MainlyMartian

That dog, still dead

Back to the Royal Institution for another -- presumably the last -- Beagle post mortem. This one was a report put together by the Beagle team itself, edited by the mission manager, Mark Sims, who, frankly, looks a bit more drawn than I remember him from back in December and January. Not that that's much of a surprise.

The Pill came too, of course, but this is at least in principle Mark's show. He has two documents to present, a slim Lessons Learned (though when I say slim, that's because it's mostly just a list - there are 288 lessons there, which is quite a curriculum) and a final Mission Report that runs to 260 pages; as of today they’re available in pdf form here. They're a sort of brain dump by the spacecraft team -- everyone's thoughts on what might have gone wrong and how it might have been avoided. An independent report would be nicer, obviously -- but at least we can read this one.

Mark went through the highlights quickly. To summarise but not traduce: something might conceivably have happened when the spacecraft got shaken during launch, or when ice crusted on or near it in the cruise phase, but the team thinks that's very unlikely. Then, during entry, descent and landing, pretty much anything could have gone wrong. Sims said his personal nightmare was that something screwed up the antenna deployment -- that the spacecraft was sitting there all ready to do its stuff for days and days and just couldn't communicate.

The suggestion that abnormal atmospheric conditions -- specifically, low densities around 20km up -- may have been a problem came up again. Mark was open in saying that in some ways the atmospheric explanation might be seen as welcome; it gets the team a little off the hook, though not all the way off, as you could definitely say their design shouldn't have been so vulnerable to slight differences between the atmosphere they got and the one they expected. Anyway, the evidence for a thinner-than-modelled atmosphere is sparse, depending on some error rich data from Mars Express's Spicam instrument and slight anomalies in the MER landing timing which NASA people are quoted as downplaying a bit. (I was later given the impression that there may be more evidence in MER descent data that hasn't been made entirely public yet; we’ll see.) All told, I thought Sims did a pretty good job of opening the possibility up but not over-playing it.

In general, though, there was an odd tension to the whole event. On one side there was the admission that lots of things could have gone wrong, and that one of them surely did, even if none of the failure modes stands out, even in retrospect, as particularly likely. On the other side there was a tendency towards a the-operation-was-a-success-but-the-patient-died sort of attitude. That still doesn't really wash. But it's good to have a long list of lessons -- whatver happens next, they should be useful to someone, somewhere, sometime. (I particularly liked the one that says don't turn off the heating when the engineers are working all night...)

Scanning the documents later, a few things jump out at me, but that's not to say they're really significant. One is that the team isn't absolutely sure the spacecraft left Mars Express pointed in the right direction, though there’s no particularly good evidence that it didn’t. Another is the static discharge while the system was still on the ground that led to some components being replaced and others not. Another is that after redesigning the parachute, the team didn't do a full reanalysis of the danger that the new parachute might be more liable to snag the spacecraft or the airbags after landing than the old parachute had been. Another is that there were problems with overseeing the work that Martin Baker, an American firm, did on the EDL systems and the gasbags; US ITAR (International Traffic in Arms Regulations) considerations apparently got in the way. (Martin Baker left the project in 2001, after which Astrium took over their part of the deal.)

As I say, these things piqued my fancy, but they're not necessarily of any great import. The reports don't particularly play them up or down, and I have no reason to second guess them. The truth is that without EDL telemetry we were never likely to have a firm answer. (The Mars Polar Lander inquest was more conclusive, but that could easily be because there was a clearer smoking gun, rather than that the inquest process was more incisive.) One interesting detail is that work was done at one point on an EDL telemetry system, but it was abandoned when ESA said that there was no orbital or earthly asset that could be made ready to receive its signals. Was that really the case, I wonder. Or was it just felt that altering the Mars Express trajectory so that the orbital insertion burn came a bit later and the spacecraft was thus free to monitor Beagle’s descent was not worth the risk. (I don’t know enough astrodynamics to say whether such an alteration was really possible; insight welcome.)

That brings up one of the lessons learned which is mostly between the lines of the reports; that ESA’s ambivalence was bad for the project. The implication is that ESA always sort of expected Beagle 2 to drop out, so it never did more than tolerate its presence. One gem in the reports is that ESA didn’t actually send representatives to one or two of the critical design reviews. Yet at the same time, ESA didn’t actually cancel Beagle either. Now whether a more hands-on ESA would have resulted in better engineering is hard to say; but a firmer commitment might have led to better management and a more even cash flow.

It’s far from obvious that Colin would have welcomed a more intrusive interest, even if it was benign -- being the outsider, and an outsider quite clearly labelled as British, clearly suited some of the narratives he was spinning about the project, and quite possibly his own self-image. But it does seem to me that these reports make clearer than ever the fact that ESA should either have cancelled the project or thrown a little more into its success. I know -- if they had, and something had still gone wrong, and as a result the orbiter had been lost too, then the calamity would have been far greater. Hard call either way. I’m just saying how it looked from the cheap seats today.

The press conference ended, predictably, with future plans. Mark talked of a possible 2009 follow on mission. Colin, typically, said it was worth pushing for 2007. He has half a point –- as he sort of said, a 2009 Beagle-based mission would look a bit puny compared to NASA’s MSL. But he has no funding, and while ESA may have a commitment to “fallen comrades”, as the phrase has it, it may be that Beagle looks more like a fallen fellow traveller -- especially when the ESA science budget is very tight.

Colin also said that he’d written to NASA about the possibility of flying a Beagle package as part of the 2009 mission. He painted a sweet picture of the MSL scurrying around picking up samples and bringing them back to little Beagle for consideration. However, it’s very hard to believe that Colin really believes this is plausible, and there’s no reason at all why any of the rest of us should, despite the fact that this, to judge by the mood of the media at the meeting, will be the idea that gets picked up in the papers. MSL is going to be packed with American instruments aiming at many of the Beagle objectives, and the chances that a large number of them will be turfed off to free up mass for an anglo-european instrument package that’s the brainchild of a man quite willing to indulge in some America-bashing when it suits his purposes seem to me negligible. As in nil. Maybe Colin’s an eternal optimist. Maybe he’s trying to pick a fight, and thinks being rejected or ignored by NASA will help his case elsewhere. Your guess is probably as good as mine.

And no formaldehyde either

In the long and intermittently very interesting comments thread to the previous post, there's some more discussion of what we should think about formaldehyde. Formaldehyde is one of the trace chemicals that Vittorio Formisano talked about looking for using the Mars Express PFS in his interview with Linda Moulton Howe. And while Formisano does not claim to have detected formaldehyde, such a claim was made a decade ago by scientists using an instrument called Auguste on the 1989 Soviet Phobos mission (Korablev, O.I., Ackerman, M., Krasnopolsky, V.A., et al., Tentative Identification of Formaldehyde in the Martian Atmosphere, Planetary and Space Science vol. 41, pp. 441-451.)

Since formaldehyde is an organic chemical that could most easily be made out of methane, you'd think that the discovery of formaldehyde on Mars would be a big thing, since it would imply the existence of methane. (Though there are other ways to make formaldehyde on Mars, they probably wouldn't make a lot, and they'd be local to the ice caps). The reason the 1993 formaldehyde paper is not seen as a big thing is that, by and large, I don't think anyone -- including the people who made it -- really puts much faith in it these days.

That's not to say no-one talks about it. As Rick Sterling points out in those previous comments, the tentative detection made in 1993 is cited in a 2003 paper by Formisano and colleagues entitled "Atmospheric Photochemistry Above Possible Martian Hot Spots", (a paper presented at COSPAR, I think, not one published in a journal). This paper asks what an instrument like the Mars Express PFS might expect to see if there's a little volcanism still going on on Mars. The authors say that one explanation for the formaldehyde might be a "large flux of methane from a localized source in the Martian atmosphere" such as a volcanic hotspot. Rick also points out a 1999 paper (Weiss B, Yung, Y & Nealson K, Atmospheric Energy For Subsurface Life On Mars, PNAS, vol 97, pp1395-1399) where the authors say "Formaldehyde, which has no known geological source on Mars & has a lifetime in the Martian atmosphere of only 13 hours, may have already been detected by the Phobos spacecraft, although this is uncertain". On the issue of uncertainty the paper references a 1997 paper by Krasnopolsky, Mumma and colleagues in which formaldehyde was not detected in earth-based measurements at anything like the level apparently seen from the Phobos spacecraft (Krasnopolsky, V., Bjoraker, G., Mumma, M. & Jennings, D. J. Geophys. Res. vol 102, pp6525–6534).

To me, though, the key is a 2002 paper in which the first author on the original formaldehyde detection paper provides a detailed summary of the Auguste instrument's measurements (Korablev, O. I., Solar Occultation Measurements of the Martian Atmosphere on the Phobos Spacecraft: Water Vapor Profile, Aerosol Parameters, and Other Results, Solar System Research vol 36, pp12-34). It's a paper written largely from a desire to dispel the idea that the Phobos mission was an ignominious failure; as Korablev argues, it sent back more useful data than all the USSR's previous Mars missions combined (though that's not as grand an achievement as it might sound). To make this case Korablev goes through the relevant data very carefully, seeking to provide a definitive account of the achievements.

In his discussion of the features in the 3.7 micron region of the infrared spectrum that he and his colleagues had tentatively described as absorption by formaldehyde, Korablev shows that in the final analysis, if you explain the data from the relevant two pixels (yes, just two) of the spectrum by invoking formaldehyde, you get a formaldehyde level of between 200 and 1,300 parts per billion, with a best estimate of 500 parts per billion. That's slightly lower than the figure they originally published -- but still a staggering fifty times the level currently being quoted for methane. And Korablev himself does not believe it.

He goes on to compare this interpretation of the data to ground based observations -- some published, some not -- including those of his co-author on the original paper, Krasnopolsky (who, in case anyone has forgotten, leads one of the teams that reported methane earlier this year). Those observations, which have much better spectral resolution, set limits on the formaldehyde level of 3 parts per billion or less. That's a two or three order of magnitude difference between what the spacecraft seems to be saying and what the earthly observers say. What's more, it's very hard to imagine what might be producing formaldehyde in the huge amounts implied by the Phobos data, since methane levels are so much lower; if formaldehyde is being made from methane, which has a longer atmospheric life, you'd expect there to be a lot less of it, not a lot more. (As I read it, the Formisano et al “Hotspots” paper mentioned above calculates that formaldehyde should be about ten million times less common than methane if it’s being made from methane in the atmosphere). Nor does Korablev think the formaldehyde hypothesis can be saved by suggesting that it's a very local phenomenon related to a hotspot or a life-bearing oasis. The relevant spectra were measured over completely different locations on Mars.

On the basis of all this, Korablev concludes that the features seen by Auguste in the 3.7 micron region cannot have been due to formaldehyde. And I conclude, yet again, that making sense of spectra is a very tough and time-consuming affair.