Highest Energy Cosmics Rays are Iron Nuclei!

March 3, 2010 at 5:44 pm 14 comments

The composition of the highest energy cosmic rays, those with energies on the order of tens of EeV (= 106 TeV), appears to include a significant number of iron nuclei – a result that I find truly astonishing. I learned about this from a nice post at Diax’s Rake.

The results come from the Pierre Auger Collaboration, and were reported recently in a paper (arXiv:1002.0699):

Measurement of the Depth of Maximum of Extensive Air Showers above 1018 eV

Pierre Auger detector elements - photo

Detector elements for the Pierre Auger observatory

Their detector has two components: a surface detector which gives timing information to measure the direction of the shower, and the florescence detector, which observes the longitudinal development. The longitudinal shower shape is key to the analysis, since the maximum of the shower energy increases logarithmically with energy (as is well known to particle physics experimenters) and decreases logarithmically with the number of nucleons, A, in the incident nucleus. The authors quote:

Xmax = α( lnE – lnA ) + β

with α and β known to be largely energy-independent at these energies. Furthermore, the fluctuations in Xmax decrease with A. Given enough showers in a certain energy interval, the distribution of Xmax relates to the composition of the incident flux.

The authors apply what sound to me like stringent fiducial and quality cuts, and arrive at a sample of 3754 events. This is not a small sample – they certainly will have a good error on the mean and rms of Xmax for several bins of incident energy. Here are their measurements:

Xmax vs. E from Pierre Auger

Mean Xmax versus E (log scale). The lines represent different air shower simulations using different hadronic interaction models.

RMS Xmax vs. E from Pierre Auger

RMS(Xmax) vs. E (log scale).

The data clearly drift from proton-like composition to iron-like composition passing from 1 EeV to about 40 EeV. (1 EeV = 109 GeV.) Although there is some spread in the predictions of various simulations, the slope of the data is like none of them, especially for RMS(Xmax). The transition occurs across a bit more than a decade in energy, and it is unlikely that the physics of air showers changes in such a strong way over only one decade. Nominally, cosmic rays with energies of 1 EeV are mostly protons, while cosmic rays with energies 50 times larger are mostly “iron” nuclei – in some average sense. All this supports the hypothesis of a transition from galactic to extra-galactic origins for cosmic rays in the energy regime. But how do such iron nuclei get produced and accelerated to such high energies, and why are they so much more numerous than protons at that energy?


Entry filed under: Astronomy.

A beautiful paper by CLEO-c: J/psi → γX Advertisement: Celebrating Woman’s Day at CERN and Fermilab

14 Comments Add your own

  • […] of Maximum of Extensive Air Showers above 1018 eV,” ArXiv, 3 Feb 2010. Visto gracias a “Highest Energy Cosmics Rays are Iron Nuclei!,” Collider Blog, 3 Mar. […]

  • 2. carlbrannen  |  March 4, 2010 at 6:29 pm

    It’s kind of suspicious when they use only a single measured parameter to distinguish between iron and protons. It’s a case of “enough parameters to fit any data”. What would happen if the curve continued to drop below the iron nuclei level? I like your version with “iron like”.

  • 3. Michael Schmitt  |  March 4, 2010 at 7:17 pm

    Hi Carl,

    yes I agree that the analysis is not sophisticated. I don’t imagine that the set of events at highest energies are all Fe nuclei, but some distribution in A with an average of 56 or so. It is a pity that they did not publish the distribution of Xmax – it would have been interesting to see the shape compared to the expectations for protons and for Fe nuclei, just to give us a feel for the data.


    • 4. Charles Jui  |  April 29, 2010 at 12:30 pm

      Beware of drawing the iron conclusion.
      Assuming a simple 2 component iron + proton model:
      If one starts with all protons, and start adding iron, the RMS in Xmax should increase at first because of the offset between the two means. At 50% proton and 50% iron the RMS would be about 50% larger than pure proton. One does not get to 1/2 the proton RMS until the distribution is nearly all iron. So the two plots above are inconsistent: the mean suggests a gradual transition starting from mostly protons at 3×10^18 eV to about 50% iron at about 3×10^19 eV. The RMS plot would indicate almost 100% iron by 10^19 eV.

  • […] composition” of cosmic rays is an interesting topic – see an earlier post about extra-galactic iron nuclei detected by the Pierre Auger […]

  • 6. Charles Jui  |  April 29, 2010 at 12:36 pm


    Have a look at the HiRes result published in PRL
    which indicates essentially pure protonic composition,
    which is consistent with the observation of the GZK cut-off

  • 7. carlbrannen  |  April 30, 2010 at 5:52 pm

    Here’s the PRL letter on arXiv: http://arxiv.org/abs/0910.4184

  • 8. Charles Jui  |  July 6, 2010 at 2:44 pm

    At the regular Friday afternoon seminar at FermiLab on July 2, 2010, there was much criticism of the AUGER analysis method. Under questioning, presenter Eun-Joo Ahn admitted that when the AUGER collaboration performed a composition analysis using a method similar to the HiRes composition study, they found a result consistent with proton dominance at the highest energies.

  • 9. Eun-Joo Ahn  |  July 23, 2010 at 8:01 am

    There is a misunderstanding in the above comment by Charles Jui on July 6 2010 which should be clarified. The statement is incorrect and was not made by the presenter during or after the seminar.

  • 10. Charles Jui  |  July 25, 2010 at 4:31 pm

    Eun-Joo Ahn was the presenter who made the statement and senior members of the TA collaboration and of FNAL scientific staff were present at the seminar and can verify this admission on her part.

  • 11. chicago medical billing  |  August 25, 2010 at 3:59 pm

    nice article i like it… WordPress is amazing…thanks for sharing.

  • 12. gribrorse  |  May 21, 2011 at 10:20 am

    Many thanks for another excellent write-up. Where else can any person get that form of information in composed in such a crystal clear approach. I’ve a demonstration next week, and I had been searching for precisely the advice you have just presented me. Thanks a lot.
    http://dieta2011.jkr30.com/ – mein blog

  • 13. stanlyklark  |  August 2, 2011 at 12:41 pm

    Это должен прочитать каждый родитель!.

  • 14. Neural Outlet..  |  February 2, 2012 at 5:40 am

    Am I right in thinking it took a little over a year to move from tests to a strong hypothesis? I find myself googling the subject after reading:

    Astrophysics has some phenomenal stuff to it!


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed

March 2010

%d bloggers like this: