CONSERT Philae location from Kofman's AGU paper.

Click to view attachment

The press conference this morning mentioned new search images taken a few days ago that should be downlinked pretty soon. These are designed to show Philae in sunlight from a range of 20km.

ROLIS images Under Philae :
(Thanks Emily)

Quick and dirty deconvolved first CIVA image. It looks that it will be possible to extract some nice
data from the full quality images.

And very quick and dirty anaglyph:
Click to view attachment
and cross-eyed:
Click to view attachment
Very cool stuff!

It would be interesting to have the Rolis image on the pan.

Yes, that is consistent with the terrain model here: http://www.bbc.com/news/science-environment-30524429 and here: http://www.nasa.gov/sites/default/files/th...n_and_cliff.jpg

It's unclear how that model was built. Perhaps they are using unreleased CIVA images to constrain topography based on shadows and sun position.
I'm curious about the terrain further out. Don't know where that data comes from.

An attempt to process the first CIVA image:
Click to view attachment

upd: more detailed version
Click to view attachment

Processed, registered, and combined CIVA1 Panorama 1 image (Perihelion Cliff) :
Click to view attachment

Again, that was stated in the press conference to be an artist's conception. They don't know where they are yet so the terrain beyond the immediate cliffs has to be pure speculation.

http://www.bbc.com/news/science-environmen..._medium=twitter

Philae comet landing 'all a blur'
By Jonathan Amos Science correspondent, BBC News, San Francisco

Good idea. Do we know the direction the top of the image points to in terms of which side of Philae (or solar panel)? The answer can probably be interpreted from this document. Meanwhile, here is a guess we can consider.

Click to view attachment
Above image has north in the center.

High resolution version has south in the center. I can further suggest the green line in this diagram is pointing south.

Just playing with a little manual clean-up of the lens flare

Click to view attachment

Looks right to me. Really brings the entire site into perspective. Great work.

That lens flare is stray light caused by them bringing the infinity-focus lens into the field of view. Earlier post-landing ROLIS images were blurred because the ground was more than the expected 30 cm away from the lander -- some of it much further!

Picture of landing site on terrain model at CNES website: http://www.cnes.fr/web/CNES-fr/11630-gp-de...-et-rosetta.php

Includes overlay of CIVA image.

Quick summary of the French news conference is that Philae has enough power now to keep himself alive, one leg is likely firmly anchored, and the cliff shadow could actually help him operate longer up to and maybe including closest approach to the Sun.

Found a little "philae glint" in the navcams...

http://mattias.malmer.nu/wp-content/upload...14/12/GLINT.gif

Wow! That seems right to be it!

The 7 ROLIS landing images, taken every 10 seconds.

Click to view attachment

Altitude ---- cm/pixel
64.7m ------- 7.1
57.9m ------- 6.1
48.5m ------- 5.1
38.6m ------- 4.0
28.9m ------- 3.0
18.8m ------- 2.0
9.0m ------- 0.95

@Malmers WOW FANTASTIC glint - right in the CONSERT blue strip zone

Malmer, you'd better get ESA on the line pronto; excellent catch!

That is a great find!
The glints look extended. Could it just be camera bloom or are we seeing the effect of reflections on the surface (covering a large area)?

Is this the Malmer Glint in the NAVCAMs from 30th October and 7th December ?
Click to view attachmentClick to view attachment

the same area in the 50km OSIRIS shot of the Landing Site from 2nd September
Click to view attachment

The glint site is also in this OSIRIS pre-landing image.
Click to view attachmentClick to view attachment
from : http://sci.esa.int/rosetta/54649-philaes-p...y-landing-site/

Where did you get them from? I thought only one like that was published.

The ROLIS images are from one of the talks at th AGU which was streamed live and is now still available on their video archive.

https://virtualoptions.agu.org/media/P34B-0...tola/0_m432jt2i

you need a (free) account to watch it.

In case this isn't posted I see Jean-Pierre Bibring's talk from AGU here:

https://virtualoptions.agu.org/media/P33F-0...ring/0_t1w3415j

As above, registration required.

I'm continuing to obsess about (exactly) how Philae got from its first touchdown point to its final resting point.

The videos below provide some information points--view of the various velocity and direction vectors atthe point of Philae's first touchdown, to see where the counce would have taken it.

1. Vectors @ moment of Philiae's landing, with *Philae velocity vector* as down direction, and giving you a view of the direction Philae would have gone if it horizontal veocity continue in the same direcction post-bounce (which it did): https://www.youtube.com/watch?v=gUeFbMZwJ1w

2. Vectors @ moment of Philae's landing, with *center of comet* as down direction: https://www.youtube.com/watch?v=WF3anN_A1mw

3. A 'great slide'down Philae's velocity vector to the impact point, and from there to the 2nd impact point on "Mount Malmer" --all from Philae's point of view: https://www.youtube.com/watch?v=qexB0qfMbs4

All these videos make the point that because of the direction of Philae's velocity vector and the exact slope of 67P at the landing point, the horizontal coponent of Philae's velocity (wrt surface normal of 67P at that point) would have continued unchanged after the bouce.

Best guess is that the vertical veocity would have been reversed and reduced in magnitude by about 2/3 (per the Rosetta team).

Put those two together and it all matches rater well with what we know Philae did do after the first bounce, per the ESA post-impact photots and Malmer's excellent simulation: https://www.youtube.com/watch?v=WF3anN_A1mw (with minor caveat that we now know that Philae had a 2nd impact at the 46 minute mark, which corresponds to the giant cliff at the 3:36 mark of Malmer's video: http://youtu.be/WF3anN_A1mw?t=36s - more info here)

Once you're at the great cliffs of Mount Malmer, it's easy to imagine one of them boinking Philae in the exact direction needed to reach its final resting spot. Several spots close by have what appears to be about the right angle, etc. More on that later.

Here is what a N00b can do, just with Malmer's model in Celestia: http://youtu.be/Tw0CnfEwqGQ

To see what's really possible, though, you have to go to the Master: http://mattias.malmer.nu/2014/11/the-mountains-of-agilkia/

It would be pretty (?!) easy to use Matlab to export Philae's position in 67P's frame, from the SPICE data, into something like a CSV file. Then bring them into something like what Malmer is doing there. (I could do the Matlab part if anybody can do the rest, fyi.)

Both of these look very good - and nice 2001ish music with the first one. I can see the potential of even greater things to come. I also enjoyed looked at these two of Malmer's works side by side:

Click to view attachment

And this comparison:

Click to view attachment

Here's another attempt at an anaglyphs, using Astro0's cleaned up image:
Click to view attachment
And cross-eyed:
Click to view attachment

Astro0's cleaned up image and Fredk's anaglyph helped give me more perspective on the ROLIS images. I've added the cleaned up image on top of the other ROLIS image in my pan. I think I can refine this based on a diagram of ROLIS being mounted on the bottom of Philae (from Stefano Mottola's AGU talk). Maybe we can clarify what is occurring with the two images. Is one prior to the rotation maneuver and thus more in sync with the CIVA pan?

Click to view attachment
Full resolution here

South is in the center. Note the second larger yellow ring - this is a diurnal track valid for today's solar declination. It's gradually expanding. Green circles (left to right) are directions of solar panels #1, #2, and #6.

Unsure if this has been posted yet, as back on November 12th this video with music by Vangelis was released with nice imagery of a simulated Philae landing, along with camera and instrument operation:

https://www.youtube.com/watch?v=W8bVOGN9jSg

This is one of a trio of videos. I wonder if they use any OSIRIS images?

These mosaics are really nice. It looks like you've got the fields of view correct? I read 60 deg - I think that's 60 x 60 - for CIVA-P, and 57.7 x 57.7 deg for ROLIS. I guess we're not sure if some of these images have been cropped.

About the two post-landing ROLIS frames, ROLIS is mounted away from the lander's vertical body axis, so a rotation about that axis is enough to provide stereo imaging. If you look at my anaglyph, there's definitely a rotation between them (I measure about 21 deg), so one had to be before and one after rotation. But you can also see that the fov of one frame is larger than the other. I think that has to mean that the body of the lander shifted a bit parallel to it's vertical axis, in addition to the rotation. If I remember correctly such a maneuver was also done with the legs?

OK, here is my complete reconstruction of the path of Philae's final descent and then bounces across the surface of 67P:

https://www.youtube.com/watch?v=3pm39qe7HyY

I am quite convinced at this point that this is the only scenario that gets us from the landing point to to the second bounce at 46 minutes to the 3rd bounce at 111 minutes, very near the final resting point. Tweaky type improvements are of course very possible, but in broad strokes, this has got to be it.

Obviously, this is based on ideas, information, and help from many of you on this forum, particularly Malmer, who has clearly identified the area of 2nd impact point and (what may be!) the exact location of final resting place of Philae. With those identified, there are only so many ways to get from the known 1st impact point to those additional points within the known time frame.

The images below show the profile of Philae's trajectory in the inertial frame. 67P is shown at the moment of 3rd impact.

The video shows how Philae and 67P move and rotate in relation to each other in the inertial frame, ending up at the point shown in the images--it's quite revealing.

Here are my thoughts about how our knowledge of Philae's trajectory across 67P could be improved over the general outline given in the video above and about an open question that still remains in my mind.

Possible points of improvement:

• More exact determination of the location of the point of 2nd impact. Perhaps spotting it in a image? There is almost certainly an impact point somewhere near "Mount Malmer"--whether it left a visible trace is an open question.
  
• More exact determination of the location of the point of the 3rd impact. We know it's near Philae's final resting place (which now (perhaps?) seems determined exactly), but it might be possible to determine it more exactly. Again, there may be evidence that can be photographed.

Of course, ESA still has a lot of data they have collected but not released, and it's quite possible some of that will shed light on Philae's wild ride.

Open question:

• Why no measurable vertical deceleration measured by ROMAP at 2nd impact? In Hans-Ulrich Auster's summary of the ROMAP data during Philae's flight across 67P, he says of the 2nd impact point: "It was not a touchdown like the first one, because there was no signature of a vertical deceleration due to a slight dipping of our magnetometer boom as measured during the first and also the final touchdown". So why did this instrument measure no vertical deceleration during the 2nd impact?

First point is, the 2nd impact is very clearly NOT a feet-first impact. If it were feet first, the feet would have absorbed a bunch of the velocity at that point and the subsequent trajectory would have looked much different (and much shorter). Rather, 2nd impact is a more-or-less elastic collision between Philae and the comet surface, perhaps edge-on and perhaps impacting the top. 67P's rotation was sweeping more or less in the direction Philae took after the 2nd impact, and since the speed of rotation is a significant fraction of Philae's velocity at that point and the impact surface was more or less vertical, it's quite possible that 67P's rotation added a bit of velocity to Philae.

At any rate, it's clear that the 2nd impact is mostly a change of direction for Philae. It might have picked up a bit of velocity, but it certainly didn't *lose* any--or it couldn't possible have made it as far as its final resting site.

That directional change in velocity could very easily have been in the 'edge-on' direction; it certainly wasn't along Philae's vertical axis.

So if the 2nd impact were edge-on, and particularly if it were oriented exactly normal to the direction the ROMAP instrument extends from Philae's side, it's quite possible to have an impact that wouldn't register on ROMAP.

If the bounce were edge-on and directly on the front foot only (parallel to the ROMAP boom and exactly opposite it on Philae) that would do it. So might edge-on, simultaneous impacts on the back two feet. Those wouldn't make the ROMAP boom dip in either vertical OR horizontal directions.

And if the ROMAP boom registers 'dips' in the vertical direction only (ignoring any horizontal deviations) it's possible that other edge-on or near edge-on impacts wouldn't register with it.

Is there any consensus here about the glint? It is quite tantalizing but the pre-landing images posted by Deepnet seem to show a similar feature.


Excellent work, I've been itching to see this kind of visualization! A couple of questions:

- Are you connecting the impact points with parabolic trajectories (or other conic sections)?
- Given that you are stitching trajectories between points at given times, does the change in velocity at the second impact make sense? (ie. considering the rotating comet frame, is the post-impact speed less than or equal to the pre-impact speed?)
- Can you share your coordinates for the three impact points in the ESA model frame (unscaled)?

Thanks - I've seen these fields of view and I think they're close in the mosaic though perhaps not quite exact yet. At this point I'm guessing/hoping the CIVA & ROLIS images may not be cropped.

Yes, it makes sense that both rotation and some vertical translation took place. I wonder if a rotation about an axis other than the vertical was involved. If we can ascertain which ROLIS image is before and which is after we can place them into the CIVA mosaic better. Also we can update the solar panel pointing directions in the mosaic to help predict future solar panel illumination.

Great video by flug as well with the explanations and all.

Click to view attachment - A Slide from Mottola's AGU talk

Click to view attachment
Stefano Mottola describes the boulders at 1st touchdown as agglomorates and suggests they may be 'primordial objects' discovered by surface erosion.
This makes me wonder if they bear any similarity to Sierk's 'Dragon's Eggs' ? -

Click to view attachment
As scalbers suggested, the 3 Vangelis ESA films( 1 2 3 ) do have some high resolution pans across OSIRIS images, youtube seems to only have 1024HD.
Is this a 'new' image ? The raking light brings up long twisty dunes about Agilka, what force made them twist ? Gravity ? And hollows appear to track the surface , rolling boulders , or surface collapse precession ?

Regarding 'dunes' etc., cannnot help but think that each perihelion passage has the nucleus erupt with a soundless wind as venting reaches peak. Not very technical I know but neat to think about.
In this microgravity environent how much force is needed to sculpt the landscape this way?

The Glint is not Philae. After looking more carefully. (I wish i had done that before posting) i have come to realize that It is just a bright boulder situated so that the protruding rock "lip" makes it look like a glint in its shadow. But it is just a trick of perspective.

I wonder why some boulders are so much brighter than the surroundings.

The Glint is not Philae. After looking more carefully. (I wish i had done that before posting) i have come to realize that It is just a bright boulder situated so that the protruding rock "lip" makes it look like a glint in its shadow. But it is just a trick of perspective.

I wonder why some boulders are so much brighter than the surroundings.

I am pretty sure the leg raising was not attempted as they decided to use the power needed for that to do science and they did not know fully at the time what the lander orientation was. The body was rotated.

A quite good animation of the Philae's adventures by Mr Asimov (sic) with works from ESA and Mattias Malmer, and sometimes real sounds from space :

http://youtu.be/mh0rM-5I6-Y

Thanks, I wasn't sure about that. We clearly see the lander translate along it's vertical axis as well as rotate. (It's also possible that it translated in its horizontal plane, but that's harder to determine because a rotation off axis entails some translation.) I think this highlights how precarious its position was, with apparently only two legs in contact with the ground/cliff surface.

There are a couple of reasons the whole lander (including legs) may have shifted when they just intended to rotate the central body. A reaction torque would have been transfered through the legs, and if all of them weren't firmly anchored the whole lander would've shifted. Also, as we noticed from the CIVA-P frames, one antenna was apparently in contact with the cliff face. So as the body rotated that would've imparted extra forces on the lander.

The SESAME team poster presentation said all three feet were touching the surface, with one leg likely touching the side of the rock or rock face they are leaning against. All 3 sensors recorded MUPUS trying to hammer it's self into the ground, and to do that they have to be touching the ground.

I should say, on reflection I'm not 100% convinced of the glint, either, though it is certainly v-e-r-y tantalizing and it matches in many ways where we would expect Philae to be. I want to believe and it has a kind of gestalt 'fits in many ways' feel to it. But rationally I'm maybe 60% or 70% certain? Definitely not 100%, until further confirmation.

But in some ways for my 'broad strokes' visualization of the path it doesn't make too much difference. Philae is definitely somewhere in that general area, according to ESA, and so the path from impact #2 to impact #3 is going to look similar to what we see here.



Well, this is where I get to reveal all the fakery behind the Great & Terrible Oz's Curtain . . . this is more along the lines of an illustration than an actual physics simulation.

What I realized, though, and the reason I went ahead & put it together, is that if we know the start and end points and exact times those were reached (to within a minute), then that completely determines the horizontal component of Philae's velocity. And we know that these were rather shallow hops, we have at least an idea of what the vertical trajectory looks like.

So, I realized that I could put together just a straight animation that would be Pretty Darn Close ™ to an actual physics simulation, and almost as valid--since there are certain parameters you're going to have to guess at/make up in a physics sim, too.

You can get an idea of how this simple 'constant velocity' sim and the real physics diverge, by watching Philae's initial descent. You can see that my little animation bug (constant speed) get a little behind, then a little ahead of of, Philae SPICE data (labelled "Philae"). That is a straight, long freefall, though, so that is where Philae's velocity will diverge the most from the animation.

So to answer your question directly, the path is just an animation path drawn with splines, hand-drawn into a more-or-less parabolic shape. I basically kept them as low as possible, just ensuring that Philae didn't intersect with the ground anywhere along its path.



I think it's reasonable but I confess I haven't worked out the math.

My STK is busy on another animation now, but I'll see if I can get it to spit these out when it's done.

To expand on this a little more, you are hitting exactly on my biggest point of discomfort with this scenario (the "Mount Malmer" to "general area of the Glint" scenario).

My impression in fiddling with this for some time in Orbiter, was that it this second, longer leg seemed to require a higher overall velocity of Philae in order to make the required time-to-next-impact and the required distance. My idea was that perhaps the comet's rotational velocity added something to Philae's velocity at 2nd impact (Philae is going somewhere between .33 & 1.0 m/s and the comet surface is rotating about .33 m/s in the approximate direction of the 2nd leg, so given the right type of impact, angles, etc it is at least in the realm of possibility for the Philae to end up with some additional velocity as well as a different direction).

But looking at this new simulation, I'm not so sure. The delta between the comet surface rotation and Philae's velocity across the surface is pretty low.

It's possible that my Orbiter simulations were wrong, or I am mixed up (I'm always in a state of confusion about whether I'm in the inertial frame or fixed frame or whatever, and I confess I've been more messing around with this haphazardly than attacking it methodically). What this new simulation seems to show is that all Philae needs to do is keep about the *same* velocity (in magnitude) before and after impact #2, in order to make it somewhere near the glint for impact #3. That, I believe, is possible.

Yes, I noticed this other 'white boulder' in all your images as well (see images attached). I wonder why it is so much brighter than the surroundings, and from seemingly a bunch of different angles as well?