Kevin and all on the Juno team--Thank you--I agree with jasedm--this is astonishing resolution and clarity

BTW, a new SCLKSET file was posted this morning. Don't know yet how big an impact it will be on positions, but I'm rerunning my processing now.

Either Ganymede has shrunk or JunoCam is taking pics from a position further away than the SPICE location.
Dots are along SPICE limb, and are symmetric between leading and trailing limb.
Click to view attachment
Now I wonder if we'll get a refined spk_pre (positions predicted kernel) before the post apogee reconstructed kernel delivery.

That's what I'm seeing in ISIS. I have a control network that should be able to fix the issue, but I ran into the issue with jigsaw and its ability to adjust s/c position I mentioned above.

EDIT: Since jigsaw can at least run without updating pointing, it seems to think that the spacecraft position is off by 23 km...

I'm also seeing errors that complicate the processing. Usually the difference between the predict and reconstructed SPKs isn't big but this time I'm expecting a significant difference. I was going to do a map-projected mosaic of all of the images but I might delay it until a few weeks from now when a reconstructed SPK becomes available.

Not much to add here other than just to echo that these Ganymede images are just phenomenal, far above my expectations. I'm certain just these few images alone will be the basis of new papers. Thanks so much to the JunoCam team and the people here processing these images. The 320 km Europa flyby on September 29th 2022 is going to blow our minds

yeah, the sharpness of those images was way better than what i was expecting
and i agree, if we can get this sharp of images from Ganymede, than imagine what we could get from Europa

anyways, here is a updated map that has more coverage derived from the images from 2 days ago


(ik some areas in the thing is a bit blurry, i was trying to fix a bad joint from where the lighting didn't line up in the 2 images that was involved to make this map...)

Ganymede... much like Snowball Earth 1 or 2 Billion years ago. A global ocean with an ice shell underlain by a warm silicate planetary body.

Thank you, Kevin - that's perfect!

And thank you to the Juno team, and all who got the extended mission going. I didn't know about it, but there will be other close flybys. Only wish there could be more!

Europa 29-SEP-2022
Io 30-DEC-2023
Io 03-FEB-2024

The wait begins...

The northern polar cap and magnetic field lines boundary crosses these pictures, north of Perrine Regio. The illumination does not allow JunoCAM to see it. I am sure that JIRAM on its side could distinguish it very well. Moreover, I wonder if the aurora could be detected on the night side. This encounter will probably give us many more surprises. So exciting !

is it possible to pull night shine from these images that could fill unmapped areas on existing maps?

Another Ganymede composite. Better color, though not perfect. Uses the first three images (PJ34-1,2,3) and uses the perspective of the third.


Ganymede - Perijove 34 Composite

Wow, these composites get better and better. Thanks a lot to all the talented people making them. It is always a pleasure to see a new one :-)
Just realize that part of this region has be photographed by Galileo (orbit G29) at about the same resolution:
https://rpif.asu.edu/Galileo/G29/fullsize/G29GSCAPCOL01.png

Nice exercise to compare and search for some changes (new impacts) after 20 years :-)

Or new furrows? Rather like comparing post-earthquake LIDAR scans to see what sort of displacement you get.

When might we expect some magnetometer and UV spectrograph data from the pass?

Is there any chance we might see some of Ganymede's UV auroras?

P

An interesting fragment of Ganymede's surface. A bit similar to Enceladus.

Click to view attachment

Going to throw my hat in the ring with my version. Not perfect yet, and probably won't be until this bug in ISIS gets fixed (or when a reconstructed trajectory kernel is released), so there there is some color fringing. Can't wait to try to do some color ratioing. There aren't a lot of apparent color variations, but you never know.

i think i can color my map with the natural colors of ganymede

not exactly natural color but this was my attempt from color processing my Ganymede grey-scale map to give it color

i did some processing to this grey-scale image ( https://www.missionjuno.swri.edu/junocam/pr...sing?featured=1 ) and created a false colored image (it was a lot of channel mixing, node blending, level and curves editing, and a lot of saturation editing

here is my result
(this is a preview below, the full resolution image was dam too big for UMSF to attach)
Click to view attachment

Close up views of the surface showing the ridges remind of a similar process happening on Europa.

if juno was able to image that spot in high detail, than how come it was unable to get that good of an image of Tros as juno did a couple days ago?
and i think that could be processed into a dem

Because Galileo had the data volume of a gnat. Given the constraint of playing everything back through the low-gain antenna, mission ops had to be very selective of what images to take and what to play back. Based on the observation name, I presume they were interested in the margin of the polar frost cap.

Moving on to Jupiter, I'm getting really bad color results out of my pipeline. Anyone else seeing this or is it just me?

Click to view attachment

A bit more uncertainty in the predict trajectory?

The color I get looks normal but the contrast in your image seems lower than it should be. Is it possible that you are not decompanding the images?

And BTW the contrast in some of the Ganymede images (in most cases not your images) I've seen here also seems somewhat lower than what I'm getting. Therefore I'm wondering if some of the Ganymede images here are maybe not decompanded but I haven't taken a detailed look yet so I might be wrong.

yeah, the contrast of some of the color processed images have very high contrast, and others have lower contrast (a bit more realistic)

i ended up getting a high contrast in my attempt
(altho this was intentional because i wanted to make a high contrast false colored image)
Click to view attachment

looks pretty realistic to me in terms of color and gamma, all though there could be a bit more contrast...

You called it! I had replaced the decompanding method for one that I use on other projected that a lot more performant. Not sure why it didn't work in this instance. Restoring the previous method worked.

Click to view attachment
Click to view attachment

now that jupiter map looks perfact

contrast of the ganymede map looks good, but the color is a bit off if your going for a natural color, and btw what reprojection program did you use to reproject juno's highly distorted ganymede images?

This is an approximately true color/contrast mosaic of images pj34_01 and pj34_02 seen from the pj34_02 viewpoint:

Click to view attachment

The color is based mainly on color computed from Ganymede's global visible light spectrum. It is a bit approximate and subjective though. The color of the two images (pj34_01 and pj34_02) wasn't perfectly identical. Probable reasons are imperfect bias removal, imperfect flat fielding, scattered light, different exposure or some combination of these factors. In theory it might now be easier to improve the flat fields and/or bias. Ganymede doesn't have an atmosphere which means that variations in emission angle are probably much less significant than in the Jupiter (and Earth) images. Ganymede also provides a nice sanity check since it is the first airless imaging target imaged by JunoCam at high resolution. This makes accurate limb fits trivial (no clouds/hazes at the limb to mess things up). I used this to make some checks and tests that are more difficult to do when processing JunoCam images of Jupiter (or the Earth). Everything worked as expected.

As discussed in earlier posts, it is clear that Juno passed slightly farther from Ganymede than intended (this is normal). The absolute error doesn't seem bigger than for a typical Jupiter flyby. The difference is that it is much larger relative to Ganymede (which is much smaller than Jupiter) and must be corrected if high quality results are desired. To fix this it turned out to be easier for me to simply shrink Ganymede slightly in my processing pipeline instead of adjusting the position of the spacecraft.

EDIT: Added different exposure as one possible reason for the slight color difference.

Does anybody ELSE notice the odd resemblance to the Earth's moon? (Well, if it were in an 80s music video?)

I look forward to somebody modeling upwelling, downwelling, and 'deccan traps' style hotspot flooding & resurfacing based on odd or even numbers of convection cells.

Thanks to Bjorn for hinting how to fix the color issues. Here is my latest attempt at the first mosaic:

Click to view attachment

Still am unhappy with the limb fit and the color fringing. But I think for now this is the best that ISIS can do until some fixes come for libale (why is alway libale...)

Thanks for diving so deeply into the ISIS problems, Jason. Anything that improves ISIS behavior for Junocam benefits not only these flybys, but Junocam processing in general.

Note that we could have the typical level of error in the first frame's start time value for each of these images, though I expect at the moment those are swamped by the error in the SPK.

Anyone know if Juno reconstructed SPICE accuracy information is available anywhere?

I didn't find a way of accessing it via SPICE API with a quick search of the docs.

Like Björn, I hope to do some validation and refinement of my model using Ganymede imagery,
and think knowledge of the SPICE position/orientation/timing accuracy will be useful for that.

I'm using ISIS (junocam2isis, cam2map, map2map, isis2std), and Blender: https://github.com/kmgill/cassini_processin...m/processing.py

this is where they are
ftp://naif.jpl.nasa.gov/pub/naif/JUNO/kernels/spk/
the last reconstructed is from 5/17/21
and
ftp://naif.jpl.nasa.gov/pub/naif/JUNO/kernels/ck/
the last rotational kernel is from 6/8/21

Local Youth Does Good... https://apod.nasa.gov/apod/ap210614.html

Nice work Kevin!

Great work, guys! This is still the place I come for the best imagery hot off the presses. Very impressive.


When I initially heard about the planetary physics science that Juno proposed to study, I was really happy, but I feared that there might be a missed opportunity with the public. Enter JunoCam and all of those fears were demolished. It's a great little camera that has far exceeded its life expectancy!

I'm also amazed that the orbital planning wizards figured out how to accomplished the Galilean Moon flybys in the 53 day orbit. Just great work seizing these opportunities and we are all better for this exploration.


One question for the group--does anyone know how much coverage the Juno Stellar Reference Unit navigation camera collected on the dark side of Ganymede? The initial public release cited a very respectable resolution on the SRU imagery, but I don't have a great feel if that was just because the closest approach was on the dark side of the moon or if it's just a high pixel density camera to precisely determine the satellite's position in space during normal operations. Best I can tell, it is a framing camera (vice JunoCam's pushbroom), so it must be programmed to acquire images at some regular (or in the case of the Ganymede flyby, more frequent?) rate. Any knowledge welcome! Thanks.

I'd like to know too. Such a tool would be SUPER useful for a Uranus and Neptune mission. The mapping potential is out there! Maybe cover some of those elusive polar areas of the Jovian moons. It's always interesting when you find another possible unintended use for an instrument.

Hi NMRguy, I replied with the SRU information I have here: http://www.unmannedspaceflight.com/index.p...st&p=253254

Click to view attachment

Taking another pass at 34C00001. Again, only the pointing was updated, not the spacecraft position due to a bug in ISIS 5.0.0. This used a larger control point network (a little over 2000 hand-adjusted points) that combined images from 34C00001 and 34C00002 and reduced the number of fixed points (fixed to a Voyager/Galileo-era basemap. Slight adjustments have been made in Photoshop to the width of each channel in the final mosaic to reduce color fringing. I think with this product I am calling it until either the reconstructed trajectory is available, or the jigsaw bug is fixed. I might try my hand at making a combined product of the two mosaics, but I haven't gotten around to that yet.

I noticed that a reconstructed SPK is now available. Not unexpectedly, the flyby distance is slightly larger in the reconstructed SPK than in the predict SPK. For the first Ganymede image (pj34_1) the difference is ~6.4 km at START_TIME.

It did reduce the offsets I've been seeing, but not eliminate them. Here's hoping ISIS 5.1 will be out soon.

Same result here. From my preliminary results Juno is still too close to Ganymede in the new SPK kernel. The error is much smaller now though but I still haven't determined how big it is.

Are you referring to the maps at
https://astrogeology.usgs.gov/search/map/Ga...ClrMosaic_1435m. (color)
and
https://astrogeology.usgs.gov/search/map/Ga...obal_mosaic_1km (greyscale)

or is there another map somewhere?

I looked at the above two, and noticed that the patchwork of images does not line up between them. Also, the IAU nomenclature (from http://planetarynames.wr.usgs.gov/shapefil...menclature.zip) aligns with greyscale version.
The descriptions on those pages are basically identical, not giving an explanation for the misalignments.

Every time you run a mosaic sequence like that you will get slight variations in how things line up depending on the exact numbers used for camera pointing and position. It's not easy to say which is going to be more accurate. To get beyond that you need a better geodetic solution from, for example, an orbital laser altimeter. We are not there yet for Ganymede.

Phil

BTW, some relaxation with so much beauty:
https://www.youtube.com/watch?v=CC7OJ7gFLvE
Have a great Sunday.
Regards,
Marc.

Just using the pointing as a guide, slower encounter than last time so there should be plenty of opportunities for JunoCAM to image Ganymede, though at a much lower resolution (31 km/px). Here is the viewing geometry at c/a