They do. The challenge is offloading the data thru the flight computer to be transmitted back up to Rosetta and from there on to Earth - all whilst the flight computer is trying to land.

This remembers me of the infamous 1201 and 1202 alarms. Unfortunately there won't be any human onboard Philae, so actually delayed transmission is better!

Anyways it's really incredible to see how electronics evolved during mission journey!

As Curiosity demonstrated, though, it is quite possible (with modern electronics) to take descent images, store them and send them back to Earth later, when the landing sequence is *not* making the lander too busy to add the image transmission to its duties.

I would hope that planetary landers from now into the future will have this capability. It's not just valuable for developing context of the landing site within the general area, it also has the highest "cool factor" of any part of the missions, IMHO.

-the other Doug (With my shield, not yet upon it)

What I really hope/expect from future missions is HD video-shooting capability, rather than 1 Frame Per DAY "videos" ! Processor power and storage space are no more a mass/space issue, even if you'd enclose processor and RAM inside a lead box 3 mm thick! (Who cares if it will take years to send HD data back to earth? One day or the other they'll eventually arrive)

What was available on the (space) market in 1993 (date of mission acceptance) for onboard processing power and data storage? 10 MHz and 10 MB?

There's a few issues with what you're saying. Firstly - Rosetta is taking a lot lot more than 1 frame per day..... that's simply what's being released at the moment. Once all this data is delivered to ESA's PSA - there will be hundreds of image of 67P that people will be able to turn into pretty amazing animations.


And who cares how long it might take to downlink an HD video? Well actually - all the scientists and engineers entrusted with operating Rosetta and doing science with it. The amount of data it's possible to return from the somewhere as far as 67P is a massive constraint on operations. It's not a question of taking an HD camera, or processing it's data...it's getting it back to Earth.

A rough calculation: 90 second of highly compressed 720p video is around 170 Megabytes. At typical data rates from Rosetta - that would take 8 HOURS to download from the spacecraft. That represents pretty much an entire day's downlink from a typical deep space spacecraft. And for what? An HD video 90 seconds long that, to be honest, would be better represented by a few 2k x 2k still images anyway - and you would have given up all the other scientific measurements Rosetta could take that day. These spacecraft do not last for ever. It's wrong to assume data will somehow magically make it to Earth eventually.


Curiosity is capable of, and indeed has taken HD video from the surface of Mars - but only on a few very specific, very select and very rare occasions, for the same reasons - the quantity of data it's possible to return from deep space is a massive, massive constraint. Kaguya at lunar distances was capable of, and indeed did return many HD movies of the moon where the data rate doesn't represent such a constraint ( typical downlink from lunar orbit is more than 2,000 times faster than Rosetta right now.

Many thanks, Olivier !

I'll be there, right in time !

I made a slopemap of site J using my terrain-model. (I'm assuming that the center of rotation is the gravityvector for the surface)

Click to view attachment

Cyan=good
Red=bad

And an anaglyph of J using the same terrainmodel.

Click to view attachment

Great work! It looks promising for Philae. Can you tell us how much percent represent the "cyan=good" (<few degree) area?

'An HD video 90 seconds long that, to be honest, would be better represented by a few 2k x 2k still images anyway'

There may well be dynamics that only video will show, something to keep in mind for a future comet probe. I think of water dumps from old live feeds on NASA TV.

'Curiosity is capable of, and indeed has taken HD video from the surface of Mars - but only on a few very specific, very select and very rare occasions, for the same reasons'

It seems a waste to have capability that isn't used.

I wonder what the minimum time between 'RGB' filtered frames is for Rosetta? If the interval is brief enough there should be color images of the comet. They can eventually be used to help add color to lander images.

Another surface feature map I made to try to understand the obstacles Philae faces.

I have no real unit for surface roughness. (is there one?) but it is clear that you want to land in the blue areas.
Click to view attachment

I'd like to make an elevation map, but I can't find any suitable free program. I tried a couple of "disparity map generators", but I get ugly results...
DMAG, StereoPlus... what else could I use?
I already 3d-printed the whole comet, now I'd like to 3d-print the landing site too, before landing happens!

You have to have a resonable disparity generator. preferrably one that understands the two cameras orientation in respect to eachother. (or you will have to bring the pictures into alignment with some other software.) then you will have to tweak and tweak until you have the best possible settings. Then you will probably get something that is resonably ok on a mid to global scale. there will still be local noise. To combat that I would suggest that you use shape from shading to capture the fine details.

Thanks a lot Mattias for this very, very good work of yours !

If fact, we are not on Mars and there is not such a big number of 'bad' slopes, because in such a low-gravity object, we may 'land' in an hi-degree slope and stay there without sliding 'downwards', because we will screw into it.

On the other hand, and because of such an eventuality, there is a strong possibility that the CIVA panorama looks very, very tilted with little horizon to be seen. Anyway, the horizon is so close that most of the interest of the CIVA images is in the looking 'downwards' to see tiny ground features around the lander that may sublimate/evaporate/move as we are closing to the Sun...

Click to view attachment

Happy to be of service!

I'm more worried about the rough boulder fields. I think those are more troublesome than just a slope.
But it is hard for me to really say much about. I would like the 10Km OSIRIS look of the area.
I made my roughness map to get a feel for the different areas.

Thankfully there is a lot of blue there. (and landing in a red area could with some luck work just as well)

There are several measures of surface roughness, most of them related to the usual statistical techniques to describe a distribution. It's mostly about amplitudes in a hipass version of a DTM, or by applying the shape by shading philosophy and assuming constant albedo, about amplitudes of the brightness in a hipass version of the greyscale image.
Dimension will be length in many cases, unit in meters, e.g. for the standard deviation; for variance it would be square meters.
You could now correlate those surface roughness values with probabilities for successful landings, to provide a probability map of successful landing constrained to the surface roughness aspect.
Adding slope etc. to a simulation of landings could provide an overall probability map for successful landing with Monte Carlo methods as the underlying principle.

Besides geometric surface properties, the consistence of the material will play a key role for successful anchoring.

Absolutely agree with you Mattias.

The big boulders are of a real concern, because :
- they may made the lander touch them and flip over upon arrival ;
- the lander may not be able to screw into them ;
- the lander body may land on the very top of a boulder (big or small) and its landing mechanism may not be able to fold down enough on its sides to get a grip on it and/or the soil...
It's really going to be a tough landing anyway !

Isn't an anaglyph enough? It is already aligned and I can extract images from it with StereoPhotoMaker.
http://www.esa.int/var/esa/storage/images/...ull_image_2.png

Is it available to the public?

Questions from an interested non-expert:

I have read that Philae may still partly function if it lands in the wrong orientation (lying on its side, or similar). Is there the possibility that Philae may bounce off the surface, or even not contact the surface? In either of these scenarios, would Philae still be able to transmit back whatever data or images it has already taken?

Is there a scenario other than spacecraft system failure where Philae would not be able to transmit data back (like landing upside down, perhaps)?

Thanks.

Good question. The landing site is so chaotic...
Franckly we don't know : so mainy possibilities of failures and so many possibility of successes...
What is certain is that as soon as it touches the ground, the lander tries to screw its footpads into it and there is a small engine on top of it that pushes it downwards to maintain a contact as long as the grip is not firmy done...
We'll know for sure in what shape the lander is upon arrival on the 12th !
Hope for the best !

NSSDC about Philae:

S-band communication usually is rather tolerant in terms of pointing.
Although I didn't find reliable information about whether the (two?) Philae S-band antennas (at least taken together) are omnidirectional.
The other instruments may shield the radio waves in some directions. But if Philae doesn't vanish in some deep gap, I'm optimistic, that communication will be possible in some directions, at least, as long as the batteries are charged.

Most descent data will be transmitted before touchdown; thus we'll get some infos of the descent phase, at least.

What do you mean by "centre of rotation"? What point do you point the gravity vectors to?

I suppose that for these purposes, and without a mass distribution model, you could choose that point such that some average slope over the area of the plot is close to zero. Of course that may not be very accurately true.

Whilst the lower frequencies are less fussy regarding pointing - the predominant factor in required pointing accuracy is antenna design - NOT frequency.

The S-Band relay from Huygens to Cassini had a beamwidth of only 2.3-2.4 degrees because it used Cassini's HGA.
http://descanso.jpl.nasa.gov/DPSummary/Des...3--Cassini2.pdf ( page 8 )

Whereas Deep Impact's probe-flyby relay used S-Band but patch-antennas that were in essence omnidirectional and managed 64kbps downlink at 8,000km range.

Great new, close-up image of that smoothe area on the "bottom" of the larger lobe. Looks like they also give us a mosaic, too.

http://www.esa.int/spaceinimages/Images/20..._October_NavCam

If tweets are any evidence, we might be getting separation images from both Philae and Rosetta! I'm assuming it would be with NAVCAM? Osiris would be tough to focus that close...

https://twitter.com/ESA_Rosetta/status/527460959657689088

And some COSIMA results: http://blogs.esa.int/rosetta/2014/10/29/co...n-called-boris/

Thank you Gerald for all the good info!



You can obviously build a disparity map from that image. (I made a shape-model of that very anaglyph a while back) but you will not know the actual height of anything unless you also know the camera positions. you get only relative heights.
you can download that old one from here if you like:
http://classic.syndicate.se/image/space/Landingsitemodel.zip



The new one is much better. I have the camera positions and all that good stuff reverse enginnered so it is an absolute map. (it matches my Global shapemodel) will post it when it is more finalized.



I'm simply aiming my gravity vector towards the center of mass of the comet (the point it is rotating around)

I did some simple tests with my global shape model. I fill it with particles and let them gravitationally affect an external point. At the top of the comet the gravity vector is pointing mostly towards the center of gravity of the comet anyway. There are areas in the neck area that have gravity vectors pointing slightly of to one side or the other but it is not by a huge deal. (I'm assuming a homogeneous comet with no mass concentrations because of lack of data) I also did not factor in the spinning. (to much math for one night)



I'm so excited about the landing. I really hope it works out well. There is so much science to be gained if it works as planned.
I made my maps to try to get a feel for the chances. I think it looks really promising. There are very few really steep slopes in the area. And there are big areas of very few boulders. (at the scale of this picture)

About those flat dust plains. Is there any reason to fear that the dust would be very soft and that the lander would just sink?

Sinking under its own weight is unlikely as it will weigh very little. (It could potentially dig itself in with its landing motor maybe?) I think a bigger problem might be attaching to dust with screws.

It's not surface 'weight' that's the problem ( about 11 grams by my rough calculations - 0.01 Newtons )

It's the impact of landing. 1m/sec to a dead stop in, say, 20cm..... is a 5m/sec/sec deceleration - which is 500N of deceleration force for the 100kg Philae: 50,000x higher than the resting surface weight.

...or in the words of xkcd:

http://xkcd.com/89/

A body doesn't rotate around a point, it rotates about an axis. So you have to decide where on that axis your gravity vectors point to.

The centre of mass (COM) calculated assuming uniform density is probably a good first approximation to the gravitational field direction. It would be very interesting to see how close the uniform-density COM is to the rotation axis. That would tell you something about how good the uniform-density approximation is, since (I believe) the COM will lie on the rotation axis for a body that actually has uniform density.

Thank you for the wonderful image in http://www.unmannedspaceflight.com/index.p...st&p=213418 vikingmars! With a bit of cropping and resizing I was able to make quite a dramatic wallpaper for my computer.

Click to view attachment

After staring at it for many days I realized that I was having a hard time grasping the scale of what I was looking at. I've spent some time looking at the scale of the many followup photos and came to the conclusion that the boulders(?) in the rift/conjunction valley(?) must be around 2 to 5 meters. Is this in the ballpark?

While awaiting more expert input, I would say the largest is about 20 meters.

Thats true, I stand corrected...

Anyway, I did not factor in rotation... so then the center of mass was the simplest point to choose.

I also did not calculate my own center of mass. I used the center of the coordinate system that ESA used for this VIRTIS map as my center of mass...

http://mattias.malmer.nu/wp-content/upload...IS_LAT_LONG.jpg

I will try to do a new map where I calculate the effects of rotation and local gravity. then I can do a dot product between the vectors and see if the difference is much to talk about... (the J site is almost on the equator so the rotation should not make all that much difference. J has most of the mass centered under it so the deflection of the gravity vector is not going to be too aggressive either.)

Interesting stuff...

Actually, to be super-pedantic about it (sorry, can't resist) rotations are best considered occurring in a plane rather than around an axis. That is why you can have rotations in a 2D world (no third axis exists to rotate around) and it is how in a 4D world you actually have 6 potential "cardinal" rotations in planes rather than just 4 about axes. Just so happens in 3D that you get one orthogonal axis per plane.

I agree with Doug about the force of contact with the comet's surface.



In fact, I predict that the Philae lander will sink out of sight.
In a paper by Schultz, et al, analyzing results of the Deep Impact mission, there is mention that the surface of Tempel 1 has a density of 0.2 - 0.5 g/cc, with a porosity of 90%!

Then how are house sized rocks sitting on the surface ?

Those "rocks" are probably not very dense. They may be more like big fluff balls, or maybe light but stiffer, like pumice.

The lander is not that dense and has three long legs to spread the weight.

Rosetta has been successfully moved to to the correct orbit for the landing.
Won't ever be as close to the comet again for the rest of the mission.

Landing orbit

I don't believe the lander will sink. Brilliant minds have been predicting sinkage ever since the first moon landings. I think the biggest risk would be in the neck of the comet, but in the actual landing area I get the sense that smooth areas have a solid underlayment. Lets hope the doomsayers are just as wrong as they were for the moon landings.

Just want to add some more details on why I think that Philae will sink "out of sight."
First, here is the link to the Schultz, et al, paper, "The Deep Impact oblique impact cratering experiment" -

http://www.planetary.brown.edu/pdfs/3589.pdf

In section 5, "Concluding remarks," there is reference to the low-density (0.2 - 0.5 g/cc) and highly porous (90%) surface of comet Tempel 1.

Here is a photo of one of Viking 1's footpads totally buried in a surface deposit -

Click to view attachment

This "burial"event occurred during the Viking landing. Similar "burials" have plagued the MER rovers on Mars.
These are examples of loosely consolidated, probably very-fine, particle piles. Even with a respectable gravity field, these Martian "traps" remain very porous.
In the case of comets, my guess is that some of the very fine particles expelled in gaseous jets fall back very slowly to the comet surface.
The extremely low gravity on a comet would mean that those accumulations of particles will remain very porous, with a very low density.
So, Tommy Gold may finally be vindicated, albeit not with respect to the Moon.
So, there is my prediction on the fate of Philae as it makes contact with the surface of Comet CG.
I may be wrong, but that is OK. Part of the business of science and exploration is the risk of predictions. I await the landing with anticipation.

Another Phil

I think it's extremely unlikely--not only is the gravity extraordinarily weak, but the vector of same might not even be parallel to local vertical at the landing site--but we'll find out in just a few days. This will be an extremely interesting event!

How are the boulders supported? Is every last one of them sitting in only a few cm of dust over a firm substrate? Whatever the boulder story is, it will probably apply to Philae as well.

(EDIT - OK, I see others said this too.)

Phil

I think this picture deserves its own post in ths thread, but I can't upload it from here right now:
http://blogs.esa.int/rosetta/files/2014/10...ts-1024x586.png

Thanks mcgyver : the very 1st post of this topic gives you a more detailed timeschedule of this event as a linked PDF (see link herebelow) :
http://www.unmannedspaceflight.com/index.p...st&p=213272

The important question is: did the boulders roll very gently into place, or come down at high velocities? As Doug pointed out, the weights are very low due to the very low gravity. But Philae will touch down at 0.5-1 m/s, so a soft and deep enough surface wouldn't allow it to decelerate without getting burried, even if the surface could support the lander's weight if it were very gently set down.

Still, they've planned for very soft surfaces - one description I read said even "one as soft as cigarette ash" (Science story).

The landing site name is Agilkia

Agilkia

I've started compiling a list of resources and background information on Philae in a new thread -- please check it out for answers to questions, and please make suggestions for additions!

Has anyone yet considered images of Philae on the surface taken with OSIRIS? I reckon that the narrow-angle camera would achieve 0.48m/pixel at altitude 29km, does that sound about right? Philae's dimensions are 1x1x0.8 meters so perhaps it could be spotted from orbit.