The Japan Aerospace Exploration Agency (JAXA) has requested that Science retract a 2006 paper about the makeup of asteroid Itokawa as observed from the spacecraft Hayabusa, the news section of Science reports.
Instead of calibrating their equipment on Earth, the scientists assumed they’d see both magnesium and silicon in the x-ray spectra, and used that assumption to assess the rest of the chemical composition of the asteroid.
The paper may be based on faulty assumptions, but the conclusions have been backed up by other published papers, according to the Science magazine report:
The Japanese-led team published a collection of seven papers in a special issue of Science on 2 June 2006 based on observations by four instruments as the Hayabusa spacecraft circled asteroid Itokawa in the fall of 2005. The craft later touched down to grab samples. The paper being retracted, by Tatsuaki Okada and colleagues, presents an analysis of x-ray spectra to determine the elements on the asteroid’s surface. The authors concluded “that Itokawa has a composition consistent with that of ordinary chondrites.” Chondrites are a type of stony asteroid.
For various reasons, the authors felt they could not rely on the calibration of the instrument done on Earth before the spacecraft was launched. To compensate, they started by assuming they would see the characteristic x-ray spectra of magnesium and silicon, elements known to be present on ordinary chondrites. They then used what they took to be the spectra of those elements to interpret the instrument’s raw data. In effect, the authors jumped to a conclusion and then based their analysis on what they expected to observe. (Explanatory materials, in Japanese, are here.)
The paper has been cited 54 times, according to Thomson Scientific’s Web of Knowledge.
Hat tip: Rolf Degen.
I find this decision preposterous. I don’t understand the need to retract the paper. Isn’t this simply a natural evolution of techniques that become disproved over time? Is this not a case of a hypothesis that becomes invalidated over time? I am really concerned that the basic investigative concept of science is being confused with incorrect conclusions drawn and the aggressive need to retract anything now. In my opinion, this would simply require an erratum, or a supplementary letter that would induce further discussion, and expand scientific discourse. Science is in a constant state of evolution. This does not mean that old science is necessarily invalid just because the conclusions become disproved by new theories or studies. After all, it is stated that “the conclusions have been backed up by other published papers”. Moreover, if this forms part of a 7-paper series, why is only one of the papers being “targeted”. I know some JAXA researchers and I think there is alot of rigor that I have seen that I sometimes have not observed in some NASA scientists, for example. Space research is so “out there”. Surely, this retraction is unnecessary (sorry to disappoint the retractophobes)?
The retractophiles?
A quick look at the table of contents strongly suggests that it’s the only one based on the XRS data.
My apologies for double-posting. I can’t find a published calibration paper for the instrument, only for the near-infrared spectrometer. It’s difficult to try to go through the relevant material using G—le Translate, but it really just looks to be a miscalibration of the instrument, maybe initially suspected on the basis of another in the same design family (take that with a wheelbarrow full of salt). But the only hypothesis to be invalidated that I can see relates to how to turn the raw output from the device into meaningful data. They may be able to rereduce it; I can’t tell whether that’s in the offing.
I think a paper of authors that belong to JAXA is retracted for their careless mistake.I was disappointed.
But They retracted their paper not for scientific misconduct but for careless mistake.
There are some of misconducter in Japan, for example H.Obokata, A.O,Y.Fujii.
But Authors of this paper should seldom be blamed.
I don’t think they are being blamed. But I do think they were correct in retracting the paper once the fault in the sensor modeling was detected.
It looks as though this (PDF) paper in Earth, Planets and Space may have to go, as well.
Sorry about the broken link. The correct one is this (PDF). Upon closer examination, it may be fine if the sulfur-channel data are OK.
^ That was sloppy. The XRS elemental calibration obviously doesn’t matter if the onboard sample itself isn’t being used. In other words, I consider my first impression to have been wrong and the EPS paper to be unrelated. Apologies.
As the author of the original report on the Science website, I would like to correct one slight misunderstanding here. Cat Ferguson writes: “Instead of calibrating their equipment on Earth, the scientists assumed they’d see both magnesium and silicon in the x-ray spectra…” Actually, as I explained, the instrument was calibrated on Earth. For reasons that are not entirely clear, the authors of the paper concluded that they could not rely on the results of that Earth-based calibration once the instrument was in the vicinity of asteroid Itokawa. They then sought other benchmarks against which to interpret the raw data from the instrument. Previous ground-based observations suggested that Itokawa was a certain type of asteroid that would have magnesium and silicon present on the surface. The authors then assumed they should see magnesium and silicon in the x-ray spectra and based their analysis on that assumption. Their assumption turned out to be a correct guess, but their methodology was flawed and so they are retracting the paper.
Dennis Normile,
Japan correspondent
Science magazine
Dennis, as the Japan correspondent for Science magazine, would you care to provide some public commentary about the state of science in Japan, why many high-profile retractions are not covered by the media in the way in which the Obokata case was, and what reforms do you know are in place that make Japanese research institutes, and their researchers, to be held more accountable not only to critics from Japan, but also from abroad. I am quite sure that in your position, you should have such in-deoth information and analysis. One of the greatest problems we are having about Japan is not only the language, but the lack of whistle-blowers in Japan coming forward about problems in the Japanese scientific literature.
I’m not so sure about this. It appears that in situ calibration was part of the plan from the outset, precisely because the solar X-ray irradiation, which is the source exciting the fluorescence, isn’t constant.
It appears that a problem was found with the Kaguya XRS that it was then suggested that the Hayabusa data be looked at. The latter did not have the same problem, but they did discover a problem with the extraction of the peaks. It seems to have to do with misidentifying the Mg channel and then using a smoothed spectrum, but again, G—le Translate.
Sorry about the failure to close the link after the word ‘constant’.