Controls on Streamflow Densities in Semiarid Rocky Mountain Catchments

Round 1

Reviewer 1 Report

This manuscript presents a study related to the dynamics of the flowing channel networks for 3 semi-arid catchments along an elevation gradient in the Colorado Front Range (USA). Moreover, field data are compared to topography derived data and national network datasets (NHD). This manuscript is focus on a topic that is of great interest to the hydrologic community. However, there are some parts that need a more detailed justification and clarifications. The study uses field data from only one year (2 periods in 2016). There is not information about how is the hydrological year 2016 (dry, wet, etc.) and the authors do not discuss the consequences of using possible data from other years. For example, find a stream dry in 2016 in these catchments, does it imply that it is dry every year for the same period?

Authors also need to introduce better why mapping flow permanence and obtain accurate maps of flow are important and its implications and applications. Also additional information of the field sampling method it would be very useful. Another bother is related to the figures, I recommend some changes and improvements in some of them as I comment later.

Other comments and recommendations are presented below:

TITLE

In my opinion “semiarid catchments” is very global and general in relation to the data used for this study (1 year, 3 catchments, …). Authors should include something more specific in the title referring to the location of the study.

INTRODUCTION

Line 37: How is the situation in other countries or locations of the world other than USA? Are the percentages of intermittent streams similar? Include some information in the introduction about the situation in other areas of the world.

Line 63: Include some examples (bibliographic references) of studies with ADD greater than GDD.

MATERIAL AND METHODS

Line 110: Include USA in parentheses at the end of the sentence to show the country location.

Line 121: Pinus ponderosa in cursive. Please check that all specie names in latin throughout the document are in cursive.

Line 122, 123: PRISM and gridMET perhaps they are unfamiliar to a reader not used to using data series from USA. You can include some information in a couple of lines or include the definition of the acronym. For example: data derived from the gridded meteorological database, gridMET.

Line 134,135,148,149. P, PET and aridity index were also derived from PRISM and gridMET for these catchments, right? Include the references or mention in the first case that both databases were used to derive the data of P, PET and aridity index in the rest of the catchments.

Table1: Include also the information of PET in a column.

Figure 1. What is the meaning of the arrows in (b)? Include the meaning on the legend.

Figure 1. Black stream network lines in Skin and Gordon are continuous according to NHD flowlines, but in Milk there are some discontinuities in the tributaries (mostly southern tributaries). Are these discontinuities temporary rivers?

Are the 3 catchments in hydrological natural conditions in relation to water management/human intervention (dams, water abstraction, etc.)? Mention it in the text so that these conditions do not influence the data taken in the sampling or explain if there are some alterations.

What is the definition used to map flowing and dry segments? (For example, there is no surface water, zero flow, etc.)

Line 163. Could you provide more information about the field-sample? How do authors follow all branches of the channel network? By foot, how many km? Or Using drones, car, etc? Were all the branching networks accessible? Was a week enough to collect information about all the stream network? If the information is very extensive, you can include it in the supplementary material.

Line 191. NHD: include the scale used.

Line 196. Geologic maps, include bibliographic reference.

RESULTS

I recommend to join the maps of figure 2,3 and 4 in one figure. I recommend to move the images (photos) of figure 2,3,4 to supplementary material.

Could you provide the hydrographs as supplementary material?

In other studies, temporal segments are represented by dashed lines. Authors need to clarify in the legend of the maps how temporary segments are represent since dashed lines represent the second sample period.

Figure 4: Why intrusive is represented in both colors, brown and pink, on the map? Which is the difference?

Figure 5: Include description of the boxplot in the legend (black line, points, etc,…)

Figure 6: Blue and yellow colors are the same color used for the sample field. I recommend to change both color in this figure to no repeat blue and yellow for a different objective on the map.

Figure 6: Using the same line thickness, there are some results that are not visible. For example, in Gordon catchments blue line (mean) is not visible. Reconsider designing the figure for its correct visualization

Line 277: I recommend using the same decimals in the text as in the tables. (pe. 1.83 km-1)

Figure 7: Indicate each one of the graphics with letters (a,b,c,d). I recommend do not use the triangle figure for the colors in the legend since triangle is used for one of the catchments. As an alternative you can use the colors in the letters. For example, field 1 in red colors, etc. (the same for figure 8).

Figure 7: Precipitation is not commented on the text. Comment on the text or remove it from the figure.

Why did not you include also in the comparation of some results the results obtained from the topographic maps? (pe Fig 5, Fig 7, etc…) Or comment if the river network obtained topographically is over or under represents drainage densities, etc.

DISCUSSION

I recommend include in the discussion how you can apply your study to other areas (if possible) or if it is not possible, why?. Field sample studies are very laborious in time, money, staff, etc… Other alternatives can be applied as support (drone, etc.)? You mention something in the conclusion, but I think you should debate it in the discussion.

Line 309. The sentence: “However, the tributaries with longer flow duration also have stable channel heads fed by seeps (Figure 4), and their flow may be sourced by groundwater that originated outside the small catchment boundaries” seems that is referenced to the Mill Creek study case since in the previous lines you cited this study case. But figure 4 is Gordon catchments, so I recommend you rephrase it.

How the land uses-land cover can influence the streamflow densities results obtained?

Author Response

R1C1: This manuscript presents a study related to the dynamics of the flowing channel networks for 3 semi-arid catchments along an elevation gradient in the Colorado Front Range (USA). Moreover, field data are compared to topography derived data and national network datasets (NHD). This manuscript is focus on a topic that is of great interest to the hydrologic community. However, there are some parts that need a more detailed justification and clarifications. The study uses field data from only one year (2 periods in 2016). There is not information about how is the hydrological year 2016 (dry, wet, etc.) and the authors do not discuss the consequences of using possible data from other years. For example, find a stream dry in 2016 in these catchments, does it imply that it is dry every year for the same period?

R1R1: We recognize that one year is a limited period of time for the surveys; this is because the work was conducted for a masters thesis. Although more data would have been helpful and preferable, we thought the findings were unique enough to warrant publication. To provide more context for the hydrologic conditions of the study year, we added text in multiple locations:

Lines 187-188: “summer 2016 [66], a year with close to average annual precipitation (Table 2).”

Line 446-447, referring to Gordon Gulch: “We re-surveyed surface flow patterns in Gordon in August 2020 and found that the locations of flow emergence remained consistent with those documented in 2016.”

Lines 460-464, referring to Mill Creek: “The patterns of wetting and drying observed in 2016 are consistent with those that have been observed in the years since. The main tributary dried between June 10 and July 2 every year from 2016-2019, while the two southern tributaries (Figure 2) maintained flow or standing water into July or longer according to Stream Tracker citizen science observations.”  

Lines 489-490, referring to Skin Gulch: “Since 2016, some of the smaller tributaries have begun to dry sooner in the summer as more vegetation returns post-fire.”

R1C2: Authors also need to introduce better why mapping flow permanence and obtain accurate maps of flow are important and its implications and applications. Also additional information of the field sampling method it would be very useful. Another bother is related to the figures, I recommend some changes and improvements in some of them as I comment later.

R1R2: We added text to the introduction about the reason for mapping flow permanence and its applications (lines 46-50): “These discontinuous streams are important contributors to biodiversity, material transport, and water supply for larger streams and water bodies [3], yet they are typically not as well mapped or monitored as larger perennial streams. Information on stream location and flow duration may be used for determining whether streams are protected by water quality regulations or land management guidelines.”

The remaining components of this comment are addressed in responses below.

Other comments and recommendations are presented below:

TITLE

R1C3: In my opinion “semiarid catchments” is very global and general in relation to the data used for this study (1 year, 3 catchments, …). Authors should include something more specific in the title referring to the location of the study.

R1R3: We have added “Rocky Mountain” to the title to be more specific about the location of study.

INTRODUCTION

R1C4: Line 37: How is the situation in other countries or locations of the world other than USA? Are the percentages of intermittent streams similar? Include some information in the introduction about the situation in other areas of the world.

R1R3: We added a reference about the extent of intermittent streams in other parts of the world (line 46, “greater than 50% globally”). Because this topic has been addressed in more detail in the reference cited (Datry et al. 2014), we chose to keep the text here brief.

R1C5: Line 63: Include some examples (bibliographic references) of studies with ADD greater than GDD.

R1R5: We have added references here (line 77).

MATERIAL AND METHODS

R1C6: Line 110: Include USA in parentheses at the end of the sentence to show the country location.

R1R6: U.S.A. has been added to the end of the sentence indicated.

R1C7: Line 121: Pinus ponderosa in cursive. Please check that all specie names in latin throughout the document are in cursive.

R1R7: We updated to italicize all Latin species names.

R1C8: Line 122, 123: PRISM and gridMET perhaps they are unfamiliar to a reader not used to using data series from USA. You can include some information in a couple of lines or include the definition of the acronym. For example: data derived from the gridded meteorological database, gridMET.

R1R8: We updated the text to provide information about the data sources:

Lines 140-142: “Mill Creek has a mean annual precipitation (P) of 464 mm, as determined from the gridded precipitation data product, PRISM [53], and mean annual potential evapotranspiration (PET) of 1245 mm, as estimated from the gridded reference evapotranspiration product in gridMET [54].”

R1C9: Line 134,135,148,149. P, PET and aridity index were also derived from PRISM and gridMET for these catchments, right? Include the references or mention in the first case that both databases were used to derive the data of P, PET and aridity index in the rest of the catchments.

R1R9: We added the references to each mention of P and PET.

R1C10: Table1: Include also the information of PET in a column.

R1R10: PET has been added as a row in Table 1.

R1C11: Figure 1. What is the meaning of the arrows in (b)? Include the meaning on the legend.

R1R11: We removed the arrows from the figure.

R1C12: Figure 1. Black stream network lines in Skin and Gordon are continuous according to NHD flowlines, but in Milk there are some discontinuities in the tributaries (mostly southern tributaries). Are these discontinuities temporary rivers?

R1R12: We have checked on this with the Colorado data steward for NHD, and it is likely an artifact of the contour crenulation method that USGS used to delineate ephemeral channels. In general that method has produced more channels than exist in our study area. However, we prefer not to comment on this issue in the manuscript since it is an issue with NHD, not with the data we collected. Most of the streams in our study area are temporary streams, not just those with discontinuous flow lines.

R1C13: Are the 3 catchments in hydrological natural conditions in relation to water management/human intervention (dams, water abstraction, etc.)? Mention it in the text so that these conditions do not influence the data taken in the sampling or explain if there are some alterations.

R1R13: We updated the methods to indicate that these are all natural (line 128): “Three natural catchments with no flow modifications were monitored along the mountain front (Figure 1).”

R1C14: What is the definition used to map flowing and dry segments? (For example, there is no surface water, zero flow, etc.)

R1R14: The methods description has been updated to clarify these definitions (lines 193-196): “Flow was mapped where surface water was visibly connected longitudinally along a channel segment, whereas dry segments were mapped where no surface water was present in the channel. Pools with standing water and no visible longitudinal connection of surface water were also mapped.”

R1C15: Line 163. Could you provide more information about the field-sample? How do authors follow all branches of the channel network? By foot, how many km? Or Using drones, car, etc? Were all the branching networks accessible? Was a week enough to collect information about all the stream network? If the information is very extensive, you can include it in the supplementary material.

R1R15: Surveys were conducted by foot (line 191), starting at the watershed outlet and systematically following each tributary upstream until its channel head. The lengths of streams are shown in the maps; around 7 km for Mill; 20 km for Skin, and 2 km for Gordon. These are all relatively small watersheds, so a week was enough time to map the stream networks. NHD networks were used as a guide to determine where to look for possible streams.

R1C16: Line 191. NHD: include the scale used.

R1R16: The scale (1:24,000) has been added

R1C17: Line 196. Geologic maps, include bibliographic reference.

R1R17: We added the reference numbers for the geologic maps to this sentence.

RESULTS

R1C18: I recommend to join the maps of figure 2,3 and 4 in one figure. I recommend to move the images (photos) of figure 2,3,4 to supplementary material.

R1R18: We considered this suggestion but thought that merging the three maps into one would make the details of the geologic maps too small to see clearly. Since we don’t think merging the maps would work well visually, we prefer to keep the images in each of the map figures. We think these are useful for readers to visualize key features in these catchments.

R1C19: Could you provide the hydrographs as supplementary material?

R1R18: We added the hydrographs as an inset plot within Figure 7.

R1C20: In other studies, temporal segments are represented by dashed lines. Authors need to clarify in the legend of the maps how temporary segments are represent since dashed lines represent the second sample period.

R1R20: We added the following text to the map figure captions to clarify the meaning of the lines: “Blue lines indicate surface flow locations for the first field survey, and dashed orange lines overlaying the blue lines indicate surface flow locations for the second survey.”

R1C21: Figure 4: Why intrusive is represented in both colors, brown and pink, on the map? Which is the difference?

R1R21: These are two different intrusive lithologies, and the color difference remained from a previous map version where we had more detailed descriptions of lithology. We updated the figure to make intrusive the same color for this simplified geologic map.

R1C22: Figure 5: Include description of the boxplot in the legend (black line, points, etc,…)

R1R22: We added a description of the boxplot to the legend: “Each box spans the 25-75% quantile range, with the horizontal line representing the median. The whiskers represent range of values, except where a value is ≥ the 75% quantile + 1.5 times the inter-quartile range or ≤ the 25% quantile – 1.5 times the inter-quartile range; values outside these bounds are shown as outlier points.”

R1C23: Figure 6: Blue and yellow colors are the same color used for the sample field. I recommend to change both color in this figure to no repeat blue and yellow for a different objective on the map.

R1R23: We have changed the colors on Figure 6, so they are not the same as in prior maps.

R1C24: Figure 6: Using the same line thickness, there are some results that are not visible. For example, in Gordon catchments blue line (mean) is not visible. Reconsider designing the figure for its correct visualization

R1R24: We modified the thicknesses of the lines to enable seeing where they overlap. We also added a clarification to the caption about why Gordon does not have a line for mean: “Gordon shows only minimum and maximum channel extents because the catchment has only two flow heads.”

R1C25: Line 277: I recommend using the same decimals in the text as in the tables. (pe. 1.83 km-1)

R1R25: We have made this change.

R1C26: Figure 7: Indicate each one of the graphics with letters (a,b,c,d). I recommend do not use the triangle figure for the colors in the legend since triangle is used for one of the catchments. As an alternative you can use the colors in the letters. For example, field 1 in red colors, etc. (the same for figure 8).

R1R26: Thank you for these suggestions. We have updated the figure.

R1C27: Figure 7: Precipitation is not commented on the text. Comment on the text or remove it from the figure.

R1R27: We added a sentence on precipitation in the text (lines 334-336): “Annual precipitation was similar for Mill Creek and Skin Gulch, so the relationship between ADD and precipitation (Figure 7b) is not as strong as those for ADD vs. elevation and snow persistence (Figure 7a,c).”

R1C28: Why did not you include also in the comparation of some results the results obtained from the topographic maps? (pe Fig 5, Fig 7, etc…) Or comment if the river network obtained topographically is over or under represents drainage densities, etc.

R1R28: Do you mean the topographic thresholds for channelization? Figure 5 gives the range of values we used to identify the topographic thresholds used to delineate channels in Figure 6. We did not then include the drainage densities from topographic thresholds in Figure 7 because no single threshold was an accurate representation of the active drainage network.

DISCUSSION

R1C29: I recommend include in the discussion how you can apply your study to other areas (if possible) or if it is not possible, why?. Field sample studies are very laborious in time, money, staff, etc… Other alternatives can be applied as support (drone, etc.)? You mention something in the conclusion, but I think you should debate it in the discussion.

R1R29: We added a discussion section to address this comment (lines 502-514):

“4.4. Implications for future research

Overall our findings illustrate how the locations of active flow in the study catchments are not easily predictable from topographic data alone. This lack of consistency with topographically-defined channel patterns is likely to be found in other relatively dry catchments where locations of flow emergence are related to bedrock lithology and structure. Expansion of active flow mapping in space and time would help further our understanding of the controls on streamflow patterns in these settings. However, conducting field surveys like these in challenging, rugged terrain is labor intensive and therefore not a feasible means of documenting streamflow permanence over large areas. While in-person field observation is often the best way to see where surface flow is present in small streams, in places where riparian canopy cover or terrain shading does not obscure streams, drones or aircraft remote sensing may be a more efficient means of monitoring flow locations. We have also found that repeat visual observations at accessible points within stream channel networks can help in documenting the variability of wetting and drying patterns over time.”

R1C30: Line 309. The sentence: “However, the tributaries with longer flow duration also have stable channel heads fed by seeps (Figure 4), and their flow may be sourced by groundwater that originated outside the small catchment boundaries” seems that is referenced to the Mill Creek study case since in the previous lines you cited this study case. But figure 4 is Gordon catchments, so I recommend you rephrase it.

R1R30: Thank you for catching this. The Figure reference should have been Figure 2, and this is now updated.

R1C31: How the land uses-land cover can influence the streamflow densities results obtained?

R1R31: See R1R13, where we clarified that all catchments are natural. We also addressed potential land use influences in the discussion section 4.3.

Lines 447-451, for Gordon: “Gordon does have several trails, and a dirt road crossing in the upper catchment has contributed runoff and sediment to the stream channels. These sediment inputs have not been quantified, but if they are substantially reworked during high flows, deposits of coarse sediments from road runoff in the channel bed could change the infiltration-exfiltration dynamics along the channel network.”

Lines 459-460, for Mill: “This catchment also has some trails, but there is no evidence of substantial runoff or sediment from trails into the streams.”

Paragraph starting line 482 for Skin addresses the influence of fire and post-fire changes.

Reviewer 2 Report

This study mapped the streams at three semi-arid catchments in Colorado Front Range through field surveys and compared them with topography-derived streams and NHD dataset, in an effort to relate activate flow to potential geologic and climatic factors. The article was well written. The streams were well mapped and analyzed. However, I have a couple of questions:

• Why the two field surveys were done so close to each other? I would assume streams change at different seasons because of changes in snow melting, precipitation, evaporation, etc. Any reasons there would be big changes in those streams within 2 months?
• I see solid results and analysis for stream maps, flow heads, discharge, ADD, GDD. However, the stated impacts from geologic and climatic factors to streams were not supported by solid number and analysis, but rather by assumptions and some literatures. The authors stated “Geology, topography, and climate all interact to drive the spatial and temporal patterns of streamflow, but it is difficult to deconstruct the relative contributions of these drivers at individual study sites.” Thus, I would not chose this part as the focus of the article. You may consider emphasizing other parts of your article in your title.

Other small issues:

Figure 1. It would be great if you could add coordinates to the location figure.

You locate the streams based on NHD dataset. What if they miss some streams? The sizes of the three catchments are relatively small, you can easily check. For large areas, streams could also be checked using high-resolution remote sensing images (e.g. world view, google earth, NAIP images).

Author Response

R2C1: Why the two field surveys were done so close to each other? I would assume streams change at different seasons because of changes in snow melting, precipitation, evaporation, etc. Any reasons there would be big changes in those streams within 2 months?

R2R1: Our initial plan was to survey high and low flow on the seasonal snowmelt hydrographs, but as this was a student project, field work was limited to summer session. The start date ended up being too late to capture peak flow. Please see also R1R1.

R2C2: I see solid results and analysis for stream maps, flow heads, discharge, ADD, GDD. However, the stated impacts from geologic and climatic factors to streams were not supported by solid number and analysis, but rather by assumptions and some literatures. The authors stated “Geology, topography, and climate all interact to drive the spatial and temporal patterns of streamflow, but it is difficult to deconstruct the relative contributions of these drivers at individual study sites.” Thus, I would not chose this part as the focus of the article. You may consider emphasizing other parts of your article in your title.

R2R2: We simplified the title to: “Controls on streamflow densities in semiarid Rocky Mountain catchments”

Other small issues:

R2C3: Figure 1. It would be great if you could add coordinates to the location figure.

R2R3: Latitude and longitude have been added to the figure.

R2C4: You locate the streams based on NHD dataset. What if they miss some streams? The sizes of the three catchments are relatively small, you can easily check. For large areas, streams could also be checked using high-resolution remote sensing images (e.g. world view, google earth, NAIP images).

R2R4: The NHD dataset is actually over-channelized for our study area (see paragraph starting on line 416). We located streams in the field by following flow pathways; it is possible that some channels were missed, but we think this is unlikely because valley bottoms are easy to identify in this steep terrain. We have not seen stream channels outside valley bottoms in this region. Unfortunately, most of these streams are so small that they are not easily delineated using even fine resolution air photos, particularly where there is near-channel vegetation. 

Reviewer 3 Report

In this manuscript, the authors investigated three semiarid catchments during the summer season of 2016 and mapped the streamflow locations for each catchment. The authors used field survey methodology to track the temporal dynamics of the streamflow based on the field observations during two trips. The active drainage density (ADD) was calculated based on the field data, which was compared with the geomorphic drainage density (GDD) calculated from the National Hydrography Dataset High Resolution (NHD HR) flowlines data. The authors also applied the D8 algorithm to the 1-m LiDAR DEM. The results indicated the overestimation of drainage densities in NHD HR dataset and the authors suggested the importance of geologic structure in affecting the flow locations. Overall, I think the manuscript is clearly written, and think it can be accepted for publication after minor revision.

Detailed comments:

Section 3. The Results section should be divided into several subsections, and discuss each point separately with a summarized subsection title.  

Line 261-275. This section needs more in-depth analysis. The difference between the surveyed results and the DEM-derived network should also be related to the geologic structures based on the authors’ assumption. In this case, it is not reasonable to use a uniform threshold given the discussion context of geologic influence. The authors have lithologic maps and can adjust the thresholds for the different geologic zones within each catchment and then compare with the field data. This may also support the author’s point about the geologic structural importance.

Line 276-289. It is not clear that if the study areas have high variations across different years due to the changes of snow cover, precipitation and other climatic factors. To make the conclusion more robust, the authors can show a graph of snow persistence and precipitation of multiple years for those three sites if data is available. In this way, the spatial variability of snow persistence and precipitation across three different sites is more convincing.

Section 4. The discussion section should also be divided into subsections.

Other comments:

line 135 add (P/PET) after index.

line 135. Missing a period after 0.45.

line 154. [52]

Author Response

Thank you for the helpful suggestions. Please see responses in blue. 

R3C1: Section 3. The Results section should be divided into several subsections, and discuss each point separately with a summarized subsection title.  

R3R1: We divided the results into subsections.

R3C2: Line 261-275. This section needs more in-depth analysis. The difference between the surveyed results and the DEM-derived network should also be related to the geologic structures based on the authors’ assumption. In this case, it is not reasonable to use a uniform threshold given the discussion context of geologic influence. The authors have lithologic maps and can adjust the thresholds for the different geologic zones within each catchment and then compare with the field data. This may also support the author’s point about the geologic structural importance.

R3R2: We analyzed whether we could identify differences in contributing area thresholds using lithology in a later part of the manuscript (Figure 8). Although the thresholds do vary with lithology, the differences are not statistically significant. Because of this we do not think we have sufficient justification to assign different thresholds for varying geologic zones.

R3C3: Line 276-289. It is not clear that if the study areas have high variations across different years due to the changes of snow cover, precipitation and other climatic factors. To make the conclusion more robust, the authors can show a graph of snow persistence and precipitation of multiple years for those three sites if data is available. In this way, the spatial variability of snow persistence and precipitation across three different sites is more convincing.

R3R3: Please see R1R1 regarding how 2016 compared to other years. The elevation patterns in precipitation and snow persistence have been similar in other years; however, we are not sure how adding this information would be helpful in the manuscript, as we only have the field streamflow maps for 2016.

R3C4: Section 4. The discussion section should also be divided into subsections.

R3R4: We divided the discussion into subsections.

Other comments:

R3C5: line 135 add (P/PET) after index.

R3R5: This has been added.

R3C6: line 135. Missing a period after 0.45.

R3R6: This has been added.

R3C7: line 154. [52]

R3R7: This has been corrected.

Round 2

Reviewer 1 Report

The authors have made good use the reviews comments. There are some minor comments:

Line 158: You do not need to include the latin name sucesives times. You had already included by the first time in line 143, 144. Only inlude new latin names, but it is not necessary repeat the names already included before.

Figure 6: Why the minimum (blue) in the legend is a point instead of a line?

Figure 7d) Include in the legend that months are represented by numbers or change in the figure the numbers by letters (eg. M, A, M, J, J, A,…)

Figure 8b) Could you explain the legend in more detail? (similar to figure 5). In the boxplot appears the point, triangle and square symbols from the 3 cathments, are they represent the mean? The point can be confused with the outliers points,…please explain it in the legend.

Author Response

Thank you very much for the helpful review. Here are our responses:

Comment: Line 158: You do not need to include the latin name sucesives times. You had already included by the first time in line 143, 144. Only inlude new latin names, but it is not necessary repeat the names already included before.

Response: We have deleted the duplicates of the latin names

Comment: Figure 6: Why the minimum (blue) in the legend is a point instead of a line?

Response: At lower thresholds for channel initiation, the locations with values above the threshold become more disconnected. The points illustrate this patchiness in the stream network for the minimum threshold.

Comment: Figure 7d) Include in the legend that months are represented by numbers or change in the figure the numbers by letters (eg. M, A, M, J, J, A,…)

Response: We added to the figure caption that the months are represented by numbers.

Comment: Figure 8b) Could you explain the legend in more detail? (similar to figure 5). In the boxplot appears the point, triangle and square symbols from the 3 cathments, are they represent the mean? The point can be confused with the outliers points,…please explain it in the legend.

Response: We added to the figure caption: ”In (b) box plots are represented as described in Figure 5, except points are shown for all flow heads.”