Skip to content

What color is beryllium-10?

October 8, 2018

This posting relates to some planned upgrades and additions to the ICE-D:ANTARCTICA website and the related parts of the online exposure-age calculators that have to do with diagrams, figures, and images. Even now, the display of some data sets via this website can produce a somewhat bewildering array of diagrams, figures, and images that are supposed to present exposure-age data in some way. Examples include the neat-looking but largely unexplained and unintelligible front page of the website:

Age-elevation plots:

Carbon-14 saturation plots:

And, in future, possibly extremely complex data-model comparison plots associated with this project. To make this proliferation of plots a little less intimidating, it seemed like a good time for myself and BGC postdoc Perry Spector, who is responsible for the data-model comparison project, to at the very least come up with a standardized color scheme for plotting measurements of different cosmogenic nuclides together on the same images. Hence the need to determine what color beryllium-10 is. And all the other nuclide/mineral pairs, which for practical purposes for this application includes He-3 in a variety of mineral targets, Be-10 in quartz, C-14 in quartz, Ne-21 in quartz, Al-26 in quartz, and Cl-36 in many possible targets. So how to do this? Here are some general principles.

  1. In reality, most of these things are either colorless (noble gases) or grayish (metals), with the exception that chlorine is a yellow-green gas if found in immediately-fatal concentrations at normal P-T conditions. So that’s not too helpful.
  2. There should be some kind of sense to the color-nuclide pairing, e.g., colors in spectral order should pertain to nuclides in mass order, or something like that.
  3. Obviously, neon-21 has to be neon  (the preceding text, which is probably unreadable, is and says “neon”). This was a critical new insight stemming from our discussion. How was this not obvious before?
  4. The colors shouldn’t look terrible (except for neon-21 as noted above), and should reproduce OK for print and screen use.
  5. Color assignments should be kind to green/red colorblind viewers to the extent possible, which may be simplest to accommodate by using one, but not both, of those colors.
  6. Black probably shouldn’t be used, because it will conflict with other plot elements, grids, etc.

So here’s our attempt to do this:



This scheme does a pretty good job of overlapping with the goals above, as follows, with a couple of additional features:

  1. Helium-3 data are the same tone (gray) but different values (light and dark) for different minerals. This is helpful in de-emphasizing He-3-in-quartz data, which are often uninterpretable due to diffusive loss of helium. If helium-3-in-quartz is one of the nuclide/mineral pairs being plotted, you probably want to pay more attention to whatever the other ones are.
  2. Neon-21 is neon.
  3. No green. No black.
  4. The colors are more or less in spectral order as you go up in mass.
  5. It’s not terribly ugly, with the exception of neon-21.

GMT is being used to make all these plots, so here are the HSV codes for these colors for use in GMT scripts:


Nuclide/mineral Face color Outline color
He-3 (qtz) 0-0-.78 0-0-.28
He-3 (px/ol) 0-0-.6 0-0-.1
Be-10 (qtz) 0-.7-.97 0-.7-.47
C-14 (qtz) 36-.85-1 36-.85-.5
Ne-21 (qtz) 65-.75-1 65-.75-.5
Al-26 (qtz) 211-.72-.90 211-.72-.4
Cl-36 (all) 291-.48-.65 291-.48-.15


Thus, expect to see this color scheme in anything generated by the ICE-D web server and online exposure age calculators in future.

One thing this does leave unaddressed is two-nuclide plots:

Obviously, you know which two nuclides are being plotted by reading the axis labels, so it is not necessary to color-code the nuclides. Instead, what is being color-coded here is the scaling method used to normalize the data to unit production rate. Red is non-time-dependent “St” scaling (Lal, 1991 and Stone, 2000); green is the “Lm” time-dependent adaptation of Lal/Stone scaling from Balco et al. (2008), and blue is the time-dependent “LSDn” scaling developed by Nat Lifton and others. This particular plot shows data from Antarctic samples with very high nuclide concentrations, so the tiny red and green ellipses out to the right in the “forbidden zone” highlight the difference in performance between St/Lm and LSDn scaling for high-elevation Antarctic sites. For these plots the R-G-B order is just inherited from the 2008 online calculators and represents the increasing order of complexity of the scaling methods, but it has the disadvantage of including both red and green. So this may have to be fixed in the near future as well.

Summary: the color scheme is a bit uglier but now standardized and hopefully more sensible.

2 Comments leave one →
  1. October 9, 2018 17:13

    How about RGB color palettes for Matlab?

  2. Greg Balco permalink*
    October 9, 2018 17:25

    MATLAB has an hsv2rgb function. The only difference is that in GMT the H goes from 0 to 360 degrees (it’s the “color wheel”, remember), and in MATLAB it goes from 0 to 1. So, for example, the GMT HSV code for neon yellow/green is 65-.75-1. In MATLAB,

    hsv2rgb([65/360 0.75 1])

    will return the RGB triple.

Leave a Reply

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

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

Google photo

You are commenting using your Google 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

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: