the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
ERA5-based database of Atmospheric Rivers over Himalayas
Abstract. Atmospheric Rivers (ARs) – long and narrow transient corridors of large horizontal moisture flux in the lower troposphere – are known to shape the hydrology of many regions around the globe. Heavy precipitation and flooding are often observed over many mountainous regions when the moisture-rich filaments impinge upon the elevated topographies. Although ARs and their impacts over many mountainous regions are well documented, their existence over the Himalayas and importance to the Himalayan hydrology have received negligible attention in the scientific literature. The Himalayas support more than a billion population in the Indian subcontinent, sustain the region's biodiversity, and play important roles in regulating the global climate.
In this study, we develop a comprehensive database of ARs over the Himalayas using the European Reanalysis fifth-generation (ERA5) fields of humidity and winds. The AR database consists of the dates and times of ARs from 1982 to 2018, their duration, major axes, and intensities and categories. We find that majority of intense ARs are associated with extreme precipitation widespread over the Ganga and Indus basins of the Himalayas, suggesting that ARs have profound impacts on the hydrology of the region. The AR database developed here is envisioned to help in exploring the impacts of ARs on the hydrology and ecology of the Himalayas. For this, we provide a few brief future perspectives on AR-Himalayas relationships.
The data developed in this study has been uploaded to the Zenodo repository at https://doi.org/10.5281/zenodo.4451901 (Nayak et al., 2021). The data is also included in the Supplemental Information for easier access.
- Preprint
(1382 KB) - Metadata XML
-
Supplement
(5301 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2020-397', Anonymous Referee #1, 11 Mar 2021
Review on ‘ERA5-based database of Atmospheric Rivers over Himalayas’ by Nayak et al.,
General comments:
The manuscript is well written and provides an interesting topic of detecting atmospheric rivers (ARs) over Himalayas. However, I think the manuscript is more suitable for a climate research journal (e.g. Journal of climate, International journal of climatology) rather than a data journal. Because Atmospheric river is a character of water vapor transport in the atmosphere, like an index, can be calculated from different data (reanalysis, simulations) and difficult to evaluate the accuracy. And data quality is one of the key standard of the current journal.
This work calculates the ARs only based on ERA5, i would like to know how it differs if calculated based on other reanalysis data, e.g. MERRA, NCEP, JRA. Do they got the similar results? Which is the best? Such questions need to be answered if the data aims to practical use.
Further, ERA5 has hourly resolution, why you use six hourly data? This has disadvantages: 1, it will reduce the accuracy of detecting durations of ARs. 2, if you use instantaneous wind speed to calculate u*q, four times a day could largely differs from 24 times a day, because the atmospheric vapor transport has strong diurnal cycle, especially during monsoon season.
There are many methods to calculate ARs, which will lead to multiple different results when calculating ARs. The method in your manuscripts itself also has some empirical treatments. For example, line 222, why you use 15 days moving average instead of 10 or 20? Is it physical mechanism dependent? The methods and the based data source selected could lead to large uncertainties in the ARs data.
Why is the threshold of IVT set to different values for different seasons? This will has disadvantages for practical application during disaster research. In such circumstance, for example, you will likely exclude some ARs that will leads to flood in wet season but include some ARs that may not leads to flood in dry season.
typing error
Line 31:
Driver-like => should it be ‘river-like’
Citation: https://doi.org/10.5194/essd-2020-397-RC1 - AC5: 'Reply on RC1', Munir Ahmad Nayak, 30 May 2021
-
RC2: 'Comment on essd-2020-397', Anonymous Referee #2, 23 Apr 2021
General Comments
The manuscript is well written and the concept for this paper is well thought out. The length and structure of the article are appropriate. High Mountain Asia could certainly benefit from more AR analysis due to its unique topography. However, I do not think that this manuscript is ready for publication in its current form. I have several suggestions for improvement of this paper (see below).
Specific Comments
It is unclear how this AR detection algorithm is unique compared to other AR detection algorithms available for Southern Asia. I agree with Reviewer 1 that spatio-temporal availability of ERA5 data could be leveraged for an improved detection algorithm (i.e. 1-hourly, 0.1° horizontal resolution) in this region. At the very least, the authors could comment on why they chose 6-hourly and 0.25° horizontal resolution.
The authors do note that there are many algorithms available that identify ARs (lines 207-208) but do not employ any comparison with their algorithm and others that are available. For example, other AR detection algorithms on a global, 6-hourly basis (e.g. Guan and Waliser, 2015; Guan et al, 2018; Guan and Waliser, 2019; Sellars et al, 2017) are freely available to the public and could be used for statistical evaluation. This would also give the authors the chance to give error estimates for their data set. Table 1 in Rutz et al (2019) would be a good place to look for available AR detection algorithms in HMA. This would improve the article greatly as it would give the authors a chance to show how novel their algorithm is and why the AR community needs yet another AR detection algorithm. Unless the authors can show that this detection algorithm is better suited for HMA compared to other available AR detection algorithms, this study does not significantly contribute to the current body of work.
I would also recommend the author review the ARTMIP articles that complete an in-depth comparison of most of the available AR detection algorithms (Shields et al, 2018; Rutz et al, 2019; Lora et al, 2020) and elaborate on why they chose to emulate the Lavers et al (2012) method over others. For example, why is this method more appropriate for HMA?
The author briefly mentions the AR study over the Bay of Bengal (Yang et al, 2018) in the results section, but it is not mentioned in the introduction paragraph where the author discusses other studies that examine ARs in Southern Asia (lines 102-118).
The readme for the AR track data seems incomplete. I’m not sure the data set would be able to be easily understood and re-used in the future. For example, what do all the columns mean in each of the files? Is there a unique ID for each of the AR tracks, or would a potential user have to join the tables on multiple columns? I would suggest clarification in the readme that describes the columns to prevent misuse of this database in the future.
Technical Corrections
Line 345-346: The sentence beginning with “The minimum” is confusing to read and should be rewritten for clarity.
Line 287 (and others) The formatting of ðð ð−1 ð −1 is off. For example, there does not appear to be a space between kg and m. This occurs in the supplemental material as well.
The folder containing the Supplemental materials is misspelled as “Sumplimetary_Information”.
References
Guan B, Waliser D (2015) Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. Journal of Geophysical Research: Atmospheres 120(24):12,514–12,535
Guan B, Waliser DE (2019) Tracking atmospheric rivers globally: spatial distributions and temporal evolution of life cycle characteristics. Journal of Geophysical Research: Atmospheres 124(23):12,523–12,552
Guan B, Waliser DE, Ralph FM (2018) An intercomparison between reanalysis and dropsonde observations of the total water vapor transport in individual atmospheric rivers. Journal of Hydrometeorology 19(2):321–337
Lavers DA, Villarini G, Allan RP, Wood EF, Wade AJ (2012) The detection of atmospheric rivers in atmospheric reanalyses and their links to british winter floods and the large-scale climatic circulation. Journal of Geophysical Research: Atmospheres 117(D20), DOI 10.1029/2012JD018027
Lora JM, Shields C, Rutz J (2020) Consensus and disagreement in atmospheric river detection: Artmip global catalogues. Geophysical Research Letters 47(20):e2020GL089,302, DOI 10.1029/2020GL089302
Rutz JJ, Shields CA, Lora JM, Payne AE, Guan B, Ullrich P, O’Brien T, Leung LR, Ralph FM, Wehner M, et al (2019) The atmospheric river tracking method intercomparison project (artmip): quantifying uncertainties in atmospheric river climatology. Journal of Geophysical Research: Atmospheres 124(24):13,777– 13,802, DOI 10.1029/2019JD030936
Sellars S, Kawzenuk B, Nguyen P, Ralph F, Sorooshian S (2017) Genesis, pathways, and terminations of intense global water vapor transport in association with large-scale climate patterns. Geophysical Research Letters 44(24):12–465, DOI 10.1002/2017GL075495
Shields CA, Rutz JJ, Leung LY, Ralph FM, Wehner M, Kawzenuk B, Lora JM, McClenny E, Osborne T, Payne AE, et al (2018) Atmospheric river tracking method intercomparison project (artmip): project goals and experimental design. Geoscientific Model Development 11(6):2455–2474
Yang Y, Zhao T, Ni G, Sun T (2018) Atmospheric rivers over the bay of bengal lead to northern indian extreme rainfall. International Journal of Climatology 38(2):1010–1021, DOI 10.1002/joc.5229
Citation: https://doi.org/10.5194/essd-2020-397-RC2 -
RC3: 'Reply on RC2', Anonymous Referee #2, 29 Apr 2021
Two corrections to the previous review:
1) ERA5 has a horizontal resolution of 0.25°, not 0.1°. So the author should only comment on their choice to use 6-hourly compared to 3-hourly or even hourly temporal resolution.
2) The technical correction for Line 287 (and others) should read: The formatting of kg m-1 s-1 is off in the manuscript. For example, there does not appear to be a space between kg and m. This occurs in the supplemental material as well.
Citation: https://doi.org/10.5194/essd-2020-397-RC3 -
AC4: 'Reply on RC3', Munir Ahmad Nayak, 28 May 2021
Thank you. Please see our responses to your previous comment.
Citation: https://doi.org/10.5194/essd-2020-397-AC4
-
AC4: 'Reply on RC3', Munir Ahmad Nayak, 28 May 2021
- AC3: 'Reply on RC2', Munir Ahmad Nayak, 28 May 2021
-
RC3: 'Reply on RC2', Anonymous Referee #2, 29 Apr 2021
-
RC4: 'Comment on essd-2020-397', Anonymous Referee #3, 04 May 2021
The authors created a dataset of atmospheric rivers for the Himalayan region derived from ERA5 data. This dataset could be useful for the community for research into extreme precipitation and flooding.The manuscript is well written and presented.
I think the manuscript could benefit from a bit more explanation on the choice of the AR detection algorithm and the chosen time step. Why did the authors decide to use 6-hourly data despite ERA5 being available on a higher temporal resolution? What made the authors choose this AR identification method over other available methods?
When looking into the dataset I think there could be a bit more additional information on how the data is organised. I am not sure that someone downloading the dataset would be able to understand it in its current form. For example, it took me a while to figure out that a detected AR has a unique id but still has multiple rows as it consists of multiple timesteps. The description in the read me file is very short and could say more about the structure in the .csv files, e.g. that there is a line for every time step in an identified AR. The manuscript and meta data say that the covered period is 1982-2018 while the first detected AR in the files is from January 1979. For one AR timestep the IVT max says one value but when looking into the columns there is a higher IVT value. It seems a bit complicated organised that the longitudes and latitudes corresponding to the AR locations are in different files from the actual IVT values.
Line 246-247: ther is "southward" twice in this sentence, while I think one of them should be "eastward".
Citation: https://doi.org/10.5194/essd-2020-397-RC4 - AC1: 'Reply on RC4', Munir Ahmad Nayak, 27 May 2021
-
AC2: 'Reply on RC4', Munir Ahmad Nayak, 28 May 2021
We have also added more information in the revised manuscript (Data Section) to justify the choice of 6-hour temporal resolution. The text there reads as:
"The 6-hourly interval is chosen for four main reasons 1.) it is a common denominator among AR detection algorithms using atmospheric reanalysis datasets (Brands et al., 2017; Bin Guan & Waliser, 2015; Mundhenk et al., 2016; Rutz et al., 2014), 2) it provides sufficient temporal information on AR events and captures the gradual changes of AR characteristics (Nash & Carvalho, 2020; Ramos et al., 2015), 3.) many studies have found minor differences in ARs based on differing the temporal resolutions (Guan & Waliser, 2015, 2017; Rutz et al., 2014; Shields et al., 2018), and 4.) as compared to 1-hourly data, it is easily-manageable on a desktop machine with small random access memory (RAM), while marginally compromising on the extent of information available on AR characteristics."
Citation: https://doi.org/10.5194/essd-2020-397-AC2
Status: closed
-
RC1: 'Comment on essd-2020-397', Anonymous Referee #1, 11 Mar 2021
Review on ‘ERA5-based database of Atmospheric Rivers over Himalayas’ by Nayak et al.,
General comments:
The manuscript is well written and provides an interesting topic of detecting atmospheric rivers (ARs) over Himalayas. However, I think the manuscript is more suitable for a climate research journal (e.g. Journal of climate, International journal of climatology) rather than a data journal. Because Atmospheric river is a character of water vapor transport in the atmosphere, like an index, can be calculated from different data (reanalysis, simulations) and difficult to evaluate the accuracy. And data quality is one of the key standard of the current journal.
This work calculates the ARs only based on ERA5, i would like to know how it differs if calculated based on other reanalysis data, e.g. MERRA, NCEP, JRA. Do they got the similar results? Which is the best? Such questions need to be answered if the data aims to practical use.
Further, ERA5 has hourly resolution, why you use six hourly data? This has disadvantages: 1, it will reduce the accuracy of detecting durations of ARs. 2, if you use instantaneous wind speed to calculate u*q, four times a day could largely differs from 24 times a day, because the atmospheric vapor transport has strong diurnal cycle, especially during monsoon season.
There are many methods to calculate ARs, which will lead to multiple different results when calculating ARs. The method in your manuscripts itself also has some empirical treatments. For example, line 222, why you use 15 days moving average instead of 10 or 20? Is it physical mechanism dependent? The methods and the based data source selected could lead to large uncertainties in the ARs data.
Why is the threshold of IVT set to different values for different seasons? This will has disadvantages for practical application during disaster research. In such circumstance, for example, you will likely exclude some ARs that will leads to flood in wet season but include some ARs that may not leads to flood in dry season.
typing error
Line 31:
Driver-like => should it be ‘river-like’
Citation: https://doi.org/10.5194/essd-2020-397-RC1 - AC5: 'Reply on RC1', Munir Ahmad Nayak, 30 May 2021
-
RC2: 'Comment on essd-2020-397', Anonymous Referee #2, 23 Apr 2021
General Comments
The manuscript is well written and the concept for this paper is well thought out. The length and structure of the article are appropriate. High Mountain Asia could certainly benefit from more AR analysis due to its unique topography. However, I do not think that this manuscript is ready for publication in its current form. I have several suggestions for improvement of this paper (see below).
Specific Comments
It is unclear how this AR detection algorithm is unique compared to other AR detection algorithms available for Southern Asia. I agree with Reviewer 1 that spatio-temporal availability of ERA5 data could be leveraged for an improved detection algorithm (i.e. 1-hourly, 0.1° horizontal resolution) in this region. At the very least, the authors could comment on why they chose 6-hourly and 0.25° horizontal resolution.
The authors do note that there are many algorithms available that identify ARs (lines 207-208) but do not employ any comparison with their algorithm and others that are available. For example, other AR detection algorithms on a global, 6-hourly basis (e.g. Guan and Waliser, 2015; Guan et al, 2018; Guan and Waliser, 2019; Sellars et al, 2017) are freely available to the public and could be used for statistical evaluation. This would also give the authors the chance to give error estimates for their data set. Table 1 in Rutz et al (2019) would be a good place to look for available AR detection algorithms in HMA. This would improve the article greatly as it would give the authors a chance to show how novel their algorithm is and why the AR community needs yet another AR detection algorithm. Unless the authors can show that this detection algorithm is better suited for HMA compared to other available AR detection algorithms, this study does not significantly contribute to the current body of work.
I would also recommend the author review the ARTMIP articles that complete an in-depth comparison of most of the available AR detection algorithms (Shields et al, 2018; Rutz et al, 2019; Lora et al, 2020) and elaborate on why they chose to emulate the Lavers et al (2012) method over others. For example, why is this method more appropriate for HMA?
The author briefly mentions the AR study over the Bay of Bengal (Yang et al, 2018) in the results section, but it is not mentioned in the introduction paragraph where the author discusses other studies that examine ARs in Southern Asia (lines 102-118).
The readme for the AR track data seems incomplete. I’m not sure the data set would be able to be easily understood and re-used in the future. For example, what do all the columns mean in each of the files? Is there a unique ID for each of the AR tracks, or would a potential user have to join the tables on multiple columns? I would suggest clarification in the readme that describes the columns to prevent misuse of this database in the future.
Technical Corrections
Line 345-346: The sentence beginning with “The minimum” is confusing to read and should be rewritten for clarity.
Line 287 (and others) The formatting of ðð ð−1 ð −1 is off. For example, there does not appear to be a space between kg and m. This occurs in the supplemental material as well.
The folder containing the Supplemental materials is misspelled as “Sumplimetary_Information”.
References
Guan B, Waliser D (2015) Detection of atmospheric rivers: Evaluation and application of an algorithm for global studies. Journal of Geophysical Research: Atmospheres 120(24):12,514–12,535
Guan B, Waliser DE (2019) Tracking atmospheric rivers globally: spatial distributions and temporal evolution of life cycle characteristics. Journal of Geophysical Research: Atmospheres 124(23):12,523–12,552
Guan B, Waliser DE, Ralph FM (2018) An intercomparison between reanalysis and dropsonde observations of the total water vapor transport in individual atmospheric rivers. Journal of Hydrometeorology 19(2):321–337
Lavers DA, Villarini G, Allan RP, Wood EF, Wade AJ (2012) The detection of atmospheric rivers in atmospheric reanalyses and their links to british winter floods and the large-scale climatic circulation. Journal of Geophysical Research: Atmospheres 117(D20), DOI 10.1029/2012JD018027
Lora JM, Shields C, Rutz J (2020) Consensus and disagreement in atmospheric river detection: Artmip global catalogues. Geophysical Research Letters 47(20):e2020GL089,302, DOI 10.1029/2020GL089302
Rutz JJ, Shields CA, Lora JM, Payne AE, Guan B, Ullrich P, O’Brien T, Leung LR, Ralph FM, Wehner M, et al (2019) The atmospheric river tracking method intercomparison project (artmip): quantifying uncertainties in atmospheric river climatology. Journal of Geophysical Research: Atmospheres 124(24):13,777– 13,802, DOI 10.1029/2019JD030936
Sellars S, Kawzenuk B, Nguyen P, Ralph F, Sorooshian S (2017) Genesis, pathways, and terminations of intense global water vapor transport in association with large-scale climate patterns. Geophysical Research Letters 44(24):12–465, DOI 10.1002/2017GL075495
Shields CA, Rutz JJ, Leung LY, Ralph FM, Wehner M, Kawzenuk B, Lora JM, McClenny E, Osborne T, Payne AE, et al (2018) Atmospheric river tracking method intercomparison project (artmip): project goals and experimental design. Geoscientific Model Development 11(6):2455–2474
Yang Y, Zhao T, Ni G, Sun T (2018) Atmospheric rivers over the bay of bengal lead to northern indian extreme rainfall. International Journal of Climatology 38(2):1010–1021, DOI 10.1002/joc.5229
Citation: https://doi.org/10.5194/essd-2020-397-RC2 -
RC3: 'Reply on RC2', Anonymous Referee #2, 29 Apr 2021
Two corrections to the previous review:
1) ERA5 has a horizontal resolution of 0.25°, not 0.1°. So the author should only comment on their choice to use 6-hourly compared to 3-hourly or even hourly temporal resolution.
2) The technical correction for Line 287 (and others) should read: The formatting of kg m-1 s-1 is off in the manuscript. For example, there does not appear to be a space between kg and m. This occurs in the supplemental material as well.
Citation: https://doi.org/10.5194/essd-2020-397-RC3 -
AC4: 'Reply on RC3', Munir Ahmad Nayak, 28 May 2021
Thank you. Please see our responses to your previous comment.
Citation: https://doi.org/10.5194/essd-2020-397-AC4
-
AC4: 'Reply on RC3', Munir Ahmad Nayak, 28 May 2021
- AC3: 'Reply on RC2', Munir Ahmad Nayak, 28 May 2021
-
RC3: 'Reply on RC2', Anonymous Referee #2, 29 Apr 2021
-
RC4: 'Comment on essd-2020-397', Anonymous Referee #3, 04 May 2021
The authors created a dataset of atmospheric rivers for the Himalayan region derived from ERA5 data. This dataset could be useful for the community for research into extreme precipitation and flooding.The manuscript is well written and presented.
I think the manuscript could benefit from a bit more explanation on the choice of the AR detection algorithm and the chosen time step. Why did the authors decide to use 6-hourly data despite ERA5 being available on a higher temporal resolution? What made the authors choose this AR identification method over other available methods?
When looking into the dataset I think there could be a bit more additional information on how the data is organised. I am not sure that someone downloading the dataset would be able to understand it in its current form. For example, it took me a while to figure out that a detected AR has a unique id but still has multiple rows as it consists of multiple timesteps. The description in the read me file is very short and could say more about the structure in the .csv files, e.g. that there is a line for every time step in an identified AR. The manuscript and meta data say that the covered period is 1982-2018 while the first detected AR in the files is from January 1979. For one AR timestep the IVT max says one value but when looking into the columns there is a higher IVT value. It seems a bit complicated organised that the longitudes and latitudes corresponding to the AR locations are in different files from the actual IVT values.
Line 246-247: ther is "southward" twice in this sentence, while I think one of them should be "eastward".
Citation: https://doi.org/10.5194/essd-2020-397-RC4 - AC1: 'Reply on RC4', Munir Ahmad Nayak, 27 May 2021
-
AC2: 'Reply on RC4', Munir Ahmad Nayak, 28 May 2021
We have also added more information in the revised manuscript (Data Section) to justify the choice of 6-hour temporal resolution. The text there reads as:
"The 6-hourly interval is chosen for four main reasons 1.) it is a common denominator among AR detection algorithms using atmospheric reanalysis datasets (Brands et al., 2017; Bin Guan & Waliser, 2015; Mundhenk et al., 2016; Rutz et al., 2014), 2) it provides sufficient temporal information on AR events and captures the gradual changes of AR characteristics (Nash & Carvalho, 2020; Ramos et al., 2015), 3.) many studies have found minor differences in ARs based on differing the temporal resolutions (Guan & Waliser, 2015, 2017; Rutz et al., 2014; Shields et al., 2018), and 4.) as compared to 1-hourly data, it is easily-manageable on a desktop machine with small random access memory (RAM), while marginally compromising on the extent of information available on AR characteristics."
Citation: https://doi.org/10.5194/essd-2020-397-AC2
Data sets
Atmospheric River Database for the Himalayas Nayak, M. A., Azam, M. F, and Vellosa, R. L. https://doi.org/10.5281/zenodo.4451901
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
1,496 | 902 | 95 | 2,493 | 115 | 75 | 74 |
- HTML: 1,496
- PDF: 902
- XML: 95
- Total: 2,493
- Supplement: 115
- BibTeX: 75
- EndNote: 74
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1