Middle east warming in spring enhances summer rainfall over Pakistan

The edge of a monsoon region is usually highly sensitive to climate change. Pakistan, which is located on the northern edge of the Indian monsoon, is highly vulnerable to heavy rainfall and has witnessed several debilitating floods exacerbated by global warming in recent years. However, the mechanisms for the frequent Pakistan floods are yet not fully understood. Here, we show that the Middle East is undergoing an increase in land heating during spring, which is responsible for 46% of the intensified rainfall over Pakistan and northwestern India during 1979–2022. This springtime land warming causes a decline in sea level pressure (SLP), which strengthens the meridional SLP gradient between the Middle East and the southern Arabian Sea and drives the changes of low-level jet (LLJ) subsequently. The impact persists into summer and results in a northward shift of the monsoonal LLJ, accompanied by strong positive vorticity in the atmosphere and enhanced moisture supply to Pakistan. Consequently, the transition region between the summer monsoon in South Asia and the desert climate in West Asia is shifted northwestward, posing significantly enhanced risk of floods over Pakistan and northwestern India.

Role of deep convecfion in regulafing the Indian summer monsoon dynamics: a regional scale modelling study.Meteorol Atmos Phys 134, 84.hftps://doi.org/10.1007/s00703-022-00917-2As a substanfial amount of work is required before publicafion, I have suggested a major revision or rejecfion as per Editor's choice due to fime restricfions.

Manuscript Summary:
This arficle invesfigates the relafionship between summer floods over Pakistan with surface warming over Middle East from 1979 to 2022.It highlights the importance of understanding the possible causes of extreme precipitafion that leads to anomalous summer floods.First, the authors explain the physical processes of moisture transport over the Arabian Sea.It is the low-level jet (LLJ) that is responsible for the channeling the moisture-rich air mass from the Arabian Sea to Pakistan and nearby western India.Over 44 years, the LLJ has been shifting northward and thus bringing more moisture to Pakistan.A stafisfical analysis shows a close relafionship between the surface warming over the Middle East and the locafion of the LLJ.Second, there is also data-driven evidence that the mean rainfall over Pakistan has been increasing from 1979 to 2022.The comparison between the data of the first period and the second period shows this increase quanfitafively.This study is interesfing and suitable for Nature Communicafions.However, I am concerned with the low correlafion coefficient (0.4) between the surface heafing over the Middle East and the low-level jet (LLJ) locafion.My final decision is major revisions specially to address the correlafion issue for this round of review.
Major Comments: 1.The fitle clearly says that the study has something to do with summer floods.In the text however the analysis is more on the increased precipitafion, not floods.I also think there is a jump from rainfall to floods.In reality, there are processes that translate rainfall into floods.This may include for instance the texture (type) of the soil in Pakistan, that maybe has low absorpfion capacity like in many semi-arid regions.I believe it is the case.The soil perhaps similar to that in the Arabian Peninsula or the Southwest US.To avoid misleading, it would be befter to revise the fitle to directly describe what is being discussed in the text.For instance, the fitle is "Middle East Warming in Spring Enhances Summer Rainfall over Pakistan".2. Lines 77 to 79.I am not sure if this sentence is fair to climate models.First, I do not think that flood is one of the variables in many climate models (unless there is a hydrological model that runs climate predicfion).Second, climate model spafial resolufion is generally coarse.Extreme rainfall can only best be captured in high-resolufion model such as a convecfive-permifting model.3. Lines 112 to 114.I am a bit surprised that the correlafion coefficient between the surface heafing over the Middle East and the LLJ locafion is 0.4.I do not think that it is a good correlafion score and I do not think we can claim this a close relafionship.Will this correlafion coeff increase if you use another level like temperature at 850 mb?Or will this be different if you use SLP?I think this is the quanfitafive evidence that mainly supports your argument.4. Lines 124 to 126.I am concerned with the beginning of this sentence ("Due to the net effect of anthropogenic forcing, …").Reader will ask how the authors know that it is the net effect of anthropogenic forcing.If you want to keep this claim, I think you will need another dataset such as CO2 emission and its stafisfical analyses in the period of 1979 to 2022.It would be easy if there is already a published study that show this anthropogenic forcing over the Middle East in that period.
Minor comments: Summary: No comment.

Main Text:
Lines 48, 49, 50: Please change USD to U.S. Dollars (and put it after the billion).Lines 62 to 64: Please add one or two references at the end of the sentence.Line 76: Please add a reference (or two) after "Unlike the wet-get-wefter paradigm".

Results:
Line 85: Please remove the ~ (filde sign) so that it is consistent with the one in Summary.Line 85: Please change to "… increased by 46% or equivalent to 0.37 mm decade-1 (Lines in Fig. 1b) …" Line 97: change per decade to decade-1.Line 108: remove the lafitude and longitude since you have it already in line 104.

Discussion:
Line 166: As menfioned in major comment, I am concerned with the statement "close relafionship".Lines 173 to 174: Please add a reference (or two) for this sentence.Line 184: Is 0.28 mm decade-1 the observed trend?If so, please make it clear.

Methods:
Line 252: Please change qudp to qu dp 1

Responses to Reviewers
Reviewer #1: I appreciate the efforts taken by the authors to find out the reason for the frequent Pakistan flood.However, the present study failed in citing some relevant works and lacking in-depth analysis.

Response:
Thank you.We greatly appreciate the insightful suggestion.The manuscript is revised accordingly.We hope all issues are satisfactorily addressed.

Please find my comments and suggestions below to improve the article 1. The present study links the middle east land warming to the increasing rainfall over
Pakistan and north-west India, which can intensify the floods over Pakistan.The proposed mechanism is the land warming causes a reduction in the surface pressure, which enhances the meridional pressure gradient between the Middle East and the Southern Arabian Sea, causing a poleward shift in the location of the LLJ.The northward shifted LLJ enhances the atmospheric vorticity and moisture supply to north-west India as well as Pakistan and thereby enhances the monsoon rainfall.However, a similar mechanism is already explained by Sandeep and Ajayamohan, (2015).How the mechanism explained in the present study differ from Sandeep and Ajayamohan's, (2015) mechanism?
Response: Sandeep and Ajayamohan (2015) indeed proposed a related mechanism for the northward shift of low-level jet (LLJ), and this paper was cited in our original manuscript.However, there are significant and meaningful differences between the current study and Sandeep and Ajayamohan (2015).We explain these here; 1) We highlight the remarkably rapid warming and the influence of surface warming over the Middle East.Sandeep and Ajayamohan (2015) argued that the LLJ shift was a response to the land-sea temperature contrast between the Southeast Asian landmass and the surrounding Indian Ocean, and we agree with that.In this study, we argue that surface warming in the Middle East is the main cause for the land-sea contrast, as the warming there is twice as significant as in all other regions over the world.
2) The seasons are different in the two studies.Sandeep and Ajayamohan (2015) focused on the summer monsoon, and we show consistent results with them during boreal summer.In addition, we extend the story to spring, i.e., we unveil the seasonal lead relationship between spring land heating and summer rainfall.
The spring heating in the Middle East is a precursor for the summer flooding, which helps to predict climate hazards on the northern edge of Indian monsoon.
For example, the Middle East was unusually warm during the spring of 2023, which has led to anomalously high rainfall over Pakistan and northwest India from June onwards this summer (Fig. A1).
3) The northward shift of the monsoonal LLJ directly leading to the enhancement of Pakistan rainfall is discussed in this study, which was not covered in Sandeep and Ajayamohan (2015).In the revised manuscript, we examine the instabilities over Pakistan and northwest India in more detail and more evidence is provided.
As shown in Fig. A2a-b In summary, in this study, we present a novel and important phenomenon, i.e., the frequent and severe rainfall in the past few decades over Pakistan, which is climatologically an arid to semi-arid region on the northern edge of Indian monsoon.
The whole process originates from the Middle East heating in spring which is likely attributable to global warming and includes the northward shift of LLJ, the increase of moisture transport from the Arabian Sea to land, and the reinforcement of atmospheric instabilities over Pakistan.The entire process of the Middle East warming influence is illustrated in Fig. A4.The mechanisms for the entire chain of processes are examined.Sandeep and Ajayamohan (2015) proposed a mechanism for the northward shift of LLJ.
Our results extend their results significantly while being consistent with theirs, and their conclusions also support our findings.The new findings are further clarified in the revision.m -1 day -1 decade -1 ).The dotted areas are significant at the 95% confidence level using the Students' t-test.2. The location and intensity of LLJ are influenced by the deep convection over India and nearby oceans (Joseph and Sijikumar, 2004;Sijikumar and Aneesh, 2022).
What is the role played by the recent changes in the convection pattern over the Indian land region and the nearby oceanic parts in the observed northward propagation of LLJ?

Response:
We thank the reviewer for the heuristic comment.Joseph and Sijikumar (2004) showed that atmospheric heating by convection over the Bay of Bengal (BoB) is able to accelerate the LLJ.Sijikumar and Aneesh (2022) re-examined the relationship with the regional model experiments, and further pointed out that the lack of convective heating in the Western Ghats and BoB weakened the LLJ over the Peninsular India.We agree that deep convection can alter the LLJ activity.
We Meanwhile, we proceeded to regress the seasonal OLR anomalies onto the LLJ index (Fig. A6).In early summer (May-July), the deep convection is roughly confined to northwest India.Theoretically, convective heating can trigger circulation adjustments and accelerate the LLJ, as reported by Joseph and Sijikumar (2004) and Sijikumar and Aneesh (2022).In this way, the enhanced LLJ can shift northward and cause deep convection over Pakistan during mid-summer (June-August).Accordingly, the deep convection will be pushed further north during later summer (July-September).
Therefore, the positive feedback between the summer LLJ and deep convection is established during the ISM season.
However, it must be recognized that the enhancement of deep convection in Pakistan is a consequence of the northward shift of LLJ.Their relationship only exists in summer, but not at the seasonal lead timescale.This study elicits the main cause of the northward shift of LLJ, that is the spring time interaction between climate change manifest in the Middle East warming and natural variability in terms of the seasonal monsoon evolution.In other words, we have proposed a positive feedback between deep convection and the LLJ that is initiated in spring itself.In all, the reason for the seasonal change of LLJ is mainly demonstrated in the manuscript.Response: We agree that Pakistan floods need sufficient moisture from both Bay of Bengal (BoB) and Arabian Sea.As the reviewer states, the major difference between the two flood events in 2010 and 2022 (Fig. A7), besides the moisture from the BoB, is the more abundant moisture supply from the Arabian Sea in 2022 (Fig. A7b).Therefore, the change in moisture supply from the Arabian Sea is the most important cause of   References:

Response:
We regret the error in the original statement.Our study focuses on the impact of the Middle East warming, rather than the variability between the Middle East warming and summer rainfall over Pakistan.As can be seen from Fig. A9, the relationship between the Middle East and LLJ becomes much stronger if the interannual variability is removed, which denotes that the impact of the Middle East on Pakistan through LLJ is mainly reflected at the lower frequency.but by applying a 7-year moving average.The dotted areas are significant at the 95% confidence level using the Student's t-test.

Line 129: How surface warming can cause ~120-200 hPa Sea level pressure change?
Response: We regret the typo.The unit should be Pa.To add further details, Figure A10 displays that the Middle East warming can decrease the sea level pressure by increasing the sensible heat fluxes.Thank you for your comment, this is explained in detail in the revised manuscript (Line 131).My final decision is major revisions specially to address the correlation issue for this round of review.

Response：
Thank you for your insightful suggestions.We have revised the manuscript accordingly.The correlation coefficient has increased to above 0.8 after a more careful examination, and the specific analysis can be found in the third response below.We hope that all changes are to the satisfaction of the reviewer.
Major Comments: 1.The title clearly says that the study has something to do with summer floods.In the text however the analysis is more on the increased precipitation, not floods.I also think there is a jump from rainfall to floods.In reality, there are processes that translate rainfall into floods.This may include for instance the texture (type) of the soil in Pakistan, that maybe has low absorption capacity like in many semi-arid regions.I believe it is the case.The soil perhaps similar to that in the Arabian Peninsula or the Southwest US.To avoid misleading, it would be better to revise the title to directly describe what is being discussed in the text.For instance, the title is "Middle East Warming in Spring Enhances Summer Rainfall over Pakistan".

Response:
We regret the inaccurate title.The title is revised according to the suggestion of the reviewer.

Response:
The reviewer is correct that flooding is not one of the model outputs.Precipitation is still the variable that we analyze in the model section.We have revised the expressions in the new manuscript, hopefully, avoiding any confusion.

Response:
As the reviewer suggests, we re-consider the definition of the LLJ index.The moisture supply and the dynamic instability for enhanced rainfall over Pakistan are much more tightly associated with the eastern side of LLJ.Therefore, we define a new LLJ index as zonally averaged between 60°E and 75°E.Accordingly, the correlation coefficient of this newly defined LLJ index for spring land heating rises to above 0.7, being significant in the Middle East region of our interest (blue rectangle in Fig. B1a).
In addition, both the new LLJ index and the originally defined index can characterize the northward movement of the LLJ.The correlation coefficient of these two indices reaches 0.9 and it is statistically significant at a 99% confidence level.Moreover, since we focus on the changes on a long timescale, the correlation coefficient is as high as 0.82 after removing the high-frequency interannual signals (Fig. B1b).Overall, we think these boosted quantitative relationships only better support the arguments in this study.Thank you once again.

Climate change and weather extremes in the Eastern Mediterranean and Middle
East.Reviews of Geophysics, 60, e2021RG000762.

Main Text:
Lines 48, 49, 50: Please change USD to U.S. Dollars (and put it after the billion).

Response:
The unit is corrected in the revision.Thank you for your eye to the details.
Lines 62 to 64: Please add one or two references at the end of the sentence.

Response:
We thank the reviewer for the suggestion.It is revised in the new manuscript.

Response:
It is corrected accordingly.
Line 108: remove the latitude and longitude since you have it already in line 104.

Response:
As per your suggestion, they have been removed for clarity.

Discussion:
Line 166: As mentioned in major comment, I am concerned with the statement "close relationship".

Response:
The correlation coefficient has been improved in the revised manuscript.In addition, it is significant at the 99% confidence level using the Student's t-test.Based on the analysis, they have a close statistical relationship with each other.
Lines 173 to 174: Please add a reference (or two) for this sentence.

Response:
This sentence was supposed to summarize the performance of the current numerical models on the simulation of summer precipitation.In the new manuscript, we have repositioned it to follow the analysis in Supplement Figure 5.
Line 184: Is 0.28 mm decade -1 the observed trend?If so, please make it clear.

Response:
Yes, 0.28 mm day -1 decade -1 is the observed trend, which is used here to compare with the simulated intensity.The expression is modified accordingly in the revision.

Methods:
Line 252: Please change qudp to qu dp

Response:
Thank you for your eye to the details.It is revised.

Table A2
Trend in the northward movement in LLJ (unit: o decade -1 ) and Pakistan rainfall (mm day -1 decade -1 ) under three scenarios.R 2 is the square of the correlation coefficient of the time series between LLJ movement and Pakistan rainfall.It denotes the contribution of LLJ movement to the rainfall changes.
3. The changes in the rainfall pattern can be due to the combined effect of anthropogenic forcing and the internal variability of the climate system.Find out the relative contribution of internal variability and the external forcing in the observed changes of Pakistan's summer rainfall.

Response:
Thank you for raising this important issue.The relative contributions of internal variability and the external forcing to the changes in Pakistan summer rainfall are estimated by the multi-models from the CMIP6.The results are included in the revised manuscript  and they are briefly summarized below.
We use historical runs from 20 CMIP6 coupled models.There are 192 members in total (see detailed in estimated by averaging the MMEM of CMIP6 members, whereas the contribution of internal variability is estimated by the difference between the observed trend and the MMEM trend.The uncertainty is evaluated with one standard deviation of the trend in each member.Besides, the linear trend in Pakistan rainfall is calculated for each model simulation and, subsequently, averaged across all ensemble members. As shown in Fig. A2, the summer rainfall in Pakistan has increased in observations (GPCP), the MMEM, and their difference over the time period of 1979-2014.It suggested that that both external forcing and internal variability contribute to the rainfall enhancement in the past decades.Meanwhile, the observed regional mean of the Pakistan rainfall trend is about 0.28 mm day -1 decade -1 (Fig. A4a).The trend is 0.1 mm day -1 decade -1 in MMEM (Fig. A4b), which means that the external forcing contributes 36% to the observed trend.In contrast, the contribution of internal variability to the rainfall trend reaches 64% as the regional averaged trend is 0.18 mm day -1 decade -1 .In addition, the external forcing and internal variability contribute 81% and 19% to the observed spring land heating in the Middle East, respectively.In summary, the relative contributions of internal variability and external forcing are quantified in the revised manuscript.
, the Middle East warming and LLJ northward shift induces a meridional gradient of quasi-geostrophic potential vorticity ( − ) to change signs within the study region.The regressed pattern agrees with the trend in − (Fig. A2c), with a zonal belt of a negative-positive-negative pattern in Pakistan and northwest India.The changes of signs satisfy the necessary condition for barotropic instability.In addition, the difference in moist static energy (MSE) between 1000 and 500 hPa is a good index for static instability.Out results show that the Middle East warming and LLJ shift, cause a positive difference in MSE in the study region (Fig. A3), which enhances the atmospheric instability over Pakistan.Overall, the changes in the instabilities due to the Middle East warming and the LLJ northward shift, favor the onset of heavy rainfall.

Figure
Figure A1 (a) Spring land heating anomalies in 2023 (unit: °C) from the climatological

Figure A2
Figure A2Regression coefficients of the meridional gradient of quasi-geostrophic

Figure A3
Figure A3 Regression coefficients of the vertical difference of moist static energy

Figure A4
Figure A4 Sketch for the impact of the spring Middle East warming on summer rainfall have also previously analyzed the relationship between LLJ and convection under climate change.In our previous study (Li et al. 2022), we revealed that the deep convection starts to move deeper into the subtropics in comparison with the climatological state (Fig. A5).It leads to the enhancement in convective rainfall over Pakistan and northwest India.Through the diagnosis of the moisture budget, we proposed that the northward shift of the LLJ is responsible for the deep convection movement in that study.The specific physical process can be found in Li et al. (2022).

Figure A5
Figure A5 Trends in intraseasonal anomalies during the summer monsoon season over extreme flooding in Pakistan, which is closely related to the LLJ shift.It inspired us to examine the changes in LLJ in recent years.In addition, the LLJ is accompanied by the strong westerly wind and pronounced positive vorticity, which could provide a track for the low-pressure systems (LPSs) to propagate (e.g., Krishnamurthy and Ajayamohan, 2010; You and Tian et al. 2021).Therefore, the northward-shifted LLJ favors the invasion of LPSs into Pakistan, and transport some moisture from the BoB to further nourish the floods.It has been further discussed on Lines 150-152 in the revised manuscript.

Figure A7
Figure A7Anomalies in precipitation (shading) and horizontal winds at 850 hPa

Figure
Figure A8 (a) The seasonal mean geopotential heights in 200 hPa during JJAS of 2010.

Figure
Figure A9 (a) Regression coefficients of the spring skin temperature (unit: °C) onto the

Figure A10
Figure A10 Regression coefficients of the spring sensible heat flux (unit: W m -2 ) onto

Figure
Figure B1 (a) Correlation coefficients of the spring skin temperature (unit: °C) onto 4 Eyring et al. (2021)  argued that the pronounced warming is related to anthropogenic activities and is mostly driven by elevated global concentrations of greenhouse gases in the atmosphere.IPCC (2021) reported that there is an increase in land warming in the Middle East and Central Asia, with high confidence in human contribution to the observed change.Furthermore, Zittis et al. (2022) pointed out that the warming trends are projected to continue and intensify depending on future tendencies of greenhouse gas concentrations which, in turn, are subject to societal and technological developments.In all, the relevant references are cited here to support the claim, as Line 127 in the revision.

Figure A2
Figure A2 Trends in (a) Pakistan summer rainfall (unit: mm day -1 decade -1 ) and (b) 77 to 79.I am not sure if this sentence is fair to climate models.First, I do not think that flood is one of the variables in many climate models (unless there is a hydrological model that runs climate prediction).Second, climate model spatial resolution is generally coarse.Extreme rainfall can only best be captured in highresolution model such as a convective-permitting model.
. Lines 124 to 126.I am concerned with the beginning of this sentence ("Due to the net effect of anthropogenic forcing, …").Reader will ask how the authors know that . Human influence on the climate system: Contribution of working group I to the sixth assessment report of the Intergovernmental panel on climate change.IPCC Sixth Assessment Report.
Table.A1).The contribution of external forcing to rainfall is