Juno Perijove 49, March 1, 2023 Options

[volcanopele]

Here is a montage of all five Io images.

Original resolution was 34-43 km/pixel but have been expanded by 10x to improve the visibility of features.

[Brian Swift]

My collection. Nominal and exaggerated processing.

[Brian Swift]

And I suppose there was something other than Io imaged.
Here's PJ49_100, exaggerated color/contrast. Altitude 20481 km.
"Losslessly" (HUFFMAN) compressed source definitely worth clicking through to full resolution.

[StargazeInWonder]

Maybe the orientation of albedo features is coincidental, but it seems like this look at Io is showing a strong specular component in the terrains in the middle of the closest images. I don't know if this has been noted in previous research. It always seemed pretty flat / lambertian to me in Voyager and Galileo imagery. Of course, Io has quite a few different terrain types, has been viewed from various geometries, and is changeable over time, so the whole discussion here could be both complex and lacking in full data.

[volcanopele]

I think those are just bright areas combined with illumination angles, not necessarily specular reflection. that being said, I should photometrically correct these images first before I am certain about that.

Brian’s stuff has a bit of a preview. the bright, VERY SO2 rich patch in NE Lei-Kung Fluctus apparently is still there based on Brian’s processing.

[Bjorn Jonsson]

A montage of the five PJ49 Io images. The viewing geometry is also shown using computer generated images that include a latitude/longitude grid. They are based on a Voyager/Galileo map of Io.



The images are enlarged by a factor of 5 relative to the original data. The color balance is preliminary. Io's color is a remarkably complex subject - hopefully this is not very far from Io's real color. More on this later.

[ugordan]

Io's color is a remarkably complex subject

Isn't that just a roundabout way of saying it's the most colorful and contrastful world apart from Earth.

[StargazeInWonder]

Isn't that just a roundabout way of saying it's the most colorful and contrastful world apart from Earth.

That's a wonderful thing about Io, but another property of surfaces is the extent to which they show lambertian ("flat") or specular ("mirrorlike") characteristics, which can itself vary by wavelength; the potential complexity is extraordinary. The Mariner 10 images of Mercury, which is one of the less colorful worlds, overall, led to research by Bruce Hapke and others on the reflectance properties of its surface, which might have seemed like an empty topic given its similarity to the Moon, but it wasn't.

With Io, and the fact that new surface of different types is deposited over time, it may not even be possible to characterize the general issues in a lasting way.

With relatively little variation in illumination angle in these (nonetheless wonderful PJ49) images, it may not be clear how much we're looking at some really bright surface, or some specular surface that would look less bright from another angle, and JunoCam's performance is itself variable, so we may be left wondering why the brightest areas here look so bright.

[volcanopele]

Uploading a new version that deals with the artifact in JNCE_2023060_49C00076_V01 much more elegantly. There is a large splotch in the flat field for the red filter at sample 1194, line 43 that I've always had a hell of a time removing. The splotch seems to be in a different spot compared to its position in PJ34-PJ35. I created a new flat for red using Jupiter images from PJ49 and that cleaned it right up so Susanoo and Mulungu no longer look red.

To create my flat, I utilized images of Jupiter from PJ49 where Jupiter fills the field of view for at least some of the frames, so for example JNCE_2023060_49C00105_V01 and JNCE_2023060_49C00104_V01. I used cubeit and cubeavg to create a single image that is the average of all of 44 frames. I need to remove the gradient caused by illumination conditions on Jupiter. So first I take a copy of that average image and I null out all the splotches seen. I then shrink the image to a full-width, single line then expand it back out to full size. This creates an image with just the gradient. Then I ratioed the average image with this gradient image. This has the effect of removing the gradient and normalizing the image.

Not the best flat for large-scale gradients in the detector since I'm effectively filtering them out, but does a good job in dealing with the splotches.

[volcanopele]

Okay, I’m not sure that the RED band splotch moves due to time. For example, the same splotch can be seen in a PJ43 image of Io and in the same spot as PJ49. I tried processing the RDR data released to the PDS for PJ43 today and I tried creating a flat using RDR data from PJ43 and the splotch was not in the same spot, about a line up. this is the same as the flat I created from PJ34 data.

Best I can tell, the characteristics of this splotch differ depending on jno:TDI_STAGES_COUNT. The Io images seem to use 2 while most Jupiter images use 1, except for the PJ49 Jupiter images.

If any of you know, maybe Brian?, of any Jupiter images where jno:TDI_STAGES_COUNT=2 and where Jupiter fills the field of view and taken prior to PJ43 so that RDRs are available, please let me know and I can create a flat from them.

[StargazeInWonder]

That would be curious if a splotch in the flats moved; it would be a question of the origin of the splotch. Any permanent change in the optics or sensors would seem to remain in place. Something due to stray light (is Jupiter a possible source?) could move. A thermal excess that excites the sensors might move, depending on the origin.

[Bjorn Jonsson]

Io's color is a remarkably complex subject

Isn't that just a roundabout way of saying it's the most colorful and contrastful world apart from Earth.

I guess so. The problem seems to be that the color of processed Io images apparently is very sensitive to the exact camera filter properties - far more so than when processing images of Europa, Ganymede or Callisto. For example the red terrain (e.g. the red 'ring' around Pele) isn't obvious in Voyager images processed in a 'normal' way where an orange filter is used as red (it is visible with extreme processing though, see e.g. this thread). The problem is that these reddish areas start to get brighter relative to other terrain when you use a red filter instead of orange and they are even brighter in the near-infrared. This brightening relative to other areas has barely started at the wavelength of orange light.

The JunoCam filters are different from the Galileo filters. There is a blue filter (Galileo had violet) and the JunoCam red filter bandpass is 600-800 nm (i.e. into the near-infrared) with an abrupt cutoff at both ends. The CCD is sensitive to light in this wavelength range. This is different from the Galileo red filter. It is apparently impossible to make Io's color in the JunoCam images very similar to Io's color in the Galileo images by simply combining the R/G/B channels into a color image. Colchis Regio turned out greenish if I made the average global color roughly identical to the global color of the computer rendered images I posted earlier (in these computer generated images I attempted to correct the color by using synthetic blue instead of violet). If I make Colchis Regio roughly gray I 'lose' most of the yellow color in areas that are yellow in the Galileo images. If desired, it might be possible to 'fix' this using channel mixing, possibly by making a synthetic red image as some a weighted average of the red and green channels.

It will be interesting to see JUICE's images of Io. Its camera has red, green and blue filters (it has a total of 14 filters). But it's also going to be interesting to see data from the MAJIS instrument. It is an imaging spectrometer that covers wavelengths from 0.4 to 5.7 µm by simultaneously collecting 480 spectra. If I understand correctly the highest resolution for Io should be something like 50-100 km/pixel (take this with a grain of salt). Not exactly high resolution but enough to be very useful for getting a better idea of exactly what Io's color looks like to the human eye.

[Bjorn Jonsson]

This is image PJ49_105 in approximately true color/contrast and enhanced versions. North is to the right and slightly up. Several haze bands are visible in the enhanced version. The central area of these images is at planetographic latitude ~65 degrees north.



The next image following this one at the missionjuno website is image 123, i.e. images 106 to 122 seem to be missing (assuming that the difference in image numbers is always 1 for adjacent images).

[Bjorn Jonsson]

In the PJ48 images, the reddening trend of the JunoCam images greatly accelerated (see the PJ48 thread). Interestingly, the reddening in the PJ49 images apparently has reverted back to the trend before PJ48. The PJ49 images actually seem to be *less* red than the PJ49 images. There are some sources of errors and uncertainty (in particular few images at PJ48 and no northern hemisphere PJ48 images) but this seems pretty certain nevertheless. Below is an updated plot showing the color correction factors I'm currently using for every perijove.



This is different from a plot I uploaded back in 2021. There I was measuring the blue/red ratio in raw images with no processing except for decompanding. Here I am interested in the color correction necessary after decompanding, flat fielding and dark current correction. Actually the flat fielding and dark current correction results in only a tiny change to the blue/red ratio (or in this case the color correction). The color correction is in % with e.g. 275 for blue meaning that I multiply blue with 2.75.

Here the measured multiplier for blue is 4.5934 at PJ48 and 4.3315 at PJ49. These values are averages from several different measurements. Even if I selected the lowest values I got for PJ48 and the highest values I got for PJ49, the PJ49 images are less red (these values are 4.55 for PJ48 and 4.42 for PJ49). In the image above, the PJ48 value is abnormal enough that I didn't include it when computing the Bezier curve.

It is very interesting that this increase in red relative to blue occurred following a JunoCam anomaly at PJ48. Maybe it is even more interesting that the PJ49 images are less red than the PJ48 images; this is very unusual as the graph above shows.

In addition to what's discussed above the data above suggests the following:

(1) The reddening probably didn't start until sometime after PJ20 (possibly at PJ22 or PJ23).
(2) This reddening trend probably accelerated near or following PJ40 (it is not clear exactly when but the trend when looking at several PJs around PJ40 suggests this).

[Brian Swift]

Best I can tell, the characteristics of this splotch differ depending on jno:TDI_STAGES_COUNT. The Io images seem to use 2 while most Jupiter images use 1, except for the PJ49 Jupiter images.

If any of you know, maybe Brian?, of any Jupiter images where jno:TDI_STAGES_COUNT=2 and where Jupiter fills the field of view and taken prior to PJ43 so that RDRs are available, please let me know and I can create a flat from them.

Yes, splotches change location in the data as TDI changes.
Unfortunately, when I went looking some time ago, I didn't find a good set of TDI=2 images that covered the full frame from which to produce TDI=2 specific flats. Instead I developed a method to smear my good TDI=1 flats for use at higher TDIs. But it doesn't work very well. 🤷