the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 30 Jun 2026
Temperature stabilization using Stratospheric Aerosol Injections minimises drying over Africa only when combined with a strong decarbonization effort
Abstract. Climate change poses severe risks to African agriculture, water resources, and ecosystems. Temperature overshoot scenarios, in which global warming temporarily exceeds target thresholds such as 1.5 or 2.0 °C before declining through mitigation and carbon removal later in the century, are plausible future trajectories. Yet, their regional impacts and the reversibility of changes during the overshoot remain poorly characterized. Stratospheric aerosol injection (SAI) has been proposed as a means to limit peak warming during overshoot; however, its effects on African climate extremes and water availability require careful assessment. This study analyses different CESM2-WACCM6 simulations to evaluate changes in temperature extremes, precipitation patterns, and surface moisture budget across Africa, using two baseline scenarios, the high GHG forcing scenario (SSP5-8.5) and the SSP5-3.4-OS overshoot scenario, which includes strong decarbonization and carbon removal efforts after 2040. In addition, three SAI intervention scenarios are assessed, targeting 1.5 and 2.0 °C (for the overshoot scenario, only) above pre-industrial levels. We compute selected ETCCDI-based climate indices, including Growing Degree Days, Warm Spell Duration Index, Consecutive Dry Days, and precipitation intensity metrics for baseline and overshoot (2060–2079) periods. Our results reveal near-universal, statistically significant changes (> 90 %) in temperature indices during overshoot, with 5–30 % increases depending on the metric. Precipitation indices exhibit more heterogeneous responses, with 40–80 % of the area showing significant changes. SAI interventions consistently reduce temperature-related indices across Africa, with the strongest cooling effects in tropical regions. However, precipitation responses to SAI display substantial spatial heterogeneity and scenario dependency: West Africa’s Sahel shows increased moisture availability under high SAI compared to SSP5-8.5, Central Africa exhibits mixed responses with regional drying in parts of the Congo Basin, and East Africa demonstrates a dipole pattern of coastal wetting and interior drying that intensifies at higher warming thresholds. All these changes are magnified under high-cooling scenarios (using the high forcing baseline) compared with cooling under overshoot, in which case many precipitation differences are reduced.
- Preprint
(2052 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-3436', Anonymous Referee #1, 15 Jul 2026
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-3436/egusphere-2026-3436-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2026-3436-RC1 -
AC1: 'Reply on RC1', Temitope Samuel Egbebiyi, 21 Oct 2026
Major comments Major comment 1: Clarify the motivation and relevance of overshoot and SAI scenarios for Africa
Reviewer comment: The manuscript addresses an important topic; however, the motivation connecting climate overshoot scenarios to SAI and their relevance for Africa needs to be strengthened.
- Scientific justification of scenarios: The authors should better explain why studying high-end overshoot and SAI scenarios is valuable despite their lower likelihood under current mitigation pathways. These scenarios can serve as stress tests to understand potential climate risks, reversibility, and adaptation challenges under extreme forcing conditions.
- Positioning within existing literature: Statements implying that African climate impacts are largely unexplored should be revised. While regional research gaps remain, there is an expanding body of work investigating African climate change impacts. The authors should acknowledge these efforts and clarify what specific knowledge gap this study addresses.
Response: We agree that the original framing understated existing African climate research and did not connect overshoot clearly enough to SAI. The Introduction now identifies the specific comparison: African responses to a common 1.5°C SAI objective on two different background pathways, with an additional 2.0°C overshoot experiment. It acknowledges prior Africa-focused studies, including Pinto et al. (2020) and Egbebiyi et al. (2025). We describe these scenarios as idealised stress tests rather than likely forecasts. They allow assessment of regional climate trajectories through a warming peak and of differences between temperature and hydrological responses. They do not establish reversibility of ecosystem impacts or isolate injection amount from all background-state effects.
Major comment 2: Improve the regional framework for interpreting impacts
Reviewer comment: The manuscript evaluates impacts using geopolitical regions; however, the study focuses on agriculture, ecosystems, and water resources, which may be better represented through climate and ecological zones.
- Use of agro-ecological regions: The authors should consider whether an agro-ecological classification (e.g., Sahara, Sahel, savanna, tropical forest regions) would provide a more physically meaningful interpretation than political boundaries.
- Link to impacts: Such a framework would allow more targeted discussions of SAI impacts on vulnerable systems, since climate responses are strongly controlled by factors such as aridity, vegetation, soil properties, and hydrological regimes.
Response: We appreciate this suggestion. We did not change the regional classification or repeat the analysis using agro-ecological zones. The revised manuscript retains the five broad African subregions, North, West, Central, East, and Southern Africa, to remain consistent with the regional masks used throughout the analysis and figures. We agree that these broad regions do not resolve the climatic and ecological variation within them, and we have made that limitation explicit. Nevertheless, it is important to mention that the regional classification gives an important overview of the major climate classification of the African continent and has also been used in different studies (e.g. Oloniyo et al., 2026) over the region
Major comment 3: Clarify interpretation of the surface moisture budget and water availability
Reviewer comment: The manuscript appears to equate changes in (P-E) with changes in surface water availability, which may overextend the conclusions.
- Distinguish moisture budget from water availability: The surface moisture budget provides information about precipitation inputs relative to evaporative losses but does not directly quantify available water resources.
- Consider missing hydrological processes: Surface water availability also depends on runoff, infiltration, groundwater recharge, soil moisture storage, drainage, and human water use. Therefore, conclusions regarding preservation of water availability should be qualified.
- Clarify the role of runoff: Neglecting runoff may be acceptable for a climatological atmospheric moisture analysis; however, the authors should acknowledge that a complete terrestrial water balance involves terms such as (P-E-R-D).
Also, the moisture budget analysis requires additional methodological clarification.
- Resolve inconsistencies in reported values: The manuscript reports moisture budget values of approximately 1-1.5 mm/day, while seasonal decomposition shows peaks of 4-4.5 mm/day. The authors should clarify whether these represent different quantities, averaging periods, or units.
- Define seasonal decomposition: The authors should explain whether values represent daily means, monthly averages, seasonal totals, or outputs from a decomposition method.
- Provide quantitative trends: Statements describing increasing or decreasing trends should be supported with numerical trend values, particularly where differences between SSP and SAI scenarios are difficult to discern.
Response: We agree. Throughout the revision, we define the surface moisture budget as precipitation minus actual evapotranspiration (P - ET). It does not directly quantify accessible water in soils, rivers, reservoirs or aquifers. A full terrestrial balance requires storage changes, runoff, subsurface drainage and relevant management fluxes, with terms defined to avoid double counting. We removed the original claim that SAI preserves surface water availability. ET is derived from actual latent heat flux, not an estimate of evaporative demand.
The seasonal analysis uses calendar-day-weighted mean daily rates in mm day⁻¹. It is a climatological seasonal average, not a decomposition algorithm. An individual wet-season mean can exceed the annual mean because lower or negative values in other seasons reduce the annual average. For the no leap calendar, annual = (90 DJF + 92 MAM + 92 JJA + 91 SON) / 365. This identity was checked for all 180 member-region-scenario-component rows; the largest numerical residual is 1.78 × 10⁻¹⁵ mm day⁻¹. Table S1 provides the regional surface moisture budget means, and Figure 6 states the units and averaging convention.
We also replace qualitative trend language with Theil-Sen slopes over the complete 2040-2099 window (Table S2). In West Africa, mean slopes are −0.0617, −0.0535, −0.0309, +0.0004 and −0.0305 mm day⁻¹ per decade for SSP5-8.5, SSP5-3.4-OS, SSP585_1.5 SAI, SSP534OS_1.5 SAI and SSP534OS_2.0 SAI. We report member ranges as descriptive ensemble spread, not confidence intervals or trend-significance tests. Missing 2029 observations prevent complete three-member full-century summaries for the two 1.5°C SAI histories; the complete later window avoids that omission.
Major comment 4: Better highlight the novelty and strengths of the study
Reviewer comment: The manuscript provides a detailed discussion of limitations, but the methodological strengths and unique contributions are less clear.
- Highlight contributions: The authors should emphasize what aspects of the modelling framework, scenario design, or analysis provide new insights compared with previous studies.
- Balance limitations with strengths: A stronger discussion of the study’s methodological advantages would improve the positioning of the manuscript and ensure that the contribution is recognized beyond the limitations discussed.
Response: The Discussion now explains the contribution after the limitations paragraph. The framework compares two forcing backgrounds at a common target, separates reference-period changes from same-period SAI effects, and calculates nonlinear indices for each member before averaging. Combining precipitation with actual evapotranspiration reveals compensating contributions that precipitation metrics alone would miss. Consistent regional weighting and explicit seasonal reconciliation strengthen internal interpretation. We avoid claiming that the design uniquely isolates aerosol amounts, that a small ensemble proves robustness, or that the analysis establishes new crop or water-resource outcomes. See Discussion section, pages 29 - 30, lines 343 - 413
Specific comments Lines 51 to 65
Reviewer comment: The authors comment on climate overshoot; however, how are they tying it in to SAI and its impacts? The paragraph seems somewhat disjointed from the next paragraph.
Response: Despite the policy relevance of overshoot scenarios, their regional climate impact, particularly over Africa, is not yet fully understood. Comparing these scenarios allows continued warming to be contrasted with a warming peak followed by cooling. The later decline does not remove the period of elevated warming, during which changes in climate extremes may affect agriculture and water resources. SAI has therefore been proposed as a means of limiting peak warming during an overshoot. Its implications for Africa must be assessed from both the temperature response and the accompanying changes in precipitation and the surface moisture budget. See page 1, lines 61-65
Line 63
Reviewer comment: Despite the policy …” I think this is not an accurate description of the current state of the science. There are quite a number of researchers working to understand this topic over the region. The expertise may not be from the region, but there are definitely quite a number of research studies being conducted over the region. This should be acknowledged.
Response: The statement has been revised to: “ Despite the policy relevance of overshoot scenarios, their regional climate impact, particularly over Africa, is not yet fully well understood”. See Page 1, lines 61 - 62.
Line 80
Reviewer comment: Change to Pinto et al. and Egbebiyi et al. demonstrated that …
Response: The sentence has been revised as “Pinto et al. (2020) and Egbebiyi et al. (2025) demonstrated that while SRM significantly reduces mean and extreme temperatures across the continent, its effect on precipitation is less linear and varies markedly across different parts of Africa, particularly between East and West Africa.” See Pg. 3, lines 79 - 81
Line 88
Reviewer comment: “These studies collectively underscore the need for targeted…” - more targeted assessments. There are some studies, but more would be better.
Response: “targeted assessments” has been changed to “more targeted assessments”. The sentence has now been revised as “These studies collectively underscore the need for more targeted assessments of how SAI interacts with specific emission pathways and warming trajectories, particularly under overshoot conditions that represent increasingly plausible future scenarios. “. See Pg. 3, lines 88
Lines 93 to 95
Reviewer comment: The authors should further clarify the relevance and applicability of this work for the African region, particularly given that the scenarios considered may represent relatively unlikely future pathways under current mitigation trajectories. The manuscript would benefit from a stronger discussion of the rationale for examining these high-end overshoot and SAI scenarios. Specifically, the authors should emphasize that, despite their lower likelihood, such scenarios provide valuable insights into the potential magnitude, spatial variability, and reversibility of climate impacts under extreme forcing pathways, serving as important stress tests for climate resilience, adaptation planning, and risk assessment.
Response: The scenarios are framed as conditional stress tests for temperature changes and regional hydrological responses, rather than forecasts. Their usefulness lies in comparing demanding forcing trajectories and intervention objectives relevant to identifying questions for adaptation and impact assessment. We avoid asserting that they bound every possible African impact.
Line 112
Reviewer comment: The authors do a good job of localizing the analysis by considering geopolitical regions. However, given that the study focuses on climate impacts relevant to agriculture, water resources, and ecosystems, it may be more appropriate to consider a subdivision based on agro-ecological zones rather than, or in addition to, geopolitical boundaries. An agro-ecological framework would likely provide a more physically meaningful basis for interpreting regional responses, as climate impacts and vulnerabilities are strongly tied to environmental characteristics such as aridity, vegetation type, and hydrological regimes. This approach would also allow the authors to make more targeted statements regarding the implications of SAI, for example, distinguishing responses in the Sahara, Sahel, savanna, and humid tropical regions rather than broader regions such as North Africa. This could improve the clarity and relevance of the discussion for climate adaptation and resource management.
Response: We appreciate the suggestion. We did not change the regional classification or repeat the analysis using agro-ecological zones. The five-region framework was retained for consistency with the existing masks, analysis, and figures. We have instead acknowledged that the broad regions do not resolve within-region agro-ecological differences.
Lines 153 and 155
Reviewer comment: L153: b should not be capitalised. L155: b should not be capitalised.
Response: We changed the bold “b” to lowercase and unbolded it. See lines 153 and 155
Line 165
Reviewer comment: ETCCDI is already defined elsewhere before this.
Response: ETCCDI is defined at the first occurrence in the Introduction and abbreviated subsequently (See Page 2, lines 50). The methodological citation is Zhang et al. (2011), and we distinguish the additional temperature indices from standard ETCCDI indices. See Page 7, line 169
Line 214
Reviewer comment: Correct “indices plateau begin”.
Response: We revised the sentence to state that, under SSP5-3.4-OS, the temperature indices level off and subsequently decline. We also removed the statement implying that WSDI extends through 2099. The revised Figure 1 and its caption show that the WSDI series end in 2091 under SSP5-8.5 and 2087 under SSP5-3.4-OS because of incomplete daily maximum-temperature records. See Page 9; lines 198-201
Lines 284 to 285
Reviewer comment: The statement that SAI shifts tropical rainfall patterns through stratospheric heating and suppresses evaporation requires further clarification and stronger support from the cited literature. The cited study (e.g., [DOI: 10.1029/2019GL085758]) appears to primarily discuss circulation and seasonal changes at higher latitudes rather than directly demonstrating tropical rainfall shifts or evaporation suppression. Therefore, this statement may overstate the current evidence. I suggest revising this discussion to provide a more general overview of the potential impacts of SAI rather than attributing the specific changes observed in this study to these mechanisms without further analysis. At present, the extent to which SAI modifies tropical rainfall distributions or suppresses evaporation remains uncertain and may depend on the background climate state, regional processes, and also the model in use. Additionally, the phrase “shift tropical rainfall patterns” is ambiguous. Please clarify whether this refers to changes in the spatial location of precipitation, changes in seasonal timing (e.g., onset/cessation), or changes in rainfall intensity and variability.
Response: We agree and have revised the Discussion to avoid directly attributing the simulated rainfall changes to stratospheric heating or reduced evaporation. The revised text presents changes in the energy balance and atmospheric circulation only as possible explanations reported in previous studies and makes clear that these mechanisms were not isolated in our analysis. We also removed the ambiguous phrase “shift tropical rainfall patterns” and now describe the simulated changes directly in terms of precipitation magnitude and spatial pattern.
Line 286
Reviewer comment: Niemeier and Simpson citations are absent from the references.
Response: Complete references for Niemeier et al. (2013) and Simpson et al. (2019) have been added to the reference list.
Lines 295 to 299
Reviewer comment: What would this mean for precipitation? Does it mean SAI could lead to higher rainfall/extreme rainfall days compared to the other climate scenarios? The sentence implies that SAI would preserve surface water availability through the century. Is this good or bad? Also, the final sentence appears to extend beyond what is directly supported by the analysis. While the surface moisture budget provides valuable information on the balance between atmospheric water inputs and losses, it does not directly quantify surface water availability, which is additionally controlled by processes such as runoff, infiltration, groundwater recharge, soil moisture storage, and human water use. Therefore, the conclusion that SAI could preserve surface water availability appears stronger than the evidence presented. The authors should either qualify this statement or provide additional analyses or supporting literature that establish a clear link between changes in P-E and surface water availability, particularly in West Africa. Given that this is a climatological analysis, neglecting runoff (R) may be reasonable for evaluating broad atmospheric moisture balance; however, it should be acknowledged that actual surface water availability depends on the full hydrological balance, including terms such as runoff and drainage (P-E-R-D).
Response: No direct inference about extremes or available water follows from a higher surface moisture budget value. We remove the preservation claim and discuss precipitation indices separately. For example, same-period continental Rx5day decreases by 8.7, 5.5 and 2.6 mm in the three SAI experiments, while the regional surface moisture budget can increase. In West Africa, the positive mean surface moisture budget differences are primarily associated with increased P: ΔP is +0.105, +0.257 and +0.308 mm day⁻¹, while ΔET is -0.005, +0.075 and +0.085 mm day⁻¹. These contrasts do not establish whether particular local rainfall extremes increase, or whether the balance change is beneficial for a given crop or catchment.
Lines 301 to 303
Reviewer comment: Can the trend values be quantified? It is hard to see the “increasing trend” for some of the scenarios like SSP3.4-OS and the SAI scenarios.
Response: Table S2 gives the mean and range of member-specific Theil-Sen slopes for all four regions over 2040-2099. For the nearly flat West African overshoot SAI 1.5°C case, the mean is +0.0004 mm day⁻¹ per decade and the member range is −0.0140 to +0.0188. We no longer describe every West African scenario as declining, or infer significance from the apparent sign of a smoothed curve.
Line 304
Reviewer comment: Are these monthly averages, monthly totals, seasonal totals or daily averages? The seasonal decomposition results appear to be inconsistent with the values presented earlier in the manuscript. Specifically, the moisture budget is reported to range between approximately 1.0 and 1.5 mm day⁻¹, yet the JJA seasonal values are stated to peak at approximately 4.0-4.5 mm day⁻¹. If these represent seasonal means of the same variable, such values would not be mathematically possible, suggesting either a difference in the quantity being analyzed (e.g., seasonal accumulation versus seasonal mean), a change in units, or a different processing methodology. Please clarify how the seasonal decomposition was performed, what quantity is being presented, and whether the reported units are correct. If these values are derived from a decomposition algorithm rather than a simple seasonal average, this should also be explicitly stated to avoid confusion.
Response: Both figures present the surface moisture budget in mm day⁻¹. Figure 4 shows a smoothed annual time series, whereas Figure 5 shows seasonal means for 2060-2079. An individual seasonal mean can therefore be higher or lower than the annual mean. We verified the annual values using the day-weighted mean of the four seasons over the same period and clarified this calculation in the Methods.
Figure 6 units
Reviewer comment: Check the units.
Response: We checked and corrected the vertical axis. The seasonal surface moisture budget and SAI-effect figures use mm day⁻¹, while Figure A3 reports seasonal precipitation totals in mm season⁻¹.
Lines 331 to 332
Reviewer comment: Beneficiary relative to what?
Response: We removed “greatest beneficiary” and now compare each SAI experiment with its corresponding simulation without SAI over the same period. The revised text reports the changes in the surface moisture budget without describing them as benefits.
Lines 371 to 373
Reviewer comment: citation needed “From an agricultural perspective”.
Response: We added Sultan and Gaetani (2016) to support the discussion of agricultural sensitivity to changes in temperature and water availability. We also clarified that the indices analysed here do not directly estimate crop yields, as other factors such as soil type and farm management practices are also major contributors.
Lines 393 to 395
Reviewer comment: The higher moisture levels… This work does not show this result, hence the authors would need to back up this claim.
Response: We removed the unsupported claim and now report only the simulated changes in the surface moisture budget and its precipitation and evapotranspiration components. We also clarify that the surface moisture budget does not directly measure water availability.
Lines 401 to 409
Reviewer comment: This paragraph appropriately discusses the methodological limitations of the study, which provides useful context for interpreting the results. However, it would also be valuable for the authors to highlight the methodological strengths and innovations that underpin the contribution of this work. In particular, the manuscript would benefit from a discussion of what aspects of the experimental design, modeling framework, or analytical approach provide new insights beyond existing studies. Placing such a discussion after the limitations paragraph would help balance the presentation by emphasising not only what the study cannot address, but also the robustness of the framework and the unique contributions it makes to understanding climate responses under overshoot and SAI scenarios. This would strengthen the positioning of the manuscript and better convey its novelty and significance.
Response: We added a paragraph highlighting the comparison of SAI experiments across two background scenarios, the calculation of indices for individual ensemble members and the combined analysis of precipitation, evapotranspiration and the surface moisture budget.
References
Reviewer comment: Alphabetise the reference list.
Response: The reference list has been arranged alphabetically.
Figures 2, 5, A1 and A3
Reviewer comment: The font size of legends and labels is too small. Also, the figures pixelate when zoomed in for details.
Response: We have regenerated the figures at a final width of 7.15 inches and 450 dpi, with fonts sized for readability at the final page width.
-
AC1: 'Reply on RC1', Temitope Samuel Egbebiyi, 21 Oct 2026
-
RC2: 'Comment on egusphere-2026-3436', Anonymous Referee #2, 08 Sep 2026
COMMENTS TO THE AUTHORS
Review of the manuscript entitled: “Temperature stabilization using Stratospheric Aerosol Injections minimises drying over Africa only when combined with a strong decarbonization effort” by Egbebiyi et al.
This study investigates the potential impacts of stratospheric aerosol injection (SAI) on temperature and precipitation-related agroclimatic indices as well as surface moisture availability over Africa. The authors use CESM2-WACCM6 simulations under SSP5-8.5 and SSP5-3.4-OS, together with SAI experiments targeting 1.5°C and 2.0°C warming levels. The study shows that temperature-related indices exhibit widespread and statistically significant changes under warming and that SAI substantially reduces these changes. In contrast, precipitation and surface moisture responses are more heterogeneous and strongly dependent on the region, scenario and SAI target. The study further highlights that the hydrological effects of SAI are generally smaller under the overshoot scenario associated with strong mitigation and carbon dioxide removal (CDR).
The topic is timely and highly relevant, particularly given the growing interest in understanding the regional implications of solar radiation modification over Africa. The combination of agroclimatic indices and surface moisture budget provides an interesting perspective that goes beyond temperature changes alone. The manuscript is generally well structured and presents potentially useful results for understanding the risks and benefits associated with SAI over the African continent.
However, several aspects of the manuscript require clarification and improvement before publication.
Below are some comments that may be beneficial for consideration:
General comments
1- The distinction between the SSP5-8.5 and SSP5-3.4-OS baselines and the different SAI scenarios is central to the manuscript. However, the comparison between the two pathways could be developed more systematically throughout the Results, Discussion and Conclusion sections. In particular, the manuscript should more explicitly highlight how and why the response to SAI differs between a sustained high-emission pathway and an overshoot pathway associated with strong mitigation and CDR.
2- The manuscript relies on a single Earth system model (CESM2-WACCM6). Although this limitation is acknowledged in the Discussion, I suggest that its implications be emphasized more clearly when interpreting the regional precipitation and moisture-budget responses. Such responses can be strongly model-dependent, and the conclusions should therefore be framed accordingly.
3- The statistical analysis deserves further clarification. The authors use a Welch's t-test with only three ensemble members for each scenario and consider p < 0.05 as statistically significant. Given the very small number of ensemble members, the authors should provide additional justification for the robustness and appropriateness of this statistical approach and discuss its implications for the reported significance levels.
4- The manuscript would benefit from a clearer distinction between statistical significance and magnitude of change. In several places, the reduction in the percentage of statistically significant grid cells under SAI is interpreted as evidence that SAI “effectively returns temperature conditions closer to the baseline state.” Statistical significance alone does not necessarily quantify the magnitude or practical importance of the change. Consider reporting and discussing the actual magnitude of changes alongside their statistical significance.
5- The title emphasizes that SAI minimizes drying only when combined with strong decarbonization. This is an interesting and potentially important conclusion, but the manuscript should make this central message more explicit and quantitatively demonstrate it throughout the Results and Discussion sections.
6- The authors should say a few more words about the uncertainties and limitations associated with the SAI experimental design itself, including the use of idealized feedback-controlled injection scenarios and prescribed global temperature targets. These simulations should not necessarily be interpreted as representing a realistic future deployment strategy.
Specific comments
Line 1: In the title, please consider whether “minimises drying” is fully supported by the results across the whole African continent. The results show important regional contrasts, including reduced moisture availability under some SAI scenarios in East and Southern Africa. The title may therefore need to better reflect the regional and scenario-dependent nature of the response.
Lines 24–32: The abstract is generally informative, but the main quantitative findings could be presented more consistently. For example, the abstract reports that precipitation indices exhibit significant changes over 40–80% of the area, whereas the Results section gives substantially lower percentages for several precipitation indices. Please check that all percentages reported in the abstract are consistent with the results presented in the main text and figures.
Line 41: All acronyms must be clearly defined from the first occurrence they are used. Therefore, please apply this instruction to "CMIP6" and all other acronyms used throughout the text
Lines 65-66: Please clarify the terminology used throughout the manuscript between “stratospheric aerosol injection (SAI)”, “stratospheric aerosol geoengineering (SAG)”, and “solar radiation modification (SRM)”. The manuscript uses these expressions somewhat interchangeably. A consistent terminology would improve readability.
Lines 111–114: There appears to be an inconsistency between the description of the regionalization in the text and the caption of Figure 1. The text mentions five sub-regions (North, West, Central, East and Southern Africa), whereas the Figure 1 caption additionally mentions Northeast Africa (NEAF). Please clarify the regional classification and apply the same terminology consistently throughout the manuscript.
Line 138: Please ensure consistency in the notation of the overshoot scenario. The manuscript uses SSP5-3.4-OS and SSP5-34-OS in different places. The scenario name should be standardized throughout the text, tables and figures.
Lines 138–157: The experimental design should be described more clearly.
Lines 165: Please clarify the citation and terminology associated with the ETCCDI framework. The text currently states “the Expert Team on Climate Change Detection and Indices (ETCCDI) Peterson (2002)”. Please check the corresponding reference and citation year, and revise the sentence.
Lines 199–208: The statistical methodology needs further explanation. It would be useful to specify exactly what constitutes the samples in the Welch's t-test: are the tests performed using annual values, ensemble-member means, or period means? Given that only three ensemble members are available, this point is particularly important for assessing the robustness of the significance analysis.
Line 214: Please revise the sentence “these indices plateau begin to decline” for grammatical clarity.
Lines 284–287: The physical mechanisms invoked to explain the precipitation response to SAI are interesting. However, the discussion could be strengthened by distinguishing more clearly between the mechanisms supported directly by the simulations and mechanisms inferred from previous studies.
Lines 295–305: The interpretation of the surface moisture budget deserves some caution. In particular, the statement that SAI “could help preserve surface water availability” in West Africa should be framed carefully because the analysis is based on P − ET, which represents a simplified surface moisture balance and does not directly quantify water availability in rivers, soils, reservoirs or groundwater.
Lines 314–329: This paragraph appears to substantially repeat the preceding description of Figure 5. Please consider removing the duplicated material and using this space to provide additional interpretation of the seasonal differences.
Lines 331–332: Please clarify the apparent contrast between the statement that East Africa is the greatest beneficiary, with an increase of approximately 0.5 mm/day, and the preceding results indicating that SAI produces lower moisture levels than the high-warming baseline in East Africa. The reference scenario used for each comparison should be explicitly stated.
Lines 396–398: The conclusion that excessive drying associated with SAI can be mitigated under a lower-emission scenario is one of the most important findings of the study. I suggest developing this point further and providing a more quantitative comparison of the hydrological response between the two emission pathways.
Lines 401–409: The limitations section is useful, but it could be expanded to include the idealized nature of the SAI experiments, the relatively coarse spatial resolution of the model for regional hydrological analysis, and the uncertainty associated with using a single model.
Tables
Table 1: The notation used for the SAI experiments should be standardized. In particular, please make clear whether “Geo SSP5-34-OS 1.5” and “Geo SSP5-34-OS 2.0” refer to temperature targets or experiments. It would also be useful to explicitly indicate which simulations are used as baselines and which are used as SAI experiments.
Table 2: Please carefully check the definitions and units of all indices. In particular, the units associated with R1MM, R20MM and PRCPTOT appear potentially inconsistent with their definitions. Please verify the table against the standard ETCCDI definitions.
Figures
Figure 2: Please clarify why only selected indices are presented in Figure 2 and explain whether the omitted indices exhibit similar temporal behaviour. This would help readers understand whether the selected indices are representative of the overall response.
Figure 3: The figure caption should clearly identify all scenarios represented in the figure. This is particularly important because the manuscript considers both 1.5°C and 2.0°C SAI targets, whereas the caption appears to focus only on the 1.5°C experiments. Please verify that the figure, caption and corresponding discussion are fully consistent.
Figure 4: The caption states that four scenarios are represented, but the manuscript considers several SAI experiments, including a 2.0°C target. Please clarify which scenarios are included in this figure and why some experiments may have been excluded.
References
Please carefully check the References section for consistency, completeness and formatting. Several references appear to contain incomplete bibliographic information or inconsistencies in formatting.
In particular:
The reference to Kravitz et al. (2017) appears incomplete and should be checked carefully.
The reference to Mupangwa et al. (2016) should be checked against the statement for which it is cited in the Introduction section, particularly the claim concerning the contribution of agriculture to the Gross Domestic Product.
The reference to Ofori & Asongu (2021) appears to concern foreign direct investment, governance and inclusive growth rather than the specific agricultural statement for which it is cited. Please verify whether this is the appropriate reference.
Please check the consistency between the citation of Emmons et al. in the text and the year given in the reference list.
Similarly, please verify the citation year and bibliographic details of Danabasoglu et al. in the text and reference list.
Citation: https://doi.org/10.5194/egusphere-2026-3436-RC2 -
AC2: 'Reply on RC2', Temitope Samuel Egbebiyi, 21 Oct 2026
Response to Reviewer 2
We thank the reviewer for the constructive comments. Line numbers refer to the submitted manuscript, while section and figure numbers refer to the revised manuscript.
General comments General comment 1:
Reviewer comment: The distinction between the SSP5-8.5 and SSP5-3.4-OS baselines and the different SAI scenarios is central to the manuscript. However, the comparison between the two pathways could be developed more systematically throughout the Results, Discussion and Conclusion sections. In particular, the manuscript should more explicitly highlight how and why the response to SAI differs between a sustained high-emission pathway and an overshoot pathway associated with strong mitigation and CDR.
Response: We revised the Abstract, Results, Discussion and Conclusion to distinguish changes relative to the 2015-2034 reference from SAI effects relative to the corresponding simulation without SAI during 2060-2079. The two 1.5°C SAI experiments are now compared consistently to show how the response differs between the sustained high-forcing and overshoot backgrounds. We also identify regional exceptions and do not attribute the differences solely to the injection amount.
General comment 2
Reviewer comment: The manuscript relies on a single Earth system model (CESM2-WACCM6). Although this limitation is acknowledged in the Discussion, I suggest that its implications be emphasized more clearly when interpreting the regional precipitation and moisture-budget responses. Such responses can be strongly model-dependent, and the conclusions should therefore be framed accordingly.
Response: We have revised the abstract, discussion and conclusion to state that the results are specific to CESM2-WACCM6. We also clarify that three ensemble members represent only part of the internal variability and do not capture uncertainty across models, while the approximately 1° resolution limits the representation of terrain, convection and local hydrology.
General comment 3:
Reviewer comment: The statistical analysis deserves further clarification. The authors use a Welch’s t-test with only three ensemble members for each scenario and consider p < 0.05 as statistically significant. Given the very small number of ensemble members, the authors should provide additional justification for the robustness and appropriateness of this statistical approach and discuss its implications for the reported significance levels.
Response: We clarified that the Welch tests compare three member-specific 20-year means in each experiment, not annual values. Welch’s test avoids assuming equal variances, but it does not overcome the limitations of the small ensemble. We therefore report unadjusted p < 0.05 and BH-adjusted q = 0.05 fractions separately, together with effect sizes and ensemble spread, and acknowledge the limited statistical power and shared simulation histories.
General comment 4
Reviewer comment: The manuscript would benefit from a clearer distinction between statistical significance and magnitude of change. In several places, the reduction in the percentage of statistically significant grid cells under SAI is interpreted as evidence that SAI “effectively returns temperature conditions closer to the baseline state.” Statistical significance alone does not necessarily quantify the magnitude or practical importance of the change. Consider reporting and discussing the actual magnitude of changes alongside their statistical significance.
Response: We removed this inference. Relative to 2015-2034, continental mean temperature changes by +2.99°C and +1.67°C in the two experiments without SAI, compared with +0.15°C and −0.10°C in the two 1.5°C SAI experiments. The corresponding WSDI changes are +192.4 and +96.2 days per year without SAI and +5.4 and −1.3 days per year with SAI. We use these magnitudes, rather than reduced statistical significance, to describe proximity to the reference period.
General comment 5
Reviewer comment: The title emphasizes that SAI minimizes drying only when combined with strong decarbonization. This is an interesting and potentially important conclusion, but the manuscript should make this central message more explicit and quantitatively demonstrate it throughout the Results and Discussion sections.
Response: The original claim is not supported across all regions. We therefore changed the title to “African climate extremes and the surface moisture budget under stratospheric aerosol injection and temperature overshoot”. The revised Discussion and conclusion report pathway-specific and regional differences without presenting strong decarbonisation as a universal condition for reducing drying.
General comment 6
Reviewer comment: The authors should say a few more words about the uncertainties and limitations associated with the SAI experimental design itself, including the use of idealized feedback-controlled injection scenarios and prescribed global temperature targets. These simulations should not necessarily be interpreted as representing a realistic future deployment strategy.
Response: The Methods now describe the global temperature and temperature-gradient control objectives. The Discussion clarifies that the feedback-controlled injections are idealised and that achieving a global temperature target does not guarantee recovery of regional precipitation or the surface moisture budget.
Specific comments Line 1 Title
Reviewer comment: In the title, please consider whether “minimises drying” is fully supported by the results across the whole African continent. The results show important regional contrasts, including reduced moisture availability under some SAI scenarios in East and Southern Africa. The title may therefore need to better reflect the regional and scenario-dependent nature of the response.
Response: We have revised the title as “Temperature stabilization using Stratospheric Aerosol Injections minimises drying over some African region only when combined with a strong decarbonization effort” to reflect the regional variation and scenario dependent nature of the response
Lines 24 to 32 Abstract
Reviewer comment: The abstract is generally informative, but the main quantitative findings could be presented more consistently. For example, the abstract reports that precipitation indices exhibit significant changes over 40-80% of the area, whereas the Results section gives substantially lower percentages for several precipitation indices. Please check that all percentages reported in the abstract are consistent with the results presented in the main text and figures.
Response: We removed the 40-80% statement. The Results now report unadjusted and BH-adjusted detection fractions separately. For example, the unadjusted PRCPTOT fractions are 30.9% and 30.3% in the two experiments without SAI, whereas the BH-adjusted fractions are 3.9% and 0%.
Line 41 Acronyms
Reviewer comment: All acronyms must be clearly defined from the first occurrence they are used. Therefore, please apply this instruction to “CMIP6” and all other acronyms used throughout the text.
Response: We have defined CMIP6, CESM2-WACCM6, ETCCDI, SAI, SRM and CDR at first occurrence and use the abbreviations thereafter.
Lines 65 to 66 Terminology
Reviewer comment: Please clarify the terminology used throughout the manuscript between “stratospheric aerosol injection (SAI)”, “stratospheric aerosol geoengineering (SAG)”, and “solar radiation modification (SRM)”. The manuscript uses these expressions somewhat interchangeably. A consistent terminology would improve readability.
Response: We use SAI for the experiments analysed and SRM for the broader class of interventions. SAG is no longer used as an interchangeable term.
Lines 111 to 114 Regions
Reviewer comment: There appears to be an inconsistency between the description of the regionalization in the text and the caption of Figure 1. The text mentions five sub-regions (North, West, Central, East and Southern Africa), whereas the Figure 1 caption additionally mentions Northeast Africa (NEAF). Please clarify the regional classification and apply the same terminology consistently throughout the manuscript.
Response: This has been resolved in the text and Figure 1. To enable robust sub-continental interpretation and to avoid continent-wide averaging that can mask localized signals, the African domain was stratified into five sub-regions as depicted in Figure 1: North Africa (NAF), West Africa (WAF), Central Africa (CAF), East Africa (EAF) and Southern Africa (SAF). See Page 3, lines 112- 115, and Figure 1
Line 138 Scenario notation
Reviewer comment: Please ensure consistency in the notation of the overshoot scenario. The manuscript uses SSP5-3.4-OS and SSP5-34-OS in different places. The scenario name should be standardized throughout the text, tables and figures.
Response: We now use SSP5-3.4-OS consistently. Table 1 defines the abbreviated SAI labels, parent pathways and temperature targets.
Lines 138 to 157 Experimental design
Reviewer comment: The experimental design should be described more clearly.
Response: Table 1 now lists all five experiments, their description, and temperature targets. The Methods explain the branch histories and distinguish the experiments from the two fixed 20-year reference periods, 2015-2034 and 2025-2044.
Line 165 ETCCDI reference
Reviewer comment: Please clarify the citation and terminology associated with the ETCCDI framework. The text currently states “the Expert Team on Climate Change Detection and Indices (ETCCDI) Peterson (2002)”. Please check the corresponding reference and citation year, and revise the sentence.
Response: We cite Zhang et al. (2011), define ETCCDI once and distinguish the standard ETCCDI indices from the additional temperature indices in Table 2. The Peterson citation is removed from this statement.
Lines 199 to 208 Statistical samples
Reviewer comment: The statistical methodology needs further explanation. It would be useful to specify exactly what constitutes the samples in the Welch’s t-test: are the tests performed using annual values, ensemble-member means, or period means? Given that only three ensemble members are available, this point is particularly important for assessing the robustness of the significance analysis.
Response: Each group contains three member-specific period means. Annual values within each 20-year period are first averaged for each member, giving n = 3 versus n = 3. Detection fractions are calculated over testable African land, and testable coverage is reported separately.
Line 214 Grammar
Reviewer comment: Please revise the sentence “these indices plateau begin to decline” for grammatical clarity.
Response: The revised sentence states that, under SSP5-3.4-OS, the temperature indices level off and subsequently decline.
Lines 284 to 287 Mechanisms
Reviewer comment: The physical mechanisms invoked to explain the precipitation response to SAI are interesting. However, the discussion could be strengthened by distinguishing more clearly between the mechanisms supported directly by the simulations and mechanisms inferred from previous studies.
Response: The revised Discussion reports the precipitation and evapotranspiration changes directly. Energy-balance and circulation mechanisms reported by Niemeier et al. (2013) and Simpson et al. (2019) are presented only as possible explanations because their regional contributions were not isolated in this analysis. We also removed the general attribution of increased West African surface moisture budget to reduced evapotranspiration because evapotranspiration increases in both overshoot SAI comparisons.
Lines 295 to 305 Water availability
Reviewer comment: The interpretation of the surface moisture budget deserves some caution. In particular, the statement that SAI “could help preserve surface water availability” in West Africa should be framed carefully because the analysis is based on P - ET, which represents a simplified surface moisture balance and does not directly quantify water availability in rivers, soils, reservoirs or groundwater.
Response: We removed the claim that SAI preserves surface water availability. The revised text explains that the surface moisture budget compares precipitation with actual evapotranspiration, while water availability also depends on storage, runoff, drainage and human use.
Lines 314 to 329 Repetition
Reviewer comment: This paragraph appears to substantially repeat the preceding description of Figure 5. Please consider removing the duplicated material and using this space to provide additional interpretation of the seasonal differences.
Response: We replaced the repeated description with seasonal SAI-minus-parent contrasts. In West Africa, the largest positive change occurs in MAM under SSP585_1.5 SAI and in JJA under the two overshoot SAI experiments. In East Africa, the largest decrease occurs in DJF under SSP585_1.5 SAI and in MAM under both overshoot SAI experiments. Southern Africa shows a decrease in DJF in all three experiments.
Lines 331 to 332 Contradictory East African claim
Reviewer comment: Please clarify the apparent contrast between the statement that East Africa is the greatest beneficiary, with an increase of approximately 0.5 mm/day, and the preceding results indicating that SAI produces lower moisture levels than the high-warming baseline in East Africa. The reference scenario used for each comparison should be explicitly stated.
Response: We removed the term “beneficiary”. Relative to the corresponding simulation without SAI, the East African surface moisture budget decreases by 0.173, 0.029 and 0.047 mm day⁻¹ in the three SAI experiments. The revised text keeps this comparison separate from changes relative to the earlier reference period.
Lines 396 to 398 Lower forcing and drying
Reviewer comment: The conclusion that excessive drying associated with SAI can be mitigated under a lower-emission scenario is one of the most important findings of the study. I suggest developing this point further and providing a more quantitative comparison of the hydrological response between the two emission pathways.
Response: At the common 1.5°C target, the East African surface moisture budget decreases by 0.173 mm day⁻¹ under SSP585_1.5 SAI and by 0.029 mm day⁻¹ under SSP534OS_1.5 SAI. The corresponding Central African changes are −0.104 and −0.047 mm day⁻¹, while continental PRCPTOT changes by −38.6 and +0.1 mm per year. We therefore narrowed the conclusion to these quantities and regions rather than retaining the original continent-wide claim.
Lines 401 to 409 Limitations
Reviewer comment: The limitations section is useful, but it could be expanded to include the idealized nature of the SAI experiments, the relatively coarse spatial resolution of the model for regional hydrological analysis, and the uncertainty associated with using a single model.
Response: The revised limitations cover the single model, approximately 1° resolution, three-member ensembles, statistical assumptions, shared simulation histories, idealised feedback control, broad geographical boxes and unmodelled impact processes.
Tables and figures
Reviewer comment: Table 1; The notation used for the SAI experiments should be standardized. In particular, please make clear whether “Geo SSP5-34-OS 1.5” and “Geo SSP5-34-OS 2.0” refer to temperature targets or experiments. It would also be useful to explicitly indicate which simulations are used as baselines and which are used as SAI experiments.
Response: The revised table lists the two experiments without SAI and the three SAI experiments, including the parent pathway, global temperature target and ensemble size for each. The reference periods are defined separately.
Reviewer comment: Table 2; Please carefully check the definitions and units of all indices. In particular, the units associated with R1MM, R20MM and PRCPTOT appear potentially inconsistent with their definitions. Please verify the table against the standard ETCCDI definitions.
Response: Table 2 now lists only the eleven indices calculated in the final analysis. R1mm is reported in days, Rx5day and PRCPTOT in mm, CWD and CDD in days, SDII in mm day⁻¹, annual GDD in °C days and GDD intensity in °C. R20mm and GSL were removed. The 10-35°C day count is described as a broad temperature indicator rather than a crop-specific optimum.
Reviewer comment: Figure 2; Please clarify why only selected indices are presented in Figure 2 and explain whether the omitted indices exhibit similar temporal behaviour. This would help readers understand whether the selected indices are representative of the overall response.
Response: Figure 2 shows mean temperature, GDD intensity, PRCPTOT and CWD. The time series for all eleven calculated indices are provided in Figures 1 and A4, so the omitted indices are not assumed to behave like those shown in Figure 2.
Figure 3
Reviewer comment: The figure caption should clearly identify all scenarios represented in the figure. This is particularly important because the manuscript considers both 1.5°C and 2.0°C SAI targets, whereas the caption appears to focus only on the 1.5°C experiments. Please verify that the figure, caption and corresponding discussion are fully consistent.
Response: The revised caption identifies the reference and all five experiments, including SSP534OS_2.0 SAI. It also defines the 2060-2079 future period, the 2015-2034 reference period and the BH-adjusted significance markings.
Figure 4
Reviewer comment: The caption states that four scenarios are represented, but the manuscript considers several SAI experiments, including a 2.0°C target. Please clarify which scenarios are included in this figure and why some experiments may have been excluded.
Response: Figure 4 now includes all five experiments, and the caption identifies them. It also states that the heatmap shows unadjusted p < 0.05 detection fractions.
References
Reviewer comment: Reference formatting; Please carefully check the References section for consistency, completeness and formatting. Several references appear to contain incomplete bibliographic information or inconsistencies in formatting.
Response: The references have been checked for consistency and completeness
Reviewer comment: Kravitz et al. 2017; The reference to Kravitz et al. (2017) appears incomplete and should be checked carefully.
Response: We completed the Kravitz et al. (2017) reference by adding the missing publication details and DOI.
Reviewer comment: Mupangwa et al. 2016; The reference to Mupangwa et al. (2016) should be checked against the statement for which it is cited in the Introduction section, particularly the claim concerning the contribution of agriculture to the Gross Domestic Product.
Response: We removed the unsupported GDP statement and the Mupangwa et al. (2016) citation from that passage. The revised Introduction uses Sultan and Gaetani (2016) to support the broader discussion of climate sensitivity in African agriculture.
Reviewer comment: Ofori and Asongu 2021; The reference to Ofori & Asongu (2021) appears to concern foreign direct investment, governance and inclusive growth rather than the specific agricultural statement for which it is cited. Please verify whether this is the appropriate reference.
Response: We removed the citation and the unsupported GDP statement.
Reviewer comment: Emmons et al.; Please check the consistency between the citation of Emmons et al. in the text and the year given in the reference list.
Response: We use Emmons et al. (2020) consistently, which is the correct publication year.
Reviewer comment: Danabasoglu et al.; Similarly, please verify the citation year and bibliographic details of Danabasoglu et al. in the text and reference list.
Response: We use Danabasoglu et al. (2020) consistently and corrected the bibliography entry.
-
AC2: 'Reply on RC2', Temitope Samuel Egbebiyi, 21 Oct 2026
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 271 | 116 | 47 | 434 | 45 | 43 |
- HTML: 271
- PDF: 116
- XML: 47
- Total: 434
- BibTeX: 45
- EndNote: 43
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
