Kevin Jardine's blog

As I write this, the scientific equivalent of a sold-out U2 rock concert is taking place at the ESTEC complex in Noordwijk, Netherlands. Some 400 scientists are gathered to hear the results of the first year of operation of the Herschel infrared space telescope. I blogged about Herschel and Planck after the two space telescopes were successfully launched last year (The king is dead - long live the king).

Because of the huge scientific interest in the conference (which is discussing results under strict media embargo), attendance had to be limited. I've been following Twitter updates from two scientists attending the meeting (Dave Clements and Brian O'Halloran) to get at least a glimmer of what is going on.

Unfortunately both scientists are attending the extragalactic sessions, so information about the Milky Way results is limited. A news conference is scheduled for later today and I hope more information will be available afterwards.

Yesterday there was a revealing Twitter exchange between the two scientists (as it happens Clements was sitting immediately in front of O'Halloran in the meeting room, but hey this is Twitter and the exchange was as much for their "followers" as themselves).

(Read the exchange from the bottom up.)

The exchange was specifically about the plan for a long delay in the public release of the Planck microwave data, but it might equally apply to the Herschel results. For example, the images collected in the Hi-GAL survey of the inner galactic plane are not scheduled for release until two years after the telescope was launched.

Clements suggests a reason for the delay, which is that the scientists immediately involved in the telescope projects are hoarding the data until they have written their papers.

Given the intense competition between astronomers, data hoarding is perhaps not surprising, but is it true that "everyone" does this as Clements states in his tweet? I'm not a professional astronomer, but it appears to me that NASA has a different policy for some of its space telescopes. In particular, Spitzer infrared data appears to be available as soon as it is acquired and archived. (Correction: Sarah Kendrew informs me that this depends upon the project and at least some of the Spitzer projects have a 12 month proprietary period.)

The fact that scientists involved with major telescope projects have a competitive advantage is not new. US scientists dominated astronomy for much of the twentieth century in part because the best telescopes in the world were located in California for most of the century. Even Jan Oort's famous paper on the rotation of the Milky Way was based largely on data begged or borrowed from US scientists.

The Internet has changed the dynamic, however, because now scientific data is routinely archived online. It is easier to share the data than not. Scientists involved with major projects need to go out of their way to password protect their data if they want it to remain private.

As the infrared astronomer Sarah Kendrew pointed out in a recent blog post (Setting free the Data), a huge number of scientific papers are now published by scientists using public data archives. In my view, this trend is a good thing and scientific data should be made available as soon as possible so that scientists without immediate access to the instruments (like Jan Oort) have the opportunity to make major discoveries.

In 2003, the Observatoire Astronomique de Strasbourg sent out a dramatic press release, Astronomers find nearest galaxy to the Milky Way. They announced a major new discovery, the Canis Major dwarf galaxy, a satellite galaxy in the process of being torn up and absorbed by the Milky Way.

Statements from the press release, as well as content taken from follow-up news stories and the related scientific papers, made their way into major Internet information sites, including Wikipedia, SEDS, David Darling and the Astronomy Picture of the Day.

The difficulty is, however, that a dwarf galaxy is only one possible explanation for the overdensity of M-class giant stars seen towards Canis Major, observed in the 2MASS data and discussed in the original 2004 paper.

The dwarf galaxy hypothesis caused a rapid response and rebuttal. First off the mark was a 2004 letter to Astronomy and Astrophysics by a group of Italian astronomers, attributing the overdensity to more local Milky Way structure (the outer galactic warp).

More detailed objections soon followed. I posted a list of 8 recent papers rejecting the dwarf galaxy hypothesis (or at least concluding that other explanations were more credible) to this Wikipedia discussion page. I think that it's fair to say that a Canis Major dwarf galaxy is currently best described as an interesting hypothesis rather than a confirmed object.

If so, then why do so many Internet-based astronomy information sources misrepresent the dwarf galaxy hypothesis as a confirmed object? Scientists often shake their heads sadly at the mass media distortion and misrepresentation of their work. In this case, however, I can only reluctantly conclude that the distortion and misrepresentation appears to have originated in the 2003 Observatoire Astronomique de Strasbourg press release, which claimed a level of certainty for their conclusions which just did not exist.

I draw three main conclusions from this sad story (and similar ones, for there are others):

• Never believe press releases, or news stories based upon them, no matter how apparently credible the source.
• Be very cautious in believing scientific papers trumpeting major new discoveries, at least at first.
• A year or two after the publication of a paper, take a look at the list of citations conveniently provided by the Astrophysics Data System. This will give you a good idea of how the paper has been received by peers of the scientists publishing the original paper. In the case of the "dwarf galaxy" paper, the list of citations contains 195 references as of today!

I think that these conclusions might be useful for scientists as well as the general public. It is interesting to see that some of the earlier papers citing the dwarf galaxy paper also seem to assume that it is a confirmed object, and it is only more recently that it appears to be considered safer to mention the controversy and describe the dwarf galaxy as a hypothesis.

I've added the Lynds bright nebulae as an optional overlay to the false colour version of Douglas Finkbeiner's all sky hydrogen-alpha map in the Milky Way Explorer.

At some point soon, I'll add some commentary on the Lynds bright nebulae catalog. In short, it's a large catalog (1125 objects) covering the same area of sky as the BFS and Sharpless catalogs but in more detail. This sounds more interesting than it actually is, because in many cases the Lynds nebulae are just parts of nebulae already in the Sharpless catalog. However, it also contains designations for fainter and smaller nebulae not in the Sharpless catalog, so is useful in some cases.

NASA released a new galactic centre image on 10 November that was supposedly constructed from three previous galactic centre images from Hubble, Spitzer and Chandra. I've now put it into the Milky Way Explorer and you can view it here.

Some of the mysteries around a few of the objects in this image have been discussed on the APOD discussion site. At least one of these objects does not appear on any of the three original images so it is a mystery where it came from.

Gregory Dobler, Douglas Finkbeiner and colleagues have released a set of all-sky gamma ray images based upon the Fermi Large Area Telescope data release back in August. I've put a colour composite of three of these images in the Milky Way Explorer.

Normally I work with the full FITS format data released by astronomers, but in this case, the FITS files use the fairly obscure HEALPix projection which unfortunately is not yet supported by a number of key software tools. For example, Aladin hangs when I try to read these FITS files. So for the moment, I've used the black and white jpeg images also supplied to create the colour composite.

I experimented with a number of data combinations, and ended up using the smoothed versions (with no point sources removed) for these energy ranges: 0.5 to 1 GeV (red), 2 to 5 GeV (green) and 10 to 20 GeV (blue). Higher energy levels up to 300 GeV are available, but the higher energy gamma rays are not particularly constrained to the galactic plane. According to my reading of the accompanying paper, it appears that these higher energy photons are mostly from relatively local objects (within a few thousand parsecs).

The Fermi data is not very high resolution, but nevertheless, it is interesting to see major gamma ray concentrations in the directions of Orion, the Taurus and Ophiuchus dark cloud complexes, and the Large Magellanic Cloud.