EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 698 (May 2025) A&A, 698 (2025) A301 Full HTML
Open Access
Issue
A&A
Volume 698, May 2025
Article Number A301
Number of page(s) 14
Section The Sun and the Heliosphere
DOI https://doi.org/10.1051/0004-6361/202554011
Published online 20 June 2025

© ESO 2025

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. Introduction

Solar eruptions are intense events, releasing a great amount of energy (1032 ergs) stored in active regions (ARs). Photospheric motions in ARs lead to the creation of twisted flux ropes, which subsequently rise and later erupt due to instability (Schmieder et al. 2015). Theoretical models and observations indicate that the footpoints of flux ropes are surrounded by hook-shaped or J-shaped current ribbons (Démoulin et al. 1996, 1997; Chandra et al. 2009; Pariat & Démoulin 2012; Janvier et al. 2013, 2014; Dudík et al. 2014a, 2019; Aulanier & Dudík 2019; Joshi et al. 2024; Thoen Faber et al. 2025). These current ribbons correspond to the intersections of quasi-separatrix layers (QSLs) at the photospheric boundary (Zuccarello et al. 2017). The classical 2D CSHKP solar flare model is based on the reconnection of the magnetic field lines under the erupting flux rope, leading to the growth of the flux rope as well as the creation of the closed flare loops, associated with heating in the chromosphere in two ribbons, where these loops are rooted. The growing flux rope is expelled as a coronal mass ejection (CME). In this classical picture, the reconnection happens in an X-type geometry, where bilateral outflows can be produced (Priest 2014). Such outflows have been observed in exploding events or brightening during eruptions (e.g., Innes et al. 1997; Cheng et al. 2015; Ruan et al. 2019; Joshi et al. 2022; Chen et al. 2024).

This 2D standard model has been extended to the 3D (Aulanier et al. 2010, 2012; Aulanier & Dudík 2019; Janvier et al. 2013). Aulanier & Dudík (2019) found that in 3D two additional reconnection geometries exist, both involving the erupting flux rope, R. These two additional reconnection geometries are: (1) RR–RF, whereby two flux-rope field lines (R-R) reconnect into another flux-rope field line (R) and a flare loop (F); and (2) AR–RF, whereby a coronal arcade (A) and a flux rope field line (R) reconnect and turn into a flux rope (R) and a flare loop (F)). More specifically, in the RR–RF reconnection, the magnetic field lines of the flux rope reconnect between them, in a manner similar to the reconnection of the overlying arcades in the 2D model, leading to a new highly twisted flux rope field line, and a flare loop. In addition to that, the erupting magnetic flux rope can reconnect with the surrounding coronal arcades in an AR–RF reconnection geometry. This latter geometry is purely 3D (Aulanier & Dudík 2019) and is connected to the drifting of flux rope footpoints. The drift of the legs is connected to the evolution of the J-shaped flare ribbons. The inner edges of the ribbon hooks sweep outward and erode the flux rope from the inside. Meanwhile, the flux rope is built from the outside as the outer parts of the J-shaped ribbons sweep additional coronal arcades, which reconnect to become new flux rope field lines. As the simulation of Aulanier & Dudík (2019) is generic, the AR–RF reconnection geometry is likely quite widespread in solar flares.

The existence of AR–RF reconnection geometry has already been shown by Aulanier & Dudík (2019), who describe the J-shaped hooks of flare ribbons moving across the solar surface, first expanding and then contracting as predicted. Zemanová et al. (2019) report drifting of flux rope footpoints via slipping motion of S-shaped loops associated with elongation and strong drift, by more than 40″, of the J-shaped ribbon. The new ribbon hook then again expands and later contracts. Dudík et al. (2019) observe all four constituents of both the AR–RF as well as RR–RF reconnections, as well as the associated expansion and contraction of the ribbon hooks. Lörinčík et al. (2021a) find that the flare loops resulting from AR–RF reconnections, located at both ends of the flare arcade, are higher and more inclined, creating the characteristic saddle-shaped solar flare arcades. Lörinčík et al. (2021b) find that the AR–RF reconnections modify the extent of the coronal dimming areas inside the J-shaped ribbon hooks.

More recently, Guo et al. (2023) quantified the rotation of the flux rope during its eruption and claimed that the drift of one footpoint of the flux rope is due to reconnections of types AA–RF and AR–RF. The former injects magnetic fluxes and twisted field lines, while the latter causes the displacements (continuous drift) of the flux rope footpoints. This effect can in some events be quite significant. For example, Gou et al. (2023) showed that the evolution of the J-shaped ribbon hooks can be so strong that the entire magnetic flux inside the erupting flux rope has been replaced during the course of an eruption. These observations indicate that the AR–RF reconnections are likely widespread and significant for the evolution of the erupting flux rope as predicted by Aulanier & Dudík (2019). However, all of the observations so far have relied on imaging datasets, which allow for the evolution of individual emitting structures to be tracked, but do not provide further information about the emitting plasma (with the exception of temperature). In particular, spectroscopic observations of magnetic reconnection in this geometry have so far been missing.

Here, we report on the imaging and spectroscopic observations of a small B-class flare accompanied by a filament eruption that occurred on December 9, 2020, in NOAA AR 12791. The Interface Region Imaging Spectrograph (IRIS; De Pontieu et al. 2014) captured very well the spectra of the reconnecting filament with eight slit positions.

2. Observations

Here, we summarize the imaging and spectroscopic observations used to analyze the AR–RF reconnection in the B-class flare.

2.1. SDO

We used the observations from two instruments: the Atmospheric Imaging Assembly (AIA; Lemen et al. 2012), and Helioseismic and Magnetic Imager (HMI; Scherrer et al. 2012), both instruments aboard the Solar Dynamics Observatory (SDO; Pesnell et al. 2012). AIA observes the full disk Sun and is useful for studying the jet evolution in lower chromospheric or coronal temperatures. The AR was recorded with all EUV (94, 131, 171, 193, 211, 304, and 335 Å) and UV (1600 and 1700 Å) channels of AIA with cadence of 12 s and 24 s, respectively. The pixel size of AIA is 0.″6 and the spatial resolution is 1 . ″ $ {{\overset{\prime\prime}{.}}} $5. HMI observes the photospheric magnetic field with a cadence of 45 s and its pixel size is 0 . ″ $ {{\overset{\prime\prime}{.}}} $5. We used these magnetograms for the AR evolution.

2.2. IRIS

The AR 112791 was observed in coordination with the IRIS spectrograph, which provides slit-jaw images (SJIs) and spectras between 00:24 UT and 01:16 UT. IRIS acquires spectra in three different ranges: far UV in C II at 1334.54 Å and 1335.72 Å (FUV1), far UV in Si IV at 1393.76 Å and 1402.77 Å (FUV2), and near UV in Mg II at 2796.4 Å and 2803.5 Å (NUV). FUV1 and FUV2 are formed in the upper chromosphere and transition region, respectively. NUV mainly consists of the chromospheric Mg II h and k lines in addition to Mg II triplet lines formed in the upper photosphere and lower chromosphere. For this flare event, IRIS provides a large coarse eight-step raster with SJI in C II with only a cadence of 10 s and field of view (FOV) of 119″ × 119″, and also provides the spectra in C II 1330 Å, Si IV 1402.77 Å, and Mg II h and k lines with a step size of 2″ and a step cadence of 9.8 s, scanning a region of 14″ × 119″. The SJIs cover an extended region (119″ + 2 × 14″), and thus a rectangle of 148″ × 119″. We generated dust-free IRIS maps to clearly observe the evolution in IRIS SJIs.

2.3. Data coalignment

The level 1 data of the full-Sun AIA instrument traditionally has an excellent pointing, although small offsets can still be present in several channels, especially the longer-wavelength ones. To correct for these misalignments, as well as to study the evolution of specific structures, we first corrected the data for differential rotation, choosing the time of 01:00 UT as the reference one. Then, we found a residual misalignment of −0.9″ in the solar Y coordinate in the 211 Å and 335 Å channels with respect to the other coronal channels (94 Å, 131 Å, 171 Å, and 193 Å), which do not show mutual misalignment, and whose pointing was thus assumed to be correct.

The 304 Å channel was found to have a misalignment of ΔX = −1.2″, but no misalignment in Y. The 1600 Å and 1700 Å channels were found to have misalignments of ΔY = −0.6″, but no misalignment in the X coordinate. All of these misalignments were corrected manually and we note that their uncertainties are typically 0.3–0.6″ as larger shifts can readily be noticed in visual (blink) examinations of morphologically similar channels, even though the resolution of the instrument is larger, 1.5″. The HMI longitudinal magnetogram was aligned to the 1600 Å channel of AIA, with no offsets being found. The resulting alignment is shown in panel (b) of Fig. 1.

thumbnail Fig. 1.

GOES B3.2 class flare that peaked at 00:53 UT, and that has a small flare peak at 00:33 as a B1.7 class on December 9, 2020 (panel a). Panel b shows the flare ribbons in the UV temperature overplotted with a magnetic field of strength of ±200 G (green for positive and blue for negative polarity) from the HMI magnetogram. Panels c–g show the evolution of the flare region in the Hα wavelength, where the filament is shown with white arrows. Panel (h) shows the magnetic field configuration at the flare site and neighborhood.

Finally, the IRIS/SJI data were also manually aligned to SDO/AIA. To do that, we used the 1600 Å channel of AIA, which shows many similarities to IRIS/SJI, such as plage polarities as well as flare ribbons. The IRIS data were found to have an extremely good pointing, with only a shift of –1.2″ being necessary in Solar Y. We note that this is only a coordinate translation; that is, the observed IRIS spectra have not been modified. The location of the IRIS slit with respect to AIA observations was determined using the Sobel edge-detection operator on the co-aligned SJI images. This works well for most SJI images, but failed in a few cases in which N–S oriented (slit-aligned) structures were present, such as continuous bright flare ribbons or large segments thereof. In these cases, the slit position was determined manually.

3. Evolution of the eruption

3.1. Context

The eruption occurred on December 9, 2020, in NOAA AR 12791 located at S15 W11. The GOES instrument registered two accompanying peaks of X-ray emission, at B1.7 at 00:33 UT and at B3.2 at 00:53 UT. Both occurred during the course of the filament eruption (Fig. 1). We believe that the small peak (B1.7) marks the onset of the B3.2-class flare, potentially indicating two distinct episodes of magnetic reconnection. The host AR 12791 is a decaying one, with a bipolar configuration consisting of only plage polarities. The following positive polarities are already quite dispersed and resemble a rather intense chromospheric network (Fig. 1 panels a–e). A small sunspot was still visible in the HMI continuum two days earlier in the leading negative polarities. A filament is detected in the GONG images along the inversion line from 23:00 UT on December 8 onward (Fig. 1).

In the coronal AIA channels, the AR shows a generally bipolar character, with a plethora of closed coronal loops as well as open fan loops at both eastern and western peripheries (see panel b1 of Fig. 3 as well as Fig. A.1.) The filament eruption leads to significant evolution (reconnection) with the closed loops in the northern portion of the AR (see Sect. 3.4).

3.2. Evolution of flare ribbons

The filament eruption is accompanied by two bright ribbons near the polarity inversion line. The development of the brightest parts of the ribbons near the filament is well captured by IRIS/SJI (Fig. 2). At 00:24 UT, an intense ribbon brightening appears, then extends along the inversion line at 00:33 UT (panels a–b). The erupting filament, parts of which are in emission, is located above this brightening (Fig. 2b). At this time, it is not easy to distinguish both ribbons and the filament in the SJI image. Later, at 00:50 UT, as the ribbons have moved away from PIL, they can be more readily distinguished. Based on their respective polarities, the ribbons are denoted as PR (positive ribbon) and NR (negative ribbon), with PRH and NRH standing for their respective J-shaped hooks, which encircle the footpoints of the erupting flux rope (see Aulanier et al. 2012; Janvier et al. 2014, 2015; Aulanier & Dudík 2019). The PR and its PRH are located in the eastern (left) part of the AR, with the PRH being quite elongated into the outlying dispersed positive polarities (see Figure 3, panels a1 and c3–c5). The NR and its much more compact NRH are located in the western (right) part of the AR. Therefore, the ribbons are quiteasymmetrical.

thumbnail Fig. 2.

Evolution of flare ribbons and the filament eruption as observed in IRIS C II SJIs is depicted. Panels (b) and (c) highlight the erupting filament, with white arrows indicating the filament’s increasing height over time. Panel (d) presents a clear view of the two ribbons forming in the flare region. In panel (e), the labeling of different ribbon regions is provided, as is described in Sect. 3.2. Finally, panel (f) illustrates the falling-back material during the later stages of the event.

thumbnail Fig. 3.

Overview of the B-class flares and eruption of December 9, 2020, as observed by SDO in three AIA filters at 94 Å, 171 Å, and 304 Å. Note that the Fe XVIII contribution has been separated from the 94 Å filtergrams (see Appendix B). The HMI magnetogram of a portion of the decaying AR is shown in panel (a), and is saturated to ±300 G. One example of an interaction between the filament with a bent arcade (loops) is shown in panels (a4)–(b4) with the evolution of the hot flux indicated by white arrows in panel (a4). An online animation is attached to this figure (Movie_overview_aia.mpg).

As the filament eruption obscures the coronal loop footpoints in the vicinity of the NR and NRH, we studied the evolution of PRH in relation to the evolution of the erupting flux rope (Sect. 3.3) as well as coronal loops and their footpoints (Sect. 3.4). The evolution of the PR and its hook PRH is rather complicated, as the hook is large and extended into the network area far away from the center of the decaying AR. We nevertheless managed to trace the ribbon PR and PRH in its entirety. This was a difficult task, as unique identification of a continuous ribbon in weak-field areas cannot be readily automated – portions of the ribbon and especially its extended hook were found to not always be continuous at a similar intensity level, or in our weak B-class event to be significantly brighter compared to the scattered plage polarities. Identification of the ribbon in AIA 1600 Å or ratio of 1600 Å/1700 Å images was also not possible for these reasons. In addition, areas previously swept by the ribbon are known to remain bright for an extended period of time (Lörinčík et al. 2019). Furthermore, the ribbon can exhibit relatively fast squirming motions (Dudík et al. 2016) as well as rapid elongations or even “jumps” to neighboring supergranules (Lörinčík et al. 2019). In our event, the squirming motions, as the ribbon searches for concentrations of magnetic flux to reconnect, were especially present in the weak-field areas. For these reasons, we resorted to manual tracing of the PR and PRH in AIA 304 Å images, relying simultaneously on both the higher-resolution IRIS/SJI data as well as difference images of 304 Å (cf., Appendix A) to identify both the ribbon and the direction of its motion.

The results of the tracing of the ribbon PR and its hook PRH are shown in Fig. 3 as the dashed green line overlaid on the AIA 304 Å observations (panels c3–c5). The evolution of the traced-out ribbon is also summarized in Fig. 4, showing the locations of the ribbon at five different times using different colors. It is seen that both the shape of the ribbons and its evolution is indeed not simple, as the ribbons contain many undulations related to the squirming motions. Sometimes portions of the PRH stay at approximately the same place, for example its far end at 00:32–00:40 UT near the vicinity of Solar X = −170″ (blue, green, orange), while at other times the ribbon or the end of its hook changes shape and moves by even several tens of arcsec (red). Although the shape of the PRH is complex, its interior is always well defined, and is denoted by the “PRH” label in Fig. 4. The relatively large change in the straight part of the ribbon at 00:54 UT (red) near the inversion line is due to the closing of the PR, as was also observed by IRIS/SJI (panel f of Fig. 2). This indicates ongoing erosion of the original filament flux rope and its gradual relocation to the far lobe of the PRH at Solar X ≲ −170″.

3.3. Evolution of the erupting flux rope

The evolution of the PRH is accompanied by the evolution of the erupting flux rope itself. Overall, the eruption starts as an erupting filament that later reconnects with coronal loops, producing a hot flux rope (“hot channel”, see, e.g., Zhang et al. 2012, 2023; Cheng et al. 2013, 2014, 2015; Song et al. 2015; Dudík et al. 2016; Hernandez-Perez et al. 2019; Liu et al. 2022). In our event, the hot channel is best detected in the AIA 94 Å channel (see Fig. 3a2–a5). Given that the GOES flare is rather weak, reaching only the B3.2 level, the hot channel is correspondingly rather weak, having only ≈5–10 DN s−1 px−1 in AIA 94 Å. Given that the 94 Å channel of AIA is multithermal (see O’Dwyer et al. 2010; Del Zanna 2013), containing contributions from Fe XVIII formed around log(T [K]) = 6.8 in solar flare conditions, as well as much cooler “coronal” contributions from Fe X and Fe XIV, we employed the Fe XVIII separation method of Del Zanna (2013) to remove contributions other than the eruption-generated hot emission. In employing this method, we took the degradation of AIA sensitivity into account (Appendix B).

Overall, the event is an eruptive one, producing a pair of J-shaped ribbons, with a rising (erupting) filament and later a growing hot channel that shows a slipping motion of its footpoints outward along its hooks around 00:40 UT (Fig. 3a4, and the accompanying animation). Subsequently, the hot channel grows and disperses outwards in the direction of the original filament eruption. This evolution of the hot channel is denoted by the dotted white arrows in panels a4–a5 of Fig. 3. We call this evolution “dispersal” rather than “eruption”, because despite the evolution being consistent with an eruption, there is a strong decrease in the signal in the AIA 94 Å channel after 00:54 UT. No trace of the erupting hot channel is left after 01:06 UT (panel a6). At this time, well into the decay phase of the B3.2 flare (Fig. 1a), only gradually cooling flare loops are left in the location of the eruption.

Finally, the event is accompanied by progressive and pronounced coronal dimmings (see panels b5–b6 of Fig. 3) that are detected in all coronal AIA channels. This indicates the removal of coronal material. In 171 Å, the dimmings are pronounced, as the majority of coronal loops in the northern part of the AR (where the eruption occurs) disappear, leaving only rather weak emission. At the location of the slipping motion of the hot channel at 00:40 UT (denoted by a pink circle in Figs. 3–5), core dimmings are later observed (downward-pointing white arrow in Fig. 3b6; see also Dissauer et al. 2018a,b, for nomenclature of coronal dimmings). These core dimmings are located in the interior of the far end of PRH (cf., panel c5); i.e., they likely correspond to the legs of the erupting structure (see, e.g., Hudson et al. 1996; Thompson et al. 2000; Gopalswamy & Thompson 2000; Attrill et al. 2006; Veronig et al. 2021; Krista et al. 2022; Wang et al. 2023). Although visible in 171 Å, these core dimmings are most pronounced in the 211 Å channel (not shown). However, no CME is ultimately detected in the SOHO/LASCO catalog1. Since this event is weak, and located close to the center of the solar disk in a decayed AR, we speculate that the resulting halo CME is also weak and ultimately not detected.

thumbnail Fig. 4.

Evolution of the ribbon PR and its hook PRH, based on manual tracing of the ribbon. Individual colors correspond to individual times, as is indicated. Pink and orange circles indicate two spatial positions of interest. The orange one is outside of the PRH at 00:30 UT, inside at 00:32 UT, and again outside at 00:36 UT. The pink one is first outside PRH, but is swept by the PRH at 00:36–00:40 UT (see text for details).

thumbnail Fig. 5.

Difference images highlighting the evolution of the filament and coronal loops during AR–RF reconnection. The rectangular IRIS/SJI field of view is shown in red, and the position of the IRIS slit in orange. Panels (a1)–(c1), (a3)–(c3), and (a4)–(c4) show times corresponding to the IRIS spectra studied in Sect. 4. Panels (a2)–(c2) and (a5)–(c5) are shown for context, indicating the footpoints of coronal loops (arcades) A1 and A2 that reconnected with the filament. At the respective times, the loops are dark structures, indicating that they no longer exist. Green arrows at 00:30:21 and 00:32:54 UT indicate locations along the IRIS slit where brightenings are observed within the filament. Animations of the CID images are available online (Movie_CID_IRIS.mpg).

3.4. Reconnection of the filament with coronal loops (AR–RF reconnection)

We now turn to the analysis of the interaction of the filament with coronal loops and show that this interaction is due to AR–RF reconnection (first theoretically described by Aulanier & Dudík 2019) between the original flux rope (filament) and the surrounding coronal loops, producing the hot flux rope (hot channel, Sect. 3.3). To study the evolution of the coronal loops, we employed the combined improved difference method (CID), which highlights the moving structures and/or changes in the observed intensity. The CID is analogous to the running difference method, but the images are scaled differently to enhance weak structures as well as partially suppress the image noise. Details on the method are given in Appendix A.

Since the coronal AIA channels observed both the moving coronal loops, the erupting filament, as well as the flare ribbons, we plot the CID from these observations alongside the CID 304 Å, which contains only the erupting filament and flare ribbons (see Fig. 5 and the accompanying animation). Since different coronal loop systems are visible in different AIA filters, we plot the CID 211 Å together with the CID 171 Å, covering the range of warm coronal loops at the AR periphery. The location of the IRIS FOV (see Sect. 2.3) is overplotted as the red box, with the vertical central line indicating the IRIS slit.

The animation associated with Fig. 5 shows that the eruption of the filament is accompanied by motions of coronal loops in all coronal channels. The first loop motions are detected around 00:26 UT, and are clearly induced by the erupting filament. Around 00:30:20 UT, a multitude of coronal loops (panels a1–b1 of Fig. 5) are seen interacting with the filament, which experiences strong brightenings (white in panel c1). One of the strongest brightenings in the filament is located directly at the IRIS slit (red arrow in panel c1). By 00:40 UT, all of the filament is in emission (see panel b4 ofFig. 3).

The interaction of the filament with coronal loops and the associated brightenings are a result of arcade-to-rope reconnection (AR–RF reconnection; Aulanier & Dudík 2019). This can be discerned by following the connectivity of, for example, two points denoted by the orange and pink circles (Figs. 3–5). At 00:24, both are outside of PRH and correspond to coronal loops (arcades A1 and A2 in panel b2 of Fig. 3). Both these coronal loops subsequently disappear, producing only dark traces in the CID(4) images. The A1 disappears by 00:31 UT (panel a2 of Fig. 5); however, by 00:32 UT this location is a flux rope (denoted R1 in panel a3 of Fig. 3), indicating a change of connectivity, A → R. This is consistent with the evolution of the hook PRH (Fig. 4: at 00:30 UT, this location is outside of PRH, indicating that it is a coronal arcade; but at 00:32 UT, the PRH has already swept it, becoming a part of the flux rope. Still later, at 01:06 UT, this location is again outside of the PRH, and at the very edge (“cantle”, Lörinčík et al. 2021a) of the flare loop arcade (F1 in panel a6). Overall, the evolution of the connectivity is A → R → F, in good agreement with the multistageAR–RF reconnection (see Sect. 4.4 and Fig. 6 of Aulanier & Dudík 2019)

The loop A2 disappears later, at 00:36 UT (panels a5 and b5 of Fig. 5), and at 00:40 UT this location is again a flux rope (R2 in panel a4 of Fig. 3), showing the slipping motions from its footpoints. Subsequently, this location is on the inside of the hook PRH (Fig. 4), but the hot flux rope visible in 94 Å moves away from this location, leaving only a coronal dimmingregion.

4. Spectral analysis

4.1. Rasters

The raster mode of IRIS observations allows us to follow the evolution of the flare ribbons and filament interaction with overlying arcades during the eruption. The AR is scanned by eight slit positions, which provided the spectra in Mg II, Si IV, and CII lines within about one minute raster cadence. We provide four examples of the eight positions of the slit locations in the C II SJI image in the left panels of Fig. 6, while the right panels show the Si IV spectra at all eight slit positions. We chose four times when the individual spectra indicate reconnections between filament and coronal loops. These are detected principally at 00:30:20 UT, 00:32:54 UT, 00:34:11 UT, and 00:35:56 UT. An accompanying animation provides a complete view of the spectral evolution at the flare site from 00:25:32 to 01:15:28 UT.

thumbnail Fig. 6.

Overview of the Si IV spectra of the reconnecting filament. The left panels show the C II SJI images, overlaid with all eight slit positions as vertical dashed lines. The right panels show the corresponding eight Si IV spectra at each slit position. Prominent enhancements of the Si IV 1402.8 Å line, which can also be Doppler-shifted, are shown by blue and red arrows, respectively. Note the similar behavior of the O IV line at 1401.2 Å. An animation of this figure is available online (SJICII_SpectraSiIV_revised.mp4).

4.2. Spectra of the reconnecting filament

By analyzing the spectra at all eight positions, we found that slit 7 effectively captures the first three reconnection episodes between the overlying arcades and the filament (see blue arrows in Fig. 6). The corresponding slit position 7 is shown in Fig. 5, panels (a1)–(c1), (a3)–(c3), and (a4)–(c4). In all three cases, brightenings are detected in the filament in 304 Å (panels c1, c3, and c4), and furthermore at locations where the filament overlaps with the evolving coronal loops. The clearest example occurs at 00:30:20 UT (panels b1–c1 of Fig. 5), when the 304 Å brightening is the strongest and the unobstructed portion of the 171 Å loop is in close proximity to the slit position.

The spectra of the reconnecting parts of filament, recorded by the IRIS spectrograph, have first to be distinguished from both the spectra of the flare ribbons in the chromosphere, as well as from the spectra of the erupting filament itself, which will appear to be blueshifted as it moves upward in the solar atmosphere. The ribbons are relatively easy to distinguish with the help of the Mg II spectral window, since the ribbons also appear bright in the extended Mg II line wings. The location of the ribbon is thus clearly indicated by a horizontal emission feature spanning the Mg II window (see Fig. 7). The spectral signature of the erupting filament is a large inclined feature crossing both the Mg II and Si IV spectra. In Mg II, the filament is easier to distinguish, since its relatively cool material is seen in absorption. The inclination of the filament in the spectrograph detector images is caused by the twisted structure of the filament, whose southern part is rising faster (≈ − 100 km s−1), while its northern part can be moderately redshifted (up to ≈ + 50 km s−1 at 00:34:11 UT, panel (i) of Fig. 7), indicating a combination of upward and untwisting motions during the filament eruption.

thumbnail Fig. 7.

Detailed Si IV and Mg II spectra of the three reconnection episodes detected at slit position 7 (see Fig. 6). Panels (a)–(c) show the IRIS/SJI observations in C II, similarly to in Fig. 6, with the position of slit 7 is shown with a vertical dashed white line. Panels (d)–(f) show the Si IV detector images, while panels (g)–(i) show the Mg II ones at this slit position at three different times. The horizontal dashed blue line indicates the reconnecting filament spectra with large Doppler shifts plotted in Fig. 8. The blue and red arrows in panels d-e-f indicate the blueshift and redshift flows during the reconnection in Si IV spectra. The white arrow indicates the filament in panel (i).

The Si IV and Mg II spectra from the first three reconnection episodes (at slit position 7) at 00:30:20 UT, 00:32:54 UT, and 00:34:11 UT are shown in detail in Fig. 7. With the help of the Mg II lines, it can be immediately seen that the position marked by the horizontal dashed blue line is distinct from the ribbon, whose emission appears further to the north, at all three times. Furthermore, the emission in the Si IV line at the location of the dashed blue line is both brighter and more blueshifted than in the rest of the filament. The strong blueshifts observed at the three times occur in all three cases in a very localized area of a few IRIS pixels, corresponding to 1–2″. The individual spectra observed at the three times are shown in Fig. 8 along with their uncertainties. The properties of these spectra are described in the following subsections. Except for the presence of strongly blueshifted components, a common feature is the relative absence of the O IV and S IV lines at 1404.8 and 1406.2 Å, which have fewer than 20 counts (DN) and are barely visible above the continuum. Only the strongest O IV line at 1401.2 Å is clearly discernible, even if it is weak in some spectra.

thumbnail Fig. 8.

Observed IRIS spectra (black) of Si IV (a–c) and Mg II (d–f) at three different times as shown in previous figures (horizontal dashed blue lines in Fig. 7). Uncertainties of the observed spectra are denoted by thin gray error bars. The fit to the Si IV spectra is shown in orange, with individual components indicated by colored lines (blue and red for blueshifted and redshifted components of the Si IV line, violet for O IV line). Reference spectra for Mg II are shown in green, and are taken in the non-flaring chromosphere to the north of the erupting filament. Vertical pink lines denote rest wavelengths for individual spectral lines. At 00:34:11 UT there are several components in both lines, in particular one corresponding to 250 km s−1 (pink arrow in panel f and the “?” sign in panel c). See text for details.

4.2.1. IRIS spectra at 00:30:20 UT

The Si IV and Mg II spectra observed in the reconnecting filament at 00:30:20 UT are shown in Fig. 8, panels (a) and (d). The Si IV line is completely blueshifted, with an asymmetric profile. The O IV line also appears blueshifted. The spectrum can be fit with four components plus a constant continuum. The overall fit is satisfactory with χred2 = 1.87 and shown by the orange curve. Although the value of the χred2 is larger than 1, most of the contributions come from very weak, unfittable features above the continuum (see Appendix B of Dudík et al. 2017a) and not from improper fitting of the spectral lines.

The relatively weak O IV line can be fit with a single, strongly blueshifted Gaussian component with vD = −170.8 ± 2.7 km s−1 (dashed violet line in Fig. 8. The Si IV line requires at least two blueshifted components, a strong one with vD = −207.9 ± 2.2 km s−1 and a weaker (less intense) one with vD = −74.2 ± 2.5 km s−1 (blue curves). A further weak, slightly redshifted component improves the fit (red Gaussian). The properties of these components are listed in Table 1, where I0, λ, vD, FWHM, and κ denote the intensity, wavelength, Doppler velocity, full width half maximum, and kappa values, respectively. A constant continuum of 10.4 ± 0.3 DN is sufficient to account for most of the weak spectral features as well as the true continuum (horizontal orange line in Fig. 8).

Table 1.

Fit parameters for the IRIS Si IV FUV spectrum. Individual components plotted in Fig. 8 are indicated.

Interestingly, the strongest Si IV component is non-Gaussian, with κ = 3.3 ± 0.5, i.e., with pronounced wings, likely indicating presence of accelerated particles (see the discussion in Jeffrey et al. 2016, 2017; Dudík et al. 2017a; Polito et al. 2018). Its width (FWHM) is also extreme, about 121 km s−1, indicating the presence of magnetohydrodynamic turbulence that can also be associated with particle acceleration (Bian et al. 2014). We find that the widths of the individual fit components, whether Gaussian or non-Gaussian, are generally rather large (see Table 1), being comparable even with widths traditionally detected from flare lines formed above 10 MK such as Fe XXI and hotter ones (e.g., Antonucci 1989; Ashfield et al. 2024). Note that the single component used for fitting the O IV line is usually wider than individual components for the Si IV line (Table 1), while it is also less blueshifted. This is probably due to the O IV line being relatively weak (40 DN), so that it cannot be reliably fit with multiple components such as the much stronger Si IV line.

Overall, we take the presence of strong Doppler shift of the Si IV line, its extreme width, and the non-Gaussian profile of the most blueshifted component as three independent lines of evidence for impulsive energy release due to magnetic reconnection at this location. The Mg II spectra also show the presence of strongly blueshifted components (see the dashed blue lines in Fig. 7, panels g–i). The most blueshifted peak occurs at about ≈200 km s−1 (Fig. 8d), which corresponds well with the most blueshifted peak of Si IV (panel a). This strong blueshift in Mg II is also quite localized, as the spectra more to the north (above the dashed blue line in Fig. 7) are represented by a dark reddish area in the core of the line. This area is more extended in panels h and i and corresponds to the escaping filament.

To better appreciate these Mg II spectra with strong Doppler shifts (Fig. 8), we overplotted the chromospheric spectra for reference (green lines in Fig. 8). The reference spectra are taken in the far northern part of the IRIS FOV, at pixel No. 700, far away from the erupting filament. The reference spectrum is characterized by the presence of a narrow central reversal, and much narrower widths. These Mg II spectra could be interpreted by a cloud model consisting of two clouds: a low-velocity blueshifted absorbing cloud representing the erupting filament (around –40 km s−1) and a strongly blueshifted (around –200 km s−1) emission cloud that originates due to reconnection similarly to the correspondingly strong blueshifted Si IV components. Note that these Mg II profiles are also similar to the profiles observed in a previous study of reconnection jet (Joshi et al. 2020, 2021). Interestingly, unlike at later times, at 00:30:20 UT the far Mg II wings of the reconnecting filament are higher than the reference profile, possibly indicating overall heating of the filament material.

4.2.2. IRIS spectra at 00:32:54 UT

At 00:32:54 UT, the filament plasma at slit position 7 is again bright and showing strong blueshifts. The Si IV line is asymmetric but fully blueshifted. Consequently, it can be fit with two components (panel b of Fig. 8). A weaker but strongly blueshifted component is found in the asymmetrically enhanced blue wing, having vD = −181.3 ± 5.6 km s−1. The main component is stronger, but less blueshifted (vD = −66.9 ± 3.0 km s−1, see Table 1). Both these components are again very broad but Gaussian, having FWHMs of 121 and 146 km s−1, respectively.

The Mg II lines are also quite broad (especially compared to the reference profile), with no central reversal, and have a pronounced, asymmetric blue wing at ≈150 km s−1, in broad agreement with the Si IV spectra. The far wings (above about 300 km s−1) are similar to the reference profile wings (green color). This non-reversed Mg II profile in Fig. 8 (panel e) is the result of filament absorption, as is evident in the spectra shown in Fig. 7 (panel h). In comparison to the Si IV profiles, the Mg II profiles are significantly affected by filament absorption. Again at 00:32:54 UT along the dashed blue line (panel i), the emission is reduced all along the Mg II blue wing due to the smeared distribution of the filament velocities during its eruption.

4.2.3. IRIS spectra at 00:34:11 UT

At 00:34:11 UT, the Si IV spectrum becomes rather complex, with several distinct components. Each of these can be fit with a single Gaussian to a good degree of accuracy. The most intense component is a redshifted one with vD = +88.1 ± 0.9 km s−1 (Fig. 8 and Table 1). There is also a relatively strong blueshifted component with vD = −52.8 ± 1.8 km s−1.

There are two other lines, one moderately strong at 1401.44 Å – i.e., in the vicinity of the rest wavelength of the O IV 1401.16 Å line) – and another weak line at 1400.01 Å. What these lines are is not immediately clear. The weak line is most likely an O IV one, either a strongly blueshifted (vD = −247.7 ± 7.1 km s−1) component of the line whose rest wavelength is at 1401.16 Å, or a redshifted (vD = +50.3 ± 7.1 km s−1) component of a weak O IV line whose rest wavelength is at 1399.77 Å (see Table 1 in Dudík et al. 2014b). The moderately strong line, denoted by the “?” sign in panel (c) of Fig. 8, can either be interpreted as a distinct Si IV component with vD = −283.1 ± 3.6 km s−1 or a redshifted O IV one with vD = +59.9 ± 3.6 km s−1. We think that the blueshifted Si IV interpretation is more likely, since if this were the redshifted O IV component, the comparatively strong blueshifted counterpart of this line would be missing. Further support that this an extremely blueshifted Si IV component comes from the Mg II spectra, which show a weak but broad blueshifted feature at about ≈250 km s−1 (pink arrow in panel (f) of Fig. 8).

In accordance with the multiple blueshifted and redshifted components being present in the Si IV line, the Mg II spectrum also shows the presence of multiple components with quite similar velocities. Furthermore, both the Mg II as well as Si IV lines are again extremely broad, with the blueshifted components typically being the broadest ones (Table 1). The nonthermal width of the two blueshifted Si IV components is 175 and 168 km s−1, respectively.

The very weak (15 DN) signal in the Mg II line and the corresponding line component in the Si IV window (97 DN) with high blueshift (−250 km s−1) likely also arise due to reconnection. However, due to the presence of multiple overlapping structures along the line of sight, it is difficult to isolate a distinct source for these highly blueshifted components in either AIA or IRIS observations (see Fig. 5, panels a4–c4). These faint signals may originate from different structures entirely and could also be explained by Doppler dimming effects (Rompolt 1980; Peat et al. 2024).

5. Discussion

Magnetohydrodynamic simulations of solar eruptions predict that reconnection in 3D happens within the volume of the QSLs, where the field lines can exchange connectivities, and not at a rather singular “X”-type location as in the 2D picture. The 3D nature of magnetic reconnection subsequently manifests itself in the slipping motion of the reconnecting structures, including the erupting flux rope (Aulanier et al. 2012; Janvier et al. 2013; Aulanier & Dudík 2019). Indeed, we have found that the drifting footpoints of the reconnecting flux rope undergoing AR–RF reconnection are slipping (see Sect. 3.3 and panel (a4) of Fig. 3).

Nevertheless, our spectroscopic observations show a strong but rather localized response to the AR–RF reconnection within the erupting filament; that is, within the original erupting flux rope. This localization is rather strict, to within 1–2″ (Sect. 4.2, and also restricted in time to several episodes. We note that even though this localization in space and time is rather imperfect on account of the IRIS being a rastering spectrometer, with the raster cadence of about one minute in our case, it is unlikely that IRIS would miss a large portion of the reconnection-induced dynamic changes within the filament plasma.

The plasma response to reconnection was found to be threefold. First, we observe rather large blueshifts (up to 200 km s−1 or more) in both the Si IV and Mg II lines. Second, these blueshifts are always accompanied by very large nonthermal widths of the Si IV line, above 100 km s−1. Third, in one instance at the beginning of the AR–RF reconnection, the strongest and most blueshifted Si IV line is also non-Gaussian, being best approximated by a rather low value of κ ≈ 3.3. Such non-Gaussian profiles were in flares detected previously only in much hotter lines, such as Fe XVI (Jeffrey et al. 2016, 2017) and Fe XXIII–Fe XXIV (Jeffrey et al. 2017; Polito et al. 2018), whose formation temperature (in equilibrium conditions) is orders of magnitude higher.

The large blueshifts have been observed previously during reconnection between jet and loops (Ruan et al. 2019) or during breakout events (Reeves et al. 2015). Our observations presented here mean that these spectral signatures extend also to the AR–RF reconnection geometry.

The large nonthermal widths, along with the strong blueshifts and brightenings in the erupting filament, likely indicate that the filament plasma is being heated by the reconnection. Supporting evidence for this can be found in the AIA imaging observations. First, portions of the filament appear in emission even in coronal filters such as 171 Å, 193 Å, and 211 Å (cf., Fig. 5). Second, the hot emission in Fe XVIII appears at about 00:30 UT (see animation accompanying Fig. 3); that is, at the time IRIS observed the first spectral signatures of reconnection. Subsequently, the hot flux rope is already well developed by 00:32 UT, and the flare arcade also starts to appear, as is shown by the panel (a3) of Fig. 3. Except perhaps for the presence of Fe XVIII, spectroscopic observations of the reconnecting filament in lines formed at temperatures of 1–10 MK would be required for direct confirmation of the rapid heating of the filament plasma to much higher temperatures, as the original flux rope turns to the flare loop (R → F) and the coronal loop arcades to new flux rope field lines (A → R).

One last question remains, of why we observe mostly strong blueshifts. The traditional reconnection scenario in an ‘X’-point (which is not entirely valid here for reasons explained above) expects the presence of bidirectional reconnection outflow jets. In our case, the filament reconnects with coronal loops that are already significantly hotter than the formation temperatures for the IRIS Mg II and Si IV lines. Therefore, the strongly redshifted counterparts to our observed blueshifts again likely occur at much higher temperatures.

6. Summary

We present a spectroscopic analysis of a filament reconnecting with coronal loops during the course of its eruption that was accompanied by a weak B3.2-class flare that occurred on December 9, 2020. Despite this event being relatively small, we observe a strong response of the filament plasma to reconnection, chiefly in terms of ≈200 km s−1 blueshifts and extremely large widths of the Si IV lines observed by IRIS.

This accompanying flare is a classical two-ribbon event, with both ribbons showing J-shaped hooks, indicating the presence of an erupting magnetic flux rope. This flux rope is first identified as the erupting filament and later a hot flux rope seen in the Fe XVIII line in the 94 Å passband of SDO/AIA. One of the ribbons is compact, located in relatively strong magnetic polarities of the AR. The conjugate ribbon is weak and extended, and its hook undergoes significant evolution, sweeping conjugate footpoints of several coronal loops. These coronal loops initially overlap with the erupting filament along the line of sight. The eruption of the filament leads to interaction with the coronal loops, which subsequently reconnect with the filament and disappear. The filament turns to a hot flux rope whose footpoints reach to the previous footpoints of the coronal loops, indicating that the erupting filament underwent significant heating as it reconnected in the arcade-to-rope AR–RF reconnection geometry (Aulanier & Dudík 2019).

The IRIS slit rastered the erupting filament in the location where it interacted with the coronal loops. Using careful co-alignment of SDO/AIA and IRIS, we identified the locations of the individual ribbons, brightenings of the erupting filament, and coronal loops along the line of sight. The IRIS spectra shows the erupting filament as slanted structure in absorption in Mg II and in emission in Si IV, while the ribbon is visible as a bright structure in the Mg II wings. On top of that, we distinguished the spectral signatures of the reconnecting filament, which are threefold:

  • The presence of strong blueshifts, up to ≈200 km s−1, and higher in some cases, as distinct components of the Si IV and Mg II lines.

  • The presence of extremely large nonthermal widths, routinely exceeding 100 km s−1 in the Si IV line.

  • The presence of a strongly non-Gaussian profile (κ ≈ 3.3) of the most blueshifted component at the start of the reconnection.

Overall, these three separate lines of evidence show a strong response of the filament plasma to the energy being released via to the AR–RF reconnection with the neighboring coronal loop arcade. Although the traditional bilateral flows are not seen in the reconnecting filament, except in the last instance, we argue that the redshifted component is not seen by IRIS because the filament is reconnecting with much hotter coronal loops, so that the redshifted counterpart cannot be seen in the relatively cool lines of Mg II and Si IV observed by IRIS. In summary, our observations of the response of the filament plasma to reconnection represent the first spectral observations of the AR–RF reconnection, revealing that this response can be quite localized within the reconnecting filament.

Data availability

Movies associated to Figs. 3, 5, and 6 are available at https://www.aanda.org

Acknowledgments

This research has been supported by the Research Council of Norway through its Centres of Excellence scheme, project number 262622 and the European Research Council through the Synergy grant No. 810218 (“The Whole Sun”, ERC-2018-SyG). Part of this work was supported by the German Deutsche Forschungsgemeinschaft, DFG project number Ts 17/2–1. J.D. acknowledges support from the Czech Science Foundation, grants No. GACR 22-07155S and 25-18282S, as well as institutional support RWO:67985815 from the Czech Academy of Sciences. G.A. acknowledges financial support from the French national space agency (CNES), as well as from the Programme National Soleil Terre (PNST) of the CNRS/INSU also co-funded by CNES and CEA. R.C. acknowledges the support from DST/SERB project No. EEQ/2023/000214. Hα Data were acquired by GONG instruments operated by NISP/NSO/AURA/NSF with contribution from NOAA. AIA data are courtesy of NASA/SDO and the AIA, EVE, and HMI science teams. IRIS is a NASA small explorer mission developed and operated by LMSAL with mission operations executed at NASA Ames Research Center and major contributions to downlink communications funded by ESA and the Norwegian Space Centre.

References

  1. Antonucci, E. 1989, Sol. Phys., 121, 31 [Google Scholar]
  2. Ashfield, W., Polito, V., Yu, S., Collier, H., & Hayes, L. A. 2024, ApJ, 973, 96 [Google Scholar]
  3. Attrill, G., Nakwacki, M. S., Harra, L. K., et al. 2006, Sol. Phys., 238, 117 [Google Scholar]
  4. Aulanier, G., & Dudík, J. 2019, A&A, 621, A72 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  5. Aulanier, G., Török, T., Démoulin, P., & DeLuca, E. E. 2010, ApJ, 708, 314 [Google Scholar]
  6. Aulanier, G., Janvier, M., & Schmieder, B. 2012, A&A, 543, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  7. BenMoussa, A., Gissot, S., Schühle, U., et al. 2013, Sol. Phys., 288, 389 [Google Scholar]
  8. Bian, N. H., Emslie, A. G., Stackhouse, D. J., & Kontar, E. P. 2014, ApJ, 796, 142 [NASA ADS] [CrossRef] [Google Scholar]
  9. Boerner, P. F., Testa, P., Warren, H., Weber, M. A., & Schrijver, C. J. 2014, Sol. Phys., 289, 2377 [Google Scholar]
  10. Chandra, R., Schmieder, B., Aulanier, G., & Malherbe, J. M. 2009, Sol. Phys., 258, 53 [Google Scholar]
  11. Chen, Y., Mandal, S., Peter, H., & Chitta, L. P. 2024, A&A, 692, A119 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  12. Cheng, X., Zhang, J., Ding, M. D., Liu, Y., & Poomvises, W. 2013, ApJ, 763, 43 [NASA ADS] [CrossRef] [Google Scholar]
  13. Cheng, X., Ding, M. D., Zhang, J., et al. 2014, ApJ, 789, L35 [NASA ADS] [CrossRef] [Google Scholar]
  14. Cheng, X., Ding, M. D., & Fang, C. 2015, ApJ, 804, 82 [NASA ADS] [CrossRef] [Google Scholar]
  15. Del Zanna, G. 2013, A&A, 558, A73 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Démoulin, P., Henoux, J. C., Priest, E. R., & Mandrini, C. H. 1996, A&A, 308, 643 [Google Scholar]
  17. Démoulin, P., Bagala, L. G., Mandrini, C. H., Henoux, J. C., & Rovira, M. G. 1997, A&A, 325, 305 [NASA ADS] [Google Scholar]
  18. De Pontieu, B., Title, A. M., Lemen, J. R., et al. 2014, Sol. Phys., 289, 2733 [Google Scholar]
  19. Dissauer, K., Veronig, A. M., Temmer, M., Podladchikova, T., & Vanninathan, K. 2018a, ApJ, 855, 137 [NASA ADS] [CrossRef] [Google Scholar]
  20. Dissauer, K., Veronig, A. M., Temmer, M., Podladchikova, T., & Vanninathan, K. 2018b, ApJ, 863, 169 [NASA ADS] [CrossRef] [Google Scholar]
  21. Dos Santos, L. F. G., Bose, S., Salvatelli, V., et al. 2021, A&A, 648, A53 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Dudík, J., Janvier, M., Aulanier, G., et al. 2014a, ApJ, 784, 144 [CrossRef] [Google Scholar]
  23. Dudík, J., Del Zanna, G., Dzifčáková, E., Mason, H. E., & Golub, L. 2014b, ApJ, 780, L12 [Google Scholar]
  24. Dudík, J., Polito, V., Janvier, M., et al. 2016, ApJ, 823, 41 [Google Scholar]
  25. Dudík, J., Polito, V., Dzifčáková, E., Del Zanna, G., & Testa, P. 2017a, ApJ, 842, 19 [CrossRef] [Google Scholar]
  26. Dudík, J., Zuccarello, F. P., Aulanier, G., Schmieder, B., & Démoulin, P. 2017b, ApJ, 844, 54 [Google Scholar]
  27. Dudík, J., Lörinčík, J., Aulanier, G., Zemanová, A., & Schmieder, B. 2019, ApJ, 887, 71 [Google Scholar]
  28. Gopalswamy, N., & Thompson, B. J. 2000, JASTP, 62, 1457 [Google Scholar]
  29. Gou, T., Liu, R., Veronig, A. M., et al. 2023, Nat. Astron., 7, 815 [NASA ADS] [CrossRef] [Google Scholar]
  30. Guo, J., Qiu, Y., Ni, Y., et al. 2023, ApJ, 956, 119 [NASA ADS] [CrossRef] [Google Scholar]
  31. Hernandez-Perez, A., Su, Y., Veronig, A. M., et al. 2019, ApJ, 874, 122 [NASA ADS] [CrossRef] [Google Scholar]
  32. Hudson, H. S., Acton, L. W., & Freeland, S. L. 1996, ApJ, 470, 629 [CrossRef] [Google Scholar]
  33. Innes, D. E., Inhester, B., Axford, W. I., & Wilhelm, K. 1997, Nature, 386, 811 [Google Scholar]
  34. Janvier, M., Aulanier, G., Pariat, E., & Démoulin, P. 2013, A&A, 555, A77 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  35. Janvier, M., Aulanier, G., Bommier, V., et al. 2014, ApJ, 788, 60 [Google Scholar]
  36. Janvier, M., Aulanier, G., & Démoulin, P. 2015, Sol. Phys., 290, 3425 [Google Scholar]
  37. Jeffrey, N. L. S., Fletcher, L., & Labrosse, N. 2016, A&A, 590, A99 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Jeffrey, N. L. S., Fletcher, L., & Labrosse, N. 2017, ApJ, 836, 35 [Google Scholar]
  39. Joshi, R., Schmieder, B., Aulanier, G., Bommier, V., & Chandra, R. 2020, A&A, 642, A169 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. Joshi, R., Schmieder, B., Tei, A., et al. 2021, A&A, 645, A80 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  41. Joshi, R., Mandrini, C. H., Chandra, R., et al. 2022, Sol. Phys., 297, 81 [NASA ADS] [CrossRef] [Google Scholar]
  42. Joshi, R., Aulanier, G., Radcliffe, A., et al. 2024, A&A, 687, A172 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  43. Krista, L. D., Manning, D., & West, M. J. 2022, ApJ, 930, 165 [NASA ADS] [CrossRef] [Google Scholar]
  44. Lemen, J. R., Title, A. M., Akin, D. J., et al. 2012, Sol. Phys., 275, 17 [Google Scholar]
  45. Liu, Y., Su, Y., Liu, R., et al. 2022, ApJ, 941, 83 [NASA ADS] [CrossRef] [Google Scholar]
  46. Lörinčík, J., Aulanier, G., Dudík, J., Zemanová, A., & Dzifčáková, E. 2019, ApJ, 881, 68 [Google Scholar]
  47. Lörinčík, J., Dudík, J., & Aulanier, G. 2021a, ApJ, 909, L4 [Google Scholar]
  48. Lörinčík, J., Dudík, J., Aulanier, G., Schmieder, B., & Golub, L. 2021b, ApJ, 906, 62 [Google Scholar]
  49. Lörinčík, J., Dudík, J., Sainz Dalda, A., et al. 2024, Nat. Astron., 9, 45 [Google Scholar]
  50. O’Dwyer, B., Del Zanna, G., Mason, H. E., Weber, M. A., & Tripathi, D. 2010, A&A, 521, A21 [Google Scholar]
  51. Pariat, E., & Démoulin, P. 2012, A&A, 541, A78 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  52. Peat, A. W., Osborne, C. M. J., & Heinzel, P. 2024, MNRAS, 533, L19 [NASA ADS] [CrossRef] [Google Scholar]
  53. Pesnell, W. D., Thompson, B. J., & Chamberlin, P. C. 2012, Sol. Phys., 275, 3 [Google Scholar]
  54. Polito, V., Dudík, J., Kašparová, J., et al. 2018, ApJ, 864, 63 [NASA ADS] [CrossRef] [Google Scholar]
  55. Priest, E. 2014, Magnetohydrodynamics of the Sun (Cambridge, UK: Cambridge University Press) [Google Scholar]
  56. Reeves, K. K., McCauley, P. I., & Tian, H. 2015, ApJ, 807, 7 [Google Scholar]
  57. Rompolt, B. 1980. Hvar Obs. Bull., 4, 55 [Google Scholar]
  58. Ruan, G., Schmieder, B., Masson, S., et al. 2019, ApJ, 865, 123 [Google Scholar]
  59. Scherrer, P. H., Schou, J., Bush, R. I., et al. 2012, Sol. Phys., 275, 207 [Google Scholar]
  60. Schmieder, B., Aulanier, G., & Vršnak, B. 2015, Sol. Phys., 290, 3457 [NASA ADS] [CrossRef] [Google Scholar]
  61. Song, H. Q., Chen, Y., Zhang, J., et al. 2015, ApJ, 808, L15 [NASA ADS] [CrossRef] [Google Scholar]
  62. Thoen Faber, J., Joshi, R., Rouppe van der Voort, L., et al. 2025, A&A, 693, A8 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  63. Thompson, B. J., Cliver, E. W., Nitta, N., Delannée, C., & Delaboudinière, J. P. 2000, Geophys. Res. Lett., 27, 1431 [Google Scholar]
  64. Veronig, A. M., Odert, P., Leitzinger, M., et al. 2021, Nat. Astron., 5, 697 [NASA ADS] [CrossRef] [Google Scholar]
  65. Wang, W., Qiu, J., Liu, R., et al. 2023, ApJ, 943, 80 [NASA ADS] [CrossRef] [Google Scholar]
  66. Zemanová, A., Dudík, J., Aulanier, G., Thalmann, J. K., & Gömöry, P. 2019, ApJ, 883, 96 [Google Scholar]
  67. Zhang, J., Cheng, X., & Ding, M.-D. 2012, Nat. Commun., 3, 747 [Google Scholar]
  68. Zhang, Q., Teng, W., Li, D., Dai, J., & Zhang, Y. 2023, ApJ, 958, 85 [NASA ADS] [CrossRef] [Google Scholar]
  69. Zuccarello, F. P., Chandra, R., Schmieder, B., Aulanier, G., & Joshi, R. 2017, A&A, 601, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]

Appendix A: Combined improved difference of SDO/AIA images

To study the evolution of various structures in time, we employed CID, which combines the advantages of both the traditionally used running-difference method and the less common log running ratio method.

The running difference (RD) of a series of imaging observations is simply constructed as

RD ( d ; x , y ) i = I ( x , y ) i − I ( x , y ) i − d , $$ \begin{aligned} \mathrm{RD} (d;x,y)_i = I(x,y)_i - I(x,y)_{i-d}\,, \end{aligned} $$(A.1)

where x and y are spatial coordinates and i denotes the i-th image in a time series. The d gives the difference between frames chosen. In practice, d is often larger than one for observations with sufficient cadence to cover the evolution of individual structures. For example, Dudík et al. (2017b) chosen d = 2 (i.e., a difference of 24 s) to highlight the evolution of coronal loops in the vicinity of eruptive solar flares observed by SDO/AIA. Meanwhile, Dudík et al. (2019) used d = 4 to better highlight coronal loops involved in AR–RF reconnections. We note that a higher value of d generally produces a more contrasting signal in the RD series, as long as d is still smaller than the typical evolution timescale of the structure.

The log running ratio method (LRR, Lörinčík et al. 2021b, 2024) is also sometimes used to highlight dynamical evolution. Its expression is

LRR ( d ; x , y ) i = log 10 ( I ( x , y ) i I ( x , y ) i − d ) , $$ \begin{aligned} \mathrm{LRR} (d;x,y)_i = \log _{10}\left(\frac{I(x,y)_i}{I(x,y)_{i-d}}\right)\,, \end{aligned} $$(A.2)

which can be rewritten as

LRR ( d ; x , y ) i = log 10 ( I ( x , y ) i ) − log 10 ( I ( x , y ) i − d ) , $$ \begin{aligned} \mathrm{LRR} (d;x,y)_i = \log _{10}(I(x,y)_i) - \log _{10}(I(x,y)_{i-d})\,, \end{aligned} $$(A.3)

meaning that LRR is essentially a running difference of logarithmically scaled images. This makes it useful for studying evolution in environments containing multiple structures of highly different intensities, where the linear RD method leads to some structures structures either being difficult to detect and different ones saturated.

Both RD and LRR methods, while useful, contain their own drawbacks. The RD method enhances noise in areas of medium observed intensity but non-moving or weakly moving structures. An example for AIA 171 Å at 00:34:09 UT can be found in panel (b) of Fig. A.1. Contrary to that, the LRR greatly enhances noise in weak-signal areas (panel c), where the photon and/or read noise is larger or comparable to the real signal. This in turn could necessitate averaging either in the image plane (for example, by a 3×3 boxcar) or in time, in each case leading to a slight loss of resolution.

Given that the noise-enhanced areas are in practice complementary in the RD and LRR methods, their product can lead to suppression of both areas of enhanced noise, while preserving real difference signal originating in moving or evolving structures Upon trial and error, we found best results for the following scaling, which we call the CID:

CID ( d ) i = | RD ( d ) i | 1 / 2 × | LRR ( d ) i | 1 / 2 × RD ( d ) i | RD ( d ) i | , $$ \begin{aligned} \mathrm{CID} (d)_i = |\mathrm{RD} (d)_i |^{1/2} \times |\mathrm{LRR} (d)_i|^{1/2} \times \frac{\mathrm{RD} (d)_i}{|\mathrm{RD} (d)_i|}\,, \end{aligned} $$(A.4)

where we omitted the dependence on (x, y) for the sake of brevity. Note the last multiplicator ensures the sign of CID(d)i is the same as that of RD(d)i and LRR(d)i. The resulting CID(4), shown in panel (d) of Fig. A.1, produces clear and contrasted images, and retains all evolving large-scale structures. Based on examination of our event, we found that the CID is especially useful for AIA filters such as 171 Å, 193 Å, 211 Å, 304 Å, and areas of higher signal in 131 Å and 335 Å. For the 94 Å, which in our event contains only relatively weak signals, the LRR leads to strong enhancement of noise. For this filter and the present event, we therefore find both LRR and CID inferior to RD.

thumbnail Fig. A.1.

Comparison of three difference methods. Panel (a) shows the original AIA 171 Å observations, while panels (b)–(d) show the running difference RD(4), log running ratio LRR(4), and combined improved difference CID(4), calculated using a time difference of d = 4.

Appendix B: Fe XVIII proxy from SDO/AIA

To isolate the hot emission of flare loops, we use the method of Del Zanna (2013), where the intensity of Fe XVIII is given by

I ( F e x v i i i ) = I ( 94 Å ) − I ( 211 Å ) / 120 − I ( 171 Å ) / 450 , $$ \begin{aligned} I({Fe\,xviii }) = I(94\,\AA ) - I(211\,\AA )/120 - I(171\,\AA )/450\,, \end{aligned} $$(B.1)

where the coefficients were determined by Del Zanna (2013) mostly for observations taken in 2010, that is, relatively close to the AIA launch date. However, EUV instruments are known to degrade in sensitivity with time (e.g., BenMoussa et al. 2013), and AIA is no exception (Boerner et al. 2014). Since its launch, the AIA instrument has degraded significantly (Dos Santos et al. 2021). Particularly affected are the longer-wavelength channels of 211 Å, 335 Å, and 304 Å, while the shorter-wavelength channels less affected. To correct the degradation of the AIA sensitivity, we use the version 10 of the AIA effective areas, available from SolarSoft. We find that the sensitivity of the 94 Å and 171 Å channels has decreased to a factors of about 0.86 and 0.72 relatively to pre-launch sensitivities. Contrary to that, the relative sensitivity of the 211 Å channel on 2020 December 9 is only about 0.44. To calculate the Fe XVIII proxy, the equation (B.1) is corrected here for decrease of sensitivity of individual filters.

All Tables

Table 1.

Fit parameters for the IRIS Si IV FUV spectrum. Individual components plotted in Fig. 8 are indicated.

In the text

All Figures

thumbnail Fig. 1.

GOES B3.2 class flare that peaked at 00:53 UT, and that has a small flare peak at 00:33 as a B1.7 class on December 9, 2020 (panel a). Panel b shows the flare ribbons in the UV temperature overplotted with a magnetic field of strength of ±200 G (green for positive and blue for negative polarity) from the HMI magnetogram. Panels c–g show the evolution of the flare region in the Hα wavelength, where the filament is shown with white arrows. Panel (h) shows the magnetic field configuration at the flare site and neighborhood.
In the text
thumbnail Fig. 2.

Evolution of flare ribbons and the filament eruption as observed in IRIS C II SJIs is depicted. Panels (b) and (c) highlight the erupting filament, with white arrows indicating the filament’s increasing height over time. Panel (d) presents a clear view of the two ribbons forming in the flare region. In panel (e), the labeling of different ribbon regions is provided, as is described in Sect. 3.2. Finally, panel (f) illustrates the falling-back material during the later stages of the event.
In the text
thumbnail Fig. 3.

Overview of the B-class flares and eruption of December 9, 2020, as observed by SDO in three AIA filters at 94 Å, 171 Å, and 304 Å. Note that the Fe XVIII contribution has been separated from the 94 Å filtergrams (see Appendix B). The HMI magnetogram of a portion of the decaying AR is shown in panel (a), and is saturated to ±300 G. One example of an interaction between the filament with a bent arcade (loops) is shown in panels (a4)–(b4) with the evolution of the hot flux indicated by white arrows in panel (a4). An online animation is attached to this figure (Movie_overview_aia.mpg).
In the text
thumbnail Fig. 4.

Evolution of the ribbon PR and its hook PRH, based on manual tracing of the ribbon. Individual colors correspond to individual times, as is indicated. Pink and orange circles indicate two spatial positions of interest. The orange one is outside of the PRH at 00:30 UT, inside at 00:32 UT, and again outside at 00:36 UT. The pink one is first outside PRH, but is swept by the PRH at 00:36–00:40 UT (see text for details).
In the text
thumbnail Fig. 5.

Difference images highlighting the evolution of the filament and coronal loops during AR–RF reconnection. The rectangular IRIS/SJI field of view is shown in red, and the position of the IRIS slit in orange. Panels (a1)–(c1), (a3)–(c3), and (a4)–(c4) show times corresponding to the IRIS spectra studied in Sect. 4. Panels (a2)–(c2) and (a5)–(c5) are shown for context, indicating the footpoints of coronal loops (arcades) A1 and A2 that reconnected with the filament. At the respective times, the loops are dark structures, indicating that they no longer exist. Green arrows at 00:30:21 and 00:32:54 UT indicate locations along the IRIS slit where brightenings are observed within the filament. Animations of the CID images are available online (Movie_CID_IRIS.mpg).
In the text
thumbnail Fig. 6.

Overview of the Si IV spectra of the reconnecting filament. The left panels show the C II SJI images, overlaid with all eight slit positions as vertical dashed lines. The right panels show the corresponding eight Si IV spectra at each slit position. Prominent enhancements of the Si IV 1402.8 Å line, which can also be Doppler-shifted, are shown by blue and red arrows, respectively. Note the similar behavior of the O IV line at 1401.2 Å. An animation of this figure is available online (SJICII_SpectraSiIV_revised.mp4).
In the text
thumbnail Fig. 7.

Detailed Si IV and Mg II spectra of the three reconnection episodes detected at slit position 7 (see Fig. 6). Panels (a)–(c) show the IRIS/SJI observations in C II, similarly to in Fig. 6, with the position of slit 7 is shown with a vertical dashed white line. Panels (d)–(f) show the Si IV detector images, while panels (g)–(i) show the Mg II ones at this slit position at three different times. The horizontal dashed blue line indicates the reconnecting filament spectra with large Doppler shifts plotted in Fig. 8. The blue and red arrows in panels d-e-f indicate the blueshift and redshift flows during the reconnection in Si IV spectra. The white arrow indicates the filament in panel (i).
In the text
thumbnail Fig. 8.

Observed IRIS spectra (black) of Si IV (a–c) and Mg II (d–f) at three different times as shown in previous figures (horizontal dashed blue lines in Fig. 7). Uncertainties of the observed spectra are denoted by thin gray error bars. The fit to the Si IV spectra is shown in orange, with individual components indicated by colored lines (blue and red for blueshifted and redshifted components of the Si IV line, violet for O IV line). Reference spectra for Mg II are shown in green, and are taken in the non-flaring chromosphere to the north of the erupting filament. Vertical pink lines denote rest wavelengths for individual spectral lines. At 00:34:11 UT there are several components in both lines, in particular one corresponding to 250 km s−1 (pink arrow in panel f and the “?” sign in panel c). See text for details.
In the text
thumbnail Fig. A.1.

Comparison of three difference methods. Panel (a) shows the original AIA 171 Å observations, while panels (b)–(d) show the running difference RD(4), log running ratio LRR(4), and combined improved difference CID(4), calculated using a time difference of d = 4.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.