Mapping of Snow Depth by Blending Satellite and In-Situ Data Using Two-Dimensional Optimal Interpolation—Application to AMSR2

Round 1

Reviewer 1 Report

The present paper investigates the snow depth two interpolation approaches applied to satellite and in-situ measurements and the results are inter-compared to GFS snow depth outputs. The results reveals that the method developed in thys study performs better for high elevation compared to lower elevation. Generally the paper is well written and the subject of it is interesting.

Some suggestions to the authors:

Section 1

L35-44: the paragraph presented herein should be moved at Section 2, between L102 and L103. Regarding Fig 1 it will be interesting to present a map with difference between snow depth for the study period and 20 years snow depth average.

Section 2

The period of the present study should be mentioned here together with the motivation of choosing this period.

L139-141 this paragraph should be moved to Subsection 2.2.  It will be interesting to provide ranging interval of the station elevation.

Section 4

In my opinion the Results and Discussion section should be better presented. Herein are presented some graphs with together with a description of them. Some discussion should be added.

L268-271 Which could be the explanation for the results presented here?

L272-329 The fig's presented here are not discussed, please discuss them or remove them.

Section 5

L391-395 The paragraph presented here is more appropiate for Introduction section than for Summary and Conclusion section.

Please try to present better which are the main findings of this study.

Author Response

Please, see the attachment. 

Author Response File: Author Response.pdf

Reviewer 2 Report

The optical interpolation approach utilized here demonstrates some advantages for generating a snow depth product in high elevation regions. This would be of value specifically in areas with limited in situ snow depth measurements as is the case across much of western Canada and Alaska.  Does the system offer any significant advantages for low elevation regions or areas with a high density of in situ sites such as the western United States?  If the advantage in the latter areas is limited that is fine. I suggest the paper has to focus more on areas where there is value.  The validation is limited to a single period during 2017, how is this likely to impact the assessment of efficacy? There is a potential for combining tables.  A snow depth difference figure focussed on high elevation areas where the model works best is needed.

14: “The production of  a snow depth product for North America is investigated by two dimensional optimal  interpolation applied to passive microwave satellite and in-situ measured snow depth.”

60: “Passive microwave measurements provide the most mature snow retrieval products of all available satellite systems, e.g., from Special Sensor Microwave/Imager (SSM/I), Advanced Microwave Scanning Radiometer (AMSR-E), Advanced Microwave Sounding Unit (AMSU) and more recently Advanced Microwave Scanning Radiometer 2 (AMSR2) instruments [2-6].

69:”… in combination with in-situ snow depth”

111: What is the source of the look up table?

113: how is the algorithm different than the NASA SWE algorithm.

193:  How many grid cells are considered?

204:  600 km is a quite a large distance, what is the justification that this is useful? I suggest looking for references.

223:  No January 10 map.

270:  Why is the RMSE at high elevation lower?

Table 1-4 should be easily combined into a single table or Figure.

374:  Why did GFS have a declining snow depth?

378:  Is the decrease noted just for high elevation regions?  Given the focus on high elevation regions and the difference in snow depth from GFS, AMSR2 and OI for the Jan-Feb 2017 period this should be the specific focus of analysis and a figure that illustrates.

382:  Is this just true for high elevation areas?

401: Was the January-February period of 2017 relatively typical in snowcover and snow depth development? 

Author Response

Please, see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Very well presented and interesting study dealing with mapping snow depth by remote sensing technique and ground measurements. Proposed method offer better temporal and spatial coverage that could not be achieved by ground measurements even if the advantage of in situ sampling is higher accuracy. 

I have only two minor comments authors might take into consideration:

1/ Legends for Figs. 3 and 4 should be increased for easy reading. This might be done also for other figures. 

2/ Some spatial features in the map of snow cover could be discussed in text. For instance, IMS show map for January 7, 2017 Fig. 3 right panel shows two prolonged latitudinal areas in central US that are without show cover. Such features might be worth of discussion. 

Author Response

Please, see the attachment. 

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

I agree with all changes and comments of the authors.