99. Preflare and flare turbulence in the transition region

Authors: Natasha Jeffrey, Lyndsay Fletcher* and Nicolas Labrosse at the University of Glasgow and Paulo Simões at MacKenzie Presbyterian University, São Paulo.

*Also at Rosseland Centre for Solar Physics, University of Oslo

<< previous nugget — next nugget >>

Introduction

From the optically thin lines emitted by hot plasmas it is possible to obtain properties such as density, temperature and flow speed. If a spectral line is formed in thermodynamic equilibrium in a plasma with a Maxwell-Boltzmann distribution of speeds then the line width can also be used to deduce the temperature of the emitting ions. However, in some cases, the line width is observed to be significantly larger than expected for a plasma with temperature consistent with the ions present [1]. This broadening is generally called “non-thermal broadening”. It could have different causes, but in this nugget we present observations made before and during a solar flare, in which the non-thermal broadening is consistent with turbulent motions in the emitting plasma, which decrease as the flare intensity increases. This is evidence for transfer of energy from plasma turbulence to heating – i.e. the kinetic energy of particles in the transition region. Evidence of turbulent energy transfer has already been seen during a flare in a coronal source [2] but here we present observations of broadening in the lower atmosphere, which clearly precede the flare brightening.

IRIS observations of SOL2016-12-06T10:36:58

The IRIS spacecraft observed a B-class flare at 1.7s high cadence, in sit-and-stare mode. This is the highest cadence yet for such observations. Figure 1 shows the location of the IRIS spectrometer slit on the background of the IRIS 1400 Å slit-jaw image. The white contours indicate the SDO/AIA 131 Å footpoints and the pink contours are the RHESSI 6-12 keV source. No higher energy RHESSI emission could be imaged.

Figure 1. LH image: IRIS slit position (green) on AIA 131 Å background, showing AIA footpoints (white contours) and RHESSI source (pink contour). RH image: close-up of the slit and flare sources on IRIS 1400 Å slit-jaw image.

The Si IV line at 1402.77Å was fitted with a Gaussian profile to obtain the line intensity, line-of-sight speed and line broadening. These values are interpreted under the assumption that Si IV, emitted at temperatures close to 80,000 K, is optically thin. Modelling suggests that this might not not be the case in some flares [3] but this event was weak enough that the optically thin assumption holds.

Time-dependent line broadening and flows

The results of the spectral fitting are shown in Fig. 2 below. In the upper and lower panels, the grey shaded area shows the intensity in Si IV, integrated over 2″ along the slit. About 2 minutes of data are analysed, and the strong flare brightening lasts less than a minute at this wavelength. Superposed in the top panel is the FWHM, or non-thermal velocity (RH axis), as well as the RHESSI 6-12 keV flux. On a timescale of around 10 seconds, the line broadening increases strongly, and does so well before the flare brightening starts. As the Si IV intensity increases, signifying atmospheric heating, the non-thermal width decreases. The non-thermal width then has another 2 peaks, separated by about 11s, before decaying back to pre-flare values.

Figure 2. Upper panel: time series of Si IV line width, shown also as a non-thermal speed (right axis), superposed on the Si IV intensity and RHESSI light curve. Lower panel: As above but showing the Si IV blueshift.
Figure 3. Upper panel: curves representing the velocity of plasma elements moving with 10 waves of different amplitude and frequency that travel through the emission region (colour = time). Middle panel: amplitude as a function of time that would be measured over a narrow range of heights. Lower panel: Resulting line broadening and line shift.

If we interpret the non-thermal broadening as due to turbulence in the transition region then the implication of this observation is significant. The timing suggests that turbulence in the region starts well before the flare heating, rather than occurring as a result of heating. Instead, the heating only happens after the turbulent broadening peaks and as it decays. This is a strong indication of cause and effect – energy is transferred from the turbulence to the particles in the transition region plasma.

The lower panel of Figure 2, plotting the Si IV line centroid, shows that the peak in the turbulence occurs at the same time as a strong blueshift. Normally, blueshifts would be associated with chromospheric evaporation caused by flare heating, but here the blueshift occurs before the flare heating indicated by Si IV intensity. So we can rule out chromospheric evaporation as a cause. However, we have found in simple modelling that the patterns of line broadening and blueshift together can be explained by traveling waves.

Using a simple model of motion of fluid roughly perpendicular to the field direction, caused by traveling Alfvénic waves, we have shown that both the line broadening and the line shift can be approximately reproduced by a superposition of a small number (around 10) of traveling waves, with an amplitude and phase varying with time, and a wavelength comparable to the thickness of the Si IV emitting region. Figure [3] shows the plasma velocity as a function of position, due to the modelled waves, with colour representing time (top panel). Then the resulting velocity amplitude as a function of time at a randomly chosen cut is shown in the middle panel, and the line broadening and line shifts that this would generate at the bottom.

The basic properties of the velocity observations are reproduced, i.e. an increase in the line broadening followed by a small number of oscillations, with the line centroid simultaneously blueshifted. A spectrum of waves entering the transition region would be very likely to reflect, interact and lead to a turbulent cascade ending in heating. How these waves are generated is, however, another question.

Conclusions

Very high cadence IRIS observations have allowed us to detect and characterise the evolution of non-thermal line broadening, interpreted as a signature of turbulence, in the Si IV transition region line before and during a small flare. The relative timing of the onset of line broadening and of flare heating indicated by the rise in the Si IV intensity shows clearly that the heating does not cause the turbulence; if anything it is the other way around. A possible interpretation is that the flare launches magnetic disturbances from the corona towards the solar chromosphere, carrying energy that ultimately dissipates in the form of ion heating as the waves travel through the transition region, setting up a turbulent cascade. This may provide a significant mode of energy transport in flares. The full results of this study are described in [4].

References