Scintillations (i.e., "twinkling") of pulsars has a particular radio-frequency and time dependence. The discovery of parabolic arcs by Dan Stinebring et al. (2001) in the doubly-Fourier-transformed spectrum was a watershed event in this field. One of the implications was that the radio-wave scattering responsible for scintillations was concentrated in a series of thin screens along the line of site, as opposed to being more broadly distributed. Observations of Brisken et al. (2010) showed that at least in some cases the scattering is strongly anizotropic. In the context of these considerations, Ue-Li Pen and I proposed a model which involved corrugated edge-on reconnection sheets as scattering screens. The overall idea was not new - Goldreich & Sridhar (2006) already suggested that edge-on sheets could be important for scattering of the radio source SgrA* (physically the latter is due to the accretion disc/jet(?) around the supermassive black hole in the Galactic Center). Moreover, in quite an old paper Romani, Blandford, and Cordes (1978), already considered edge-on planar structures as a possible source of scintillations. What was new in our work was the (somewhat trivial) realization that the corrugations of the reconnection sheet naturally produce scattering centers and caustics, and that no major density perturbation along the sheet was required.
Pulsar scintillations from corrugated reconnection sheets in the interstellar medium