Solar Magnetized Tornadoes: Rotational Motion in a Tornado-like Prominence

From a paper describing "Solar Magnetized Tornadoes" one can expect a cartoon with curlicues. Here they represent the winding-up of prominence material, both hot and cold, by vortices at the photospheric level. The cartoon suggests that the twist may therefore build up until the prominence erupts, which sounds familiar. But these prominences apparently rotate physically, so does that mean that the twist is being reduced without eruption? We note that different ∇ x B conditions at the two footpoints of a coronal flux tube would disagree on how twisted it should be, and so one would expect torsional Alfvén waves to flow back and forth to mitigate this conundrum. That's a common conundrum for many cartoons that follow the Parker problem in having loops anchored in the photosphere - that is, pretty much all of them.