the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Future dust concentration over the Middle East and North Africa region under global warming and stratospheric aerosol intervention scenarios
Seyed Vahid Mousavi
Khalil Karami
Simone Tilmes
Helene Muri
Abolfazl Rezaei
Abstract. The Middle East and North Africa (MENA) is the dustiest region, in the world and understanding the projected changes in the dust concentrations in the region is crucially important. Stratospheric aerosol injection (SAI) geoengineering aims to reduce global warming, by increasing the reflection of a small amount of the incoming solar radiation to space, and hence reducing the global surface temperatures. Using the output from the Geoengineering Large Ensemble Project (GLENS) project, we show a reduction in the dust concentration in the MENA region under both global warming (RCP8.5) and GLENS-SAI scenarios compared to the present-day climate. This reduction over the MENA region is stronger under the SAI scenario, while for dry season (e.g., summer with the strongest dust events), more reduction has been projected for the global warming scenario. The maximum reduction of the dust concentrations in the MENA region (under both the global warming and SAI) is due to the weakening of the dust hotspots emissions from the sources of the Middle East. Further analysis of the differences in the surface temperature, soil water, precipitation, leaf area index, and near surface wind speed provides some insights into the underlying physical mechanisms that determine the changes in the future dust concentrations in the MENA region. We also conduct wavelet analysis using the time series of the monthly, seasonal, and annual climate changes under the SAI simulation to identify the dust relationship with the considered variables. Our findings show that a stronger reduction of the dust concentration in the MENA region under SAI relative to the RCP8.5 scenario is a complex interplay with temperature reduction, precipitation, soil water and leaf area index enhancement, as well as weakening of near surface winds compared to the present-day climate.
- Preprint
(1905 KB) -
Supplement
(100 KB) - BibTeX
- EndNote
Seyed Vahid Mousavi et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-370', Anonymous Referee #1, 12 Aug 2022
The paper presents a study of future dust concentration in the MENA region under the RCP8.5 scenario and a corresponding geoengineering experiment
designed to keep the global temperature at 2020 levels. The experiments used are from the ensembles generated under the GLENS project. The paper first present a cross-coherence analysis between dust concentrations and variables like temperature, precipitation, soil-moisture etc. It then goes on with a detailed presentation of the changes in the different variables in the RCP8.5 scenario and the geoengineering scenario at the end of the century.
I think the topic is interesting and a study of the changes in the future should be welcomed. However, I think the conclusions from the cross-coherence analysis are questionable and I find that the rest of the paper is too much of an 'atlas' over the changes and don't really answers the questions of what factors that drives the changes in dust concentrations. Because of these concerns, I can't suggest that it is accepted in its present form.
Major concerns:
a) The cross-coherence analysis:
I don't find that the method is very well explained in section 2. What are the connections to the axes in Fig. 2? How do you come from the equations to the quantities (amplitude, phase?) shown in the figure? More importantly, I am also confused about the physical interpretation. There are probably annual cycles in all meteorological variables. This means that there will always be coherence between them. As the annual cycle probably is different from a pure sinusoidal, there will also be a signal at 1/2-year. So what do we actually learn from Fig. 2? In the discussion section (l365) it says that the dust is 'substantially influenced' by the changes in the other fields. But I don't think you can conclude that from the analysis. What we learn is only that there is an annual cycle in all the fields including the dust but nothing about the physical interpretation.
b) The rest of the paper seems to me to be too much focusing on presenting the details about the changes in the different fields. I think many of the panels basically shows the same and that the number of plots and panels could be reduced. I really miss some solid analysis and results about what drives the changes in the dust. The dust generally decrease in the RCP8.5 scenario but it decreases further in the geoengineering scenario. Perhaps I am missing something but I could not find an explanation. The correlations in Table 3 could be a beginning, but the physical connection between the variables requires that the trends - which I guess determines most of the correlations here - are removed.
Minor comments:l54: reginal -> regional
l97: So this is an ensemble based on a single climate model? How are the different ensemble members generated?
l103: What is 'interhemispheric temperature gradient'?
l115-130: Is this a new method adopted for the present study? Is it described in the literature before? If it is new perhaps it should be described in more details and more background given. As it is now it is not transparent for me. For example what is a transport bin?
l148: composite analysis? Is this the right word? You calculate the difference of temporal means.
l160, Table 3: Are the correlations averages over all the ensemble members? It should be mentioned in the caption that this is annual means.
As mentioned I have problems with the presentation of the wavelet coherence.
In line 171 why is [(n'-n)dt/s] the complex conjugate? Is omega_0 a constant? If it is how is it selected?
l172: The sentence 'In this approach .. ' seems misplaced here and should be moved down near line 184.
More importantly in Fig. 2 the coherence is shown as function of time (x-axes) and period (y-axes). It is not clear from the text what these
correspond to in the formulas.
Furthermore, the figure caption mention both the power and the phase which is not described in the text. The same goes for the cone of
influence.
Eq. 6: Should there not be some smoothing here too?The discussion of Fig. 2, page 7-8:
It should be pointed out more specifically in the text that Fig. 2 is for SAI. Does it look the same for the RCP8.5? Why focus on the SAI here?
The 22-years variability and variability larger than 16 years seems to be outside the cone of influence. Also, it is not significant in the GWTC. In general the two regions in Fig. 2 look identical to me. I don't think you can say that there are significant differences.
And I don't really see any change after 2040. Perhaps just presenting the GWTC would be better.l207: 'Out of phase'. Does this mean -180? Is it just difference in sign?
l246: How does this indicate that the model is consistent with observations? There are no observations used in the present study.
Table 3: Why the big difference between RCP8.5 and SAI for temperature correlations? Is this table only discussed in l258?
Section 3 should be split in two or more subsections. Perhaps not start with the coherence?Citation: https://doi.org/10.5194/acp-2022-370-RC1 -
AC1: 'Reply on RC1', Seyed Vahid Mousavi, 08 Dec 2022
Publisher’s note: this comment is a copy of AC2 and its content was therefore removed.
Citation: https://doi.org/10.5194/acp-2022-370-AC1 -
AC2: 'Reply on RC1', Seyed Vahid Mousavi, 08 Dec 2022
Publisher's note: a revised supplement was added to this comment on 12 December 2022.
Dear Sir/Madam,
The point-to-point response to all the comments (the comments are rewritten in black color and the replies in blue) is attached. We appreciate the opportunity to revise our paper. We believe that the manuscript is much improved after positively addressing all the requested revisions. The main changes that have been made in the new version based on the referee’s comments/suggestions are as follows:
- We replaced Fig.2 with two new figures, a new figure (Fig. 9) for detailed analysis of the correlation between dust and considered variables, and a second figure (see next point).
- We provide a new figure for annual trends (Fig. 10) of all considered variables over dust hot spots to interpret the positive and negative correlation considering ascending or descending trends
- New Table for the correlation coefficient over dust hot spots, shows which variable would have more effect on the change of future dust concentration in different regions
- New figures for monthly trends using box plots (Fig. 11 and Fig. S1), to give a better view of the statistical analysis and standard deviation of different scenarios
- Rewrite the result section with three subsections to increase readability
- Rewrite the result, discussion, and conclusion sections based on new findings and figures
- To magnify the parameter’s changes over the dust hotspot regions, these regions are specified by dashed lines and overall contour plots.
Best Regards,
Vahid Mousavi
-
AC1: 'Reply on RC1', Seyed Vahid Mousavi, 08 Dec 2022
-
RC2: 'Referee Comment on acp-2022-370', Anonymous Referee #2, 09 Sep 2022
This paper is a straight forward comparison of predictions of model ensembles using one model with two scenarios, global warming (RCP8.5) and stratospheric aerosol injection (SAI), over 80 years compared with a control run of 20 years. The variable of interest is dust and its correlation with surface temperature, leaf area index, precipitation, soil moisture and wind speed. The region of interest is north Africa and the middle east with various dust hot spots identified. The bulk of the paper rests on describing Figs 3-9 which show the spatial and temporal variation of each of these parameters under the two scenarios for monthly and annual means. The spatial differences are shown variously ad absolute value or percentage depending on the variable. It is not clear why they are not all shown as percentages.
The authors then make some conclusions about the differences between the RCP8.5 and SAI scenarios, a number of which are difficult to believe if the error bars are included in the discussion of the annual differences or trends in for example soil moisture, wind speed.
Error bars should be included on all the figures showing mean values: monthly, annually, or spatially. Currently error bars are included only on the annual means. The same should be applied to Fig. 10. Then the authors discussion of notable differences can be placed in the context of how well any one variable is known.
One of the results which is rather striking, but which the authors largely ignore, is how little difference there is between the various variables, except for surface temperature and leaf area index, for the two scenarios, see e.g. Fig. 10. Similarly for most variables there is primarily little difference between the two scenarios and the control. Isn’t this surprising given one scenario is global warming as usual, whereas the other is to deal with global warming. Are we to conclude that only primarily temperature will be affected?
The paper would be improved if some discussion along these lines was added and if the authors treated the supposed differences and trends more carefully to put them in the context of the uncertainty in the knowledge of variable in question. If differences or trends are small fractions of the uncertainty, there cannot be much confidence in such predications.
More detail on these and other points follow in paper order, including a couple of minor points.
44 From remote “regions?”
80 dioxide
166 Isn’t it the cumulative LWTC averaged over time? Or is there a new variable WTC?
Fig. 2 Some general comments should be made to explain the similarities of all the figures no matter the variable being correlated, particularly for readers not accustomed to such plots. For example, why is there always a strong annual cycle? Is this just the strong annual seasonal cycle? Why is there a definite semicircle traced out delineating the bright and dim colors in all plots? Is this an issue with the period versus the year, i.e. there can’t be an eight year correlation for times less than 16 years beyond the start date? Presumably this is the cone of influence. But if that is the case why are there any correlations outside this cone shown on the figure?
Fig. 2 caption is unclear. 1) Isn’t the cone of influence denoted by the more intense colors? If that isn’t the case then it suggests the cone of influence is only from 2-20 years before 2050 and after 2070 with no influence in the center of the figure? 2) What is meant by the whole MENA region. Is that different than the MENA region? Also in the text line 199, and similarly confusing whole middle east. These regions were defined clearly earlier, now there seems to be a confusion about what they mean.
218 Again the whole MENA compared with the Middle East. Is this now not the whole Middle East?
Fig 3 c-q. Consider using percentages. The average reader may not know if 45 ug/m3 is a lot or a little. But checking Figs 3a,b indicates that 45 ug/m3 is 50-100% above or below the mean value, so it is a lot.
Figs 6-9 q) which depict the annual mean value. Don’t all of these figures, except fig. 6q) show that considering the error bars there is no difference between RCP8.5 and SAI. The difference in the means is a small fraction of the range of differences mapped out by the error bars. The differences shown in the monthly mean value figs p) appear at first more significant, but where are the error bars on this figure? If they were included the picture might be just as difficult in concluding a difference between RCP8.5 and SAI. Of these figures the only two that show a distinct difference outside the error bar range are surface temperature and TLAI.
Thus the authors conclusions such as at lines 311-, “Figure 7q further shows … and under SAI, the wind speed reduction is gradually stronger than RCP8.5 starting from 2050.”, or 324, “Fig 8q shows that a moderate positive trend of the annual mean value exists in the soil moisture under the SAI scenario.” are deeply flawed. There is no trend that would stand under any statistical test given the size of the error bars on the data. The authors must be much more careful about what can be concluded from these monthly and annual mean values.
Similar comment can be made about Fig. 9r), a slight difference appears in the mean values east of 50 degrees, but would this appear significant if the error bars were included on this figure? The error bar range is on the order of plus/minus 100 mm/year.
Fig. 10. Error bars should be included on this figure, just as they have on all the annual means shown. This is needed to put the differences noted in the context of the overall uncertainty in the predictions.
Citation: https://doi.org/10.5194/acp-2022-370-RC2 -
AC3: 'Reply on RC2', Seyed Vahid Mousavi, 08 Dec 2022
Publisher's note: a revised supplement was added to this comment on 12 December 2022.
Dear Sir/Madam,
The point-to-point response to all the comments (the comments are rewritten in black color and the replies in blue) is attached. We appreciate the opportunity to revise our paper. We believe that the manuscript is much improved after positively addressing all the requested revisions. The main changes that have been made in the new version based on the referee’s comments/suggestions are as follows:
- We replaced Fig.2 with two new figures, a new figure (Fig. 9) for detailed analysis of the correlation between dust and considered variables, and a second figure (see next point).
- We provide a new figure for annual trends (Fig. 10) of all considered variables over dust hot spots to interpret the positive and negative correlation considering ascending or descending trends
- New Table for the correlation coefficient over dust hot spots, shows which variable would have more effect on the change of future dust concentration in different regions
- New figures for monthly trends using box plots (Fig. 11 and Fig. S1), to give a better view of the statistical analysis and standard deviation of different scenarios
- Rewrite the result section with three subsections to increase readability
- Rewrite the result, discussion, and conclusion sections based on new findings and figures
- To magnify the parameter’s changes over the dust hotspot regions, these regions are specified by dashed lines, over all contour plots.
Best Regards,
Vahid Mousavi
-
AC3: 'Reply on RC2', Seyed Vahid Mousavi, 08 Dec 2022
Seyed Vahid Mousavi et al.
Data sets
the Socioeconomic Data and Applications Center (SEDECA) Alex de Sherbinin, Deborah Balk, Karina Yager, Malanding Jaiteh, Francesca Pozzi, Chandra Giri, Antroinette Wannebo https://sedac.ciesin.columbia.edu/
Stratospheric Aerosol Geoengineering Large Ensemble Project - GLENS Simone Tilmes, Jadwiga H. Richter, Michael Mills, Ben Kravitz, Douglas G. MacMartin https://www.cesm.ucar.edu/projects/community-projects/GLENS/
Seyed Vahid Mousavi et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 371 | 123 | 17 | 511 | 43 | 5 | 6 |
- HTML: 371
- PDF: 123
- XML: 17
- Total: 511
- Supplement: 43
- BibTeX: 5
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1