Context. The term "solar tornadoes" has been used to describe apparently rotating magnetic structures above the solar limb, as seen in high resolution images and movies from the Atmospheric Imaging Assembly (AIA) aboard the Solar Dynamics Observatory (SDO). These often form part of the larger magnetic structure of a prominence, however the links between them remain unclear. Here we present plasma diagnostics on a tornado-like structure and its surroundings, seen above the limb by the Extreme-ultraviolet Imaging Spectrometer (EIS) aboard the Hinode satellite. Aims. We aim to extend our view of the velocity patterns seen in tornado-like structures with EIS to a wider range of temperatures and to use density diagnostics, non-thermal line widths, and differential emission measures to provide insight into the physical characteristics of the plasma. Methods. Using Gaussian fitting to fit and de-blend the spectral lines seen by EIS, we calculated line-of-sight velocities and nonthermal line widths. Along with information from the CHIANTI database, we used line intensity ratios to calculate electron densities at each pixel. Using a regularised inversion code we also calculated the differential emission measure (DEM) at different locations in the prominence. Results. The split Doppler-shift pattern is found to be visible down to a temperature of around log T = 6.0. At temperatures lower than this, the pattern is unclear in this data set. We obtain an electron density of log n e = 8.5 when looking towards the centre of the tornado structure at a plasma temperature of log T = 6.2, as compared to the surroundings of the tornado structure where we find log n e to be nearer 9. Non-thermal line widths show broader profiles at the tornado location when compared to the surrounding corona. We discuss the differential emission measure in both the tornado and the prominence body, which suggests that there is more contribution in the tornado at temperatures below log T = 6.0 than in the prominence.

We investigate the properties of a 'solar tornado' observed on 15 July 2014, and aim to link the behaviour of the plasma to the internal magnetic field structure of the associated prominence. We made multi-wavelength observations with high spatial resolution and high cadence using SDO/AIA, the IRIS spectrograph and the Hinode/SOT instrument. Along with spectropolarimetry provided by the THEMIS telescope we have coverage of both optically thick emission lines and magnetic field information. AIA reveals that the two legs of the prominence are strongly absorbing structures which look like they are rotating, or oscillating in the plane of the sky. The two prominence legs, which are both very bright in Ca II (SOT), are not visible in the IRIS Mg II slit-jaw images. This is explained by the large optical thickness of the structures in Mg II which leads to reversed profiles, and hence to lower integrated intensities at these locations than in the surroundings. Using lines formed at temperatures lower than 1 MK, we measure relatively low Doppler shifts on the order of ± 10 km s −1 in the tornado-like structure. Between the two legs we see loops in Mg II, with material flowing from one leg to the other, as well as counterstreaming. It is difficult to interpret our data as showing two rotating, vertical structures which are unrelated to the loops. This kind of 'tornado' scenario does not fit with our observations. The magnetic field in the two legs of the prominence is found to be preferentially horizontal. arXiv:1512.04727v1 [astro-ph.SR] 15 Dec 2015 Figure 1. Filament observed at 13:37 UT on July 11 2014, in H-α with the Meudon survey instrument.

Spectropolarimetric observations of prominences have been obtained with the THEMIS telescope during four years of coordinated campaigns. Our aim is now to understand the conditions of the cool plasma and magnetism in 'atypical' prominences, namely when the measured inclination of the magnetic field departs, to some extent, from the predominantly horizontal field found in 'typical' prominences. What is the role of the magnetic field in these prominence types? Are plasma dynamics more important in these cases than the magnetic support?We focus our study on three types of 'atypical' prominences (tornadoes, bubbles and jet-like prominence eruptions) that have all been observed by THEMIS in the He I D 3 line, from which the Stokes parameters can be derived. The magnetic field strength, inclination and azimuth in each pixel are obtained by using the Principal Component Analysis inversion method on a model of single scattering in the presence of the Hanle effect. The magnetic field in tornadoes is found to be more or less horizontal, whereas for the eruptive prominence it is mostly vertical. We estimate a tendency towards higher values of magnetic field strength inside the bubbles than outside in the surrounding prominence. In all of the models in our database, only one magnetic field orientation is considered for each pixel. While sufficient for most of the main prominence body, this assumption appears to be oversimplified in atypical prominence structures. We should consider these observations as the result of superposition of multiple magnetic fields, possibly even with a turbulent field component.

Context. High resolution movies in 193 Å from the Atmospheric Imaging Assembly (AIA) on the Solar Dynamic Observatory (SDO) show apparent rotation in the leg of a prominence observed during a coordinated campaign. Such structures are commonly referred to as tornadoes. Time-distance intensity diagrams of the AIA data show the existence of oscillations suggesting that the structure is rotating.Aims. The aim of this paper is to understand if the cool plasma at chromospheric temperatures inside the tornado is rotating around its central axis.Methods. The tornado was also observed in Hα with a cadence of 30 seconds by the MSDP spectrograph, operating at the Solar Tower in Meudon. The MSDP provides sequences of simultaneous spectra in a 2D field of view from which a cube of Doppler velocity maps is retrieved. Results. The Hα Doppler maps show a pattern with alternatively blueshifted and redshifted areas of 5 to 10 wide. Over time the blueshifted areas become redshifted and vice versa, with a quasi-periodicity of 40 to 60 minutes. Weaker amplitude oscillations with periods of 4 to 6 minutes are superimposed onto these large period oscillations. Conclusions. The Doppler pattern observed in Hα cannot be interpreted as rotation of the cool plasma inside the tornado. The Hα velocity observations give strong constraints on the possible interpretations of the AIA tornado.