The PASTEL catalogue (Soubiran et al. 2010, Soubiran et al. 2016) lists stellar atmospheric parameters calculated from spectroscopic measurements, in particular the effective temperature T_eff, surface gravity log(g), and metallicity [Fe/H]. Lower surface gravities correspond to giants (they're much larger for a given mass, so the surface is much further away from the centre of gravity), and higher values to dwarfs. A metallicity of zero corresponds to a star with the same iron abundance as the sun.

I don't have a ready reference to explain the plot – I'm learning by asking friends who did astro for variables to try out, and then looking at the results. Looking at [Fe/H] versus log(g), there's a noticeable fall in the density of points at around log(g) = 3.5, separating the giants from everything else (the 2016 paper linked above puts it at 3.2 rather than 3.5). There's also a population of metal-poor giants, and a few very hot dwarf stars(?). Most of the stars fall close to a straight line in the log-log plot of surface gravity against effective temperature.

Mouse controls: Left-click and drag to rotate; alt (Mac)- or ctrl (Windows)-click-drag or middle-click-drag to pan; scroll or shift-click-drag to zoom. Touch screen controls: one finger to rotate; two-finger scroll to pan; pinch to zoom. (If the points don't appear, try giving it a quick rotate and hopefully it'll fix itself.)

You can see how the plot is constructed in the HTML source.

• The default_point_height is set to 1, for pretty pixels.
• The x_scale_bounds and y_scale_bounds are manually set so that they can be reversed, to follow with astro convention.
• The default tick format didn't work for log(T_eff)! I set it to "0.1f" to get the decimal place.
• I don't know why, but when I ask for 1-pixel-tall points, the plot doesn't render properly when it first loads. After the call to three_d.make_scatter(), I run window.setTimeout(function () { return three_d.update_render(0); }, 500); to try updating the render after 500 milliseconds.

The PASTEL catalogue can be downloaded from here. It's in fixed-width format, and the R script I used to turn the relevant fields into JavaScript (after deleting the non-data rows) is here, which sources my main JS-writing function write_three_d.R.

Posted 2016-10-03.