MainlyMartian

No methane on Mars

Or at least very, very little. Observations made with the Tunable Laser Spectrometer on Curiosity (paper|press release) rule out an atmospheric level above 1.3 parts per billion with 95% confidence: the data is fully consistent with no methane at all. As the authors say:

Our result greatly reduces the probability of significant methanogenic microbial activity on Mars and recent methane production by serpentinization or from exogenous sources including meteoritic, interplanetary dust and cometary infall.

There will be some background from meteoritic sources, I assume, at the parts per trillion level (where I imagine they will stay undetectible even if Mars passes through the coma of comet C/2013 A1 (Siding Spring)). But these observations leave the idea of a methane-producing cryptic biosphere with no supporting evidence at all. This would feel like the end, if it didn't feel like things had ended a while back -- see this post and links therein. I rather regret that Mars looks deader still than it did before, but I don't regret having been so excited by the subject back in the day -- it's the nature of the game.

My only real loss is that I made a bet on the subject with Chris McKay, right at the beginning of the story, and consequently owe him a meal. But that loss is offset by the far greater gain of, hey, having a meal with Chris McKay. 

And another gain is a greater appreciation of the power of theory. The observational case for methane on Mars was pretty good -- good teams, different techniques. The theoretical case against it, though, as brilliantly articulated by my friends Kevin Zahnle and David Catling and their colleague Richard Freedman in this article (pdf), was really persuasive. That insight has broader applicability, as I noted here:

Thomas Huxley, Darwin's ally in the fight to get evolution accepted, spoke warmly of the facility with which ugly facts can kill beautiful theories. But that fatal ability should not hide the fact that well-applied theories, beautiful and otherwise, can play a crucial role in deciding which observations get treated as facts in the first place. 

 

The unlikelihood of methane on Mars

One of the early results I think we're expecting from Curiosity is an analysis of trace gases in the Martian atmosphere -- including the level of methane. There was a time when I, and this blog, were obssessed with the news of methane on Mars and what it meant for the likelihood of a cryptic microbial biosphere in the Martian subsurface (full archive). Indeed it is possible, I can't say for sure, that I may have been the origin of the martian-methane-expressed-as-a-number-of-cows meme found recently in an excellent piece by Dick Kerr (summary|paywalled full text). In the excitement, and newly enamoured with the fun of blogging, I was quite the believer.

More recently I have become more sceptical, very largely as a result of this article (pdf) by my friends Kevin Zahnle and David Catling and their colleague Richard Freedman. My take on the article and its arguments, by which i was and remain broadly convinced, appeared a couple of years ago in The Economist. I won't recap it here, but if you're interested it's worth a look. And I liked the conclusion: 

The debate carries a worthwhile scientific lesson in itself. Observations, which to an outsider might sound like simple things, are often remarkably difficult, and depend on complex models to make any sense at all. Thomas Huxley, Darwin's ally in the fight to get evolution accepted, spoke warmly of the facility with which ugly facts can kill beautiful theories. But that fatal ability should not hide the fact that well-applied theories, beautiful and otherwise, can play a crucial role in deciding which observations get treated as facts in the first place.

Pretty soon after landing, insha'Allah, we should know whether the level is in parts per billion or parts per trillion. It may, indeed, be one of the mission's crucial results (unless something satisfies the "Knoll criterion"...)

Mumma's methane

At the September meeting of the American Astronomical Society's Division of Planetary Science, which is rather pleasingly in Cambridge this year, Mike Mumma will be presenting some very interesting sounding methane data. The abstract is here. The key idea is that the rate of methane production is a great deal higher than calculations based on the theoretical atmospheric lifetime would suggest

Our differential extractions for methane implied strong latitudinal gradients - these are contrary to predictions if photochemistry limits the lifetime of methane on Mars, but instead require local release and a much shorter lifetime. The lifetime against destruction cannot be much longer than equator-to-pole transport times imposed by the Hadley circulation (weeks). A shorter lifetime requires that estimated production rates be revised upwards commensurately.

If the methane had a long lifetime, it would be spread evenly across the face of the planet. It doesn't, so it has a short lifetime (possibly because of catalysed interactions with surface oxidants, or oxidants on dust). And for a given total amount of atmospheric methane a shorter liftime means a higher rate of production. The 2,000 cows have reinforcements.

Neither the absolute rate or the implication that production must be regional in themselves argue for or against biological production. Geothermal sources are regional too (though if the rate is really high, you'd have to wonder why the hotspot doesn't show up in the infrared). Sushil Atreya will be looking into some of the possibilities at the same session of the meeting. Interestingly, Atreya's abstract suggests that some of Mumma's data show a significantly higher level than PFS or the Krasnopolsky team reported.

Also interestingly the Mars Express PFS team (of which Atreya is a member) has no abstracts at the meeting reporting any more methane data, or indeed any trace gas data at all. That said Vittorio Formisano is part of the team reporting on Cassini's intriguing bright spot on Titan.

I'm looking forward to September 6th quite a lot right now.

For anyone needing a refresher, the summary from the end of last year's methane excitement on this blog is here, and the most recent take on the PFS work was here. (And if anyone knows why some of my archives seem to be in a strange blue format, I'd be grateful for the info). There's been a recent series of articles on astrobio.net, including this one on possible methane sources.

MSL shakeup -- and a possible methane mission?

The rumours about MSL, mentioned a while back, seem to be getting stronger and stronger. Since the most recent MEPAG meeting a number of people have started more or less assuming that MSL is going to slip back at least one launch opportunity, to 2011; there's also talk of it being doubled up into an MSL A and an MSL B, like MER. It's not been announced yet and may not be a done deal, but the delay, at least, seems very much to be the way that the wind's blowing. It's possible that a formal decision is going to wait until NASA's new administrator, Mike Griffin, weighs in.

An ancillary rumour suggests that if this happens there will be serious consideration given to moving the next Scout opportunity for a small mission up from the 2011 launch opportunity to the 2009 opportunity. Four years is a very short time in which to go from a sheet of paper to a spacecraft, and a truncated selection process might be put in place, in which only the scouts that were shortlisted last time need apply. Those three were: Scim, a mission to pull dust from the upper atmosphere; Ares, an aircraft that would do magnetometry, inter alia; and Marvel, an orbiter with which to measure atmospheric trace gases, including methane, based on the bus for Mars Odyssey. If it were done that way, then the case for Marvel would be extremely strong: the science basis is all the stronger now we have reports of methane, and the technical risk has to be low compared to the other two, since it's a fairly straightforward orbiter, content to look at the atmosphere from above rather than rip through its edges with ramscoops flaring, or extend its wings and try and fly through the stuff. If you wanted to be really practical about the matter, you could save time by just selecting Marvel straight out of the gate -- but I'm not sure that that's actually legal for a scout mission.

Marvel would get really accurate data on methane levels, it would measure any seasonal variations, and it would pick up a bunch of other trace gases if they're there. It would also have a good chance of measuring the carbon-12/carbon-13 ratio in the methane, a measurement which, if it is different from the isotope ratio in the carbon dioxide, would offer strong evidence for a biological source. (The idea is that the methane-making organisms would feed on a well-mixed global reservoir of carbon dioxide, but would discriminate between isotopes as they did so, just as methane makers and photosynthesisers do on the earth; most non-biological processes don’t make such discriminations.) But in an ideal world -- say, for the sake of argument, a world in which the aim was not to fly an already accredited Scout design, but to fly the best methane mission you could -- you'd want to at least consider an alternative mission that looks for the same data a different way.

Marvel would be in an orbit that kept it close to the terminator; its spectrometer, instead of looking straight down, as most spacecraft instruments do, would look at the horizon -- the planet's limb -- in order to see the sun through the atmosphere. This means it gets a strong signal, because it's got a very bright light source shining through a lot of gas. An alternative, which Mike Mumma at Goddard has thought about, is to put a telescope in a halo orbit around the Martian L1 point, which is on a line between Mars and the sun. At L1 a telescope could spend all its time looking at the whole sunlit face of Mars -- and paying attention to any region that was of particular interest.

The challenge of an L1 mission is that the signal is a lot weaker. Your light source is the surface of Mars, which is a lot less bright than the sun. And the light passes through much less of the atmosphere. Light coming to Marvel on a tangent to the planet's surface goes through something like 40 times as much of the atmosphere as light reflected from the surface to the L1 point, and thus the spectral features due to methane, or anything else, will be considerably stronger.

The advantage of L1 is that you could look at specific places -- at resolutions down to a few tens of kilometres -- and look at them at any time of day that you chose. Marvel has to look only at sunsets or dawns, and it can't choose to focus down on a particular region (though it will see all the regions of the planet on a fairly regular basis). If methane levels are strongly variable in space and time, that would seem to be an advantage for the L1 mission. The lower signal at L1 should be something that can be compensated for in principle simply by taking data for longer; in principle, an L1 mission could stare at a particular spot all day.

I really don't know which mission might be better. If I knew for sure that Marvel's signal to noise ratio made it significantly more capable in terms of isotope analysis, I think I'd go with that. There again, the idea of being able to look down wherever and whenever you want is certainly appealing, especially if it were to turn out that there are point sources and complex surface chemistry. But developing a mission from scratch to fly in 2009 would be very, very hard, and I think that the greater amount of design work already done on Marvel would probably trump all other factors. So if there is to be a methane mission at all, it may well be a case of love the one you're with - not too hard, since Marvel is pretty damn loveable.

Formaldehyde again

As Rick Sterling has kindly pointed out in the comments, there's been movement on the Formisano front. Jenny Hogan at New Scientist reports that he is now claiming very large amounts of formaldehyde have been picked up by the Planetary Fourier Spectrometer that he runs on Mars Express. He thinks the formaldehyde is being made as methane is oxidised, and that the formaldehyde level he sees is equivalent to an annual production of 2.5m tonnes of methane. It's near impossible to see how that much methane could be made by any process other than life. This is presumably the stuff he was keeping quiet about on Ischia. He's going to be talking about these results at an ESA meeting this coming week.

I've posted on the formaldehyde story before. And, even more now than then, I think Formisano is making a mistake. As Jenny points out, so do a number (quite possibly, from what I hear, all) of his colleagues on the PFS, including those who have more experience modelling atmospheric chemistry and interpreting spectrometer data than Formisano has. I don’t want to rehash everything in the earlier post on the subject, but the gist is that a) formaldehyde is expected to have a very short lifetime in the atmosphere, and thus it is very hard to explain how there could be so much of it and b) earth-based telescopes have looked for the stuff and found no evidence for it even at levels far lower than those that Formisano appears to see.

In addition, I’ve spoken to someone who’s seen the data and has a quite good theory as to what might be up. While no one else thinks there’s any formaldehyde in the atmosphere, everyone agrees that there’s carbon dioxide -- the atmosphere is almost entirely composed of the stuff. And carbon dioxide has some complexities. Since there are two stable isotopes of carbon (12 and 13) and two stable isotopes of oxygen (16 and 18) there are in fact six different possible types of carbon dioxide molecule, all with very slightly different spectral features and present in the atmosphere at very different levels. In the area of the spectrum that Formisano is looking at, around 3.7 microns, you’d expect some little features due to minority isotopes, and it seems plausible that this explains what Formisano is seeing. I believe this has been pointed out by reviewers of a paper Formisano has written on the subject, which may be why it hasn't appeared anywhere yet.

Another relevant point here is that there are two different regions where formaldehyde signals might be looked for in the PFS data. The 3.7 micron region is where one would expect a weak signal; the 5.1 micron region is where one would expect a stronger one. But Formisano seems to have said nothing about a signal in the 5.7 micron region, even though, at the sort of formaldehyde levels he’s talking about, you’d expect something quite visible there.

I think Formisano has put together a very good instrument, and that he’s also a nice guy. But on the issue of formaldehyde, he seems mistaken.

I’ll try and post on the Carol Stoker and Larry Lemke story (mentioned in comments here) a bit later, but for the time being it seems to me that that comment thread sorts out the story pretty well. Take note of this, a comment on the matter by Carol, relayed to us by Penny Boston and Phil Plait, and provided in the comments here yesterday by Bruce.

And back from Ischia

I got to Naples about nine o'clock Tuesday night, took the train to Pozzuoli, and first thing in the morning made my way to the quay, admiring the catches of the octopus fishermen as I waited for the ferry over to Ischia. This was probably a fairly daft trip to make, but what the hell -- there are worse places to see the day in than the deck of a ship cutting across the Bay of Naples, Vesuvius behind you, Capri to port, and the smell of wood smoke on the air.

I caught up with Vittorio Formisano at lunchtime, and asked him what was going on. Here's the gist. His presentation was basically what is now up on the ESA website: the data from the Mars Express Planetary Fourier Spectrometer (PFS) seem to show three patches of enhanced methane and enhanced water vapour in the martian atmosphere, one of which is over Arabia. Formisano puts some weight on the idea that these three areas are also the areas where the Mars Odyssey sees hydrogen in the shallow subsurface; if the hydrogen is a layer of ice, he suggests, water vapour and methane from could be pent up beneath it and leaking out, or something along those lines. Having talked to Bill Boynton -- who’s the principal investigator on the Odyssey gamma ray spectrometer -- a bit in the afternoon, I found him not terribly supportive of that interpretation. For one thing, the regions Formisano is talking about aren’t a terribly good match to the regions that Boynton and his colleagues have mapped out. For another, Boynton tends to think that the hydrogen they’re seeing in those low-latitude regions is more likely to take the form of hydrated minerals than ice. One piece of evidence for this view is that the same regions have high chlorine levels, suggesting the rocks there have been weathered, which would be a way of getting hydrated minerals to form.

Formisano’s notion that methane, like NASA, follows the water is, as I understand it, a new analysis, one that probably goes beyond what's said in the paper on martian methane that he and various PFS colleagues have submitted to Science. That paper, various people have now told me, has already been through one round of revisions at the behest of the journal's referees, and the revised and resubmitted version has still not satisfied all of those referees. It has not yet been accepted.

The talk Formisano gave was quite different from the talk one might have expected from his abstract, which the Enterprise Mission has posted on its website. That abstract, he told me, is "all wrong" and "a mess". That's not to say it's not the abstract that's in the abstract book for the meeting -- it is -- but he says it at no point reflected what he really wanted to say. I suspect the big issue here is that the abstract talks about formaldehyde, and that Formisano at the moment doesn't want to talk about formaldehyde, because he has a paper on the subject under review at Nature. Another factor could be that the abstract talks about a weirdly high level of formaldehyde (much higher than the level of methane) and Formisano may no longer think that that level is real. (The general assumption is that formaldehyde is made from methane and much shorter lived, so you’d expect it to be present at lower levels; and observations form earth tend to bear out those expectations.) Given the way in which he’s not talking about it I’d be very surprised if Formisano doesn’t think he has a formaldehyde measurement of some sort or other, but he's adamantly saying nothing to the press or, apparently, to his colleagues on the subject for the time being. As always, it’s worth noting that the referees at Nature may not think the data strong enough for publication, and even if it is published other scientists may not buy the story. We’ll see.

For what it's worth, Formisano gave absolutely no indication, nor any impression, that he was under any sort of duress or pressure to change his presentation.

One other thing about the abstract is that it talks about hydrogen fluoride. Formisano didn't mention it in his presentation, but when we chatted it seemed to me he was saying that he thought he had indeed spotted some. If so, it appears that that would be pretty good evidence of current igneous activity. He wouldn't talk about levels, but he did say that there had been a misprint in the abstract, and we’re talking parts per billion not parts per million.

I asked Formisano about the ammonia story, and here's his version. He says they were asked by the Beagle team to look for ammonia, since it was the gas used to inflate Beagle's airbags. Ammonia seen in the atmosphere over Isidis very shortly after the Beagle loss would have been likely to have come from the airbags, and that would have been interesting to know. I remember Colin talking about the possibility of such a search in the days right after the loss. However, Formisano says they didn't do it. He also says that the relevant markings on the spectrum that was posted on the ESA website were there as indicators of where one might look for particular compounds, not as tentative identifications. That seems plausible to me. I don’t know if it’s the whole story, but it covers a fair bit of the ground.

After lunch (a very nice lunch indeed, shared with Friend Of This Blog Leo Enright) the highlight of the day's talks was another methane presentation, this one by Formisano's collaborator Sushil Atreya of the University of Michigan. The talk wasn't about the evidence for methane; it was about how a methane level of 10 parts per billion in the martian atmosphere could be explained. (Just to recap; ten parts per billion is the average level that the two earth-based teams which have reported observations of methane say they see, as well as the average level the PFS people are talking about. Much more here.)

The gist of the talk was that a steady state level of 10ppb requires a flux of 126 tonnes of methane a year (in the units used back here, that's about a thousand cows. I think Atreya uses a longer lifetime for methane than I did, and thus needs a smaller supply). That said, there's no requirement that Mars be in a steady state. Specifically, the insertion of a large amount of methane at some time in the past by a comet impact could explain a much lower level now -- we'd be seeing the 10ppb tail of something that had once been much higher and would eventually settle down to zero again. If my notes are right, Atreya suggested that a 200-metre-radius comet that hit 500 years ago would do it, assuming a 600 year lifetime for methane. Years of kibitzing on debates over the hazards posed by near-earth objects make me think that that sounds pretty unlikely. Most of the hazard on earth comes from asteroids, not comets, and a hit by a comet that size would be expected to be a pretty rare event. (There is a minority opinion in the NEO community which would say that comet chunks of this size are actually more common than the rest of us think -- Hi Duncan -- but let's let that slide for the time being).

A comet impact is the least interesting of the mechanisms for providing martian methane, but it's also nicely testable. If this is what happened then a) the methane levels are going to be homogeneous the whole planet round and b) they'll be dropping over time.

The other possible exogenous source -- meteoritic dust -- doesn't deliver remotely enough methane to make a difference, according to Atreya's calculations.

Since there's no way to make methane in the atmosphere, that leaves us with the subsurface. Atreya pointed out that both high temperature and low temperature reactions between water and basic rocks in the presence of carbon dioxide could produce methane. He cited work by others which suggested that a warm wet early Mars could conceivably have produced 10^15 tonnes of methane through aqueous alteration of basalt. If a small fraction of that had been squirreled away as methane clathrates, it could be enough to give the flux needed today. On the other hand, if there's subsurface igneous activity in the presence of water today, the stuff could be being generated fresh. On the gripping hand, it could be being made by methanogens, using either carbon monoxide and water or hydrogen and carbon dioxide.

What about the spatial variation in the methane? At first blush it looks almost impossible. Methane given off at a particular point on the martian surface would be spread through the whole atmosphere in weeks or at most months, after which it would stick around for centuries. This implies that almost all the methane in the martian atmosphere will be well mixed at any given time. But both Atreya and Formisano point out that it's possible methane has a much shorter lifetime than has previously been thought (something they were already hinting at back in Nice). The calculations that give lifetimes of a few centuries assume that the methane is destroyed only by chemicals in the atmosphere. If chemicals on the surface play a role too, then a) the life time could be a lot shorter and b) you could get more regionality if the chemicals to blame were concentrated in particular regions. It’s also worth bearing in mind, Atreya pointed out, that atmospheric chemicals are not in fact evenly spread out: he showed a recently made map of hydrogen peroxide in the martian atmosphere in which the concentrations varied quite a lot from place to place.

One particularly interesting thing Atreya said came almost as an aside. Without going into detail, because it's other people's ongoing research, he seemed to hint that there was a possibility that some abiogenic ways of making methane could also make formaldehyde. I'll try and find out a little more about this; if it’s true it may be important in the months to come. I suspect significant HF, some formaldehyde and a plausible story about an abiogenic source for that formaldehyde might strengthen a purely igneous story.

As I mentioned in the previous post, this all makes the prospect of seeing under the surface with Marsis amazingly tantalizing. But there's still no definitive word on the safety of the deployment. Apparently some simulations of the deployment have the long thin radar antennae springing out then springing back and hitting the dish of the high gain antenna -- the antenna through which data get sent back. That's really not something you want to happen. However other simulations show the antennae just springing out and staying put with grace and decorum. It seems no-one yet knows which models to trust. There's going to be a big meeting about this at JPL on October 5th, apparently. If deployment goes ahead, the likely date would be November 15th.

The politics seem to be rather as I'd imagined. The Marsis people are keen to be up and running as soon as possible, and very keen indeed to be up and running by next March, when their radar will be in a position to probe the foundations of the north polar ice cap. People working on other instruments are less gung-ho, and don't see why they should risk the whole mission this way. Why not wait a bit first, and deploy Marsis when the spacecraft has already taken a Mars-year’s worth of data. Among other things, that would make the case for an extended mission over a second Mars year a sure thing (though things would have to get pretty damn dire for ESA to close down a functioning spacecraft round Mars just because its primary mission had ended). But every day in orbit is another day that things might go wrong regardless of whether Marsis is deployed, and if the mission is lost to something else first we won’t see the deep subsurface until 2012 or so. Which option would you choose?

There were various other interesting things -- including the fact that MRO is a real spacecraft these days, with some of the instruments integrated to the bus and all. (And the interesting insight that although the O stands for Orbiter, in one crucial sense it’s a lander. In order to get the best data back -- a staggering 34 terabytes of the best data, according to current plans -- MRO is in a low orbit, so low that at some point after the mission is over it will fall to the surface. Because of this, in planetary protection terms it’s being treated as a lander and sterilised appropriately.) But having now hydrofoiled back over the bay (hydrofoils are neat, but they’re also fundamentally disappointing, because though the idea of flying over the water is cool, the sheer amount of work the engines do to make it possible kind of rips the poetry apart – flying’s meant to be easy); flown back over half of Europe; and struggled home from Gatwick, and what with it being unpleasantly late, I’m going to post and sleep.

Ischia

The Ischia news (which Alex Blackwell has been most kindly feeding into the comments threads) is definitely interesting stuff. At the Mars meeting on Ischia Vittorio Formisano has reported that methane and water vapour are concentrated over the same three parts of Mars -- Arabia, Elysium and Arcadia. The suggestion is obviously that in each place they have a common source. One possibility, presumably, might be igneous activity that both melts ice and releases methane. Another, as Mike Mumma points out on BBC Online, is that the source could be the breakdown of clathrates -- ice with methane already trapped within it.

If you buy into the clathrates, that still leaves the question of where the methane came from open. But here’s a possibility. Imagine that under these regions there are deep aquifers with liquid water, of the sort that Steve Clifford and Tim Parker have written about. Maybe they’re the last remnants of a planet-spanning aquifer. And imagine that in these aquifers there are some jolly little methanogens. This means that in the pore spaces above the aquifers there will be both water vapour and methane. The vapour and the methane will rise up until they reach the cold rocks of the shallow subsurface, at which point they freeze, together, into clathrates. Now imagine that this is happening a little while back, in one of the ice ages when the planet’s obliquity was higher and its lower latitudes colder. Back then, clathrates just below the surface might well have been stable; today, with the lower obliquity and the warmer low latitudes, they might be breaking down.

It’s just a scenario. I’ve no idea if it’s feasible (I don’t know if you can make clathrates out of water vapour and methane in this way: on the earth you start with liquid water). And you can imagine something pretty similar going on with an abiogenic methane source. But it’s kind of fun.

Five thoughts. One is that we have to remind ourselves that this data still hasn’t been published, hasn’t been peer reviewed and hasn’t been made available to the scrutiny of other workers in the field. Doesn’t mean it’s wrong, but it’s worth bearing in mind. (Given that they were writing this up in April, there’s definitely been time for it to be submitted). Leonard David quotes Jim Garvin, NASA’s head of Mars science, saying that widespread inspection of the data will be “the real test”.

Two is that there’s no mention in the reports that I’ve seen of the purported temporal changes in methane concentration mentioned back in Nice. That said, the chart on the ESA page shows a level of 35 parts per billion, which is high. Maybe the temporal differences were actually spatial differences.

Three is that the ESA release also says there is a paper on other trace gases (ammonia? Formaldehyde?) currently in the works.

Four is that all this makes it more important than ever that the Marsis antenna gets deployed. Seeing what the subsurface looks like in these three areas has suddenly become on of the highest priorities.

Five is that I think I’m going to go and get a plane ticket to Naples and hear some more.

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.

Or rather, no ammonia

This weekend I heard something that made me considerably more dubious about the ammonia story, and which fits pretty well with the article by Stephen Strauss in the Globe and Mail that Alex Blackwell (Hi Alex) linked to in the comments for the previous post.  Strauss reports that ESA denies categorically that PFS has detected ammonia, to the point that the agency's spokesman, Guido De Marchi, said bluntly "it's a hoax". The BBC told Strauss that they stood by their story.

So at this point, it's not entirely clear what if anything Vittorio Formisano, the principle investigator on the Mars Express Planetary Fourier Spectrometer, said or meant to say to David Whitehouse, the BBC reporter. It's certainly possible that he may have said something ESA didn't want him to have said. After all, while Mr Strauss reports the ESA spokesman as saying "The instrument aboard the Mars Express has not looked for ammonia", in an earlier interview with Lynda Moulton Howe, published  on her website a couple of months ago, Formisano is clearly quoted talking about benzene, formaldehyde and ammonia as targets being searched for. So either Formisano was misquoted in that interview, or his answers were misleading, or the ESA denial over-reaches. It may be, for example, that different factions in the PFS data-analysis team are reaching different conclusions, and ESA officialdom is on the conservative side. And if so they may well be right to be there; as mentioned in the last post, ammonia is a really tough identification for a system like PFS.

In general, that web interview seems consistent with Whitehouse's story, giving the impression that Formisano might have thought he was finding something but wouldn't talk about it as yet. The interview also gives the impression that he's more willing to speculate freely on life as a source of trace gases than I have found him, though I wouldn't claim he's ever opened up to me that fully. (Some readers may want to take on board the fact that Lynda Moulton Howe has a track record in writing about crop circles and cattle mutilation. On this matter I'll report and you can decide.)

If Formisano were a source for David's story, one might imagine he'd said something ESA, and quite possibly also his PFS colleagues, didn't want him to say, and the agency is now engaged in what it sees as damage control. But looking again at David's story, it's not clear that Formisano was a source (and according to Strauss, ESA says he wasn't). Indeed it's not clear who any of the sources are -- they're all anonymous "researchers"; the only one quoted directly is just a "US space agency (Nasa) scientist".

The only clearly discernible basis for the story is Formisano's abstract (pdf here) for this week's Cospar meeting in Paris, which says "The high spectral resolution [of the PFS] allows us also to identify a number of small signatures which possibly will bring us to the identification of minor compounds (at the moment a good candidate is ammonia)." David didn't specifically cite or quote that abstract in his story, but the BBC pointed to it when Strauss at the Globe and Mail asked them if they stood by the story. Again, that makes ESA's "never looked for ammonia" hard to believe, and Strauss quotes ESA backing down a little, saying there had been plans to look that weren't followed through on.

At the same time, it's a pretty tenuous basis for the BBC story; that abstract doesn't sound to me like a "detection", even a "tentative detection", nor does it sound like "evidence for the gas was seen by". You'd really have to hope that David -- who I know on casual but friendly terms -- was working with more than that abstract and some responses from other researchers not directly in the know. Surely a story like this needs comment from one or more of the PFS workers (especially if there's dissent in the team). Strauss, at the Globe and Mail, says he got no response when he repeatedly called David to ask him about this. Maybe the comments section here will be luckier.

In short, it doesn't look like we should expect anything ammoniac at Cospar this week. Indeed, David's comment on the previous post suggests as much. A pity, since as Rick Sterling pointed out in another comment, David's story if anything underplayed the significance of ammonia. According to an email from Mark Allen of JPL that's up on the web here "At present there are no known abiotic processes that would result in ammonia being present in the atmosphere without the existence of life." Allen knows a lot about this, as he is the principle investigator on the proposed Marvel mission to measure trace-gas abundances, a mission I talked about a little in this post on next steps for methane detection. (And in the dog-that-didn't-bark category, I wonder why Formisano's abstract says nothing about methane, since that's obviously the big PFS story so far; maybe the abstract was submitted a very long time ago. Intelligence from anyone who's actually at Cospar most welcome)

Given all this, I think I'll avoid talking about the tangled history of the possibility of formaldehyde, which Formisano also mentioned in that web interview, for the time being. A pity, since it means passing up some more perfectly good dead shark jokes.

Correction: When I first posted this entry I mistakenly typed "methane" for "ammonia" at a crucial point, as Alex Blackwell notes in his comment below. To avoid confusion I fixed the mistake.

Ammonia

At BBC Online David Whitehouse is reporting that the Mars Express PFS is going to report the detection of ammonia in the martian atmosphere at the COSPAR meeting next week. Obviously very exciting if the observations stand up; ammonia would be very unstable in the martian atmosphere and so would have to be constantly replenished, or if only present intermittently produced in contemporary fits and starts. It could clearly come from a biological source -- if Mars life is remotely like earth life it needs a nitrogen cycle. And the abiogenic sources of ammonia are probably scantier than those for methane: no serpentinisation, and no clathrates, I think (though ammonia can be mixed into water ice). So ammonia could be a strong signal of life.

But there are reasons to be a good bit more cautious about this detection than the methane detection (about which one should also be moderately cautious, though I must say I haven't been). The most important difference is that in the case of the methane, the news was particularly exciting because three different teams using a number of different instruments all seemed to see compatible results; here there is only one team reporting anything. Like the methane results, this ammonia result hasn't been published yet, and other scientists elsewhere haven't even seen the spectra at conferences yet. Indeed, as far as I know, the PFS team hasn't published results on anything as yet, and so there's no way for outside experts to assess the fine details of the ways in which they're analysing their data. As David notes in his piece, "[members of the PFS team] are still coming to terms with the complexities of the PFS as well as coping with some nagging power problems on Mars Express."

By a slightly freakish coincidence, I read the news while sitting next to Michael Mumma of NASA Goddard, one of the original methane observers (and thus, in some senses, a PFS competitor, though also a fellow striver after truth). Mumma -- who tried to detect ammonia in the martian atmosphere with earth-based instruments a few years ago, to no avail -- was clearly intrigued, but pointed out that if the ammonia detection was at ten microns (which the BBC report doesn't quite say, but which would be quite likely) then it would be very close to a quite striking carbon dioxide feature, and that if you were using a comparatively low resolution spectrometer such as PFS you'd want to be very careful indeed to make sure that the two bands weren't getting mixed up. Make of that what you will.

So if confirmed -- or simply published and scrutinised and found highly plausible -- this is extremely hot stuff: right now a very interesting thing to keep an eye on.

Incidentally, here in Iceland there is a delicacy called hakari made by burying sharks and leaving them to rot. Apparently it smells quite strongly of ammonia. Since Mars's ocean, if it has one, is undergound, it stands to reason that its sharks come ready-buried, and will thus be a potent ammonia source. Yet another way in which Mars is just Iceland writ large...