^ some very good points, well put.

From this abstract, Philae
The sky must be at the upper left side of the frames. Here I've pointed to a few potential particles:
Click to view attachment

Very interesting. From my interpretation of Philae's location and position, what we see on the left is more of Perihelion cliff below what we saw in the CIVA image. The edge appears to have a dust halo just like we see in the recent NAVCAM images, only on a very small scale. It looks as if it is created by the Sun striking the other side of the cliff, the sublimating gas raising micron size dust particles and a number of larger particles in the mm size range, from the surface. The sunlight reflecting from Philae would appear to have changed the surface character this side of the cliff, bottom left. Any guesses on the bright glare top right, reflection from a foot maybe? The other linear reflection I suppose might be a CONSERT antenna, the curved one at the bottom from the ROMAP probe, not sure about that though.

That upper left region of the image being sky is consistent with scalber's proposed placement of the rolis image in this post.

After discussions off thread with Scalbers, our interpretations are in close agreement derived from different approaches, so I also think his visual interpretation is a really good approximation. In simple terms Philae is pointing, by my reckoning, only about 10-15 degrees differently from the direction that she would have been pointing at Agilkia, hence the maintenance of the RF link. The comet's surface at her final landing spot is about 90 degrees to Agilkia, so the 80 degree inclination he uses seems well justified to me.

Unfortunately, unlike the AGU presentations, it would appear the EGU ones will not be available to view online. The abstracts make a really appetising read and give a few clues as to what direction the teams findings are leading them in.

https://www.flickr.com/photos/124013840@N06...in/photostream/

Nice to see a more consistent picture of things here. If I could double check this, are the presently released ROLIS images without the IFL and we will see a future release of ROLIS showing a sharper horizon with the IFL?

From the abstract I linked to:

So one with and one without. But we don't know which is which. Both we have seen are subsampled from their original 1kx1k pixels.

Rosetta will attempt to contact Philae in the next days. The DLR posted an interesting article on their page. A signal is expected as early as March 12th, but we would be really lucky.

Interestingly that article cites a different power requirement for turning on the Tx unit, and a different minimum temperature requirement, than this previous DLR article. 19W vs 17W and -45°C for "turning on" vs -40°C for "lander electronics".

Possibly those 2W difference could be needed to heat the lander internals from the -45°C needed for the electric system to the -40°C needed for electronics operations? 3.5W average then to boot up the system including periodically switching on the receiver, and 13.5W on top to activate both Rx/Tx continuously?

According to a pdf I've read (ftp://ftp.irit.fr/IRIT/CSC/4218.pdf) the threshold for stand-by would be -45°C along with a voltage requirement. It seems that the thermal control has a preferred temperature that is above -40°C however, below which additional heaters are used. So it may be that in the range -45°C to -40°C it simply doesn't have the power to do anything useful except trying to warm up even further.

I have been playing with Fridays NAVCAM image which shows the Philae landing area well illuminated. Initially I was just looking to see how much sunlight was falling on where I've concluded Philae is hidden, but I got a bit carried away. At least I am happy its slap bang in the middle of the refined CONSERT defined region from the AGU meeting that basically covered the region between the two outcrops either side.

https://www.flickr.com/photos/124013840@N06...in/photostream/

This one is where I got carried away, working on the assumption that Philae's albedo is way higher than the surrounding comet and so she should appear a lot bigger than she actually is.

https://www.flickr.com/photos/124013840@N06...in/photostream/

If it is Philae, she sure is shining brightly and thats got to be good news.

Have you seen this image? It highlights the features close to the CONSERT search area both on the shape model and on a cut-out of the comet. It might not be 100% correct but, the area you believe Philae is roughly the purple area on that image and thus outside the CONSERT area.
The blue "bird foot" hill is probably the most recognizable feature so it's a good place for orientation. The CONSERT area starts roughly between the left and the middle "toes". It then continues along what almost looks like a ridge until it ends another small bump. http://i.imgur.com/sZvw7pI.png is my quick take on the 9th March NAVCAM. The reddish area is the CONSERT search area, the black arrow obviously points to the "bird foot" hill and finally the orange arrow at the small bump/hill where the CONSERT area ends.

It's great new picture of the comet, I'm really looking forward to see if the current dark side is completely devoid of the large pits of the type that we've seen in for example Seth and Ma'at.

If the position you suggest is right, and if the brigthness helps to identify, then philae might be the little white spot in this image:

Click to view attachment
2,4m/px
(http://dlr.de/dlr/en/desktopdefault.aspx/tabid-10726/#gallery/18459)

But I dont believe in it myself.

The area you've circled gamp is a place with material that's been visibly brighter all the way since August.

Philae is ~ 1 metre across; it would span a maximum of ~ 1 pixel in that image at native resolution.

I'm aware about that. "Sherbert" wrote, that maybe the lander looks bigger, because it reflects much more light than the surrounding comet material. So I found this image with this white spot, which is at least near the point, Sherbert has maked. It is nothing serious.

It wouldn't appear bigger as much as it would appear saturated; methinks. I think there should be one pixel that is both very bright and clearly brighter than the neighbouring pixels. This one pixel would be the one that has 'filled up' with electrons and is now letting many additional electrons past to neighbouring pixels where they don't belong. If the original point is very bright, neighbouring pixels could start to saturate as well, but there should still be some sort of central area of very bright pixels. In a pixel-saturation scenario, Philae should look like a glint, basically.

Thanks for all your comments. Firstly I should say that the close up image is ridiculously upscaled and zoomed and as such I did not mean for it to be taken that seriously. It is blatantly obvious that the original resolution is woefully insufficient to show Philae except in the extremely unlikely scenario of a lucky very bright glint from a metallic surface. As Habukaz points out this would be a point source covering only at most one pixel, possibly bleeding into a few more. A defined shape would seem highly improbable. The intention was to prove to myself that the alcove I could faintly see on the original NAVCAM image, was indeed the same one as I have suggested as Philae's hiding place, in previous posts. If anything the triangular shape is more likely to be the bright rock seen in the CIVA 1 & 2 images. The bright spot that Gamp spotted is "Malmer's Glint", a feature Matthias picked out from the NAVCAM movie frames released at Christmas.

DoF your positioning of the CONSERT search area is most appreciated. From the AGU presentation I picked the two "bumps" to be the left hand one marked with the orange arrow and the other further to the left. The problem with this search area and how it relates to the actual features on the comet, has always been its position was only defined publicly on a shape model that was based in this particular area on a minimal number of poorly illuminated images, a fact that the team acknowledges. I certainly can't argue with your interpretation and interestingly there is a very similar alcove/cave formation in the middle of your red area, which I considered a possibility from it's appearance in a couple of other images. I ruled it out mainly because it appears too wide for Philae to have got wedged half way down, as all the evidence from the CIVA and ROLIS images would indicate and the OSIRIS image specifically filtered and with the resolution to show up Philae, does indeed show a feature that looks quite plausibly like Philae. Whether it actually is Philae is a matter of interpretation and one assumes spectral data from the OSIRIS image could determine that.

I am not out to prove to the world that this is actually where Philae is, but to highlight the fact there is some decent circumstantial evidence that makes it worth a more detailed investigation by those with access to more images, data and better software tools than I have.

Indeed, it would be nice to see the CONSERT area on an updated shape model. An improved shape model should also reduce the CONSERT area as an added bonus. I'd love to hear what they've deduced so far on possible positions, although I suppose I'd like it even better if Philae managed to wake up and tell us. With how much erosion they are predicting I at least started to ponder whether the lander will end up being buried or moved, so in a worst case scenario what they have so far is the only thing they'll get to determine the final landing spot.

As an aside, the bright areas such as "Malmer's glint" are very intriguing. Assuming you accept one pixel features, which maybe you shouldn't, then it could be seen all the way back in August. Granted, you have to be careful with interpreting anything as actually bright in these pictures, but it underlines how something that looks bright isn't necessarily Philae.

Since both the production of shape models and the search for Philae are in the hands of the OSIRIS team, we might be waiting a while. It might seem prudent to some, to wait until more of the "Dark Side" is revealed, say around the equinox in May. Even if Philae does reawaken, I wonder how easy it will be to actually determine her location without a more complete shape model. Close in imaging is now extremely limited and geared towards other science goals and so triangulation using CONSERT data would still be the only method available.

Judging from the sunlight orientation I can guess which way north is in the Bird Foot region. This helps to correlate with the Phliae imagery and identify Perihelion Cliff for example. In other words, Perihelion Cliff could be at the Bird's Feet in this image referenced above in DoF's post, specifically the rightmost toe off to the east northeast? It seems plausible that the Bird Foot has been the main feature in the northeast shading Philae during the morning hours leading up to noontime.

Also, ledges at varying distances to the southwest (recalling Malmer's distance cues) lead up to the tip of the orange arrow. Might be interesting to try and simulate the Philae point of view now (from a location just right of the center of the red area) with a 3-D shape model using this new Rosetta image. Perhaps it will show a reasonable match?

Hi everyone,

I prepared some visualizations of Philae's "landing" site. It's just my luxuriant imagination ofc , I assume it has probably nothin common with actual situation (especially bearing in mind XYZ orientaton), but well. I gave it a try ;-)



And here goes a case when Philae "decided" to land properly to save us some heart palpitations




Maciej, PL

Catching up on this thread, I'm seeing some 'Philae is less than 1 pixel across at this distance, so if it shows up as a bright spot, it will only be one bright pixel, possibly with bleeding into adjacent pixels if it's bright enough' talk recently.
I'll point out that there are a few things wrong with claims like that.
First, an object would only be limited to a single pixel at that distance if the entirety of the object were contained within the particular area covered by that pixel. Assuming it's true that Philae is 'less than one pixel wide' at this distance (I haven't checked, and don't intend to), if there isn't a pixel centered on Philae, which there won't be in most cases, then Philae may occupy 2-4 different pixels. It's worth noting that those pixels will have a smaller percentage of Philae contribution to them than you'd see from a pixel that happened to be centered on Philae.
Second, the 1-4 pixels possible, as described above, assumes a perfect imaging system with a nonexistent point spread function. I don't know the point spread functions for any of Rosetta's instruments, but I know they aren't 'perfect' and I would guess they've been published. Actually, I'll take a moment to see if I can find them... After a minute of searching, I didn't find any detailed technical specs for NAVCAM, its lenses, and its filters (I still think they're out there somewhere), but I did find some description on PDS of the extent of the point spread function of the system for two of the filters. One was 1.3 pixels FWHM and the other was 1/4 pixel FWHM (0 would be perfect). I also found a document on PSA that said of the point spread function for NAVCAM that, depending on the configuration of the optical system, 30-50% of light from a point source with a pixel centered on it would end up in a different pixel.
Third, the 1-4 pixels possible also assumes a stationary camera-target system, which is not the case here. Rosetta is 'orbiting' 67P. The range of exposures possible for NAVCAM is apparently 0.01 seconds to 30 seconds. I don't know what exposures are typical for images we've seen, nor do I know what, if any, method Rosetta may use to maintain a target pointing during image acquisition, so I can't begin to work out what effect the motion has on the distribution of a target point across the sensor in an image. Suffice it to say it has 'some' effect. It's possible that effect is negligible; I wouldn't know.
Bleeding is probably possible on the CCD. The conditions for, nature of, and extent of bleeding, depend on the CCD used (and the conditions under which it operates), about which I know nothing, so I can't say anything about that. I would guess that the CCD used does not bleed significantly within the operating and target conditions expected for the mission.

Herobrine--I think you are correct that more than one pixel is likely to show up, but such small signals are still much harder to recognize than if Philae was 4-8 pixels across in an image.
Station--From my understanding of what we know, Philae is in a cave/hole/vent with a significant overhang that currently keeps it out of sunlight most all of the time. Your visualizations seem to have it on an exposed surface which we know it is not.

@Herobrine: You have to remember the context in which this was posted. It was not a pixel-by-pixel analysis, but about what looked like an object in a specific NavCam mosaic. If anything less than one pixel in size according to the resolution shows its real shape, that would be more or less a coincidence.

Nice visualization Station, although I'm about to suggest more work for you . How about trying to somewhat recreate the situation when the CIVA images were taken? Some restrictions on the local area can be made. CIVA2 is staring into space, CIVA1 (and CIVA3?) and ROLIS sees the horizon, the CONSERT antenna in CIVA4 might be touching the ground etc. You can also determine (roughly) where the sun was due to the shadows cast on Philaes feet. It should also be possible to attempt to work out very roughly some local topography that way as well, CIVA5-6 was looking into shadows after all.

Well, easier said than done, but it would be interesting to see an attempt at it. I had considered doing it myself but I'm not up to par for it.

Well, ESA did something like that here:
http://www.esa.int/spaceinimages/Images/20...n_visualisation

That one doesn't look very pretty, though.

I would like to follow up on the mosaic at the link below that is mentioned just up at the top of this page in post #1321, and in that posts's links to earlier posts. It suggests there are some viewing directions that are clearer than others. Things will get more favorable as the sun rises up more. It would be interesting if this CIVA/ROLIS view can be translated (with the Bird's foot Rosetta image) into a fuller view of Philae's environs. This fuller view could be either from Philae, or from orbit?

http://laps.noaa.gov/albers/scratch/philae...olled_1221a.png

I think the zenith can be seen from Philae, so it wouldn't be that type of overhang mentioned 5 posts up.

I had actually hoped someone would do that scalbers. One problem is that there isn't really many well lit high resolution images of the area, the ones we have are either shadowed or too far away to have a lot of detail. Making an accurate 3d model from it is probably difficult. As you say though, CIVA/ROLIS images plus a good shape model could probably constrain its orientation/position quite a bit.

In the most recent video of ESA Euronews, there's some updates on post-wake-up drilling ideas.
https://www.youtube.com/watch?v=_GCscfE1h0k

"Last November the drill couldn't reach the surface, but that could change."
..
"The images that we took show if we slightly turn Philae on its feet, the drill will have access to these materials, which the images show to be the most fabulous we could imagine, because in these materials, there are all these original blocks of ice filled with organic molecules within our grasp."

Sounds good DoF. It seems to me the new Birdfoot image may just have enough detail to start some attempts at this, at least to get the main features visible from Philae. I wish I presently had the time to dig more into this.

So some new information about Philae from the EGU2015 press conference http://client.cntv.at/egu2015/PC1, with the Philae part starting at ~16:50.

I honestly don't understand how the attitude of Philae with respect to 67p they presented, shown at ~21:35, is supposed to work. The best example would be CIVA2. The CIVA2 camera should see the ground regardless of how the local topography looks with the attitude they show. However in the actual image taken by CIVA2 only space is seen. Am I misinterpreting the attitude visualization somehow?

Apart from that they also show a recreation of the area that can be seen by the CIVA cameras. Also it is now clear that the illuminated solar panels were on the same walls as CIVA2 and CIVA3 along with the one on the lid.

Thanks for the update DoF.

Solar panels on the walls of CIVA2 and CIVA3 would make sense for getting a relatively short period of glancing sunlight late in the day. The upper left green dot in the updated image is solar panel #2. Solar panel #3 (left green dot on the western horizon) is underneath that. It looks like these panels are somewhat under 90 degrees away from the left portion of the inner solar diurnal arc, allowing for glancing illumination mainly late in the day. Panel #2 would even have some illumination at noon-time if there wasn't any obstruction by Perihelion Cliff.

Top panel is the green dot on the northern horizon near the center of the panorama. At 25:00 into the new EGU video, a plot is shown where the top panel gets even briefer illumination than the two panels on the side. I'll have to think about whether this alters the best orientation solution, or if it can be explained by a parallax effect if we're quite close to Perihelion Cliff. However I agree with Stephan's comment in the video that the upper lid heating/solar power will be absent beginning about at equinox time, while other panels get stronger illumination.

Click to view attachment
North is in the center. Full resolution at http://laps.noaa.gov/albers/scratch/philae...posite_0417.png

The outer solar diurnal arc shows the sun's track for the present time and looks quite a bit more favorable to get more insolation on the panels. Even panel #1 (upper right green dot) gets more into the act. It should be possible to draw circles with varying angular radii around each green dot to estimate the intensity (i.e. irradiance) on each panel as the day progresses. This could also be calculated numerically with a little spherical geometry program.

Another flyby during the next weeks with an option to spot Philae has been cancelled due to dust grains disabling star tracking.
Next try not before October 2015.

Nicely done Scalbers.

It took me a bit to get orientated, but indeed putting this together with Stephan's comments, at the moment Philae's top plate is probably receiving it's best illumination and even if power is not increasing enough, the temperature inside Philae should. It was mentioned there are also heat absorbers on the top plate. This also shows how Philae is slightly tilted backwards towards the cave wall behind. As 67P approaches its equinox the angle of the sun will no longer be perpendicular to the top plate but start to illuminate the side panels more. There might be a fly in the ointment from the large cliff below "Spider Hill". This sits slightly below Philae's alcove and so nearer Perihelion its shadow may be cast "up" the comet and across Philae. This could also be a blessing, as "Perihelion Cliff" has proved to be, protecting Philae from the heat at the height of Perihelion.

As far as the temperature of Philae goes, if the surface upon which it rests is extremely cold, wouldn't that tend to encourage any heat it develops within the lander to follow the second law and basically flow to the cold surface?

It would very much depend on how much of Philae's structure is in direct contact with the surface, and the thermal characteristics of both that portion of Philae's structure in contact as well as the thermal characteristics of the surface. But I know that if you set a metal box on a very cold surface, heat it draws in from above will flow into that surface and keep the temperature in the metal box colder than if it box was not in contact with that cold surface. Would heat be depleted faster from Philae into the comet's surface than it would radiate into vacuum?

-the other Doug

Overheating isn't considered as a risk anymore.
But there is no information about what's above Philae, including a possible ceiling.
From ROLIS images it's known, that the final turn has been successful.
At the moment Philae might be able to boot (at about 4 W) and to execute "blind" commands, but not necessarily able to send before June (more than 19 W needed).
Temperatures of the environment after landing have been about -145 °C, -45 °C are necessary for boot.
The -145°C are no problem for the exposed exterior instruments, but it's well below the design temperature of the thermally insulated interior electronics and instruments. So there is some risk due to thermal stress.

Considering the ultra-low gravity, the legs may be contacting the surface only just barely, ie only a very small surface area would need to be in contact to hold up the lander. So thermal conduction to the ground may be very poor.

I clarified the wording a bit in my post #1334 from yesterday. My present solution has the initial tilt of Philae at around 80 degrees toward Perihelion Cliff, so the top panel was pointed just above the northern horizon. The CIVA images are including the zenith in the field of view and it looks clear, so there isn't an overhang directly overhead.

The ROLIS images suggest a clear southern horizon, so it seems that in summertime there wouldn't be much shadowing, depending on the azimuth of the cliff below Spider Hill. I'm assuming Perihelion Cliff is the large obstruction to the north and east, so it is blocking the sun at landing time and would be out of solar arc near perihelion.

Perhaps green dots could be added with the post-rotation solar panel orientations in the existing panorama. The ROLIS images should help to specify the rotation event more quantitatively.

http://blogs.esa.int/rosetta/2015/04/29/ma...s-800-to-30-km/

I see this French news story that Rosetta will start to listen for Philae beginning tomorrow.

A similar story appears in Nature.

They already had one listening period, but nothing was heard, as expected. This next one might have a better chance as Philae warms up.

So they know where it is within 50 metres! Not enough anyway to modelize the exact environment where it ended up (or down)

Hypothetical/theoretical (certainly not practical) question here: if contact isn't established with the lander during the Rosetta mission, could Philae revive itself in the future? Could we send another spacecraft out to the comet to re-attempt contact in six years? If it doesn't get bumped around by an outgassing from the comet, how long can it remain re-bootable?

At last, we found Philae ! She has landed in front of the Grand Palais, in Paris.



More pictures :
https://plus.google.com/photos/112434140547...=CJTm7qGD4oiMTg

Pity they'll take that exhibit down again in two weeks...

(couldn't they wait another three till i'm there?)

a nice interview (in French) with the guy in charge of operations at the Philae control cemter here in Toulouse. there is a new communication window opening


http://www.cnes.fr/web/CNES-fr/11924-gp-le...l-de-philae.php

Interesting, they've completely redesigned the scientific activities to work without charging the rechargeable battery, but using the solar power directly, a way to operate which hasn't been anticipated before landing:

Rosetta has identified a "promising candidate" for Philae

http://blogs.esa.int/rosetta/2015/06/11/th...-find-philae-2/

Pros:

- visibile only in post-landing images, not in pre-landing ones
- good size match
- good solar illumination match
- good radio visibility match
- region was poorly illuminated -> relatively inactive -> unlikely that it's something "natural"
- current shape model isn't perfect, so the landing ellipse could be slightly adjusted and include this spot
- it's visible in two images, so it's a real feature, not a dust grain or an image artifact

Cons:
- 7 weeks between images, plenty of time for a rock to fall off a cliff or something
- just outside landing ellipse