New Horizons: Awaiting the data

Update July 20, 14:43 PDT: Contradicting what I was told about on July 17 about the image release policy, some new images (from the "First Look A" downlink) were posted to the raw image website this morning. I don't know what this means for future image release schedule; I do not plan to ask further questions about it now, as the situation appears to be evolving. --ESL

New Horizons is speeding beyond Pluto and Charon, sailing a million-plus kilometers farther away each day. It's still taking photos and other data. Now that we're more than a half-Pluto-day beyond encounter, the spacecraft is seeing again the terrain it observed closely during the encounter, but at a great distance and under challenging lighting conditions; we don't know yet how scientifically useful those pictures will be, although I am sure they will be visually striking. New Horizons will keep being able to point optical instruments at Pluto until the end of July, after which they begin spinning the spacecraft to increase the data transmission rate from about 1 to just under 2 kilobits per second.

It's time to take stock of what images we have on the ground, and what we can expect to see in the future. Here is what has hit the New Horizons raw image website so far. A few more pictures from the flyby have come out as processed press photos, but are not on the raw website yet. By the end of the day Monday, July 20, New Horizons will have finished transmitting to Earth all of the "First Look" data set, and that will largely bring image transmission to an end until mid-September. But when will they get to the raw images page for the public to play with?

I've been eagerly awaiting the posting of the post-flyby images to the New Horizons raw image website, but alas, it became clear on Friday morning (when the images from the first post-flyby transmission, "E-Health 1", should have hit the website) that the mission has decided not to keep updating it the way they did before the flyby. The policy enshrined in the mission's Data Management Plan (PDF) is that "There are no proprietary data rights for the New Horizons mission. Selected uncalibrated (CODMAC Level 2) data, particularly image data, will be publicly released by the New Horizons project over the Internet in close to real time." In January of this year, when it became time to interpret that policy for the Pluto flyby, Alan Stern told me that images would be released within 24 hours of arrival on Earth; at the end of April, when they began raw image release, that became 48 hours of arrival on Earth, and that's the policy that applied throughout Pluto approach.

The 48-hour mark came and went early Friday morning, my time, and no pictures appeared, nor have any new ones appeared since. So I used my opportunity to ask a question of the panel at Friday's briefing to ask what was going on with raw image release, and learned that it will henceforth be less rapid and less frequent. I've exchanged some emails with Alan Stern to find out about the new policy. Here it is:

• Raw images will be released in weekly batches around 4:00ish on Friday afternoons, eastern time (1:00pm-ish PT, 20:00-ish UT).
• The raw image releases will include everything downlinked through "close of business Tuesday" preceding that Friday (which I interpret to mean 14:00 PT | 17:00 ET | 21:00 UT).

Given the fact that downlink of images pauses on July 20 and resumes mid-September, that means Friday, July 24 is going to be the one major raw image release day for the summer, and then there may not be much further until September 26. (However, I have heard it rumored that there may be a couple of images that sneak in to image downlinks between now and then.) So, amateur image processors: Mark your calendars for Friday the 24th, let's hope the pictures do get released then, and then people around the world can spend the weekend processing the pictures. Post them far and wide, to your blogs, Twitter, Flickr, unmannedspaceflight.com, or whatever, and encourage your friends to join in and see what the science team was seeing this week. On Monday, the 27th I'll do a great big image roundup!

Why the change in policy? One reason I've heard is that it's turned out to be somewhat harder than anticipated to turn the downlinked data into images. The near-encounter data set is a little weird compared to the slightly more routine data transmissions that happen before and after the near-encounter phase, and the software tools built on Earth to automatically turn the downlinked data into science data sets are not working in the automatic way that they should. One significant obstacle is the fact that the spacecraft has not yet had a chance to downlink a lot of the telemetry that documents where the spacecraft thought it was pointed when it took the data, and that kind of telemetry data is used by the automated routines. The images are fine, but without that telemetry, the processing pipelines don't always know what to do with them; it takes a lot of manual interaction, and that takes time.

As for the other reason, I believe that it has to do with the team having a desire to caption and release images themselves before they go out on the raw website. They had evidently thought that 48 hours would be enough to do this, but that process has not gone as smoothly this week as it could have. While the flyby happened perfectly at Pluto, there were several problems on Earth this week with image releases, which have to be coordinated between the Applied Physics Laboratory and NASA. None of these problems are particularly important; they certainly don't reflect any problem with the spacecraft, which is operating exactly as planned. But because of these problems I don't think that as many captioned images are being released as were initially planned, and I think the mission desires more time to work with the images and put their own stamp on their interpretation before they get posted. There are certanly plenty of people inside the mission to do such caption writing, as they actually hired an additional six science journalists to help with communications during the encounter. I was offered the opportunity to be one of them, but turned it down in part because I figured that, especially with the raw images being released, I could more effectively (and voluminously) communicate the exciting encounter from the outside.

It's been harder to do that than I anticipated, and not just because of the halt of raw image release. There are several images from the Friday press briefing, including the ones that I most wanted to write about, that still have not made it on to the NASA or APL websites: especially, a high-resolution, three-image mosaic of the southern end of Tombaugh regio, and a detail view of the dark potential "wind streaks." That's one reason I never wrote up that press briefing. I was waiting for the photos and refuse to post ugly screen caps from NASA TV. If I had had the raw images available to me, this would have been no problem; I could have made my own mosaic and written about it. While I wait for those images, I hope you'll enjoy the excellent coverage of Friday's briefing provided by Nature, Science, National Geographic, NBC News, The New York Times, and others. Anything I wrote at this point would just repeat what they said, because we were all at the same briefing and shown the same pictures.

What do mission scientists fear from raw image release? Based on past assurances, I had thought that this fear was not shared by the New Horizons mission team. But there's a common fear shared by many scientists on many missions: They're afraid of being scooped. Not so much by the public (although there are certainly scientists on some space missions that I will not name who like to try to exert control over what the public may say about their photos), but more by other scientists. I can't explain why this suddenly became a worry the day after the flyby. It's an old story that I've heard on nearly every mission: the New Horizons team has waited nine and a half years to get to Pluto, and fear that some scientist may do work on publicly released images and get it to a journal and published before the science team -- which is still spending most of their time working spacecraft operations -- gets a chance to reap the reward of their decade-plus of work. Even if it's just posted on somebody's blog, there's the boogeyman of the Ingelfinger Rule; despite Science and Nature's repeated assurances to the contrary, scientists fear that some mission outsider posting to a blog will be interpreted as "prior publication" by a journal editor, and their work will be rejected on that basis. There's also the problem that, like any NASA mission science team in the throes of a high-profile event, they are under a gag order that prevents them from disclosing exciting new science without notifying and involving the NASA public affairs bureaucracy first. They could talk about the stuff that has been released, but to reveal stuff that hasn't been through the public affairs process will make for surprises, and NASA doesn't like being surprised. The very people who can tell us the most about what's going on in the pictures are currently under the tightest constraints about what they may say about them. Why did the New Horizons team not consider this a problem in January, but do now? That one, I can't answer.

I don't think the threat of academic scooping is as serious a problem as people fear, at least not in planetary exploration. I do not believe that it has harmed the science output of the Mars Exploration Rover, Cassini, or Curiosity missions. But the stakes are high; I do empathize with scientists' fears, and I also empathize with their frustration about NASA not allowing them to speak freely about what they see. I do see scientists not on the New Horizons team excitedly discussing the press-released images in social media. You may notice that New Horizons team members do not participate in these conversations, because they know things that are not yet public, and can't bring those things to bear on the arguments. Perhaps, if the raw images were out there, the team members would be more free to participate in the public conversations about their mission. Of course, there are many other instruments whose data sets aren't amenable to raw image release.

I think that the release of raw images with lossy JPEG compression and auto-stretching is a compromise that facilitates conversation and inhibits scooping. The compression and stretching munges the data in a way that reduces their usefulness for science while still allowing the public to ride along with the mission. Any publication that would publish an outside scientist's work on poorly-documented, uncalibrated, lossily JPEG compressed image releases is not a good publication, and meanwhile, we can all enjoy the excitement of seeing the encounter unfold in near real time.

Alan Stern has always engaged closely with the amateur image processing community; indeed, he has been a member of, and active participant on, unmannedspaceflight.com for more than a decade. He has vocally supported raw image release as a great way to engage the public, a public whose advocacy for a Pluto mission is partially responsible for the successes of this week. I know I'm not the only member of the amateur image processing community that has lauded him in our public talks for that support. So I'm not going to lie to you; I'm disappointed by the change in raw image release policy, and it hurt that the sudden halt of raw image release came as a surprise to me after I had been cheering for the mission's openness ever since the Jupiter flyby in 2007. It left me unsure what to do with myself for my last three days at APL after the flyby, a time that I had planned to be working on showing the public what they could do with the data that had been released to the Internet.

On the other hand, release of images within four to eleven days of their receipt is still quite fast, and given how much is happening elsewhere in the solar system, it's enough. Meanwhile, the weekly rate for New Horizons image release will encourage us space fans to turn our attention away from Pluto and back toward some of the other active missions. All of the Opportunity cameras, Cassini's ISS, and all Curiosity cameras have always shared their data with the world within hours of acquisition on Earth. Mars Express VMC also auto-posts data, when the camera is operating. Those four represent the gold standard for raw image release. New Horizons LORRI will be right behind them, with chunkier, slightly later, weekly releases of all their data, provided they stick to the most recently articulated plan. I expect Juno Junocam to join the list of raw image releasers when they arrive at Jupiter next summer.

Most other active planetary missions from NASA and ESA release selected images with text captions on a daily or weekly basis: Dawn FC, Mars Reconnaissance Orbiter HiRISE, Mars Odyssey THEMIS, Lunar Reconnaissance Orbiter Camera, and Rosetta Navcam, for instance. Eventually, every mission's data sets -- from all the instruments, not just the cameras -- get formally released into NASA's Planetary Data System (for best results, start with the Rings or Imaging or Geosciences nodes) and/or ESA's Planetary Science Archive. How long that takes depends on the mission; it's as short as three months (for HiRISE images) and as long as twelve months (for Cassini ISS). Not every mission abides by the rules, though (I'm still waiting for the scheduled release of Rosetta OSIRIS data that was supposed to happen in May). The New Horizons Project Data Management Plan (PDF) appears to stipulate that the full browse data set will be released just 2 months after it's received on Earth, though the plan (written in 2005) includes a long out-of-date calendar for data downlink.

It's only once the data are in these public archives that scientists outside the science teams are supposed to do actual science work with it, because it's only then that the data are certified, calibrated, and so on; one reason we pay scientists to work on missions is to make sure that these calibrations are correct and that the data sets are well documented and properly understood, to enable other scientists to work with them. However, there are fears that opportunistic scientists will work on press-released or raw images and get that work published, scooping mission scientists who've worked so hard and not gotten their results out yet. Over the years, I have asked many scientists for examples of these, and have come up with only two examples in more than fifty years of solar system exploration; both are borderline cases that involved people inside the missions that the work was supposedly scooping. I think that is not very many, and not enough to justify all the fear, but it's quite likely that I'm not aware of other examples of professional scooping. If you know of such examples, please let me know in the comments or by email!

And now, I'll go check out Mars or Ceres or Saturn while I wait very impatiently for more from Pluto...