Topographic Wetness Index as a Proxy for Soil Moisture in a Hillslope Catena: Flow Algorithms and Map Generalization

Round 1

Reviewer 1 Report

Dear authors, you are presenting a very nice study related with Topographic Wetness Index (TWI) and its relationship calculating soil moisture with a study case of the USA.

My detailed corrections are in the attached PDF but my general corrections are the following ones:

1) Add "in Fayetteville Arkansas, USA" in the Abstract

2) check what the journal indicate about the number of keywords, I think they are 10 maximum

3) Page 2 paragraph 3: Very good. There other recent lines where TWI have been used. Please check and add the following references according the topics.

Floods:

Improving regional flood risk assessment using flood frequency and dendrogeomorphic analyses in mountain catchments impacted by tropical cyclones

Wildfires:

A geomorphometric model to determine topographic parameters controlling wildfires occurrence in tropical dry forests

4) Last Introduction paragraph: put your research questions,  hypothesis, and aims clearer in this paragraph

5) Figure 1: improve figure quality, coordinates bigger, add a noth arrow, why those blanks? and why the weird upper right corner?

also, it will be nice to add a small map of the location of the site in the USA

6) Discussion:

1) Indicate the uncertainties of your model compared with other applications of TWI in a general and detailed scale.

2) Enhance the Discussion of your Results with previous studies with similar results than yours.

3) Moreover, try to indicate how your results can improve the soil moisture knowledge wordlwide depending on their topographic and climatic characteristics

All the best.

Comments for author File: Comments.pdf

Author Response

"Fayetteville, Arkansas, USA" was added to the abstract.

The number of keywords has been reduced to 10.

The two references indicated were added to the paper. 

In the last introduction paragraph, the following text was added to address the need for contextualization elaborated by the reviewer: 

"The research asks the question: Can the TWI, a stationary representation of surface runoff, be used to assess soil moisture conditions that are dynamic with respect to time and depth. We hypothesize that the TWI can be used as a proxy for soil moisture dynamics at a soil landscape scale. Further we hypothesize that factors such as TWI algorithm, grid resolution, temporal resolution, and placement of soil moisture sensors affect the relationship between TWI and soil moisture."

To contextualize the study to better meet the reviewer's comments we added the following: 

"The sensitivity of the correlation values to moisture conditions during the transitions from wet to dry and vice versa, increases the usefulness of TWI especially for topographically responsive areas. For example, the TWI could guide management decisions such as schedules for fertilizer applications, planting and use of machinery, especially in rainfed agriculture on sloping areas around the world. However, for very dry and wet conditions as well as flat areas, TWI usefulness for soil moisture predictions at finite temporal resolutions should likely be considered as complimentary to other soil moisture monitoring devices and techniques."

We believe we have discussed the literature thoroughly in the introduction section. We also mention literature in the Discussion section to contextualize our work and believe that no additional information is needed for the reader to understand the context of our study.

 

Reviewer 2 Report

This manuscript reported an interesting result. I think the experiments and the data presentation are sound. The research problem has been well resolved and discussed.

1. All equations should be numbered.

2. Shortages of the TWI method should be discussed.

Author Response

All equations have now been numbered.

Shortcomings of TWI have now been discussed, particularly with respect to using a static index to represent dynamic processes in the following new paragraph:

"The sensitivity of the correlation values to moisture conditions during the transitions from wet to dry and vice versa, increases the usefulness of TWI especially for topographically responsive areas. For example, the TWI could guide management decisions such as schedules for fertilizer applications, planting and use of machinery, especially in rainfed agriculture on sloping areas around the world. However, for very dry and wet conditions as well as flat areas, TWI usefulness for soil moisture predictions at finite temporal resolutions should likely be considered as complimentary to other soil moisture monitoring devices and techniques."

Reviewer 3 Report

Dear Authors

 

I read the manuscript titled "Topographic Wetness Index as a Proxy for Soil Moisture in a Hillslope Catena: flow algorithms and map generalization". The results and the methods followed can be helpful in cases where soil moisture parameter information is needed from a practical point of view. Although it is generally well-organized, it should be developed by clarifying the issues I mentioned below. I am not a modelling person or a computer expert, but as a soil physicist, I needed help understanding the smoothing of some topography-related values when calculating the Topographic wetness index. In an equation with one unknown where two parameters are used, it can be mathematically aimed to increase the index value by keeping the numerator constant and decreasing the denominator. However, I would like to see the logic in estimating the soil volumetric moisture value more clearly from the point of soil physics.

 

Major comments

 

I want to state that the purpose of the article is not clear.

 

As the authors and many others agree, soil moisture content can vary temporally and spatially in any soil landscape depending on many factors. Topography and local roughness are among the first. As I understand from the paper, only two parameters were used: the area of a portion of land and the slope angle. The authors found that most TWI algorithms performed poorly on DEMs, at first and then applied to filter to improve the TWI performance. In this case, DEMs with larger pixels (lower resolution DEMs) have yielded higher values of TWI, which seems better than lower results of smaller pixels (higher resolution DEMs).  This contains a contradiction and needs to be explained. Unfortunately, I could not find such an explanation, although I missed it due to the length of the article.

 

 

Author Response

The reviewer helpfully points out "The results and the methods followed can be helpful in cases where soil moisture parameter information is needed from a practical point of view. "

This is the point of the paper, as well as to numerically describe the goodness of fit between TWI and measured moisture when certain practical decisions are made regarding elevation model preparation and algorithm choices terrain modelers have to make. Some of these choices can be much better for estimating variability of field moisture than others. 

The contradiction mentioned between pixel resolution and TWI performance is not meaningful, as it is not the magnitude of the TWI values that is at issue, but the spatial variability within a given context, and whether that variability matches the observed moisture in similar sites. The reviewer perhaps does not have a sufficient background to understand map generalization, the smoothing of surface roughness and how this improves flow models. We tried to explain this in the paper. The reviewer states: "Unfortunately, I could not find such an explanation, although I missed it due to the length of the article." The fact that the reviewer missed the explanation due to the article being lengthy implies a lack of careful reading. It also implies that the article may be too long as it is written. As we believe the article does already contain an adequate explanation for the relationship between map generalization and TWI performance, we have opted not to repeat ourselves by adding more text on this issue. Doing so would lengthen an already long (perhaps too long!) article.

Reviewer 4 Report

My comments about the article are as follows.

 

·         The number of keywords is high, the number of keywords should be reduced

·         “15-cm and 60–75-cm” spelling should be corrected

·         Equation 2 is mentioned on page 8, but the names of the equations are not written The names of the equations should be written

·         The TWI formula and explanation should be in the material and method section.

 

This article can be accepted after minor correction.

Comments for author File: Comments.pdf

Author Response

The number of key words is now reduced to 10. 

The notation of sensor depths was corrected.

Equations have now been numbered for reference. These are now referred to throughout the text by their number. 

Equation number 1 is maintained in the Introductory section because it is appropriate there. While it is customary to discuss equations in the methods and materials, equation 1 belongs in the Introduction because it is fundamental to the concepts being discussed. Other authors have included Equation 1 in their introductions as well. See, for example, Riihimäki, H.; Kemppinen, J.; Kopecký, M.; Luoto, M. Topographic Wetness Index as a Proxy for Soil Moisture: The Importance of Flow‐Routing Algorithm and Grid Resolution. Water Res 2021, 57, doi:10.1029/2021WR029871.