Philae Wakes Up! Options

[Malmer]

I was thinking about adding some kind of indicator of comunicatability on my rosetta now page. a dotproduct between rosettas position vector and philaes local surface normal would be one part and then perhaps a dotproduct between the surface normal and the sun vector to give an indication of philae energy budget... that gives two scalars... if i multiply them you get a sortof communication factor...



Would that be interesting?

[fredk]

That would be interesting. Toss in the dot product between the lander body's +z axis and the lander-orbiter vector, for ability to receive signals from the lander, and you've got something very interesting. I believe someone mentioned a roughly 60 deg transmission cone about z+. Perhaps scalbers can "point you in the right direction" for z+.

Of course this would be with the caveat that the +z axis direction is not known very well and may have changed since landing. There's also the caveat that the very local normal may not agree well with the normal determined from your shape model, and of course cliffs/boulders will block sun and communications. But a plot of these three indicators vs time (and perhaps their product) would be very cool to see since it would give an indication of when communication might be possible...

[climber]

From Twitter, another brief contact last night.
https://twitter.com/philae2014/status/614093166166298624

[climber]

And another one from Mopus itself: https://twitter.com/philae_mupus/status/613990454460325888

[scalbers]

But a plot of these three indicators vs time (and perhaps their product) would be very cool to see since it would give an indication of when communication might be possible...

Sounds pretty interesting. I do have the green dot at my estimated z axis location. However I may want to ask opinions about the latitude of Philae at its location. Also, do we know the longitude and (solar) hour angle of Philae when the CIVA mosiac was taken? My diurnal solar track, horizon, (and green dot) is presently assuming Philae is on the equator. Being over the edge of Hatmehit it may really be a somewhat southerly latitude.

The ultimate would be if we can consider the shadow of Perihelion Cliff in the calculations, though this would only be relevant if Rosetta were to move in to higher north latitudes.

[chuck0]

Adding this measure would really be pretty nice. Though as others have mentioned using the local surface normal at Philaes suspected location might be not really robust. Instead you could take a sphere around the position and compute your measure for all vertex normals in there. After that the measures can be averaged (maybe weight values at the center of the sphere higher and at its boundary lower? ). And right, the most exact thing would be to actually reconstruct the landers actual orientation from the CIVA pano.

Btw, another interesting thing could be to mark the locations of rosetta when contact was established. Though to get this right we would have to know what time the timestamps in the telemetry blips actually relate to. As others have suggested it is probably either Rosetta or Ground time...

Anyways malmer your work is really really impressive!

[Malmer]

I calculated a very good estimate of the suns direction vector in Philaes body coordinate frame using the CIVA panorama. Problem is that I do not have the exact timestamp when the CIVA pano was taken. I do not exactly know how philae is rotated around that vector but it is fairly well constrained. The horizon is in frame and we know Philaes position (or a good candidate pos at least)

sun vector:
http://www.unmannedspaceflight.com/index.p...st&p=215678

Problem is that they moved Philae after that Pano was taken... but that was a rotation around its Z axis right? so that would make no real difference in comms? Unless it also tipped over?

I really would like that Civa Timestamp...

[Herobrine]

Chuck, I haven't checked the timestamps of the battery data points against reported times of contact (when I'm not on a phone, I'll take a crack at it) but assuming the actual contact times for any of the post-wakeup comms have been reported, it shouldn't be too hard to check if the timestamps match the contact time or the DSN downlink time or neither. I posted the battery data received yesterday 7 or so posts ago and the timestamps were:
2015-06-24 17:24:11
2015-06-24 17:28:35
2015-06-24 17:33:00
2015-06-24 17:37:24
The MUPUS tweet climber posted shows a timestamp 1 minute after the last battery data point, so they're likely the same type of timestamp. I'll take a crack at seeing if I can correlate the timestamps with the contact times later tonight.

[scalbers]

Problem is that they moved Philae after that Pano was taken... but that was a rotation around its Z axis right? so that would make no real difference in comms? Unless it also tipped over?

I really would like that Civa Timestamp...

Indeed it's reasonable to assume a rotation about the Z axis and it wouldn't impact the communications, nor the direction/azimuth of tilt of Philae. I think a reasonable estimation of the latitude corresponding to Philae's horizon can help correlate with the solar diurnal arc, if the exact CIVA pano time is unavailable.

[Herobrine]

After looking through these times and timestamps, I can only conclude that the timestamps in battery data from November 2014 are UTC SCET, but that the timestamps in battery data from June 2015 are in UTC ground receipt time.

I can see a clear indication of the separation event in the power data at timestamps corresponding to the time of separation, down to the minute. I can also see a clear indication of the initial touchdown and of the penultimate touchdown, both with timestamps matching the time those events are said to have occurred, down to the minute.

When I look for post-wakeup communication, on the other hand, timestamps in the power data match reported ground receipt times instead. For example, the first post-wakeup signals "were received at ESA's European Space Operations Centre in Darmstadt at 22:28 CEST on 13 June", so 2015-06-13 20:28 UTC, which matches (down to the minute) the timestamp of the first post-wakeup battery data point, which is 2015-06-13 20:28:46.025.

[alphasam]

Some new info from the Rosetta blog;

http://blogs.esa.int/rosetta/2015/06/26/ro...-a-good-signal/

Edit:

And DLR;

http://www.dlr.de/dlr/en/desktopdefault.as...9/#/gallery/311

[scalbers]

Good links - and interesting that they are varying the latitude to assess the communications. Somewhat surprisingly it seems (so far) to be a better signal at lower latitudes suggesting not so much Philae tilt to the north. It also might fit into the notion that Perihelion Cliff helps to block the signal. Rosetta is flying over the morning terminator, so Philae is really just waking up when Rosetta is passing over the meridian. As the sun rises higher in the east southeast, Rosetta would be retreating lower to the east northeast in the sky and disappearing behind Perihelion Cliff. I wonder if they will consider flying over a bit later in the morning?

Judging from what I heard some time ago, ESA may know Philae's orientation in absolute terms and that should suffice to predict in assessing antenna alignment.

(EDITED June 27 1616UTC)

[Herobrine]

That new Rosetta Blog post contradicts the conclusion I reached regarding the latest timestamps. It says "the contact on 24 June started at 17:20 UT (on board Rosetta)". That is only 4 minutes before the first power system data point, so if the contact with Rosetta started at 17:20 UT, the timestamp couldn't possibly be the time it was received on the ground, because the signal travel time from Rosetta to Earth is around 16 minutes right now.
Either the two reported ground receipt times I used last night (to determine the nature of the timestamps) erroneously reported Rosetta contact times as ground receipt times, or the latest blog post is wrong, or the meaning of the timestamp changes for some reason.

I'm inclined to believe that the latest blog post is correct and that the earlier reports simply misreported spacecraft times as ground times, because that makes the most sense, since the timestamps in the power system data from November 2014 are definitely spacecraft times and I don't have any other reason to think they would change.

If they are Rosetta times, then that is good news because they indicate times that Rosetta was in stable-enough communication with Philae. I've never touched a SPICE kernel before, nor do I know much about them, but I may just have to play around with them later to see if I can figure out where Rosetta was during those contacts with Philae.
The time spans of power system data points from the post-wakeup contacts are (all times UTC):
2015-06-13 20:28:46.025 - 2015-06-13 20:29:29.839 (84 Philae power system data points)
2015-06-14 21:22:51.609 - 2015-06-14 21:23:14.022 (only 2 data points)
2015-06-19 13:20:36.425 - 2015-06-19 13:35:58.438 (8 data points)
2015-06-20 13:55:29.277 - 2015-06-20 14:04:05.573 (62 data points)
2015-06-20 15:59:17.222 - 2015-06-20 15:59:31.859 (113 data points)
2015-06-21 02:32:53.429 - 2015-06-21 02:33:10.115 (20 data points)
2015-06-21 02:43:50.338 - 2015-06-21 02:44:16.879 (47 data points) (There was a MUPUS timestamp of 2015-06-21 02:44:16 tweeted as well)
2015-06-24 17:24:11.336 - 2015-06-24 17:37:24.992 (4 data points) (You could extend that range to the MUPUS timestamp tweeted yesterday, which was 2015-06-24 17:38:43)

[chuck0]

Thanks a lot for looking at those timestamps so thoroughly!
And the two blog posts are really interesting... i wonder if they stay at the morning side of the terminator because that is the zone of lowest activity? Also getting to the night side would probably take quite a while since they are so far away. Anyways i think that the information that distance to the comet seemingly did not improve connection quality is good news. This way they can try to optimize the orbiter/lander geometry in a much safer zone. Also it is interesting that one of the landers antennas might be damaged. I really dont envy the flight ops team. Now they have another really complex issue to juggle around. This mission is really a nail biter from start to (probably) finish.

Oh, and yeah i think ill take a look at those spice kernels as well

[Herobrine]

While I'm thinking about timestamps and communication windows, here's a chart of the amount of time between data points in the post-wakeup power system data.

Note: The vertical axis is logarithmic and the horizontal axis is not linear time; it's evenly spaced data points.
The time ranges I mentioned a couple posts back, for the different Philae-Rosetta communication sessions are indicated by the red, vertical dividers. For each of those ranges, the last time is missing from the graph because it represents the time until the next range (so it's off the chart, high).

I keep looking at the time between data points to see if it can tell me anything about the stability of the communication. So far, I haven't been able to conclude anything other than that there is no simple relationship. There are many deltas that clearly occur often (98 ms, 117 ms, 234 ms, 1400-2000 ms, and 264500 ms, to name a few) but, without knowing a lot more about what data is being transmitted and how the communication between Philae and Rosetta works, I can't really draw any conclusions from it. It seems like there might be different modes of transmission, or perhaps just different types of data at different times. One thing I do notice is that the times between data points are much lower, on average, than they were in November 2014 (most of which were around 264500 ms). I'm assuming the times are shorter now because there isn't a lot of other data being transmitted, just "housekeeping" data.