Re-review of: Impacts of reductions in non-methane short-lived climate forcers on future climate extremes and the resulting population exposure risks in eastern and southern Asia
Overview:
The authors have mostly addressed the concerns I raised in the previous review, with a few exceptions. My major concerns are outlined below, followed by line comments.
Major concerns:
1) Consistency of communication of findings:
In my initial review, I had suggested the authors stipple (add dots to) regions where the models disagree, but the authors only made this change to one figure. The other figures have stippling where the models agree- these dots cover the colors/shading in the locations where the changes are agreed upon by the models, effectively visually masking the relevant information. I suggest consistently using stippling throughout (in other words: stipple on all maps where models disagree, not where they agree on some maps and where they disagree on other maps- this can be confusing to the reader if the stippling changes from figure to figure).
The authors added country borders/boundaries to some of the maps, but not all of the maps- why did the authors only include this information on some of the maps and not others? I find the political boundaries/borders helpful for identifying which regions are impacted in which country when the authors discuss the findings in the text.
The authors are still using a very low significance criteria for stippling on the maps- most of the models must ‘agree on sign of change’ as the significance test. As I stated in my last review, I think in addition to agreeing on sign of change, some sort of signal:noise (or coefficient of variation or similar) would be a more robust/effective metric, particularly in the maps showing warming in ssp370 vs ssp370lowntcf- most of the globe warms almost everywhere over land in both experiments, so showing where the models all agree it warms as an agreement metric isn’t particularly useful in my opinion. Agreeing on magnitude of changes would be more robust and meaningful.
2) Significance and uncertainty of findings, especially as they relate to precipitation changes.
In the text, I had suggested in my previous review that the authors present not only the multi-model mean/median when reporting findings, but also the spread or +/- 1 or 2 sigma. I appreciate the authors added this information in the Abstract and in some locations in the text, but they left out the uncertainty/model spread in many locations in the text. I suggest showing this uncertainty/spread throughout, both for consistency and transparency in terms of model agreement.
The authors have chosen to report several findings related to precipitation changes in the text with minimal discussion/mention (mentioned maybe once in the Discussion?) of the fact that the maps indicate most of the models don’t even agree on the sign of the change (let alone the magnitude). Several of the precipitation results are reported in the text that seem to not be stippled in the maps, suggesting to me that the models don’t agree, and I question how much we should trust the results. If the models don’t agree on the sign of change in a region, I strongly suggest the authors mention this along with the results to warn the reader that the results should be interpreted with caution.
Related to this, the authors use wording such as ‘significant increases are found’ - If the authors are going to suggest that the results are ‘significant’, what is the significance test? If no test is used, can the authors use a test, such as a Mann-Whitney/Rank Sum test to determine if the medians/distributions are statistically distinguishable?
3) Proofreading/changing text after authors made changes:
There are several instances where the authors made changes to figures, but the figure numbers did not change (for example, there are two Figure 2s). Also, colors in line plots are changed from green to grey, but all captions still reference green lines that no longer exist. Please double check all of this, and that updated figure numbers in text match figure numbers in captions after fixing figure numbering.
Line comments:
Line 17: ’the entire study area’- suggest 'entire study area of Asia' or similar more specific wording to remind reader
Line 21: ‘For temperature extremes…’ To prepare reader for sub-regional reporting, I suggest starting with wording such as 'In terms of sub-regional changes in temperatures extremes, the largest increases in...'
Line 45: ‘North-East of India’ - suggest 'North eastern' or 'The North East of'
Line 63: ‘using the five Earth system models (ESMs)’ - which five? (author use of the word 'the' seems to imply these were introduced previously?)- suggest removing ‘the’ unless authors have previously introduced these models
Line 69: ‘global warming 1.5 °C and 2 °C,’ – suggest 'global warming of 1.5 °C and 2 °C'
Line 70: ‘Plateau and when’ – there are two complete phrases run together with 'and'- please separate the phrases into different sentences.
Line 90: ‘on human body’ - suggest 'to human health and well-being' or similar wording
Line 93: ‘is lacked’ – suggest ‘is lacking’
Lines 76-93: thanks to the authors for adding in more detail about previous work, how this work represents a significant advance- I was much more convinced of the novelty/importance of this work after reading this paragraph.
Line 95: ‘increasing’ – suggest ‘will increase’ or better 'will increasingly become a threat to...'
Line 107: ‘have become more’ - can authors be more specific here about what is better in these ESMs that make it worthwhile to do the analysis? they suggest/hint at this earlier but would be helpful to be more explicit here.
Line 113: ‘in eastern and southern Asia using’ – suggest adding in: 'Asia (here defined as 0-60°N, 70-150°E. ) using' and then eliminating the sentence at the end of the paragraph because the last sentence doesn’t seem to fit, and this information can be included in this sentence here.
Lines 146-147: ‘thus, we directly used the CMIP6 multi-model ensemble (MME) mean’ - please specify that the authors are referring to the sub-selection of models here so the reader is not confused thinking authors are using the 'entire' ensemble of ~40-50 CMIP6 models
Line 151: ‘high level of resolution’ - this is subjective- I suggest removing and simply describing the resolution.
Line 182: ‘were are’ – typographic error/fix wording
Lines 209-210: ‘Regionally, the greatest warming of more than 1.5 K is found in central and northern Asia and in northern North America, particularly in the Arctic, where warming is greater than 2.5 K.’ - might be helpful to cite a paper here that has shown this previously as I’m sure Arctic amplification of warming is not a new finding here
Lines 216-217: ‘The reduction of non-methane SLCF emissions results in an average increase of 0.19 K in global mean SAT in the MME results, ranging from 0.06 K to 0.29 K across different models (Table 3).’ – thanks to authors for including the spread in model results to give the reader a sense of the uncertainty- please do so elsewhere where reporting mean model results.
Lines 227-233: Where is the model spread/range? Only one number is reported (“27.7%”- this makes the answer sound so certain, but I’m sure there’s disagreement in the models in terms of magnitude, correct?)
Lines 239-247: In my opinion, this description of previous work and comparison with results from other regions outside the study region seems to fit more in the Introduction or Discussion
Lines 256-257: ‘Concomitant with the increases in TNn and TN90p, TR is also projected to increase over the entire study area in the future.’ As far as I can determine from looking at the maps, this is not true- there are blue areas on the map, indicating a decrease in days with tropical nights on the Tibetan Plateau, unless I am misinterpreting the map and colorbar- either way, TR does not consistently increase over the entire study area.
Lines 281-284: please include the model spread here, not just the mean
Lines 284-285: ‘In general, the SCB is most strongly affected by extreme temperatures under both scenarios’ - based on which metric? looking forward two sentences, the authors indicate that the other regions are most strongly affected. This wording seems to directly contradict what the authors say later in the paragraph.
Lines 293-294: ‘Figure 7 shows the time series of changes in annual mean precipitation indices averaged across the entire study area under the SSP3-7.0 and SSP3-7.0-lowNTCF scenarios from 2015 to 2050 relative to the reference period’ - can authors please show range of difference in gray shading around the grey line? Or does this make the figure too complicated?
Lines 296-300- here and elsewhere, please include range in model results, not just the mean/median
Lines 304-306: ‘Under the SSP3-7.0 scenario, R10 decreases by approximately 2 days in central India, the Indo-China Peninsula, and by more than 4 days in the southeastern Qinghai-Tibet Plateau, SWC, and parts of Indonesia, while it increases in all other regions of the entire study area (Fig. 8a).’ - do the models even agree on the sign of change in all these locations? How confident are the authors in projected changes in local precipitation if the models don't even agree on the sign of change?
Lines 310-311: ‘significant’- how is this significant? Based on what metric? Most of the boxplots in Figure 9b show nearly complete overlap in the 25th-75th percentile estimates.
If the authors are going to suggest that the results are ‘significantly different’ can they at least use a significance test, such as a Mann-Whitney test to determine if the medians/distributions are statistically distinguishable?
Line 320: ‘significant increase’- see earlier comments on significance
Lines 331-334: again, how is significance determined?
Line 337: ‘However, CDD increases in NIN, the southeastern Tibetan Plateau, and the southern Yangtze River (Fig. 8a and b).’ – do the models even agree on the sign of change across all of these locations, let alone the magnitude? I don’t see stippling on the maps in many of these areas.
Lines 338-341: ‘Future reductions…’- as far as I can determine, this is the first time that sign agreement has really been acknowledged for precipitation changes- I think the significance and sign agreement needs to be better addressed elsewhere, not just here.
Lines 343-344: ‘Notably, the increases in CDD in India and eastern China are accompanied by increases in the frequency and intensity of extreme precipitation, which may be related to the probability distribution of future precipitation.’ – see previous comments about model agreement on sign of change, significance considerations.
Lines 349-351: ‘In general, along with the increase in average temperature, future reductions in the emissions of non-methane SLCFs will increase the intensity and frequency of extreme precipitation and decrease the occurrence of extreme droughts in the entire study area.’ This and other similar statements seem to be worded as incredibly certain, but don’t seem to be backed up by clear evidence from my reading of the figures. For example, when I look at this figure, it looks like this is somewhat true in southern China for R10 and RX5day, and R95p, but not true for CDD, which shows mixed results, and very little model agreement on sign of change.
Lines 355-356: ‘along with a decrease in CDD in northern China’ - I see some light blue in Figure 8 with some stippling in isolated areas in eastern China, but a lot of China shows no stippling, which gives me little confidence in the sign, let alone magnitude, of these projections, unless I am misinterpreting the results, if so, please explain.
Lines 357-359: ‘Consequently, our results indicate that future non-methane SLCF reductions may alleviate the observed precipitation anomaly pattern of “southern flood/northern drought”
over eastern China, but may not have apparent mitigating effects on the precipitation reduction in India.’ – I am not familiar with the southern flood/northern drought research, but when I look at this figure, I see a mostly east/west, not north/south pattern here with little sign agreement, so I don't know what evidence this statement is based on.
Lines 390-391: ‘The climate factor is also the largest contributor to the increase in population exposure, followed by the climate- population interaction factor and the population factor in NC, the SCB, and SC’ - please be more specific- to precipitation extremes?
Lines 401-403: ‘Our results show that non-methane SLCF reductions will exacerbate the warming effect caused by GHGs, resulting in increases in extreme temperature and precipitation events compared to the standard SSP3-7.0 scenario.’ - if reduced alone, correct?
Lines 420-421: ‘Notably, large differences in extreme precipitation changes are found at the regional scale in response to reductions of non-methane SLCFs and only aerosols, with changes in opposite directions observed in some regions’ – yes, I appreciate the authors noted this, and I think this should be noted when individual results are presented if the models don’t consistently even agree on sign of change (in the Results section).
Line 431: ‘contrary’- opposite in sign?
Lines 433-435: ‘The latter is usually used as an indicator of flooding, suggesting that heavy precipitation associated with natural disasters will be aggravated in the future due to non-methane SLCFs reduction.’ True in southern/SE China, but really not much agreement on sign anywhere else if I am interpreting the maps correctly.
Line 435: what is ‘well assessment’?
Lines 436-437: ‘The reduction of non-methane SLCFs will result in the exposure of millions of people to extreme events, and up to tens of millions in densely populated areas’ – This sentence gives me the impression that these extreme events wouldn't have happened in the SSP3-7.0 scenario without NTCF reductions. However, my understanding looking at the SSP3-7.0 vs lowNTCF figures is that under SSP3-7.0 extremes also occur, but the elimination of NTCF just makes them more frequent/likely/worse. Is this correct? If so, can the authors more carefully word these summary findings please?
Line 443: ‘that some of the extreme indices were not well fitted’ – what does this mean?
Lines 452-455: thanks to the authors for contextualizing this.
Lines 458-461- this description seems more accurate to me than what is stated on lines 436-437 (see comment above)
Line 486: ‘has’- have
Lines 493-494: ‘which may lead to underestimation of the impact of SLCF
emissions reductions in China’ - wouldn't this lead to an overestimation if simulated SLCF emissions are higher than they actually are in reality (so future emissions reductions would be less impactful because they already happened, so the models would overestimate the real impacts of SLCF reductions bc we’ve already experienced these impacts?)? Or am I misinterpreting the impact of lower emissions? Or what the authors intend to say?
Lines 496-499: ‘What policymakers, the public, or the media need to know’- I will leave this wording choice up to the authors, but in my previous review I was not intending that the authors needed to explicitly address the conclusion to policymakers and the media- instead, I was suggesting the authors contextualize their results as they have done in the Discussion/Conclusion so as not to be misinterpreted by science reporters etc. In my opinion, the authors can eliminate the phrase ‘What policymakers, the public or the media need to know is that’ and start the sentence with ‘Air pollution…’
Suggestions for figures:
For figures with maps:
In my previous review, I suggested the authors use one colormap for showing changes through time, and another colormap for showing differences among experiments so they can be quickly/easily visually distinguished. The authors combined color bars for the climate change maps, but kept the same color maps (the same colors) to show the differences among experiments- I suggest using different color maps (a different divergent color map for showing differences vs changes through time) so the reader can easily determine that the maps are not showing the same things.
For all line graphs: authors reference ‘green’ lines in figures, but lines are now grey- please double check figure captions and change where appropriate.
Line 775/776 (Figure 1 caption): may be helpful to add a line something like ‘if red line only is shown (g,h), this indicates that ssp370 and ssp370-lowNTCF overlap” or other wording to let reader know why there is only one line in bottom panels (g,h).
Figure 2 (line 780): figure panels are so small, I can’t see text, etc.
Instead of this being a figure that is 2 (columns) x 8 (rows), can the authors make this two sets of double columns (4x4) so precip is next to temperature with a little space in the middle to separate them visually? This might help fill the white space on the page and improve readability bc the maps could be larger?
Figure 2 (line 785): The stippling in the maps changes from figure to figure - please consistently use stippling- are significant or insignificant regions stippled? Also, the authors changed one figure in response to my comments, but the same comment could be made for all figures showing warming- showing where models agree on warming in warming scenarios is a low bar that doesn't show meaningful information, and obscures the coloring underneath- this is the case in not only Figure 3, which the authors changed, but also Figure 2, (for ERF changes not warming), Figure 5, Figure 8, etc.
Figure 3: what about showing the range of differences in grey shading around the grey line as well? Or does this make the plot too complicated to interpret? Also, as noted above, please change ‘green lines’ to ‘grey lines’ in caption here and elsewhere where necessary.
Figure 5: I suggest adding units on colorbar label, or add in units in the figure caption as in earlier figures- the labels in the upper right are hard to find/read
Figures 6,9,10: Why are country/political boundaries only shown in Figure 5 and 8, and not in other figures with maps (such as maps in Figs 6,9,10)? Can the authors please be more consistent with showing these borders for reader orientation unless there is a specific reason to exclude them?
Figure 6, 7 and other graphs with y axes that cross zero: here and in other figures, a horizontal reference line marking zero on the y axis can be helpful.
Figure 11, Figure 12: legend has an error- ‘climte’ |
