A recent paper by Andrew Kirk and a group at CCFE Observation of lobes near the X-point in resonant magnetic perturbation experiments on MAST may not sound all that promising, but it contains what has to be one of the most amazing photos in tokamak physics for a long time.

Resonant Magnetic Perturbations (RMPs) are non-axisymmetric magnetic fields which affect the plasma edge, and are being studied as a means to control eruptions called ELMs. For years field-line following codes have predicted that RMPs should produce intricate structures (with the fancy title of "homoclinic tangles") with lobes, shown below:

For just as long, most researchers have thought that these could never exist in a real plasma. Surely if you tried to do this to a plasma, it would respond to reduce the effect, and then diffusion would smear these things out...right?

Wrong!

The images in this paper from the Mega-Amp Spherical Tokamak, taken using a fast camera capable of taking hundreds of thousands of images per second, shows what looks suspiciously like the predicted lobe structures

The plasma response does appear to reduce the size of the lobes, but doesn't remove them. Furtive pre-publication viewings of these images at a workshop last year were greeted by something like "Holy **** they're actually real!"

The implications of this are still sinking in, but these lobes could have a big impact on the power handling in tokamaks: If one of these lobes hits a part of the vessel, it will produce a large, localised hot spot. This is fine in current machines, but wouldn't be great news in a large fusion reactor.

Exciting times, and it's nice to know that the models aren't completely wrong after all!