The problem is that the radial distance of Pluto from the Sun isn't determined as well, because all our observers sit nearly in line, right? Could observations from other probes help "triangulate" Pluto's radial distance? It looks like Cassini and Dawn, along with the Mars orbiters, are the furthest from the Sun right now. Cassini is more than 30 degrees away from the Pluto/Sun line, and Dawn is close to 5 degrees on the other side.

Dumb Q: Were there any searches done in pre-discovery star surveys? I see from http://seds.org/~spider/Spider/Misc/star_cats.html that there were some surveys done in the 1890's capturing ~500,000 stars. Any chance they might have inadvertently captured Pluto?

Could a solar-orbiting probe like Ulysses get an instrument into play?

Ulysses doesn't have any imaging instruments (but, amazingly, is still going!).

Cassini MIGHT be usefull.

I doubt if Cassini could help. We're talking about resolving the position of Pluto to within a good bit less than one Pluto radius as seen from Saturn. I doubt if any planetary probe camera can do that. The only hope would be a lucky occultation. Of course, I'd like to be wrong.

I agree about Cassini not being useful. The only way I can imagine Cassini's vantage point would help is a precise timing of a star occultation which is unlikely. VIMS has an occultation mode that's fast enough IIRC, but what are the odds of running into a star bright enough?

Besides, Saturn's orbit provides only modestly different vantage points to Pluto anyway, I'd hazard a guess at something under 20 degrees off Earth's line-of-sight, not a very marked improvement.

Good points both.

ugordan,
Actually, it's more than 30 degrees off LOS:

http://space.jpl.nasa.gov/cgi-bin/wspace?t...=1&showsc=1

Saturn right now is near furthest elongation from the Sun for Pluto's vantage point.

But, doing some back-of-the-monitor (envelopes? What are those? ) calculations, I think that Cassini, at best, can get an accuracy of about +- 25,000 km. That's just for a single shot though. I assume multiple images will improve the astrometric accuracy, but I doubt it could get in the sub-100km range.

Alan - could you elaborate on that? Do you have any particular candidates in mind, or is it just that if there's something down there glowing, you'll see it?

The earliest observation of Pluto that I know of was in 1915. IIRC, it was Pickering (or one of the folks working for him), who took that photo. I'm not sure when that image was discovered to have Pluto on it, so Pickering may never have known. I can only imagine how he would have felt if he had known.

I suppose it would be too risky to allow New Horizons to engage in any automated real-time course correction maneuver as it approaches Pluto?
Just in case there is anybody left on the team still without some gray hair nine+ years after launch.

You're right- an automated course correction would be way too risky and is in any case not needed- the trajectory is fine if we know where to point the instruments, and it's easier to change the instrument pointing than the trajectory. I think we kicked around the idea of an automated update to the pointing based on where Pluto appears in the images (probably over a beer), but it's too complex and potentially error-prone. We have explored in more detail the idea of an automated "trigger mode" where we take images continuously as we scan across where the target might be, and only store the images when the camera detects the presence of the target in the frame, which would have saved some storage space. But we don't have the resources to make such a technique bullet-proof in time for the Pluto encounter, and it would have only helped in a few cases.

The mantra for the Pluto encounter has to be "don't mess up". We only have one shot at this, and getting too fancy always introduces the risk of introducing a fatal error. So we're keeping our strategies simple and bulletproof, and a having to store a bunch of blank images is a small price to pay.

John

Not a happy question but -

Does NH have any special features or strategies to guard against a badly timed safe mode event such as we have occasionally seen on other spacecraft? (Cassini just after the Iapetus flyby comes to mind.) The overriding imperative must be rather different for multiple flyby and orbiter missions: something like 'survive to try again'. Is the different mantra reflected in different design and practice, or does NH just have to live with a higher risk to it's science return from this cause?

Alan Stern already covered that, I asked about that same thing a while ago. Apparently, the approach will be to pretty much disable onboard fault protection and just keep going.

Also, don't forget that they're using the cruise time to conduct rehearsals & refine NH as a total system before The Big Day. That's one good thing about a 9-year transit, anyhow!

Thanks - I knew somebody must have asked that before! And presumably the effects of disabling fault protection (in different ways) whilst deliberately inducing faults can be studied here on Earth. You wouldn't want to rehearse it in transit I presume. As computer hardware gets even smaller perhaps future single flyby missions will carry 3 or 5 identical computer systems that are able to proceed on a majority vote.

Well, computer/memory hardware faults are just a subset of possible fault scenarios. Take Dawn during the Mars flyby which went into safe mode due to a software bug regarding the star tracker. Or the inaugural Ariane V rocket launch for that matter - dual redundant computers both running the same buggy software and the backup unit actually failed split seconds *before* the primary unit.

If all computers are running the same software and the software is at fault, majority voting won't help anything. It would help with cosmic ray faults, but modern rad-hard processors probably already take care of that by themselves.

IIRC, the Soviet Phobos 2 spacecraft contained 3 identical computer systems, with majority rule. En route, one failed. In Mars orbit, another failed, leading the third and still-functioning computer to be unable to override the broken ones...regardless of the number of computers it seems a full fail-over might be preferable to a vote.

I take it then that, because the "Approach" vector, the "Sun-Pluto" vector and the "Pluto-Earth" vector are all close enough to parallel that the various eclipses and occultations are not put at risk by the inaccuracies in your knowledge of where Pluto is?

-Kevin

That's correct. There's some "cross-track" error that we need to worry about for the occultations and eclipses, so there's some uncertainty in their precise geometry, but no concerns that they will fail to occur. There will be course corrections on approach to make sure of that.

John.

A little early to think about it (since the objects are yet to be discovered and many scenarios are still possible), but what is the feeling about navigation accuracy issues on KBOs ? Uncertainties will be higher, targeted object(s) far smaller, so we may perhaps end up with uncertainties at C/A tens or hundreds of time bigger than the object...

I can't think of a reason, on a unique flyby, to even consider going into safe mode. A vote could concern which of several redundant systems to use, but not whether to go into safe mode.

....You know, this is about keeping our minds and eyes open to anything and everything possible. On the science team, we make flip comments about Plutonian cities and forest fires in this context, but more likely would be geologic hot spots, aurorae, or some kind of luminescence-- or maybe nothing at all.

In any case, Nature will have her Plutonian wares on display; we only have to be alert to the possibility of these possibilities so we can hope to learn a little more as we get a brief glimpse of the ninth planet's night side up close, that second Tuesday in July six summers hence, in 2015.

Alan

cool thanks for all those infos..

I think I'll make a resolution to have my landscaping finished and pool in by then.

I hesitate to offer suggestions, since you guys know your system and I don't, but just in case it's useful, here's a thought:

Depending on how you encode them, I'd think it might be possible to use very few bytes to store the blank images. Is storage the limiting factor here, or is it exposure time? That is, will the cameras be taking pictures just as fast as they possibly can throughout the encounter, or is storage space for the photos likely to limit how many pictures you can take? If it's the latter, you could consider a strategy of taking more images over a larger field of view and then deleting the smallest ones if you start to run out of storage space -- on the theory that the smaller images will be the ones with little or no planet in them. (Yeah, you'd have to make an exception for pictures of Nix and Hydra that you KNOW will be small; I'd apply that algorithm per-target, not across all images.)

In fact, even if you don't actually compress the images (perhaps because of computational limitations), you could fairly quickly compute the entropy for each picture and choose to keep only the highest-entropy images.

But this all presupposes that the biggest limit is storage space, not camera speed.

--Greg

Greg, do I correctly understand your idea is the software would not delete a picture (even a blank one) till the memory is full, and then it only starts deleting in order of most blank to least blank (frame average), and would not save a new 'blank' frame that was 'blanker' than anything already in the memory ??

Yes, that's the idea, where "blankness" is defined as "how much space would this picture take up if we compressed it?" Obviously there are many ways to compress pictures, but for this purpose I expect something relatively simple would do. In fact, exploring such option is something folks at UMSF could play with, if there's any point in it.

But it all hinges on storage space being a limiting factor and not camera speed (or something else that I didn't think of).

--Greg

So as long as no one has their thumb in front of the lens, this seems to yield a maximum of potentially useful images. You might want to allow the software to allow flagging some images to be exempt from the screening like Alan's night side of Pluto shots. A blank frame in that case does tell us something, no aurora or incandescent volcanic eruptions for instance.

Interesting idea.

What might make more sense would be to apply the algorithm only to each set of photos, but not apply it between sets. So, for example, if they're not sure exactly where Pluto is, they take (say) 20 shots instead of 10, but just keep the "least-blank" (highest-entropy) 10. Then they do the same thing on the unlit side. You'll end up with 10 pics from each side, even if the least-blank nightside pic is actually MORE blank than the most-blank sunside picture. For something like Nix or Hydra, you might take four or five images but just keep one or two, on the theory that all but one will be blank (unless you split the moon between two frames).

Again, though, someone from the New Horizons team needs to tell us whether they expect to be storage-limited or exposure-time-limited.

I assume the raw pictures from the Jupiter encounter are available? That'd probably be the data to practice with. It'd be interesting to see which shots would have been discarded.

--Greg

If looking for such ethereal phenomena as "geologic hot spots, aurorae, or some kind of luminescence"(Alan Stern) on the unlit side, it's probably better not to eliminate any images. The darkest of all the images may be where these traits are most likely to be evident.

Yep, that makes sense too. As long as there ARE some sets where it's appropriate, though, it could still be a useful technique.

--Greg

Hi, I've been in hibernation regarding NH for as long as the space craft has been sleeping so I read all posts here since last year and tried to check out NH's website for a trajectory and position update but I was surprised when the website didn't open. I thought it was a temporary problem but it is three days now and I always get an error message saying "the website is not available". Anybody else experiencing the same problem?

BTW one stupid Q: suppose no KBO is discovered on or near NH's trajectory? What follows next?

I'm guessing not much.

Though, another whole-solar-system portrait like the Voyager spacecraft did would be nice

Spirit,

All the APL web sites are down due to IT problems. I suspect they will be up soon.

As to finding KBOs, the Monte Carlo models predict we should find 4-8 to choose from and our pilot search effort confirmed the search feasibility. So I am confident. It the search does go bust, we will try harder with bigger guns. If we never find anything, then we will not have an extended mission, but I consider that unlikely.

-Alan

To answer that question, I'd say (greatly simplifying) that we're limited more by observation time near closest approach, and by storage space further from c/a. It's near c/a that we are taking all the blank images because that's when the uncertainties in the object locations are largest compared to our fields of view, though storing fewer images also saves us a bit of time, because time is needed to allocate the memory for the images before they are stored.

Your suggestion could still be useful, however, in an ideal world- as I mentioned, we've been studying a similar scheme where we only store images that exceed a certain brightness threshold. But even our somewhat simpler scheme will probably not be implemented at Pluto because of limited resources for the exhaustive testing that's needed for any change in our software, particularly changes in how we store images.

John.

Thanks for letting us know. My first thought was "but you've got six YEARS to do it!" But then I realized you probably don't have many programmers and testers on your team now, nor any budget to hire more. In that case, it doesn't matter how much time you have. Since the code has to be absolutely bullet-proof, I well understand that it'll take a lot more time and effort than most of the code we wrote when I was at Microsoft. (In fact, I'll admit the idea of having to write code with so little margin for error is absolutely terrifying.)

Out of curiosity, could you answer a couple of questions about the code? What langauge is it written in? How many lines of code is it? And do you compress the images, or are they stored raw?

Thanks again. It's great that you guys monitor us and actually send us replies!

--Greg

http://pluto.jhuapl.edu/ is back online.

On Tuesday, New Horizons will awaken from its long (202 day) hibernation that began in December, and begin almost two months of checkouts, calibrations, cruise science, and navigation tracking.

-Alan

What will cruise science take during the NH's awaken times?

SpaceListener-

This year we're limiting ourselves to space plasma measurements by SWAP and PEPSSI. Next year we'll do the same and also make phase curve studies of Uranus and Neptune with LORRI and Alice measurements of H in the interplanetary medium, as we did last year; also, whenever we're in hibernation, SDC operates to map out the interplanetary dust density.

-Alan

Alan, I assume that attitude control is one of the few systems that stays pretty much live all the time. Does that include the IMU, or do you have any sort of secondary rate and/or displacement (or star!) sensors that eat less power & just keep NH at the last known attitude?

Nprev,

Actually, Guidance and Control and IMUs are off in hibernation. We just set up an inertial spin and leave it.
When we need to change the pointing (e.g., Earth moves out of the antenna beam), we wake her up-- which is only a couple of times per year.

-Alan

Alan, Thanks for the comments.

One thing that I interest is about the interplanetary dust density. Is there any insight about this? How does the dust density varies between the trajectory between the Earth and the present position. I am assuming that the asteroid ring between Mars and Jupiter has the highest dust density, doesn't it?

Thanks, Alan. That's quite cool! Clearly she's an extremely well-balanced machine.

Yes, balanced right down to the state quarters we used as trim masses!

Our dust data are not published yet. The previous data is from Pioneer 10/11 that operated out to about 18 AU and showed constant density dust as you go outward beyond Jupiter. The only viable explanation, we now understand, is that there must be an outer source to counteract the dilution effect of the volume as you go farther out-- so in effect, this is evidence for the Kuiper Belt, but no one recognized that at the time (early-mid 1980s).

Alan

*head smack* An opportunity missed! But then the KBOs weren't really detectable until those searches in the early 1990s were they?

The topic about the dust space is interesting. The SDC has the aim to prove that the source of dust space is originated from KBO (I am afraid I didn't interpret clearly well about the phrase: .

What would you be the outer source supposing?

As the NH goes farther from our solar system, the dust space must be decreasing since it is leaving the dust zone originated by the formation of our solar system and also from the orbit zone of comets and asteroids. They are source of dust.
Anyway, I am supposing that the distribution of dust in space is not uniform but its dust density average must be consistent with a constant and long direction. Does the SDC compute the dust density in intervals of time?

"...this is evidence for the Kuiper Belt..." -- Alan Stern

Flash report for today: New Horizons is awake after 202 days of hibernation that began way back on December 16th, and all systems telemetry looks good! Cruise checkout 3 has begun. It will continue until late August.

-Alan