[X] My Assistant

[Exploitcorporati...]

This was an experiment done in an attempt to get a grip on the shape and identifiable surface features of a world that gets to be a whole box of rocks in itself (and keeps making me feel as dumb as one.) The ranges and resolutions vary widely, and it's more or less an effort to follow rotation crudely rather than being chronological. Hope it's somewhat useful and informative.

Perhaps the upcoming NT flyby in October will sharpen up some of these perspectives.



Below is a color-coded key with features tracked from frame to frame:

[CAP-Team]

I hope this can result in a new model and map for Hyperion that can be used in (for instance) Celestia

[ugordan]

Now you're talking, EC!

It looks like there's plenty of angles here to be able to derive a simple shape model of the moon. The second composite is very useful in navigating around this lump, did working this stuff out take a lot of time for you?

* /me wishes he knew how to make shape models... *

[Paolo]

I hope this can result in a new model and map for Hyperion that can be used in (for instance) Celestia

I fear that there may be some problems in figuring out the rotation phase, spin axis orientation etc

[ugordan]

He said model, not accurate spatial orientation

[volcanopele]

Well, the chaotic rotation does make shape modelling significantly more difficult. I trust Peter Thomas can do it, but I certainly understand that it will take quite a while. With most other moons, you can know the lat and lon of a point on the surface and you can use the limb profiles to derive a shape model. For Hyperion, Lat and Lon vary depending on the spin state and the orientation of the spin axis.

[Phil Stooke]

The recent paper about Hyperion included a bit about Peter Thomas's shape model. A full paper in Icarus can be expected fairly soon, I think.

Phil

[Exploitcorporati...]

I'm particulary curious about the processes applied when mapping an irregular body, especially in terms of applying a digital photomosaic over the (future) shape model. Imaging coverage of Hyperion is fairly extensive now globally, but I imagine the relatively low resolutions over much of the surface and radically different lighting conditions complicate things. The Bond-Lassel facet seems to fare best, as it did in Voyager 2 imagery (the addon shape model I'm using in Celestia was marginally useful with respect to topography).
@#%*!!! I wish I could hold this daffy rock in my hands and run my fingers over it...come to think of it, every time I hold a chunk of basalt, I imagine Hyperion.

The rocks with holes are warm in my hands...apologies to Kimya Dawson. I hope the XM will have a few decent passes.

[Phil Stooke]

I don't know much about the process of draping an image over a shape model, but it seems to be well established these days. I'm more involved with developing shape models in the first place. For those who are interested, here's a bit of history of the field.

The first seriously non-spherical worlds studied were Phobos and Deimos, imaged by Mariner 9. The first shape models were best fit ellipsoids. Tom Duxbury of JPL wrote papers in 1974 about using control points - small craters or other features - in multiple images to define coordinates of those points and hence rough shapes. Viking allowed many more control points to be defined, so shapes (at least for Phobos) were well defined.

But many worlds were seen later in very inferior images - Amalthea, Hyperion, Prometheus, Janus and so on, even Halley's nucleus. Control points were either impossible to identify (low resolution) or never seen in more than one image, so useless for 3D analysis (e.g. Voyager images of Janus). So for my PhD work starting in 1985 I devised a new modelling method. I started with an ellipsoid, and sculpted its shape digitally to match the positions of lumps and hollows in the limb topography in images. To a lesser extent the terminator could be adjusted to fit as well. If you have several views you get a reasonable shape.

In about 1989 I gave a talk about this at Cornell. Peter Thomas and his computer programmers were developing a much more sophisticated shape modelling method which actually combined my method with Duxbury's. They called it SPUD, and Damon Simonelli (alas, no longer with us) published a paper on it in Icarus in about 1993.

The new Hyperion model, apparently, was created out of separate partial models made from a variety of stereo sequences, giving global coverage between them, and then added together.

Phil

[Exploitcorporati...]

Thank you, Phil. SPUD is a marvelous acronym for these purposes. This all seems like a potential growth industry with so many newly (or better) imaged aspherical worlds to consider. Hyperion looks in this light to be the outstanding candidate next to Eros and Phobos for the clearest possible modelling of shape prior to the Dawn and Rosetta results in years to come.

Condolences for the loss of Mr. Simonelli.

[Mariner9]

I can't find reference to a NT flyby of Hyperion in October. If I recall correctly the Non-targeted term tends to be used when the flyby is less than 100,000 KM, but more than 5,000 (or 10,000 ?).

If less than 100,000 KM, that means images with resolution better than 700 m. That sounds cool. Could you elaborate?

[elakdawalla]

I think "nontargeted" is used a bit more loosely than that. Basically, it's an encounter where they plan to do multi-instrument observations, but which is not set up by a burn of the engines, so the targeting has to account for the fact that the pointing may be slightly off of the nominal trajectory (hence "nontargeted"). Typically they're within 100,000 km but they can be from farther away, especially for Iapetus.

--Emily

[Exploitcorporati...]

This blurb from the significant event reports was the basis for the idea:

Tuesday, July 17 (DOY 198):

The final sequence development process for S34 kicked off today. The sequence is unusual in that it will be composed of four parts. The first part is similar to a normal background sequence but will run for only two weeks. This is followed by uplink and checkout activities for CDS version 10 flight software that will last for about a week and a half. During this time period there will be no science observations. After the conclusion of the CDS activities, a mini-sequence devoted to Hyperion observations will run for about four days. The last piece of S34 is again like a normal background sequence with full science activities. This will run for a little over a week, concluding on or about November 1. The Hyperion mini-sequence base products and stripped subsequences and the sequence products for the pieces of the S34 background sequence have been released for team review.

The (admittedly inaccurate) Cassini XYZ trajectory used in Celestia gives a closest approach on the order of 120,000km 21 October, which would likely produce images about good as those obtained 10 June 2005 (but properly exposed). The orientation of The Hyperion model is also way off, so I have no idea as to the viewing geometry.

[ugordan]

Solar System Simulator also gives around 120 000 km C/A, but the phase angle looks somewhat unfavorable at 100 degrees.

[belleraphon1]

All...

today's CHARM presentation is all about Hyperion.

Here is the link to the pdf.
http://saturn.jpl.nasa.gov/multimedia/prod...ARM_verH_FC.pdf

If you wait a month, the transcript of the presentations are usually available as well.

Have not had the chance to look at this yet... working.

Craig