is not spatially resolved.
- 3: why did the authors select 1% as the threshold?
- Figure 1 & 2: Please explain why the simulations in S1 and S2 show very different trends in the global average and regional values. Please add the corresponding LUC analysis and clearly explain it in the main text.
- Line 157: Considering that there is a big difference between LUH1 and GLASS, replacing LUH1 with GLASS in 1982 can induce some uncertainties. Please discuss it. How did the authors harmonize these two LUC datasets?
- The work neglected the impacts of LUC on surface roughness, which deserves some discussion.
- In the methods section, the authors mainly introduced the sensitivity analysis. Also need to introduce how to use two LUC datasets for the analysis of albedo-induced RFs. Please also introduce the objective of the sensitivity analysis. In the sensitivity analysis, the authors define multiple new variables. However, some of them are not easy to understand. Please make them easier to follow.
- There is a large spatial variation of surface albedo. Surface albedo is dependent on the vegetation structure, leaf/soil albedo and surface topography. I am curious how OSCAR considers the spatial variation of surface albedo.
Minor concerns:
- L67: Please provide the citation.
- L163: to2010 -> to 2010.
- L267: This equation can be moved to methods section.
- Figure 3: Effective area and RF have different units. Why did the authors put them together?
Â
Citation: https://doi.org/10.5194/egusphere-2024-1497-RC3
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
110 | 43 | 8 | 161 | 47 | 3 | 1 |
- HTML: 110
- PDF: 43
- XML: 8
- Total: 161
- Supplement: 47
- BibTeX: 3
- EndNote: 1
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1