Dear authors, dear editor

This is a very interesting article on sediment dynamics in ephemeral streams. It is based on two different measurement techniques that are currently very popular in our field of research and which are combined here.

This parallel use of smartrock and seismic measurements is new and extremely exciting in bedload transport research. Quantifying seismic energy in the range where (artificial) bedload particles transition from vibration to actual displacement will certainly also be helpful for future studies using seismometers along torrents and mountain rivers outside semi-arid areas. The authors' focus on the moment in the discharge hydrograph when the water level/discharge rises sharply is clearly emphasised. Such rises occur, for example, during flash floods. Experts in the field are well aware that studying sediment transport dynamics during such rapid rises in discharge is of fundamental importance. This is because, compared to steady flow conditions, sediment transport rates are higher during the arrival of flash floods, when there is an intense and rapid increase in water levels. The fact that high sediment transport rates also greatly increase the risk of damage along watercourses could be explicitly mentioned (in the Introduction or Discussion sections) in order to also convince non-experts of the importance of this research.

I strongly support the publication of this article in Earth Surface Dynamics. It represents a step forward and will be important for the research community dealing with bedload transport. I can't think of any significant reasons why this manuscript shouldn't be published quickly. And below, I would like to list only two points of a more general nature.

However, as is well known, the devil is in the detail. This is evident in the considerable number of specific questions and comments attached as a supplement. However, I am confident that the authors will be able to respond to these comments and provide clarifications relatively easily, thereby strengthening their manuscript further.

[1]    In the Introduction section (L44-46), the authors state that "Rapid increases of stage are often unpredictable, infrequent, and short-lived, posing challenges for traditional bedload monitoring methods." Further down the page (L59-61), several of these traditional measurements are listed in the text (samplers, impact sensors and hydrophones). In an innovative approach, the authors decided to use smartrock technology alongside seismic measurements. This combination is very well suited for improving the description of bedload transport dynamics during rapid increases in discharge in ephemeral streams.

However, in addition to a detailed description of the dynamics, it would also be interesting to be able to make quantitative statements about bedload transport rates or transported volumes during a bore. It might therefore be interesting for readers to learn briefly about the challenges or limitations of traditional methods of measuring sediment transport in semi-arid and arid regions especially in very unsteady flows. A reference to the relevant literature would be helpful. Perhaps to the work carried out in the nearby Nahal Eshtemoa (e.g. Halfi et al. 2018; https://doi.org/10.1051/e3sconf/20184002036)? Could the seismic-smartrock combination also be used along Nahal Eshtemoa, where both direct and indirect sampling methods were employed? Provided, of course, that the slot samplers and pipe/plate microphones are still opera­tional. I realise that there is not much space for this in the Introduction section of the article. I would therefore ask the authors to consider whether this point could be covered in a new, brief sub-section of the discussion.

[2]    The authors introduce water level stages (1–3) for both monitoring techniques applied in this study. Stages 1-3 based on gyro velocity ratios are shown in Figure 5 (and introduced in the text in lines 213-214). Stages 1-3 based on PSD ratios are shown in Figure 9 (and introduced in the text in line 272). The only place in the Results section where the authors point out that there are two differently defined sets of stages (1–3) is in the caption of Figure 9, lines 289–290.

To avoid confusion among readers, I strongly recommend that these two differently defined sets of water level stages (1 to 3) also be named/labelled differently. This is all the more important given that yet another set of stages is introduced in Figure 10 (and in the text on line 280). These are also defined based on the PSD ratio. However, they refer to the relative water depth and apply to both study sites.

Figure 11 concludes the presentation of the results by showing how the PSD varies with smartrock gyro velocity. Here, the authors once again introduce stages (this time two, Stage 1 and Stage 2). These have no direct connection to the set of “stages” mentioned above. They should definitely be designated differently.

I do realise that the authors clarify this right at the start of the Discussion in section 4.1 and distinguish between the three stages determined using the PSD ratio and gyro velocity ratio. Nevertheless, I would like to insist on my comment regarding the need for greater clarity in the Results section. I think the manuscript would benefit from this.

In summary, I would be glad if this article could be published in ESURF. I suggest the manuscript be re-evaluated following moderate revisions. Attached is a document containing specific comments on the text, figures and tables, as well as technical corrections. I would ask the authors to consider these.

Summary: The paper titled 'Integrating smartrock and seismic monitoring to investigate bedload transport dynamics during rapid increase of stages in ephemeral streams' by Matanya Hamawi et al., investigates the bedload transport in two different ephemeral streams using smartrocks and seismic measurements. As main results, the authors observed three different scenarios:

At shallow depth, a rapid increase of stage enhanced higher bed material and vibrational movements compared to steady flow [1], while at intermediate relative depth, this rapid increase in water level decreases the effect on bedload activity [2].  However, at greater relative depths, the bedload movement is enhanced again [3]. The paper is well-written. It tackles an interesting problem related to sediment transport dynamics and measurements during different floods/unsteady flows. I believe this study is suited for publication after some improvements.

[1] From lines 55-57. What are the limitations presented in the previous studies performed in ephemeral rivers (in arid and semi-arid regions)? It would be better if the authors could briefly explain them in this part.

[2] Line 73 – The authors used the term 'different morphologies'. Can you be more specific? What are the differences between the two studied sites?

[3] Lines 72-73: ‘we aim to develop a more comprehensive understanding of how rapid increase of stage influences bedload transport processes’. Which processes are you referring to?

[4] Lines 75-76. I would suggest rephrasing the last sentence of the introduction.

[5] I suggest adding the timeframe (instead of winter) to lines 80-81.

[6] Lines 161 - 162 – The authors compared the results with the steady flow. How do you define/measure the data on the steady flow condition?

[7] Lines 193-194. It was not clear to me how you selected the threshold values for each site. I suggest briefly explaining it.

 [8] Line 210 – I suggest to be more specific on the amount of (or %) of increasing compared to steady flow conditions.

[9] Figure 5 and lines 213-214 :  After line 214, the authors need to explain each one of the stages, and how they separate those regions (even though you discuss it later).

[10] Figure 6 – How did you define the quasi-steady flow conditions?

[11] From figure 9, it seems to have very similar values compared to steady flow conditions on stage 3. I suggest adding some thoughts on that in the results section as well.

[12] Why not add the same two images for both sites in Figure 11? Instead of adding in the SI.

[13] In the last part of the discussion, I would consider adding some explanation on the limitations using the gyro velocity information, as it adds an external source of motion (vibrational).

The paper presents an expertly crafted and well written study that directly investigates bedload transport, a core process in Earth surface dynamics, focusing on the critical but understudied context of unsteady flow. It tackles a recognized methodological challenge: monitoring sediment transport during unpredictable, high-energy flash flood events where traditional methods fail, which is fundamental to geomorphology and fits perfectly within the scope of ESurf.

The authors present here a novel application of two key tools used for assessing bedload dynamics: seismic sensors and smartrocks. The choice of a using these during flash flood bores in ephemeral streams, is quite challenging and the authors deserve compliments for demonstrating this application in such difficult conditions.

The authors find that rapid stage rises enhance bedload transport in shallow and deep water, but not at intermediate depths, and that these transitions are governed by relative depth (`h/d₈₄`).

However I am wondering if this substantial conclusion that has been validated for these conditions (characterised by these highly unsteady flows in these few ephemeral streams) can confidently be extended to broader flow conditions with implications for scaling these processes across broader and more different river types.

The methods for data collection (seismology, smartrocks, stage logging) are state-of-the-art and clearly described. However, the key assumption that the frequency band best correlated with total gyro velocity is the best proxy for bedload transport might need more clarification (because the total gyro velocity for a rock in a natural setting accounts for both vibrations and downstream transport not only the latter). For example one cannot reliably distinguish between different types of movement that have similar rotational speeds with the gyro velocity alone. Indeed there are studies in controlled (uniform, steady flow and fixed particle pocket) environments that may use this measure as a surrogate that may suitably work, but in highly unsteady flows in the field where the particle can be found in any random pocket geometry, there may be less confidence that the total gyro velocity is either a rock rolling smoothly downstream (linear transport by rolling), or a rock violently spinning in its resting pocket (high-energy vibration - without transport). For instance, using equation 1 discards all the directional and temporal pattern information that is essential for distinguishing movement types - maybe a caveat can be given to discuss this limitation and that perhaps as a next step such cases, if any exist, can be filtered out, by direct consideration of the gyro information from each individual axis. As a result the "Displacement" category might be contaminated by such high-energy vibrations and localized rearrangements that do not contribute significantly to downstream bedload flux. This may imply that the gyro-based evidence for enhanced "transport" during rapid stage rises (especially in Stage 1) might be partly evidence for enhanced "agitation”.

Further since the "total gyro velocity" used to calibrate the seismic signal includes these undistinguished movements, the seismic method is calibrated to detect general bed activity and energy dissipation, not pure bedload transport (but also any high energy vibrations not resulting into transport).

The 30-second median filtering of gyro data is a valid smoothing technique, but the assumption that it adequately captures transport dynamics overlooks the loss of short-duration high energy transport events. Of course for larger rocks, the longer the duration for gyro-data is meaningful but it would be nice to demonstrate the utility of the current choice (and optionally how it may vary for different rocks).

The data processing steps for both seismic (PSD calculation, response removal) and gyro (magnitude calculation, thresholding, binning) are described with sufficient detail for reproduction.

The authors use relatively easy to access field data eg the stage, based on the methodology described in the manuscript (primarily in Sections 2.1 and 3.1, and Supplement S2) and acknowledge that stage alone is inadequate for predicting a bed surface grain-scale process due to the turbulent flow field. The assessment of steady flow versus rapid increase of stage is performed through a quantitative analysis of the stage hydrograph data, using a fixed, absolute rate-of-rise threshold: this can introduce an asymmetric relative error, disproportionately affecting low-flow stages.