Gravitational lensing is one of nature’s more useful gifts to astronomers. At a very simple level (the kind I can reach given what time I got up this morning) the path of light is bent by the presence of mass, as best seen in galaxy clusters like this one.  The arcs are distant galaxies whose image has been distorted.

Galaxy Cluster Abell 1689 as seen by the Hubble Space Telescope.

Of course, this is an extreme example, but by studying the shear - a small distortion of the shape of many galaxies - it’s possible to reconstruct the distribution of the mass by which the light has passed. In recent years, this has been done for large chunks of the Universe, producing a rather blobby map of the distribution of dark matter.  

To improve on these results, Rachel McInnes of the University of Edinburgh reckons we need to take advantage of new surveys like Pan-Starrs (more on which later in the week). To do that and include many many more galaxies in the analysis she needs to automate lots of the tasks currently carried out by humans, such as cleaning up the images to avoid confusing the program; work’s underway, and the techniques are being tested on another, smaller survey while we build up to Pan-Starrs at the end of the year.