Water Storage Variations in Tibet from GRACE, ICESat, and Hydrological Data
Round 1
Reviewer 1 Report
This work estimated changes in total water budget in the Tibetan Plateau between 2002 and 2017. Honestly, there is no originality/novelty of this research to me. Lots of researches have been published by using same (or at least similar) data for the same region. Neither new data nor new methods were applied in this manuscript. The authors did do a fair amount of work analyzing different types of data to demonstrate a conclusion that “the integral downtrend of the total water budget at the Tibet Plateau caused by melting of glaciers prevails over the investigated period”, which is not convincing to me. My reasons are given as follow.
1) The authors claimed that they applied a “forward modeling technique” to reduce signal leak. To me, the difference between Fig 3a and Fig 3b (without and with forward modeling applied) is more like with and without filtering applied. And it’s getting more suspicious by comparing the time series in Fig 5, as I do not think applying “forward modeling” will make the signal noisier. So I urge the authors to give a further demonstration about their “forward modeling technique” they applied. A perfect case to demonstrate would be apply it to the “ground water depletion” signal in Northwest India (72o ~ 80oE, 28o ~ 32oN) in Fig 3, to see if the method can really recover the leaked signal. Otherwise, I can’t be convinced that the authors really applied an “forward modeling technique”.
2) If there is not an appropriate “leakage reduction” method applied, we have enough reason to question the decreasing trend in “Indus”, “Salween” and “Brahmaputra” basins were caused by glacier/ice melting caused by increased temperature as the authors claimed. The decreasing trend in “Indus” could be just the leakage from the “groundwater depletion” in Northwest India. And by plotting the time series for the block (88o ~ 92oE, 25o ~ 28oN) which is not part of Himalayas, and has no glacier covered, you will see the same pattern as of “Brahmaputra” and “Salween” basins. A reasonable explanation for this decreasing trend over those regions would be decreased precipitation in some monsoon season. Without this part well explained, but considering the decreasing mass purely as glacier/ice melting caused by increased temperature would be wrong.
3) In Line 380 Page 13, “the surface temperature was increasing at the annual rates up to 0.9 degree in the western part of Himalayas” and in Line 381 “a similar increasing trend in the surface temperature at the annual rate of 0.6 to 0.8 degree was also detected”. Have the authors thought about how scary it is that the temperature in Himalayas increases like they estimated, 0.9o/yr * 15 yrs = 13.5o ?
Besides the major concerns and questions I showed above, the English is really poor for being published, the authors should have the manuscript proofread by native English speakers. Moreover, all the citations listed should have consistent format. Overall, I recommend to reconsider after major revisions.
Comments for author File: 
Comments.pdf
Author Response
This work estimated changes in total water budget in the Tibetan Plateau between 2002 and 2017. Honestly, there is no originality/novelty of this research to me. Lots of researches have been published by using same (or at least similar) data for the same region. Neither new data nor new methods were applied in this manuscript. The authors did do a fair amount of work analyzing different types of data to demonstrate a conclusion that “the integral downtrend of the total water budget at the Tibet Plateau caused by melting of glaciers prevails over the investigated period”, which is not convincing to me. My reasons are given as follow.
1) The authors claimed that they applied a “forward modeling technique” to reduce signal leak. To me, the difference between Fig 3a and Fig 3b (without and with forward modeling applied) is more like with and without filtering applied. And it’s getting more suspicious by comparing the time series in Fig 5, as I do not think applying “forward modeling” will make the signal noisier. So I urge the authors to give a further demonstration about their “forward modeling technique” they applied. A perfect case to demonstrate would be apply it to the “ground water depletion” signal in Northwest India (72o ~ 80oE, 28o ~ 32oN) in Fig 3, to see if the method can really recover the leaked signal. Otherwise, I can’t be convinced that the authors really applied an “forward modeling technique”.
We now applied a forward modelling over the region with large negative signal in northwest India. Results and short discussion is given in Section 4.1.2 signal leakage. We also added some explanatory text in Abstract. In attached document, we provide a more detailed review of the forward modelling technique including results.
2) If there is not an appropriate “leakage reduction” method applied, we have enough reason to question the decreasing trend in “Indus”, “Salween” and “Brahmaputra” basins were caused by glacier/ice melting caused by increased temperature as the authors claimed. The decreasing trend in “Indus” could be just the leakage from the “groundwater depletion” in Northwest India. And by plotting the time series for the block (88o ~ 92oE, 25o ~ 28oN) which is not part of Himalayas, and has no glacier covered, you will see the same pattern as of “Brahmaputra” and “Salween” basins. A reasonable explanation for this decreasing trend over those regions would be decreased precipitation in some monsoon season. Without this part well explained, but considering the decreasing mass purely as glacier/ice melting caused by increased temperature would be wrong.
We also applied a forward modelling to this region with expected decreasing precipitation in the western Himalayas. Results and short discussion for this region as well as the region situated in northwest India is given in Section 4.1.2.
3) In Line 380 Page 13, “the surface temperature was increasing at the annual rates up to 0.9 degree in the western part of Himalayas” and in Line 381 “a similar increasing trend in the surface temperature at the annual rate of 0.6 to 0.8 degree was also detected”. Have the authors thought about how scary it is that the temperature in Himalayas increases like they estimated, 0.9o/yr * 15 yrs = 13.5o ?
Such large values are provided by GLDAS surface and air temperature datasets. We provide a detailed report in attached document. In the latest version or article, we used the air temperature data provided by China National Meteorological Information Center.
The whole text was checked and corrected by a native speaker from England who is a member of editorial board to this journal.
Besides the major concerns and questions I showed above, the English is really poor for being published, the authors should have the manuscript proofread by native English speakers.
Moreover, all the citations listed should have consistent format.
Citations were reformated.
Overall, I recommend to reconsider after major revisions.
Author Response File: Author Response.pdf
Reviewer 2 Report
I would like to suggest to accept the manuscript based on the revisions.
Author Response
Thank you.
Reviewer 3 Report
You improved the manuscript quite a lot. The discussion is short but should be OK.
Some minor comments:
rate instead of trend
l29: delete 'typically'
l30: lakes occured.
Author Response
Thank you. We corrected the sentence as: "Both, GRACE and ICESat results confirm a substantial retreat of mountain glaciers, while an increasing trend of major lakes occurred. "
Round 2
Reviewer 1 Report
Again, the so-called "forward modeling techniques" has no contribution to reducing signal leak at all in TP area, or more generally to inland areas. It might work for Greenland and Antarctica because those regions are surrounded by oceans, but not for reducing leak among regions on land. So the authors should not mislead readers that their estimates over TP had been corrected for signal leak from outside basins. If the authors have no capability to do a real "signal leak reduction", at least they need to make it clear to the readers that the "forward modelling techniques" they applied is solely reducing overall leak between land and ocean, but not among basins on land. Frankly, this kind of research has really low merit, I intended to reject this article. However, I realized that this is the work from a Ph.D. student or a fresh Ph.D. graduate, I still want to encourage young researchers to do some solid research work by agreeing to accept after minor revision.
Author Response
We thank reviewer for his critical view. In the updated version of the article we now clearly state that a forward modelling technique suppress overall leak between land and ocean, but cannot mitigate signal leakage among different basins on land.
All relevant changes in text are indicated in red.
