Observations and Modelling of the Pre-flare Period of the 29 March 2014 X1 Flare

Abstract

On 29 March 2014, NOAA Active Region (AR) 12017 produced an X1 flare that was simultaneously observed by an unprecedented number of observatories. We have investigated the pre-flare period of this flare from 14:00 UT until 19:00 UT using joint observations made by the Interface Region Imaging Spectrometer (IRIS) and the Hinode Extreme Ultraviolet Imaging Spectrometer (EIS). Spectral lines providing coverage of the solar atmosphere from the chromosphere to the corona were analysed to investigate pre-flare activity within the AR. The results of the investigation have revealed evidence of strongly blue-shifted plasma flows, with velocities up to $200\text{km}{\text{s}}^{-1}$ , being observed 40 minutes prior to flaring. These flows are located along the filament present in the active region and are both spatially discrete and transient. In order to constrain the possible explanations for this activity, we undertake non-potential magnetic field modelling of the active region. This modelling indicates the existence of a weakly twisted flux rope along the polarity inversion line in the region where a filament and the strong pre-flare flows are observed. We then discuss how these observations relate to the current models of flare triggering. We conclude that the most likely drivers of the observed activity are internal reconnection in the flux rope, early onset of the flare reconnection, or tether-cutting reconnection along the filament.

Keywords: Flares; Magnetic fields; Models; Pre-flare phenomena.

Figure 1

GOES light curve of the soft X-ray flux from 29 March 2014 14:00 UT. Joint Hinode/EIS and IRIS coverage starts at 14:09 UT. The flare at 14:30 UT occurs outside the spectrometer field of view. At 16:24 UT a C1.1 flare was observed by both spectrometers, followed by the X1 flare at the peak time of 17:48 UT.

Figure 2

Panel (a) shows the fields of view of the IRIS (blue) and Hinode/EIS (red) spectrometers overlaid onto the active region seen in the 193 Å AIA channel, presented with an inverted colour table. All images have been differentially rotated to the time of the closest AIA 193 Å exposure at the time of the central Hinode/EIS raster step. Panel (b) shows the three regions of study and their positions within the Hinode/EIS Fe xii field of view. The position of the filament is clearly seen in the 193 Å and 304 Å AIA channels, with panel (c) showing the 193 Å data. The path of the filament is marked in panel (d) by the purple line for emphasis.

Figure 3

Evolution of the central portion of AR 12017 between 26 March 2014 and 29 March 2014. Flux emergence is observed in panel (b), and subsequent panels show the formation of a highly sheared polarity inversion line. Plots are scaled between $\pm 550\text{Gauss}$. Panel (e) also indicates the AIA field of view used in Figure 2, marked by the red shaded area, as well as the three regions of study.

Figure 4

Evolution of mean intensity (red) and mean $V_{\mathrm{nt}}$ (blue) calculated for Fe xii in time for Regions A, B, and C, as shown in Figure 2. In Region A, shown in panel (a), the two profiles broadly match, in particular showing a clear peak during the 16:24 UT C1.1 flare. The location of $V_{\mathrm{nt}}$ enhancement during the C-flare is shown in Figure 5 panel (a). In Region B, panel (b), we can see that at 17:00 UT, non-thermal velocity peaks at a value of $\approx 70\text{km}{\text{s}}^{-1}$. This non-thermal velocity peak occurs in the absence of a corresponding intensity increase. The location of this region of $V_{\mathrm{nt}}$ enhancement is shown in Figure 5 panel (b). Region C, panel (c): both intensity and non-thermal velocity profiles match well, with no irregularities as in the case of Region B. A smaller response to the C1.1 flare is observed in this region. Prior to 17:00 UT, increases in intensity and non-thermal velocity are observed. Region C also exhibits the earliest response to the X flare of the three regions studied. Figure 5 panel (c) shows the location of non-thermal velocity enhancements.

Figure 5

This figure shows non-thermal velocity contours overlaid onto corresponding AIA 193 Å images to exhibit the location of the non-thermal velocity enhancements. $V_{\mathrm{nt}}$ contours are plotted between 70 and $200\text{km}{\text{s}}^{-1}$, and the AIA colour scale is inverted to improve clarity. Panel (a) shows the location of these enhancements during the C-flare at 16:24 UT. Panel (b) shows that the $V_{\mathrm{nt}}$ enhancements seen at 17:00 UT are located in a region in the centre of the filament. Enhancements seen at the start on the X-flare, at 17:35 UT, are shown in panel (c).

Figure 6

Spectral time profiles recorded in Region A at 16:16 UT. Panel (a) shows He ii spectra, panel (b) shows Si iv spectra, and panel (c) shows the evolution of the Mg ii k line during this time period. Broadening of the lines is observed at 16:24 UT in response to the C-flare. The response to the X-class flare can be seen in all lines from $\approx 17\text{:}35\text{UT}$.

Figure 7

Spectral time profiles recorded in Region B at 16:16 UT. Panel (a) shows He ii spectra, panel (b) shows Si iv spectra, and panel (c) shows the evolution of the Mg ii k line during this time period. Little activity is observed until the onset of strong blue shifts is seen from $\approx 16\text{:}52\text{UT}$. The onset of the flare can be seen to occur in this region from 16:40 UT.

Figure 8

Spectral time profiles recorded in Region C at 16:16 UT. Panel (a) shows He ii spectra, panel (b) shows Si iv spectra, and panel (c) shows the evolution of the Mg ii k line during this time period. A small response to the 16:24 UT C-flare is observed in this region. From 16:45 UT until the onset of the X-flare, blue-shifted line broadening is observed.

Figure 9

Stills detailing coronal activity observed by SDO/AIA in 193 Å, shown using an inverted colour table, between 16:42 UT and 17:04 UT. Panel (a) shows the filament prior to any observed activity. Overlaid are the three sub-regions of study. Panel (b) shows the appearance of a bright feature to the west of the filament. Panel (c) shows a loop feature crossing the filament at the site of the peak coronal blue shift as observed by Hinode/EIS. This activity is marked by the black arrow in the diagram. Panel (d) shows an extended bright feature lying along the southern edge of filament. The path of this brightening is marked by the overlaid black arrows.

Figure 10

Here we see the same AIA 193 Å (colour table inverted) field of view as Figure 9, over-plotted with Hinode/EIS Fe xii data at $-100\text{km}{\text{s}}^{-1}$. No strong blue shifts can be seen in the EIS data in panel (a) at 16:42 UT. By 16:55 UT (panel (b)), we can see that there is a strong area of blue shift located over a bright region in the AIA data. There is also a less intense region of blue shift located in the centre of the filament. At 17:00 UT (panel (c)), both these blue-shifted regions have increased in intensity. Panel (b) shows the situation at 17:04 UT, where the strong blue shifts have dissipated greatly with the feature in the centre of the filament being no longer visible.

Figure 11

For the same region detailed in Figures 8 and 9, this figure displays Si iv emission at $-100\text{km}{\text{s}}^{-1}$ observed by IRIS, overlaid onto AIA 193 Å data (colour table inverted). In panel (a), at 16:42 UT, we see a blue-shifted region centred on the region of the earlier C-class flare. In panel (b), 16:54 UT, blue shifts in the region of the C-flare have waned in intensity. In the region of brightening in the AIA data there is a region of strong blue shift, as well as discrete areas of blue shift extending along the filament. At 17:00 UT, panel (c), these discrete blue shift features are located along the extended bright feature visible in AIA data. By 17:04 UT, panel (d), blue shifts have decreased in intensity, but are still present along the extended bright feature.

Figure 12

(a) Graph of the variation of total flux (solid line), positive flux (dashed line), and absolute value of negative flux (dotted line) over the time period of the simulation. (b) Initial potential field configuration used in the simulation corresponding to a start time of 16:30:31 UT on 27 March 2014.

Figure 13

Connectivity of the field lines along the location of the filament at 09:30:31 UT on 28 March 2014, (a) and (b), at 00:00:31 UT on 29 March 2014, (c) and (d), and at 16:30:31 UT on 29 March 2014, (e), (f), and (g). Panels (a) to (f) show the field lines from above superimposed on the magnetogram (white represents positive flux, and black represents negative flux), while panel (g) shows the field lines from panel (f) viewed from the side. Finally, panels (a), (c), and (e) do not include the additional helicity injection term, while panels (b), (d), (f), and (g) include additional helicity injection at a rate of $3.75\text{–}5\times {10}^{16}{\text{Mx}}^2{\text{cm}}^{-2}{\text{s}}^{-1}$.

Figure 14

Comparison of the position of blue shifts observed in Hinode/EIS Fe xii data at $-100\text{km}{\text{s}}^{-1}$ with flare ribbons identified in IRIS slit-jaw imager 1400 Å channel (an inverted colour table has been used). Panel (a) shows the position of the blue shifts at the time they occurred. These areas of blue shift are aligned with brightenings also visible in AIA 193 Å data. Panel (b) shows the blue shifts superimposed onto SJI data during flaring. This image is chosen as it is the closest image to the flare peak that is not saturated. Flare ribbons are visible and appear to lie close to the brightenings seen in earlier SJI and AIA 193 Å data. Panel (c) shows the situation post flare. We can clearly see the flare ribbons have expanded outwards from their initial positions.

Figure 15

Evolution of positive flux with time over-plotted on the GOES light curve. We can see that over the period of joint EIS/IRIS observation, there is sustained flux emergence into the region of study. This emergence continues until 16:10 UT, highlighted by the dotted line. The nadir of this drop is coincident with the C-class flare at 16:24 UT. The flux then stays constant until 16:46 UT, where a second smaller drop in flux is observed (dashed line). This drop in magnetic flux is broadly coincident with the onset of the flows described in Section 3.2. This decrease in flux recovers quickly to its previous level. Again the flux remains roughly constant until 17:30 UT, where a large increase in the flux is observed prior to the steep drop coincident with the X-flare at 17:35 UT.