Deriving River Discharge Using Remotely Sensed Water Surface Characteristics and Satellite Altimetry in the Mississippi River Basin
Round 1
Reviewer 1 Report
General – Well written and complete.
Specific –
Lines 48 and 50 - Include sentences here that address 1) the fact that hydrologic models that estimate river discharge derive the discharge through water balance accounting and therefore require as input precipitation, evapotranspiration, and storage changes as well as routing parameters that apply at the temporal and spatial scale of the model; and 2) that the hydrologic input data and parameters required for these models can often be difficult to obtain at the necessary scale, and may result in significant uncertainty particularly for small spatial and temporal scales.
Line 67 – add at end of sentence: “or rapidly varying flow on the scale of a week or less”.
Line 74 – Reference 13 should reference to the following paper which is not included and needs to be added to the reference list “Bjerklie, David M., Charon M. Birkett, John W. Jones, Claudia Carabajal, Jennifer A. Rover, John W. Fulton, Pierre-André Garambois, 2018, Satellite remote sensing estimation of river discharge: Application to the Yukon River Alaska, Journal of Hydrology no. 561, pp. 1000–1018.”
Line 80 – Reference should be to Bjerklie et al., 2018 (see above).
Line 141 – There are many examples of Manning n values exceeding 0.07 (e.g. see Hicks and Mason, 1998, Roughness Characteristics of New Zealand Rivers) especially at lower flows in smaller rivers. I would qualify this statement to say that for larger rivers observable by SWOT or comparable satellites (i.e. widths greater than 100 m) Manning n values would not be expected to be larger than 0.07. Additionally, I would add a qualifying statement that most rivers exhibit varying Manning n values with flow level generally increasing with decreasing depth of flow for most rivers.
Line 154 – Reference Bjerklie et al., 2018 (see comment line 74) which discusses use of a parabolic cross-section shape, and reference “Bjerklie, D. M., Fulton, J. W., Dingman, S. L., Canova, M. G., Minear, J. T., & Moramarco, T. (2020). Fundamental hydraulics of cross sections in natural rivers: Preliminary analysis of a large data set of acoustic Doppler flow measurements. Water Resources Research, 56, e2019WR025986. https:// doi.org/10.1029/2019WR025986.” Which shows with data that rivers tend towards cross-sections exhibiting max to mean depth ratios that are consistent with a parabolic shape.
Line 225 – It is not clear what TWSA is or for that matter GRACE TWSA. This needs to be defined as it is used in the optimizations.
Line 252 – A new section should start here titled “Discharge Estimation and Parameter Optimization”. This section needs to explain how the optimization was accomplished and what if any objective function data or assumptions were used for optimization.
Line 268 – It is not clear to me how GRACE TWSA is used to define h0.
Line 298 – I recommend assessing the discharge method by examining the estimate absolute and log residual distribution in order to say something about error and goodness of fit with regards to relative flow level. In many cases, as shown in figure 3, errors are larger for lower discharge. This makes sense due to errors associated with remote sensing resolution and in respect to modeling increasing flow resistance and flow complexity as depth decreases. Likely, high flows are also more difficult to model accurately because of the influence of increasing flow resistance due to bedload transport, occupation of bars and vegetation along the banks and overbank flow (among others).
Paragraph beginning on Line 346 – Why is the triangular channel shape assumed for the region below the lowest observed WSE rather than parabola (see comment Line 154 and statements made in line 152 – 154 or the report).
Paragraph beginning Line 405 – The satellite observational record can be extended temporally and spatially by including satellite altimeters and optical observations – although this would require reconciling observations from the different platforms.
Paragraph beginning on line 442 – The observed spatial trend could be at least partly the result of lower flow conditions in the lower Missouri and Arkansas region if lower flows are more dominant. This would be consistent with more bias at low end of discharge range as pointed in comment Line 298.
Reference 13 in 2nd and 3rd paragraphs section 4.2 (page 15) – My previous comment that reference 13 used on page 2 should be for Bjerklie et al., 2018 rather than Bjerklie et al.,2003 does not apply here – the Bjerklie et al. 2003 reference is correct for the references on this page.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The paper presents the use of Manning's Equation with the help of remotely sensed water surface characteristics and satellite altimetry to quickly estimate river discharge in the Mississipi River Basin. It is a suitable topic for Remote Sensing MDPI journal. The paper is well presented, written, and discussed. I recommend the manuscript to be accepted as it is presented.
Author Response
Thank you for your support of the manuscript. We have attached a document with our responses.
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear Authors,
The proposed study “Deriving river discharge using remotely sensed water surface characteristics and satellite altimetry in the Mississippi River Basin” is quite interesting, and the results are helpful. It used several datasets from different space-borne sensors to estimate river discharge based on the optimization of unknown parameters such as surface roughness and bathymetry. The overall results of the estimated discharge based on different experiments in the framework of Manning’s equation showed the feasibility of the methodology. One of the study's main limitations is perhaps the need for observed discharge, which might be unpracticable over many rivers. Many other aspects must be addressed before a potential publication in Remote Sensing. I have listed them below, and I hope the comments can help you improve the ms.
1. Authors’ affiliation, please, add the country and city.
2. Line 24: “…other environmental factors…” Please, add some examples of such factors.
3. Line 34: Please add some sentences saying that the number of freely available datasets is also reducing globally for many reasons, as reported by (Sneeuw et al., 2014).
4. In the first paragraph, add some information about the importance of such datasets to calibrate and evaluate the hydrological model.
5. Lines 39-42: Ok, this agrees with my comment 3. Perhaps the most appropriate place for such sentences would be in the first paragraph.
6. Line 50: Delete “global”
7. The sentence between lines 51-53: Add a couple of references that have used different missions and concepts for estimating discharge in other parts of the world. For example, satellite gravimetry (Ferreira et al., 2013), satellite altimetry (Tourian et al., 2013), satellite imagery (LeFavour & Alsdorf, 2005), and satellites altimetry and imagery (Tourian et al., 2017) and so on.
8. Line 72: “…in-situ data [11] whereas few exceptions are possible (Lamine et al., 2021).”
9. Line 97: It is better to explain which characteristics are those to clarify the contribution of the current study to the previous ones.
10. Although the two research questions are valid, they have been addressed in previous studies. Therefore, it is necessary to introduce the main objectives of the current research and the novelty. That would help to place the present study among the previous ones.
11. Figure 1: Please, consider adding a legend to the figure. It also is necessary to say “not to scale” since the values for with, and depth could lead to misinterpretations.
12. Lines 146-148: Why not generate a “real” cross-section above the minimal WSE by stacking several widths (imagery) and elevations (altimetry)? That would allow one to develop the “bathymetry” above the minimal WSE. Although altimetry and imagery missions acquire datasets at different epochs, it is still possible to generate such values. Actually, DAHITI applies such an approach for deriving lakes/reservoirs volume storage changes above a minimal observed WSE, such an approach could be used for rivers as well. Refer to https://dahiti.dgfi.tum.de/en/products/volume-variation/#processing_strategy.
13. Line 124: S is assumed to be the water surface gradient; please clarify.
14. Sub-section 2.1 does not clarify the new method and contribution from the study.
15. Sub-section 2.3: How about potential biases among the water level retrieved from the different missions? How have the datasets been harmonized to minimize the effect of potential biases?
16. Sub-section 2.3 could be sub-divided according to the different products reported there.
17. Line 228-229: What the authors would like to say here is unclear. Has GRACE (TWS datasets) been downscaled somehow? Actually, the nominal resolution is approximately 300-400 km at the equator. Not clear how in some areas, GRACE could present a more satisfactory resolution without downscaling or applying the scale factors (gain factors) as proposed by (Landerer & Swenson, 2012).
18. So, the optimized parameters rely on the observed discharge limiting the method to be applied at many ungagged river basins. The introduction has reported that there are so many ungagged river basins, hence the importance of estimating discharge based only on remotely sensed datasets.
19. Some methodological aspects are mixed with datasets description making it hard to follow the author’s contribution. Please, organize Section 2 into sub-sections describing the methodology and the datasets, respectively. A flowchart summarizing the text between lines 233-251 would also be helpful.
20. Line 356-357: Why would it imply in higher Q Missing?
21. Line 358: The Latin expressions e.g. and etc. cannot be used together. Please choose one of them and avoid redundancy.
22. Figure 4: TWSA is not cm/mo, it is just cm since it is storage and not flux.
23. It's not clear how h0 has been estimated based on both Q vs P and Q vs TWSA. Again, the methods need improvement to explain the caveats better. How about the epochs between Q and TWSA? (They are probably not aligned.)
24. Lines 407-408: Stacking different missions would not help on that?
25. Line 511: Perhaps both, width and water level.
26. Lines 538-539: That has also been demonstrated in several other studies.
27. Lines 544 and 558: Please, check whether Remote Sensing allows references on the conclusion.
28. Lines 550-551: The impact of the shape above the WSE would be worthy of investigation. Please, refer to comment 12.
29. The text between lines 552-566 is engaging. However, it is not directly related to the present study's results and discussion. It might be better to keep it on your “to-do list” and provide some suggestions for improvements, perhaps in relation to the limitations of the present study.
References:
Ferreira, V., Gong, Z., He, X., Zhang, Y., & Andam Akorful, S. (2013). Estimating Total Discharge in the Yangtze River Basin Using Satellite-Based Observations. Remote Sensing, 5(7), 3415–3430. https://doi.org/10.3390/rs5073415
Lamine, B. O. M., Ferreira, V. G., Yang, Y., Ndehedehe, C. E., & He, X. (2021). Estimation of the Niger River cross-section and discharge from remotely-sensed products. Journal of Hydrology: Regional Studies, 36(June), 100862. https://doi.org/10.1016/j.ejrh.2021.100862
Landerer, F. W., & Swenson, S. C. (2012). Accuracy of scaled GRACE terrestrial water storage estimates. Water Resources Research, 48(4), n/a-n/a. https://doi.org/10.1029/2011WR011453
LeFavour, G., & Alsdorf, D. (2005). Water slope and discharge in the Amazon River estimated using the shuttle radar topography mission digital elevation model. Geophysical Research Letters, 32(17), 1–5. https://doi.org/10.1029/2005GL023836
Sneeuw, N., Lorenz, C., Devaraju, B., Tourian, M. J., Riegger, J., Kunstmann, H., & Bárdossy, A. (2014). Estimating Runoff Using Hydro-Geodetic Approaches. Surveys in Geophysics, 35(6), 1333–1359. https://doi.org/10.1007/s10712-014-9300-4
Tourian, M. J., Elmi, O., Mohammadnejad, A., & Sneeuw, N. (2017). Estimating river depth from SWOT-type observables obtained by satellite altimetry and imagery. Water (Switzerland), 9(10), 1–22. https://doi.org/10.3390/w9100753
Tourian, M. J., Sneeuw, N., & Bárdossy, a. (2013). A quantile function approach to discharge estimation from satellite altimetry (ENVISAT). Water Resources Research, 49(7), 4174–4186. https://doi.org/10.1002/wrcr.20348
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 3 Report
Dear Authors,
Thanks for properly handling my concerns. I'm satisfied with the current version and support its publication in RS as is.
