Calibrating JANUS: an overview of the calibration pipeline and what we learned from the Moon-Earth flyby data

Agostini, Livio; Matz, Klaus-Dieter; Tubiana, Cecilia; Della Corte, Vincenzo; Penasa, Luca; Bilotta, Thomas; Sarti, Francesco; Read, Matthew; Hviid, Stubbe; Mottola, Stefano; Aboudan, Alessio; Trauthan, Frank; Politi, Romolo; Aye, Michael; Hueso, Ricardo; Lucchetti, Alice; Kersten, Elke; Belgacem, Ines; Dietz, Angela; Evill, Ry; Bell III, Jim F.; Zinzi, Angelo; Castro-Marin, Jose María; Roatsch, Thomas; Schmitz, Nicole; Patel, Manish R.; Lara, Luisa M.; Portyankina, Ganna; Palumbo, Pasquale

doi:10.5194/egusphere-2026-2032

Preprints

https://doi.org/10.5194/egusphere-2026-2032

Preprints

14 Apr 2026

| 14 Apr 2026

Calibrating JANUS: an overview of the calibration pipeline and what we learned from the Moon-Earth flyby data

Livio Agostini, Klaus-Dieter Matz, Cecilia Tubiana, Vincenzo Della Corte, Luca Penasa, Thomas Bilotta, Francesco Sarti, Matthew Read, Stubbe Hviid, Stefano Mottola, Alessio Aboudan, Frank Trauthan, Romolo Politi, Michael Aye, Ricardo Hueso, Alice Lucchetti, Elke Kersten, Ines Belgacem, Angela Dietz, Ry Evill, Jim F. Bell III, Angelo Zinzi, Jose María Castro-Marin, Thomas Roatsch, Nicole Schmitz, Manish R. Patel, Luisa M. Lara, Ganna Portyankina, and Pasquale Palumbo

Abstract. JANUS (Jovis Amorum Ac Natorum Undique Scrutator) is the high-resolution multispectral camera on board of the ESA Jupiter Icy Moons Explorer (JUICE). The mission, launched in April 2023, is currently traveling towards the Jovian system where it will characterize Jupiter, its environment and its icy moons, primarily Ganymede. During 19–20 August 2024 JUICE flew by the Moon and the Earth performing a Lunar-Earth Gravity Assist maneuver (LEGA) that provided a unique opportunity to test the instruments in mission-like conditions. In this paper we describe the status of the calibration pipeline applied to JANUS data acquired during LEGA, the lessons learned by the analysis of the calibrated data and the open points.

How to cite. Agostini, L., Matz, K.-D., Tubiana, C., Della Corte, V., Penasa, L., Bilotta, T., Sarti, F., Read, M., Hviid, S., Mottola, S., Aboudan, A., Trauthan, F., Politi, R., Aye, M., Hueso, R., Lucchetti, A., Kersten, E., Belgacem, I., Dietz, A., Evill, R., Bell III, J. F., Zinzi, A., Castro-Marin, J. M., Roatsch, T., Schmitz, N., Patel, M. R., Lara, L. M., Portyankina, G., and Palumbo, P.: Calibrating JANUS: an overview of the calibration pipeline and what we learned from the Moon-Earth flyby data, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-2032, 2026.

Received: 09 Apr 2026 – Discussion started: 14 Apr 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-2032', Anonymous Referee #1, 09 Jun 2026
General comments
The manuscript nicely describes the calibration process for the JANUS data, supported by an overview of the existing calibration pipeline flow focusing on the radiometric and geometric calibration. In addition, the study provides an outlook into future already identified improvements to the JANUS calibration pipeline, as well as identifies clearly investigations to be continued on the calibration following effects/artifacts detected in the LEGA data. This is an important work to understand and analyse JANUS data.
In my opinion, the manuscript goes beyond providing an overview of the calibration pipeline and illustrates the calibration process for the JANUS data, which is why 1) I believe that the manuscript would benefit from expanding slightly some of the sections describing by the underlying input data and/or processes to the study (sections 3.2, and 3.5 for example), 2) I would encourage the authors slightly change the wording for the study to refocus the scope on the calibration “process” (rather than “pipeline”, e.g. in the title, main text). Unless the calibration pipeline is released to the science community, I also do not personally see the need to mention “in the next version of the pipeline” in several places in the manuscript.
Please find below my detailed comments.

Comments on main text
Lines 1-2: “on board of the” should be “on board the”.

Line 6: expand on what those “open points” are exactly about.

Lines 12-13: “limited in-flight data volume” should be “limited in-flight data volume available for science observations” (or similar)?

Line 23: The “space environment and operating conditions” are certainly more “similar to those expected during the science phase” than during on-ground testing. Yet, those conditions are significantly different than those that will be encountered in the Jovian system (radiation, temperatures, operational constraints, ...). I would suggest to rephrase slightly.

Line 25: “[…] observed by a large number of different sources […]”. What is exactly meant by “sources”? (Instruments on other missions?)

Lines 29-32: This paragraph is unclear. The paper describes “the current status of the JANUS calibration pipeline” (line 29) but the “detailed description of the current calibration pipeline and how the products have been generated” is not within the scope of this work (lines 30-31). I would suggest to rephrase, since I believe that the pipeline is actually conceptually described, the calibration process is exposed, but the technical description of the pipeline (code) is not the scope of the paper.

Line 79: unit of the calibrated data should be W/m^2/sr (written as W/m2/sr).

Section 3.2: Too little is said about how the saturation level mask is obtained. The manuscript would benefit from more information on this aspect,

Lines 90-91: the meaning of the last sentence is unclear to me. Where in the pipeline/process are the saturated pixels actually masked? How those are identified in the data? What exactly is the “flag” mentioned in the sentence?

Section 3.3: How is the bad pixel map looking like? How many pixels are affected (hot pixels, dead pixels)?

Section 3.4: I am missing the actual temperature range of the “first dataset” and “extended dataset”.

Missing section between 3.4 and 3.5? Figure 1 refers to a filtering step called “out of knots range” which is, I believe, not described in the manuscript. Please expand.

Section 3.5: “The exact method and considerations used for the determination of calibration products are not the subject of this work”. I would strongly encourage the authors to provide more information about such methods and considerations used to derive the radiometric calibration. Since a significant part of the manuscript deals with calibration uncertainty of various sources, it would fit perfectly well here and set the calibration context clearly for the rest of the study.

Line 163: “this step is left to the user according the objective of their research”. In order for a user to perform such correction for the Earth, the instrument transfer function must be provided, as well as the spectral radiance of the calibration light source.

Section 3.6: The spectral correction of the Moon data is performed using a full disk irradiance spectrum of the Moon computed using the LIME model. It is not fully clear to me how was the spectral correction process performed. JANUS observations were performed either close to the Moon or at distance (Earth-Moon observations), but in either case with a relatively high phase angle (line 236). It would be interesting to expand a bit more on the method used to determine the spectral correction factors given the particular geometry of observation (partial illumination of the Moon) and also the fact that the disk itself potentially integrates lunar terrain of different natural types. How is this taken into consideration?

Line 238: How does the spatial resolution of the Selene DTM compares with JANUS’?

Line 239: Remove “already” (redundant with “previously”)

Lines 243-244: “JANUS acquired 19 August” should be “JANUS acquired on the 19th of August”.

Line 334: “ESA SPICE operational” should be “ESA SPICE JUICE operational”

Line 344: “in bottom part” should be “in the bottom part”

Line 345-346 refers to the quantum efficiency degradation due to the silicon material, but silicon is not explicitly mentioned in Section 2. To be added to Section 2?

Section 5.1.8: Please provide the ranges of exposure times and radiance levels for completeness. See also comments for Figure 22.

Line 411: Missing dot.

Line 415: “Figure 26 and 25” should be “Figures 25 and 26”

Line 422: “ops kernel” should be “juice_ops kernel” for consistency.

Line 430: “executed the science planning” should rather be “executed the LEGA science operations plan” (or similar).

Line 431: ”84 of the Moon-Earth system” should be “84 Moon-Earth system images”.

Line 441: “Jupyter” should be “Jupiter”.

Comments on figures
Final dots are missing in most figure captions.

Figure 1:

Recall the processing levels in the figure caption.

In the caption, “openness” should rather be “clarity”?

I am also missing information on the pipeline inputs expected at every processing step (similar to Figure 8).

I am missing the step described in section 3.6 in this figure.

Figure 2: Is there an explanation for the column gradient seen in the saturation map?

Figure 4: Is the content of figure relevant to LEGA only, or does it cover JANUS images prior to LEGA? From the Conclusion section, it seems that 400-500 images have been acquired during LEGA but the acquisition ID shown here goes up to 600+. If the data are from different periods of time, it would be good highlight those in the figure.

Figure 5: Fonts are small, please increase these.

Figure 7:

I am having a hard time to interpret this figure. Is the resulting distortion a lower left - upper right “shear” of the image? I think it would be useful to illustrate the distorted VS undistorted shape of the image related the optical distortion.

The pixel origin of the distortion maps seems different from some of the other images shown in the manuscript (see Figure 2 for instance).

Figure 8: Based on the pipeline sketch, the “geo” products do not include the quantities associated to an intercept with a surface (“geo” being before the dashed line). Is this correct, knowing that the geometry cube in the figure includes such quantities? (Or am I misinterpreting the figure?)

Figure 13 and 15: Upper case “P” missing for “Partially calibrated”

Figure 13: Why choosing the central column for the vertical profile? It should be close to a tile edge subject to variation in intensity. Wouldn’t it be better to extract 4 profiles (one per tile in the middle of the tile, or one average per tile)? Profile locations could be shown on the image and actually display on the right hand side of the figure (outside the image).

Figure 14: Add legend for the dashed line (terminator?)

Figure 16 could be annotated to better show the multiple artifacts visible in the image (SDBS, ghost, straylight)

Figures 19 and 20: Those are very busy plots.

I would suggest to move the profiles outside the image.

Profiles color is cyan.

Add a title with the filter number, and the central wavelength (+ exposure and median signal currently on the image).

Figure 21: Missing Y-axis label.

Figure 22:

Symbols are hard to read, use one color per test case? It would be easier to read panel a.

Legends need to be explained (in the caption and/or text). I guess that COLD refers to the setpoint (based on the text), what do HOT and MIDDLE stand for?

Panel c has a different test case (SPECIALFILTERS) which is not explained.

There is no COLD setpoint case for ITF vs TEXPO, why?

Figure 24 needs to be better explained, at least in the caption (acronyms not defined). I do not understand it as it is now.

Figure 25 image quality needs to be improved. Also please note the missing opening bracket in the titles.

Figure 26 caption: “coherent” should be “consistent”.

A number of figures could be combined in a single figure for better readability. This is the case (in my opinion) of the following figures:

9 and 10 address both the rolling shutter effect.

13 and 14 address both the straylight effect (one is the effect and the other illustrates the geometry).

17 and 18 address both the absolute radiometric calibration

19 and 20 address both the diffusion, using different filters.

Comments on tables
Final dots are missing in some of the table captions.

All tables: Use the same naming convention for the filters between the text and the tables (1 -> F1)

Table 1: I think the parameters refer more to the JANUS specifications than actual performance parameters.

Table 3: Recall the meaning of S_moon and C_filter in the caption.

Table 4 (and data generated):

Should the distance to the target limb be one of the geometric data provided by JANUS? For limb observations, this would be very relevant.

Should there be pixel corners lat/lon coordinates provided as part of the JANUS geocube? (I assume that the coordinates mentioned are those of the pixel centers).

Table A1. Possible changes to the figures to be reflected in this table. Will those calibrated and partially calibrated data be available in 2029 with the same Image ID?

Bibliography
Please update the bibliographic references with published article DOIs

Please use consistent citation style throughout the bibliography (all authors or a subset of authors).

Please correct the citation to the ESA SPICE Service (which contains twice the DOI).

Please complete the references to books or book chapters by including (publisher, DOI if any)
Citation: https://doi.org/10.5194/egusphere-2026-2032-RC1
RC2:
'Comment on egusphere-2026-2032', Anonymous Referee #2, 10 Jun 2026
General comments:
The manuscript presents the statute of the calibration pipeline for JANUS data as it is now, with nice demonstrations including examples of what physical effects and parameters are taken into account, what are the ones not yet taken into account, and what are the effects that have a yet undiscovered origin for further calibration. The paper do not focus on the theoretical calculations behind the calibration, but on the general principes and results of the calibration process. This is an important work for the analysis and interpretation of JANUS data.
I believe that the abstract could be more detailed, including the main findings about the state of the calibration pipeline. The figures are also often not described and explained enough in the text, nor the leaps from the figures to the conclusions to get from them. The paper would gain a lot by being more pedagogic in that sense. The readers of this papers that will work on already calibrated images of JANUS later on may not have all the theoretical knowledge necessary behind each step of the calibration process, so it is important to explain them in a pedagogic way. I understand that the detail of the calibration is to be written in another paper, but it was frustrating at time not to have the details of how many "models" were obtained, to better understand some of the figures.
Please find below my detailed comments:
Comments on main text:
line 1: There may be a comma after Jovis in the sentence Jovis Amorum.....

line 21-24: It's better to cut the sentence in 2

line 88: genuine question : Why not lowering the threshold value described in line 85 then ?

line 103: I believe a comma between the words "exposure time" and "generating" is necessary, depending on the meaning of your sentence.

line 109: genuine question :Is there any physical reason why the dependence is assumed to be linear ?

line 111-113: Please explain each terms of the equation in the text (x0;w1;t0;t1,T...) event if it seems obvious. Also, add the units of the variables in the text if there is one (ex: [K] for T - or [°C])

line 115: Figure 4 appears in the text before Figure 3. Therefore the figure numbers should be switched. Please, also precise how the average residual is calculated.

line 117-119: how do you see the contribution of the dark current ? Guide your reader in the interpretation of the figure and its results.

lines 123-124: on line 123 you divide the proxy by the exposure time. On line 124 it is explained that the proxy is the radiance multiplyed by the exposure time. So in fine you are just using the radiance ? It was very confusing, and the need, use and meaning of a proxy value needs to be better explained

line 130: what are the spline parameters ? (how many consecutive data points taken into accounts ? is that what the 5 polynomial pieces means ?)

line 150-151: it is uncertain what is meant by "the latter" (the detector ? the band integrated radiance ?). In general I have trouble getting how it is necessary to correct from the target spectral shape, when it's part of the result

line 156-158: please add units of S, L, lambda when it applies.

line 160: Table 3 cited in the text way before table 2. Please change table 2 number, as it is used last.

line 162: Please write what LIME stands for.

line 167-169: Please explains what is the I/F, and add the units of each terms in the text. In the equation, [AU] should not figure there, but in the text after d: d is the distance in Astronomical Units (AU).

line 170: what does MODTRAN stands for ? can you tell more on this solar spectrum ?

line 173: you could write "and C_filter=pi.d/Int(E(lambda)dlambda)" of course using the proper mathematical symbols.

line 174: Table 3 should become Table 2.

line 236: This is a high phase angle, could it create problems for some photometric corrections models ?

line 238: genuine question: Should the DTM resolution be adapted to the resolution of the image ?

line 248: "all lines are captured simultaneously" the formulation is confusing if the meaning is something along "all pixels forming a line are captured/acquired" simultaneously.

line 251: what is the "attitude" ? Please define it as you talk about it also latter in the text without explaining what that is.

line 294: Please describe and explain the figures in the text.

line 298: please explain and define the terms in-field, near-field and out-of-field.

line 304: "previous case" -- It is unclear if you talk about figure 13 or 14.

line 305: "same signal on the surface" -- I don't see the surface ? Or what do you means by it ?

line 306: "0.0012%" - genuine question: is it really linear between acquisition time and DN ?

line 314: please explain how do you see the 12000DN in fig 16, and how you obtains 0.0083%

line 323: Describe and explain the figure in the text.

line 345: please define QE.

line 353: Table 2 here should have another number as it is its first appearance.

Comments on the figures:
In general, all captions need to be written in a way that the main text is not necessary to understand the figure. A lot of figures are missing labels on the y and x axis, and sometime on the colorbar. There is also a lack of consistency in the direction (toward the left or toward the right) of the x axis in the maps. Please ensure that all figures also are colorblind-friendly (especially the lines plots, and fig. 25). Figures treated in the same section could be regrouped in 1 main figure with subplots.
Figure 1: The font sizes need to be increased, especially for 'par' and 'cal'. These terms should also be redefined in the caption.

Figure 2: missing y and x axis labels. The font size needs to be increased. Do you know why the tiles has such a difference in saturation ?

Figure 3: what are "hot" pixels ? Increase the font size of everything, but especially the legend. The fluo color for the green can be difficult to see on screen. The caption needs to be more detailed: simulated background of what ? simulated profile of what ?

Figure 5: The figure is way too small, and so is the font.

Figure 6: add pixels lines and rows in axis labels. Please also recap what is included in the partial calibration.

Figure 7: I have trouble understanding the full image. What are the white lines ? What is the module of distorsion vector ? And why is the unit in pixels ? when speaking about coordinates, is it a distorsion of the pixel coordinate (row, column) ? or probed space coordinates ? It really need to be better explained.

Figure 8: please increase the font of "der" and "geo". Define also DEM, and C-kernels.

Figure 9: suggestion: the word "footprint" may fit better than "case". Remove the line crossing the round and cross data points in the legend

Figure 10: please increase the font size

Figure 11: please increase the font size, and add all axis and colorbar labels and units if it applies. for the right plot please check if the chosen color combination is colorblind-friendly. Check consistency of the direction of the x axis with other figures

Figure 12: add axis labels and a colorbar. Check also the x axis direction. In the caption there is an inversion of blue and orange line. the blank spaces ? Please write it explicitly that the blank spaces are the saturated pixels

Figure 13: Add axis labels, check x axis direction. join with figure 14. How do you know if the DN values before seeing the surface is due to straylight and not something else ?

Figure 14: Place the x_sc ( and same for y and z) in more convenient positions that do not hide lines or texts.

Figure 15: Axis labels + x axis direction

Figure 16: Axis labels + x axis direction. In the legend, the exposure time is 22.15 ms, which is not consistent with the text ( 0.2215 ms). Is there a target in the image ?

Figure 17: Increases font size. In the legend, the last 3 colors are the same as the first 3. Check if the image is colorblind-friendly

Figure 18: Axis labels + x axis direction + font size

Figure 19 & 20 : Uses the same scale in both DN axis. Labels missings in axis and colorbar. Also check for x axis direction consistensy with other figures

Figure 21: increase font size, and add y axis label. The dotted lines could be larger

Figure 22: all words in the legends need to be explained.

Figure 23: write CAD in full letters

Figure 24: needs to be explained in details as I do not understand it.

Figure 25: Checks if the colormap is adapted to colorblindness, and for one of the black line, usedotted lines instead to differentiate them easily. What are the axis labels ?

Figure 26: Axis labels + font size

Comments on the tables:
Table 3: Please recall what are S_moon and C_filter in the caption
Citation: https://doi.org/10.5194/egusphere-2026-2032-RC2

Viewed

Total article views: 331 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
181	128	22	331	15	20

HTML: 181
PDF: 128
XML: 22
Total: 331
BibTeX: 15
EndNote: 20

Views and downloads (calculated since 14 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	112	44	14	170
May 2026	31	69	1	101
Jun 2026	38	15	7	60

Cumulative views and downloads (calculated since 14 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	112	44	14	170
May 2026	31	69	1	101
Jun 2026	38	15	7	60

Viewed (geographical distribution)

Total article views: 330 (including HTML, PDF, and XML) Thereof 330 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 16 Jun 2026

Short summary

JANUS is the high-resolution multispectral camera on board of the ESA JUICE mission. During 19-20 August 2024 JUICE flew by the Moon and the Earth performing a Lunar-Earth Gravity Assist maneuver that provided a unique opportunity to test the instruments in mission-like conditions. In this paper we describe the status of the calibration pipeline applied to JANUS data acquired, the lessons learned by the analysis of the calibrated data and the open points.


Total:	0
HTML:	0
PDF:	0
XML:	0