I have revied the manuscript entitled Temporal patterns and potential drivers of CO2 emission from dry sediments of a large river. The authors perform several GHG measured with different techniques in the riparian area of a river, during different periods along a year. They complemented these measures with several other variables related to the sediment characteristics. Although I think the content of the study is novel and interesting and the topic is relevant for this journal, my mayor concern relays in the spatial scale of the study. It was performed in a specific reach of a large river, with very specific environmental characteristics. I think this limitation in the spatial scale should be acknowledge and taken in account while discussing the results.

General comments:

#1. In the introduction, the objective of the study was presented as the determination the origin of the CO2 emitted by the dry sediment, saying that a possible source would be the seeping of ground water. When saying “seeping of ground water” the first thing that came to my mind was the presence of a ground water source (aquifer) in the catchment.  

However, in this case, the term “seeping water” wanted to refer to the distance of the dry sediment to the flowing water in the river channel and to the saturated layer. I think this should be clear since the very beginning, to clearly understand the purpose of the different measures that were performed and the experimental design in general.

#2. The title of the study says “…CO2 emission from dry sediments of a large river” but the spatial scale of the study was small, measures where not taken all along the river but in a specific reach. Taking in account the spatial scale of the study, I would change the title to make it a bit less pretentious.

#3. The aim of the study was to elucidate the origin of the CO2 emissions. The response to this question was discussed taking together all different measured variables (section 4.1). However, the relationship of all the measured variables with the aim of the study was not totally clear for me while reading the methods and results section.  Due to the high number of variables, I suggest adding a small explanation of their propose in the methods section. Moreover, finally not only the origin of the emitted CO2 (ground water or respiration) was addressed, but also the drivers of the magnitude of the fluxes. I think that taking in account the extend of this other aspect, it should be also mentioned in the introduction and aim of the study.  

Specific comments:

Introduction

L41: Large rivers with high-flow are also susceptible to seasonal dry (i.e., Albarine river catchment in sud-west France, where more than 80 km representing ~25 % of the catchment and including the most downstream part are intermittent)

L55: and what about long term dynamics?

L65: see my general comment #1 about the use of the term “water seeping water”.

L75-77: related again to the general comment #1, make clear what you mean with “ground water” to clearly understand the aim of the study

Methods

L87: Could you specify the length of the reach? Maybe it’s included in figure S1 but I was unable to download the supplementary material.

L96: I don't understand what do you mean with “the chambers measured hourly CO₂ fluxes”. Did you measure 5 minutes intervals during a period of 1 hour?

L111: Why were these periods chosen? Looking at Figure 1, there were also other time-frames in which the water level was lower than in those selected periods

L152: The significance of the acronym “Bq” is not specified

L154: How/Where (flowing channel? ground water?) were these water samples taken?

L189: The significance of the acronym “g-dw” has not been specified

L207: For how long were the drying and the ignition?

L252-253: As the variability of the data shows a temporal pattern, it would be interesting to analyse the potential drivers of this changes.

L268: And what about precipitation? Why didn't you include it in the mixed model?

Figure 3. The way the hour is indicated in legend (0-20) seems a bit estrange

Results

Related to the general comment #3, at a first reading only the section 3.2.1 of the results seems directly related to the current aim of the study (source of the CO2).

Discussion

The first point of the discussion (4.1) perfectly addresses the aim of the study and summarises all the results obtained. I really like the way it’s written, very clear and direct.

In relationship with the general comment #2, along the discussion, the influence of different factors (dependence of temperature, sediment characteristics, thickness of the unsaturated layer...) in the CO2 emissions were discussed. However, although performed in a big river, the spatial scale of the study was small. Those factors can substantially change along the river, from the headwaters (typically at higher altitude, with more forested and closer riparian areas) to downstream areas (wider reaches, less forested riparian areas, more exposed to solar radiation). I think this spatial scale (together with the already addressed temporal scale) should be at least mentioned in the discussion.

Typing errors:

L207: repeated words: loss after

L243: May

General comments:

The authors investigate CO2 emissions from dry sediments at one site in a large German river. High frequency automatic flux measurements provide an excellent view into the temporal dynamics of CO2 emissions. Additionally, measurements across transects provide information on spatial variability and the contribution of groundwater is assessed using Rn as a tracer. The CO2 emissions are primarily driven by microbial respiration. Furthermore, there interesting descriptions of hysteresis and dark CO2 uptake. The study appear thorough, methods appropriate, and results are well presented and discussed. Unfortunately, I was not able to access the supplementary material.

Specific comments:

• I miss some explicit hypothesis. The aims (1.4) are presented in a broad sense, and test of the groundwater hypothesis is mentioned but so much more data is presented in the manuscript which is why a think specific hypothesis should be included.
• How are the flux chamber data quality checked (L 103)? I think this should be described.
• L 216, following the ANOVA test I would have expected something like a Tukey post hoc test adjusted for multiple comparisons and not repeated pairwise t-tests.
• Regarding LME, how was model selection performed? In general, I miss some more details on the modeling procedure.
• Also regarding LME, I miss a more detailed description of LME results. Currently, only the R2 values are presented but a table (supplementary perhaps) with model coefficients etc. would be welcome.
• Figure 5, I had a difficult time understanding this figure. Could this alternatively be shown using lines in a CO2 flux (y) vs distance (x) type plot. Something is also wrong in the legend, i.e. “NA” values.
• I think the hysteresis results (L 430-432) should be presented the Results section. The hysteresis is interesting and could potentially be further elaborated in the discussion, where there any differences between sites?
• An admittedly minor thing perhaps, but please be consistent with capitalization of axis and legend labels in all figures. Also for figure references, e.g. Figure 5 (L 304), figure S1 (L 91) and Fig. S1 b in (L 135). Please correct throughout the manuscript.
• Date formatting in tables and figures differ, e.g. month-day in figure 5 and day.month.year in table 1, at least month-day or day-month order should be consistent. Please correct throughout the manuscript.

Technical comments:

L28 Replace “largely” with “greatly” or other.

L55-57 Awkward sentence, please rephrase.

L64 Is something missing e.g. “In contrast to respiration”? Please rephrase.

L71 Replace over-saturated with super-saturated

L131 Replace “manual” with “Manual”

L186-188 and 232-234 Same paragraph occurring twice

L230 regard log-transformation, there are also negative fluxes how were they treated.

L242 +/- what – standard error? Please write.

L243 Replace “Mai” with “May”

L260 Just write LOESS smoother with span 0.1. The gray confidence region around the smoothers are confidence intervals or standard errors? And not SD?

L262 What is “HF” in title?

L266 Details regarding modelling, e.g. chamber as a random effect should be in methods.

L267 What are the R2 values for the mixed models? Often they are conditional/marginal depending on whether they include random effects or not.

L291 Replace “spatial” with “Spatial”

L306 Awkward sentence, please rephrase.

L318 Description of texture method should be in Methods.

L550 “Short term temporal dynamics” Maybe replace “dynamics” with “variation”?