Improving the AMSR-E/NASA Soil Moisture Data Product Using In-Situ Measurements from the Tibetan Plateau
Round 1
Reviewer 1 Report
Dear authors,
thank you for considering my comments; I can see that you put a lot of effort in the revision of the manuscript.
There are still some points that I would like to be clarified:
Fig. 7 plus accompanying text: my suggestion to scale the AMSR-E data to the in situ SM was not carried out. Can you explain why not?
l441: I guess my comment wasn't clearly formulated. To my understanding you only have one shortcoming here, since the high RMSE values are directly caused by the low dynamic range. So, I don't see why you are talking of TWO, not ONE shortcoming. Please explain this again.
l517: To me it looks as if SM was highest in July. Precipitation is highest in June. So, it seems to me that the SM patterns and intensities do not match precipitation. Do you have an explanation for this? In general, please add some more discussion on the spatial patterns.
Fig. 9: Please consider changing the colors for the precipitation maps (blue for zero rainfall is not very intuitive).
In general: there are some spelling/formulation and formatting errors throughout the manuscript; e.g., line 398 "that the, , SM content". Please check this again.
Thank you for considering my comments!
Author Response
Dear reviewer, thank you very much for your comments. These are our detailed responses to comments:
There are still some points that I would like to be clarified:
Fig. 7 plus accompanying text: my suggestion to scale the AMSR-E data to the in situ SM was not carried out. Can you explain why not?
Authors’ response: Thanks very much for your comment. To our understanding the core of your previous comment was to suggest that the NASA and / or the JAXA SM data might capture correctly the dynamics of actual SM. We have illustrated this by adding time series of the relative SM anomaly(deviation from the mean divided by the mean, see Eq.16) in Fig.7 , which shows that indeed the JAXA SM data capture correctly the dynamics of in-situ SM, while the NASA SM data fail to do so. We have also added the mean-std. dev values of the NASA and JAXA SM data in Table.2.
l441: I guess my comment wasn't clearly formulated. To my understanding you only have one shortcoming here, since the high RMSE values are directly caused by the low dynamic range. So, I don't see why you are talking of TWO, not ONE shortcoming. Please explain this again.
Authors’ response: Thanks very much for your comment. The high RMSE values of AMSR-E/NASA are not only caused by the small dynamic range of AMSR-E/NASA soil moisture, since the RMSE might have been smaller, if the NASA SM would have been close to either high or low in-situ SM, The high RMSE shows that the accuracy is poor, i.e. RMSE=0.10 cm3cm-3, because the NASA SM is different from the in-situ SM throughout the year. So, we think the issue is two-fold.
To explain it more clearly, at L 441 we added the sentences “The high RMSE values of AMSR-E/NASA are not only caused by the small dynamic range of AMSR-E/NASA soil moisture. The RMSE might have been smaller if the NASA SM would have been close to either high or low in-situ SM. The high RMSE shows that the accuracy is poor, i.e. RMSE=0.16 cm3cm-3, because the NASA SM is different from the in-situ SM throughout the year. ”
l517: To me it looks as if SM was highest in July. Precipitation is highest in June. So, it seems to me that the SM patterns and intensities do not match precipitation. Do you have an explanation for this? In general, please add some more discussion on the spatial patterns.
Authors’ response: Thanks very much for your comment. We apologize for a mistake in the previous version of our manuscript. The precipitation is highest in July in Naqu study area. We calculated the mean precipitation and improved AMSR-E soil moisture in the Naqu area in June and July. In June, the mean precipitation was 157.47 mm. In July, the mean precipitation was 170.55 mm. In July, the mean soil moisture was 0.285 cm3cm-3, higher than the mean soil moisture in June i.e. 0.175 cm3cm-3. So, we wrote that in Naqu the temporal pattern of soil moisture does match precipitation.
About the spatial pattern we have added in the previous revision of our manuscript Fig.11 and related comments about the relationship between SM and MDPI, see L545 – 550. In addition we have added at L520 the following comments in the revised version:
“There are large differences in the spatial SM pattern (Fig.9) between the AMSR-E/NASA, AMSR-E/JAXA and I_AMSR-E SM. AMSR-E/NASA SM is very flat and uniform in spatial distribution in the Naqu area. Compared with AMSR-E/NASA SM spatial distribution in Naqu area in June, the spatial dynamic range of AMSR-E/JAXA and I_AMSR-E SM is larger from 0.05 to0.5 cm3cm-3. Compared with the spatial pattern of precipitation in June, the precipitation is higher in the NE (Northeast) than in the NW (Northwest) portion of the Naqu area. On the other hand both the AMRS-E/JAXA and I_AMSR-E SM are lower in the NE than in the NW portion of the Naqu area. The difference in the AMSR-E/NASA SM between the NE and NW portion is small. In comparing the spatial patterns of SM with precipitation, however, it should be taken into account that the NW portion is flatter with a large presence of water bodies, which explains the differences observed in Fig.9. We further note that our improved SM estimates (I_AMSR-E) reflects correctly the terrain, with a lower elevation catchment in the NW portion of the Naqu area.”
Fig. 9: Please consider changing the colors for the precipitation maps (blue for zero rainfall is not very intuitive).
Authors’ response: Thanks very much for your comment. The color scale applied to the precipitation maps in Fig.9 has been revised as suggested,
Figure 9. Spatial and temporal comparison of the monthly AMSR-E/NASA, AMSR-E/JAXA, I_AMSR-E SM and precipitation in 2011, Naqu area (upper panel); Digital Elevation Model (DEM, lower panel).
In general: there are some spelling/formulation and formatting errors throughout the manuscript; e.g., line 398 "that the, , SM content". Please check this again.
Authors’ response: Thanks very much for your comment. The spelling/formulation and formatting errors have been revised at L 398, 404, 418 and everywhere else needed.
Reviewer 2 Report
Dear Author,
The paper has been sufficiently detailed to warrant its presentation in a scientific journal.
Some comments are pointed in the text.
The methodology needs to include more detailed information about the resampling of the data used.In the results section, I suggest improving figure 5. Therefore, I recommend publishing the article after some suggested corrections in the text. Yours sincerely Luciana
Comments for author File: 
Comments.pdf
Author Response
Dear reviewer, thank you very much for your comments. Following are detailed responses to comments.
The paper has been sufficiently detailed to warrant its presentation in a scientific journal.
Some comments are pointed in the text. The methodology needs to include more detailed information about the resampling of the data used.
In the results section, I suggest improving figure 5. Therefore, I recommend publishing the article after some suggested corrections in the text. Yours sincerely Luciana
peer-review-5477026.v1.pdf
Authors’ response: Thanks very much for your comments and suggestions. We corrected the words according to peer-review-5477026.v1.pdf. We revised the marks of X-axis and Y-axis in Fig.5 to make it clearer.
In the Methodology at L340-343, some details about the resampling procedure have been added: “Dataset 2 is the AMSR-E Level 2A brightness temperature data that was resampled to a grid (i.e. the Equal-Area Scalable Earth, EASE-Grid) of approximately 25km×25km using the distance-weighting method applied to AMSR-E L1 brightness temperature data. EASE-Grid is a global, cylindrical, equal-area projection, with 1383 columns×586 rows.”
Round 2
Reviewer 1 Report
Dear authors, thank you for taking into account my comments and providing another revision of your manuscript.
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Regrettably, this manuscript requires "significant" English language revision before any comments or feedback can be given on the scientific or technical discussion. Overall, the authors discussions, methodology, and explanation are very convoluted and I find it hard to prove an honest assessment.
I highly encourage the authors to take their time to revise and improve their presentations and explanations. I also suggest having the paper's language certified.
Reviewer 2 Report
The paper is well written. It address a difficult scientific problem, namely the low dynamics as seen from remotely sensed soil moisture. It is well written and to the point.
Reviewer 3 Report
The authors compare two different SM datasets (NASA and JAXA algorithms) derived from AMSR-E observations for a region in the Tibetan Plateau. They evaluate differences between the two products as well as to in situ measurements. As it turns out that especially the quality of the product derived using the NASA algorithm is poor in the study region, they provide an improved retrieval method by building a linear relationship between SM and MPDI (Microwave Polarization Difference Index).
The manuscript has a clear structure, and the authors support their methods and findings with helpful figures and flow charts.
Some questions/comments to the contents:
Abstract
1. line 32: also mention what is the original value of A1, or, instead of the value, describe the effect of a higher value compared to a lower value (i.e. steeper slope)
Introduction
2. lines 40f.: how do you define a long-time series? Also be careful with the spelling: both "long-time series, long-time-series" is used
3. general: large parts of the introduction are a description of the dataset algorithms. This is important, but should mainly be explained in the dataset section. In the introduction however, I miss a description of how you are actually going to improve the NASA product. This is in your title, so it should be explained in the introduction.
Study Area and Data
4. lines 135/136: I assume you are using the mean SM value for each pixel because your pixels are relatively homogeneous in land cover and topography, and the mean value is thus a good representation of the entire pixel. Have you tested this? Please add in your text.
5. lines 137-140: Please explain why you only use Pixel 1 for the validation of the linear model.
6. line 166: "and so on" - please reformulate.
7. lines 180f: good explanation of the algorithms, but repeats parts of the introduction. See comment 3
Methodology
8. line 286: (when) does the special case occur in real observations?
Results
9. Fig. 5: Can you change the axis so the dateticks show e.g. always January and July? This would facilitate the identification of seasonal SM patterns. In addition, please add grid lines so it is possible to compare Figures a and b to c and d (e.g., does high std. dev. occur during high or low SM periods).
10. line 345: Figures 6 and 7?
11. line 365: The SM increase seems to start in April already.
12. Figure 7: NASA_A and NASA_D have a narrow dynamic range, but they seem to show certain dynamics that are also present in the in situ data, e.g. the SM drop followed by an increase in August 2011. The metrics (especially the high correlations) in Table 2 also suggest that the dynamics are captured well, just the absolute values are different. But in lines 391-392 you say "On the whole, the accuracies of both SM data products are poor". Please explain how you define "accuracy". Perhaps add figures showing the NASA/JAXA time series scaled to the in situ datasets, e.g. using a mean-std. dev. scaling, and anomaly time series to be able to better evaluate the representation of SM dynamics in NASA/JAXA SM. In Table 2, metrics of not only the raw time series but also the anomalies would help.
13. lines 401-405: You mention two shortcomings, but to my understanding, the different dynamic ranges of NASA and in situ (shortcoming 1) directly leads to high RMSE values (shortcoming 2). Concerning "the accuracy is poor" and "such poor performance": see comment 12. Please elaborate on this.
14. lines 406f.: partly repetition from study area section
15. line 454: Please add what TOA stands for.
16. Figure 9: Can you add maps of monthly rainfall for comparison?
17. line 489: replace "is better than" with something like "compares better to the in situ data than"
Discussion
18. line 500: what is the evidence? Please mention in the text.
19. lines 500f: Do you find this narrow range also in other regions of the world? Please add more literature examples that use the AMSR-E/NASA SM dataset.
20. lines 503f: How specific is this value of 8 to your study area? How could AMSR-E/NASA SM be improved in other regions where you have no in situ data?
Thank you for considering my comments.
