Graphs of spatial 3D data

These figures show some results of 3D simulations of a collapsing gas cloud. The rows are different simulations, and the columns are different length scales, zooming in towards the right.

The figure below shows gas density projections, meaning you’re essentially looking at whole columns of gas. For such figures that require colormaps, often a ‘default’ rainbow colormap is used. However, there are a number of problems with the rainbow colormap: it is confusing to interpret, it can obscuring features in the data, and it can even be misleading, indicating features where there aren’t any. This is discussed at length for example here and here. Therefore, I have used the cubehelix colormap, which ensures a continuous increase in perceived intensity, the same way a grayscale colormap from black to white does, but with added color hues to make it easier to identify specific values. The fun thing about cubehelix is that it’s not just a single colormap, but a whole family where you can easily adjust how many and which hues you want, how bright they should be, and more, to tailor it to your needs. Examples here, play around with it here (uses javascript).



The figure below shows slices through the simulation box, centered on the densest object and looking at the x-y plane. The first and third column show the density field, overlaid with streamlines that indicate the velocity in the plane (so you can see gas moving towards the central object), where the thickness of the line is a measure for how fast the gas is moving. The second and fourth column show the gas temperature, using a different colormap more intuitive for heat maps.