This manuscript uses a novel data set from the ESA Swarm mission to examine the extent to which similar magnetic perturbations on different scales, which the manuscript relates to field-aligned currents (FACs), were observed by two of the spacecraft as there azimuthal and along-track separations were varied. The results back up previous results that meso-scale perturbations (~100 km scale) are temporally steady up to a few tens of seconds and beyond 20 km azimuthal extent. There is a more in-depth study of small-scale perturbations, which are shown to be far less steady, particularly when the magnetic perturbations are large and/or there is larger solar wind driving or geomagnetic activity. I find that the results are largely compelling, however I have two major concerns: (1) the study uses a 60 min window to capture the variability of the magnetic field which has been pre-filtered in various pass bands from 1-3s to 39-60s - this fixed window length may introduce an inadvertent bias to the results; (2) the short-period variations that are not well correlated between the two spacecraft passes are discussed as unstable field-aligned currents, but as this is a time-varying magnetic field measurement rather than a direct measurement of the current, it is unclear that this is the only possible interpretation of these results. Given that point (1) above requires some study into the methodology, I recommend that this manuscript be considered after major revisions.

Major Comments:

1) On the methodology

As noted above, I have a concern with regards to the filtering and windowing of the data in the analysis performed. The manuscript details that various time-scales of magnetic field fluctuations are considered in order to examine different scale sizes of potential field-aligned currents (derived from the spacecraft motion in that time). However, the window used to calculate the cross-correlation between the two spacecraft and RMS values of the field remains fixed. As a result, for the smallest scales, there are between ~20 and 60 possible perturbations within the 60s window whereas for the largest scales it is 1-1.5 perturbations. If a some of the small-scale perturbations vary, then the cross-correlation will drop, whereas the whole perturbation has to vary for the largest scales. This, I feel, naturally results in fewer good correlations at the smaller scales. I suggest that the authors explore the sensitivity of their results to varying the length of the window over which they do the cross-correlation. For example, are does the success ratio for the 1-3 s scales increase for a 15 s cross-correlation window. The authors may also need to consider whether the correlation coefficient threshold needs to be increased with the smaller window size such that the P-value is consistent.

With regards to the correlation coefficient used, I note that a correlation coefficient of 0.75 means that only 56% of the variability observed by one spacecraft is observed by the other. This feels quite low to say that the observed current systems are stable. There is also no consideration of changes of amplitudes of current that might result in relatively high cross correlations but differences in RMS values.

(2) Discussion of results

This study examines whether or not magnetic perturbations seen by one of the Swarm spacecraft were also seen by a second spacecraft passing through at a slightly later time and at some azimuthal separation. If this is the case, then it is reasonable to assert that these perturbations arise from the quasi-stationary field-aligned current system. However, when the conditions set are not met it does not necessarily follow that it is an unstable current system. I feel it is better to described the observations as temporally or spatially varying magnetic field perturbations. Previous studies using data from Swarm and other spacecraft (e.g. the cited Ishii+ 1992 study and the Pakhotin+ 2018 doi: 10.1002/2017JA024713 study) have shown that large amplitude small-scale magnetic field perturbations may be associated with Alfven wave activity. 

Minor comments:

Line 44: "at the ionosphere plays a role"

Line 76: please provide a reference for the Swarm L2 data product

Line 77: Swarm A and C do not fly side-by-side, as is a key point of the manuscript. I believe that for the L2 product, one of the datasets is lagged so that it is treated as if they are side by side.

Line 86: As noted above, I was under the impression that the dual spacecraft product is a 2D curlometer, not the mean of two single spacecraft FAC estimates as implied by this line.

Line 90: "the range dual-spacecraft FAC estimates are valid" ?

Line 116: Swarm A and C are not side-by-side by are lagged by a few seconds

Line 122-128: The Zhou et al figure should be referenced here. In fact, I think the Zhou et al figure (or similar) should be included in the manuscript as it is crucial to the study.

Line 178: " For these example passes, we use a 60 s sliding window...". When I first read this I was confused as I thought the whole interval shown was 60 s.

Line 238: "look at the variable magnetic field signal". Given that the introduction discusses the separation of waves and FACs, the current wording was confusing

Line 246-256: It would be helpful to mark some of the intervals of interest on the figure.

Line 248: "practically all magnetic fluctuations above 20 s can be..."

Line 395: give the zonal length in km as well as seconds

Line 397: I don't understand what is meant by "the time  between samples is more decisive for the occurrence ratio"

Line 414: My reading of the figure is that d_cross < 3 km is where the ratio exceeds 50%. Please confirm

Line 426-437: Consider also the Pakhotin et al (2018) study that examined Alfven waves using Swarm

Line 444: At line 415, delta-t was noted as 16 s, not 18 s. Please confirm and be consistent

Line 476: specify the size range where it says "in this size"

Line 536: how were the quiet days selected for the calculation of the mean merging field and what is the value of this mean?

Line 540-542: Are these results from another paper? The results shown do not show large amplitude FAC structures are prone to instability nor the the FAC current density largely depends on driving.

Figure 10 and discussion: It would be useful to include an indication of the range of values at each epoch and whether the higher values for "deselected" events are statistically significant.

Line 603 - 607: it is not evident from this study that large amplitude currents are unstable, just that large amplitude small-scale magnetic perturbations do not meet the criteria for stable FACs. They may be signatures of wave activity, something which is not examined in this manuscript.

Line 625-627: this is speculation with no citations and no discussion in the preceding manuscript, so I suggest removing this sentence.

General comments:

The manuscript presents a study of small- and meso-scale field-aligned current (FAC) structures using Swarm satellites. The study presents interesting statistical properties of these currents obtained from the close and unique positions of the Swarm constellations. The study finds that merging electric field can affect FAC structures and densities.  The statistical properties reported in the manuscript may be useful for researchers of FACs.  However, there are serious concerns, which shall be divided into two parts: (a) presentation and (b) technical.

Presentation: The labels on all the figures are difficult to read because they are too small.  Please increase the label font sizes on all figures.  The manuscript contains many typographical errors, some of them are pointed out below, but the authors should go through the manuscript carefully to check the English grammar and errors. Finally, the organization can be improved.  Section 5 (discussion) does not read like a discussion, but rather a continuation of the Section 4 (results/analysis). The discussion section usually provides context, interpretation, physical insights gained from the results (see below). 

Technical: The manuscript does not provide much physical interpretation of the results, which usually goes into the discussion. What causes the small- and meso-scale FAC structures?  How does merging electric affect the FAC structures on the dayside and more curiously on the nightside? What is the mechanism? What causes the local time variations?  It would be nice if the manuscript can discuss some these questions. Perhaps, the following example may help.

FACs of different spatial scales at different MLTs may be caused by different processes. Generally, large scale FAC structures such as R1 and R2 (Iijima and Potemra, 1976) are fairly stable, but superimposed on these large FAC structures are small and meso scale FAC structures that may be transient in nature.  For example, in the afternoon near the open-closed boundary, large scale upward FAC with spatial scale hundreds of km (but can be as small as several tens of km or as large as 1000 km in some cases depending on the solar wind condition) can be attributed to the velocity shear between the solar wind and magnetospheric plasma at the magnetopause boundary layer (Lyons, 1980, Siscoe, 1991, Echim et al., 2008, Johnson and Wing, 2015; Wing and Johnson, 2015).  This large scale upward FAC structure is nearly always present because the velocity shear at the magnetopause boundary is always present (but the FAC thickness and strength may vary depending on the solar wind condition).  Superimposed on this large scale FAC are small/meso scale FAC structures of tens of km (~50-70 km) in the afternoon sector can be linked to the KH vortices at the magnetopause boundary layer at dusk flank (Petrinec et al., 2022; Johnson et al. 2021).  These KH vortices are not static but rather they move anti-sunward with the solar wind and hence the small and meso scale FAC structures associated with these vortices are not static either. Perhaps, a discussion along this line in the introduction and/or discussion section may help the readers appreciate how the statistical results presented in the manuscript may help improve the physical understanding of the magnetosphere and ionosphere and help provide context.

Specific comments:

1. Line 45. “dominate” should be “dominant”
2. lines 70-71, the sentence is a bit awkward. Would the following capture the meaning better?

The longitudinal extension of the small FAC sheets on the dayside was found to be comparable to the latitudinal width, but 4 times larger than the latitudinal width on the nightside. 

1. line 91, “erose” should be “arose”.
2. lines 134-136, What are the assumptions here? Would this technique only work at high latitude, e.g., in the auroral oval? If so, please state. 
3. lines 147-148, do the authors mean FAC intensity (unit = A/m) or density (unit = A/m^2)? FAC density is usually obtained by dividing FAC intensity by the width of the FAC (see for example, Ohtani et al., JGR, 2005).
4. line 184, should FAC density be closer to 15 microA/m^2 (100/7.5) rather than 10 microA/m^2?
5. line 240, “base”  should be “basis”
6. Figures 3-5 (and other figures), rather than displaying the six period ranges in sec, would it be more useful and practical to display them in spatial scale, i.e., km? Most readers would care more about the spatial scale of FAC structures rather than delta t.
7. lines 279-280, how do the authors remove the local time and seasonal effects?
8. lines 282-283, “larger ratios of stable FACS are obtained in the afternoon sector”. Perhaps, something is missed. Where can we see this in Figure 5? Figure 5 shows that afternoon sector has lower ratios.
9. lines 310-311, would 10-20 km is more precise for “1-3 s periods “than “10 km”?
10. lines 313-319, In Figure 5 there are local time variations as well as temporal (UT) variations. The authors would like to attribute the minimum FAC ratios to 4 Nov 2021 storm. However, the minimum is located in the afternoon sector or morning sectors. Is there an ambiguity in local time vs. UT (storm) effect? 
11. line 502 and elsewhere in the manuscript, the terms “selected” and “deselected” are a bit confusing. What do the authors mean by “deselected” and “selected”.  Can these terms be described more clearly?
12. line 538, “cause” should be “caused”
13. Figure 8, what causes the gap near the top right? Is it the result of the constraint on the selected periods? Please explain. 
14. Section 6. The manuscript provides many statistical results, which, at times, are hard to keep track.  It would be nice if the authors can provide a table, which can summarize the various properties of the FAC structures.