This work from Nishiyama and co-authors describes an in-flight calibration campaign of the Ganymede Laser Altimeter aboard JUICE. The team used reflected sunlight from the Earth and the temporal and spatial magnitude of the signal on the GALA receiver APD to assess the pointing of the GALA receiver itself. They conclude that based on comparisons with JUICE telemetry, there is a large (700 microrad) between the pre-flight and in-flight pointing of the GALA receiver. This has tremendous consequences for the use of GALA during the JUICE mission, as the  GALA receiver FOV is only 550 microrad, thus if the receiver pointing is independent of the transmitter (unlikely) the transmitted beam could be completely outside the receiver FOV, rendering the instrument unable to range or provide active reflectance measurements. In general, the paper is very well written and the figures and analytical methods are of high quality. However, there are two main issues I see with this work at present. First, the (unavoidable for this observation) paucity of telemetry data, especially at the key time period related to the onset and falloff of the passive radiometry signal make the analysis wholly reliant on potentially larger errors in the assumed pointing of the spacecraft. The magnitude of the differences between the three interpolation methods (timing delays) are of the order of the overall pointing offset, and the expected temporal behavior for a fixed offset (positive or negative delay in both onset and falloff) is not observed using any of the three methods. The other part of the analysis rests on the structure in the passive data during Earth viewing and the magnitude of the signal during that period. The authors use very small signal variations (0.1 mV change on a 3.1 mV signal) to describe an optimal offset in pointing. This ~3% level of change is smaller than the residuals between the best fits and the data, the change in the ratio of the JANUS and GALA radiometry due to terrain type, and the change in responsivity due to detector temperature effects. This makes both of their lines of argument (temporal onset and falloff, and quantitative signal strength and correlation) suspect. I wholeheartedly agree that there is a strong need for a better round of calibration with high-rate telemetry at other targets, but this paper does not provide strong evidence for the unexplainable 700 urad offset they conclude is present.

Specific comments are below:

Line 22: Paper under review shouldn’t be cited here, especially a self-citation.

Section 2.1: A mechanical description of the GALA instrument would be useful to assess how this offset could have occurred and if it would affect the transmitter and receiver together.

Section 2.1: Table 1 should be moved much further up, or at least the FOV radius of GALA should be mentioned in this section since it is relevant to Figure 1.

Section 2.1: it would be useful to include further details of this Earth-observation campaign including slew rate, targeting pattern, etc. Even though the telemetry is only at the 1-minute level, you should be able to use the commanded slew rate and onboard time to ascertain when exactly the FOV should leave the disk.

Figure 1b: Y-axis should probably changed to “angle between Earth and GALA-Rx-LOS (urad)” rather than “distance between…”

Figure 1b: the switch from linear to log within a single y-axis makes it very hard to assess the transition between the two scales (and is also something I’ve never seen done before). My recommendation would be to break up Figure 1 into two figures. The new Figure 1 will keep Figure 1a as is, change Figure 1b to be fully on log scale, and add a Figure 1c which is the same as 1b but on a linear scale and with Y-limits between 600 and ~1000. The original Figure 1C would then become the new Figure 2, which it perhaps should be already.

Figure 1b: Are the large arrow in the figure, the shading, and the arrows on the left hand side all giving the same information? If so I think you only need one method. Perhaps adding a horizontal line at the angle where the GALA LoS is within the Earth disk. Also is that the FOV fully within the disk or is that the start of the GALA FOV inside the disk?

Figure 1b: Although the three spline methods agree quite well with each other it is very hard to trust the large step between 10:07 and 10:08.

Figure 1c: This portion should come in Section 4, not in section 2 before any of the details of the analysis are discussed. Also, what are the yellow lines and yellow bar in the left hand portion? Is that the slew track?

Section 3.1: You should provide the spectral curves of the Filter 12 and GALA BPF bandpasses in this paper rather than just referencing the Enya and Palumbo papers. That will give a check on whether or not the value from Equation 1 was properly calculated.

Line 170: “Noise Level” should also include units. “N” suggests photon number or rate, but then Nds is described as current.

Section 3.2: Was the temperature of the APD monitored during this experiment? That may have an effect on both the noise and the responsivity (see Xiaoli Sun, J. Bryan Blair, Jack L. Bufton, Marcela Faina, Sigrid Dahl, Philippe Bérard, Richard J. Seymour, "Advanced silicon avalanche photodiodes on NASA's Global Ecosystem Dynamics Investigation (GEDI) mission," Proc. SPIE 11287, Photonic Instrumentation Engineering VII, 1128713 (2 March 2020); https://doi.org/10.1117/12.2545203)

Figure 3: Titles to each Row should be added to indicate which interpolation method was used.

Section 4.1: The use of RMS to quantify the offset does not seem quite right. Assuming you had perfect knowledge of the spacecraft telemetry, the geometry of a fixed offset of the boresight should result in a constant magnitude but inverse sign offset for the onset and falloff of the noise. In fact your results don’t show that (Figure 2). A table of the timing delays for the 6 cases (Three interpolation methods and onset and fallow) should be given somewhere in the paper as well, along with uncertainties. Using the RMS makes it seem like the onset and falloff measurements should be taken combined, but what it actually seems like is the telemetry uncertainty is overwhelming any pointing offset. Also, the RMS values for the three interpolation methods should be given in a table. If the size of the timing offset is of the same magnitude as the spread of the three interpolation methods (it looks like it does from  Figure 3c,f,i) this raises questions about the feasibility of the analysis using such coarse telemetry.

Figure 4: Is there an assumption here that the JANUS radiometry is “perfect”? You are looking at very small changes here (several percent level), so the JANUS radiometry being off at that level seems possible. Was the JANUS radiometry verified during this Earth-imaging campaign?

Line 235: The increase in the noise after 9:47 is quite subtle, as is the decrease in the simulated noise in the pre-launch boresight curve. Based on the generally poor fit between all the simulated curves and the GALA data in general the claim that this portion of the time series is indicative of a boresight offset is not strongly supported.

Line 260: The 0.1 mV difference (out of 3.1 mV total) is only 3%, which is much less than the 9% described earlier from changes in the spectral sampling of different terrain types. Doesn’t this present a problem where the offset could be the radiance ratio of GALA and JANUS rather than the boresight?

Line 260: Again, the temperature effects on the detector need to be taken into account here. A 0.1 mV offset is well within the range of the variability of the detector responsivity as a function of temperature (see Sun et al paper above). Figure 6 in that paper shows the responsivity changes about 1% per degree C, so a 3% offset in voltage could easily be explained by a 3 C difference in detector temperature that is not taken into account here.

Line 267: Aren’t the mean offset and correlation coefficients correlated with each other? They are being discussed as being two independent lines of evidence in the text. If you first normalize all the simulations to the same zero mean offset do the correlation maps change?

Line 277: The qualitative match of the clouds at 9:39 is completely absent for the interpolation case, and the magnitude of the match is very poor and much larger (0.25 mV) than the offset of the pre-launch boresight vector to the GALA noise data.

Line 287: The LOLA boresight issue stemmed from force applied on the transceiver by the thermal blankets under varying thermal conditions. Were the TVAC tests in Hussman et al 2025 performed with the final spacecraft thermal blanketing installed or not?

Line 307: Figure 1C is for the Earth observation, not the lunar observation, correct? Where are the lunar passive radiometry data from GALA?

Line 309: Where is the 110% number coming from? I can’t determine where in Figure 6 this number originates.

Figure 6: The caption should explain what the vertical dotted lines are.

Section 5: It is misleading to call this a “noise” measurement, since it is actually a passive reflectance measurement, which is distinct from the noise from the detector itself. I understand that in the context of pulse detection and ranging the solar background acts as a noise source, but it is confusing (especially in section 5 where you start talking about “maximum possible noise increase”. You should carefully discriminate between passive radiometry measurements and noise during ranging measurements.

Section 5. Can the authors provide a timeline for when future calibration activities will occur? Is there another Earth flyby happening? When will the lunar data be available and used for calibration?

The manuscript provides a unique data set obtained from GALA Earth flyby and describes how they used the data to verify the GALA receiver bore-sight after launch. The result shows a potential bore-sight shift of 700 micro radian compared to the pre-launch measurement. The timing offset between the GALA data and the spacecraft attitude is assumed to originate entirely from the GALA bore-sight misalignment with respect to the spacecraft coordinate system.

However, the manuscript does not give the latencies of the GALA and JANUS camera data which could also cause a time offset in the received data. The direction of the bore-sight shift is not provided (along or cross the bore-sight scan), which could have significant effect on the result.

The timing resolution of the data, ~1 mrad over 25 minutes, or 40 urad per data point, should be fine enough to solve for the bore-sight offset without interpolations. Three interpolation methods are used in the data analysis which give conflicting results.

The bore-sight shift of JANUS after launch could also play a role in the time offset of the data. Although the JANUS bore sight can be solved separately from the time of the spacecraft attitude data and the Earth image, it is not included in the data analysis presented in the manuscript. 

The use of LOLA in orbit bore-sight shift as an example, which is not relevant to this study. LOLA bore-sight offset was mostly in the laser transmitter, which is pulled by the thermal blanket. It was an anomaly rather than a typical case. We hope GALA does not have the same problem.

I feel there are too much uncertainties in the result for the publication. A bore-sight offset of GALA would have a major implication to the JUICE mission. It warrants a more thorough study with all available data to obtain a more certain solution. 

# Review of "In-flight receiver calibration of the Ganymede Laser Altimeter (GALA) 

by passive Earth observations" by Nishiyama et al."

## Overall Assessment

This is an important paper that presents a novel approach to calibrating the GALA instrument using passive Earth observations during the JUICE mission's Earth-farewell campaign. The methodology is scientifically sound and has significant implications for the accuracy of future GALA measurements at Jupiter's moons. The paper is generally well-structured and addresses a critical calibration need for the mission. There are however, some shortcommings that need to be addressed. The first thing is the insufficient description of the observation campaign. The second issue is the rather qualitative approach to error handling. I recommend **major revisions** before acceptance.

## Major Strengths

- The paper clearly articulates the scientific question and presents a well-designed methodology to address it

- The approach of using Earth observations for boresight vector calibration is innovative and valuable for the mission

## Major Concerns Requiring Revision

### 1. Spacecraft Attitude Uncertainties

The most significant concern is that the authors assume spacecraft attitude is accurately known at 1-minute intervals without 

adequately addressing potential uncertainties in these measurements. The paper states: "during the Earth-farewell campaign, 

the Juice attitude data were acquired only at one-minute intervals" (line 117), but does not quantify the expected error in these attitude measurements.

The manuscript mentions that "discrepancies between simulations and observations remain (Fig. 4), perhaps due to uncertainties in spacecraft attitude knowledge" (line 280), but this needs to be quantified rather than merely acknowledged.

The authors need to:

- Explicitly state the expected accuracy of the spacecraft attitude determination system

- Incorporate these uncertainties into their error analysis

- Demonstrate how attitude uncertainties propagate through their interpolation methods and affect the boresight vector determination

- Address whether the observed discrepancies between simulations and observations could be explained by attitude uncertainties rather than a boresight offset

### 2. Clarification of the Earth-farewell Campaign

The paper refers repeatedly to the "Earth-farewell campaign" without adequately defining it. 

Readers need to know:

- What specific dates and times this campaign encompassed

- The spacecraft's distance from Earth during observations

- The spacecraft's velocity relative to Earth (dV)

- Whether the spacecraft was rotating to keep Earth in the field of view

- Why is it called "farewell" when JUICE will return for additional Earth gravity assists in 2026 and 2029?

This information is critical for understanding the observational context and should be added to Section 2.

### 3. Inconsistent Terminology

The paper uses "Juice" throughout the text. As JUICE is an acronym (JUpiter ICy moons Explorer), it should consistently appear in all capital letters. 

The manuscript refers to the GALA background observations as "noise". Although this term makes sense for laser observations, it is inappropriate in this context, where the GALA background represents the signal to be measured.

## Minor Issues Requiring Revision

### 1. Grammatical and Typographical Errors

- Line 110: "5 degree" should be "5 degrees"

- Line 317: "may be also performed" should be "may also be performed"

- Line 331: "unknown uncertainty in the analyzes" should be "unknown uncertainties in the analysis"

- Line 340: "GEochemistry,and" should be "GEochemistry, and"

- Line 351: "instrument teams proprietary period" should be "instrument teams' proprietary period"

- Line 355: "aquirement" should be "acquisition"

### 2. Figure Interpretation

Figure 2 shows an asymmetric response with the GALA noise increasing slowly (~20s) after the first limb crossing but 

dropping rapidly (~5s) during the second limb crossing. The authors should explain this asymmetry, which may relate to the spacecraft's motion relative to Earth or instrument characteristics.

### 3. Methodology Clarification

The paper would benefit from additional details about:

- The specific SPICE kernels used for attitude reconstruction

- The exact interpolation routines applied to the quaternion data

- How were the criteria in Figure 5 combined (line 268). Were they weighted equally?

    - I gather it should be something like this:

    Where max( corr(sim(t),obs(t))) at boresight timing residuals RMS < 15s AND |Mean noise difference| < 0.25mV 

    - That does not fit with Fig.5. The point of highest correlation does not coincide with marked "most likely point" within bounds given above

### 4. On Discussion and Conclusion

The authors state that their method can be applied to other laser altimeter instruments used in different missions. In particular, they mention BELA. This made me wonder, as BELA could not perform any observations during the Earth fly-bys as it was pointing towards the transport module. At Mercury, timing limb crossings could help to narrow down any boresight offsets; however, the much lower albedo range of Mercury's surface compared to Earth would likely present an obstacle to the correlation part of the authors' algorithm.

## Conclusion

This paper presents an important calibration technique for the GALA instrument that will improve the scientific results of the JUICE mission. While the methodology is fundamentally sound, the manuscript requires revisions to adequately address uncertainties in spacecraft attitude and provide clearer context for the Earth farewell campaign. Once these revisions have been made, the paper will make a valuable contribution to the field of planetary laser altimetry. I recommend acceptance following revisions to address the above concerns.