Two-sided Loop Solar Jet Driven by the Eruption of a Small Filament in a Big Filament Channel

Similar to the cases of anemone jets, two-sided loop solar jets can also be produced by either flux emergence from the solar interior or small-scale filament eruptions. Using high-quality data from the Solar Dynamics Observatory, we have analyzed a two-sided loop solar jet triggered by the eruption of a small filament. The jet occurred in a pre-existing big filament channel. The detailed processes involved in the eruption of the small filament, the interaction between the erupted filament and the big filament channel, and the launch of the two-sided loop jet are presented. The observations further revealed notable asymmetry between the two branches of the jet spire: the northeastern branch is narrow and short, while the southern branch is wide and long and accompanied by discernible untwisting motions. We explored the unique appearance of the jet by employing the method of local potential field extrapolation to calculate the coronal magnetic field configuration around the jet. The photospheric magnetic flux below the small filament underwent cancellation for approximately 7 hr before the filament eruption, and the negative flux near the southern footpoint of the filament decreased by about 56% during this interval. Therefore, we propose that the primary photospheric driver of the filament eruption and the associated two-sided loop jet in this event is flux cancellation rather than flux emergence.

Morphologically, solar coronal jets are categorized into two types (Shibata et al. 1994a;Shen 2021): straight anemone jets and two-sided loop jets.Anemone jets have been discussed extensively in past decades.Sterling et al. (2015) refer to them as "single-spire jets" since they only exhibit a single spire.The spire of anemone jet often vertical to the solar surface approximately, and a bright point often occurred at the edge of the jet base.In contrast, two-sided loop coronal jets are bidirectional jets.They consist of two roughly antiparallel spires, which usually develop symmetrically from the eruption source region and extend horizontally to the solar surface.The trigger mechanism of both anemone jets and two-sided loop jets is traditionally explained by the emerging-flux model (Shibata et al. 1994a;Yokoyama & Shibata 1995).In this model, a magnetic bi-pole emerging from the solar interior undergoes magnetic reconnection with the ambient pre-existing open field or farreaching loops, resulting in the formation of jet spires along that open (or far-reaching) field.If the ambient coronal field is oblique or vertical to the solar surface, the produced jet will be an anemone jet; while if the overlying coronal filed is horizontal, a two-sided loop jet will be initiated.

3
Early magnetohydrodynamic (MHD) simulation of emerging-flux model is 2D (Yokoyama & Shibata 1995).Subsequently, Moreno-Insertis et al. (2008) presented the first 3D numerical experiment of the model, and Moreno-Insertis & Galsgaard (2013) extended the work further.Observations also revealed that some anemone jets were triggered by the emergence of magnetic flux and thus validated the emerging-flux model (Shibata et al. 1994b;Zhang et al. 2000;Chen et al. 2008;Li et al. 2015).Furthermore, Moreno-Insertis & Galsgaard (2013) and some other MHD simulations illuminated that a parasitic bipolar magnetic field emerging from the solar interior would naturally evolve into a sheared arcade.This is consistent with observations that flux emergence is often accompanied by photospheric shear motions, allowing for the transfer of magnetic helicity to the chromosphere and corona (Xu et al. 2022).This sheared arcade would subsequently reconnect with the pre-existed ambient field, and produce a so-called blowout jet (Moore et al. 2010).

Yang et al.
On the other hand, numerous observations have revealed that the generation of anemone jets does not always involve the emergence of magnetic flux, but is closely associated with the eruption of filaments/mini-filaments in the source regions (Hong et al. 2011;Yang et al. 2012a;Sterling et al. 2015;Hong et al. 2016;Shen et al. 2017;Duan et al. 2022).Specifically, jets triggered by minifilament eruptions typically manifest as blowout jets (Moore et al. 2010), and magnetic flux cancellation rather than flux emergence often occurs before the eruption of mini-filaments and the launch of jets (Adams et al. 2014;Panesar et al. 2016;Joshi et al. 2018;Panesar et al. 2018;Duan et al. 2019;Chen et al. 2020;Yang et al. 2020).Based on the observations of 20 randomly selected solar coronal jets in polar coronal holes, Sterling et al. (2015) found that each of them was driven by the eruption of a mini-filament, and proposed a revised picture to interpret the production of jets.
Simulations also support the scenario that filament/mini-filament eruption is the source of coronal jet (Wyper et al. 2017(Wyper et al. , 2018)).Using ultrahigh-resolution 3D magnetohydrodynamic simulations, Wyper et al. (2017) demonstrated the production process of a coronal jet through the eruption of a mini-filament or flux-rope structure embedded in open unipolar field, and its subsequent evolutionary stages.Wyper et al. (2018) further studied in detail three realizations of the model to explain different observational features.
As we mentioned above, theory models for producing solar jets are applicable to both anemone jets and two-sided loop jets.However, comparing with anemone jets, the observational studies of two-sided loop jets are insufficient till now.Some cases have shown magnetic flux emerging from below the photosphere, reconnecting with a pre-existing overlying horizontal field, and producing a two-sided loop jet (Kundu et al. 1998;Jiang et al. 2013;Zheng et al. 2018;Tan et al. 2022), thus supporting the emerging-flux model.Beside these, there are also some other high-resolution observations revealed that the eruption of mini-filaments plays an important role in the formation of two-sided loop jets, aligning with the perspective of Sterling et al. (2015) for anemone jets.Tian et al. (2017) presented observational analysis of two successive two-sided loop jets which were produced not by the emergence of magnetic flux but by the reconnection of two filamentary threads.Sterling et al. (2019) presented an event to show strong evidence that a two-sided loop jet resulted from an erupting mini-filament, and the magnetic trigger of the eruption is apparently flux cancellation.The examples from Shen et al. (2019) and Yang et al. (2019) further advanced our understanding of two-sided loop jets.Beside confirming the role of mini-filaments in triggering two-sided loop jets, Shen et al. (2019) revealed that the birth of the two-sided loop jet involved two reconnection processes, while Yang et al. (2019) inferred that the mini-filament could reform at the same neutral line after its first eruption and produce recurrent two-sided loop jets.Particularly, Yan et al. (2021) reported that continuous magnetic cancellations could produce a series of unidirectional and bidirectional jets with the eruption of a hot channel rather than a mini-filament.
The current study focuses on a two-sided loop solar jet that appeared in a big filament channel on the eastern part of the solar disk on March 22, 2012.This event provides a good opportunity to investigate the triggering mechanism of two-sided loop solar jets from a perspective of observation, which has been relatively scarce until now.Moreover, this jet deviates from the typical symmetrical spires seen in many two-sided loop jets, exhibiting obvious asymmetric jet spires.Through an examination of the evolution of photospheric magnetic field and the eruption process of the jet, we aimed to identify the trigger of the jet and the reason behind its distinctive appearance.The paper is arranged as follows: the instruments and the data we used are described in Section 2, Section 3 presents the main observational results along with our explanations for the event, and in Section 4 the summary of the event and a brief discussion are given.
The 304 Å and 171 Å images are especially important when analyzing the eruptions of the jet in this event, but the data from other AIA EUV passbands are also examined.Full-disk line-of-sight Yang et al.
(LOS) magnetograms obtained by the Helioseismic and Magnetic Imager (HMI; Scherrer et al. 2012) on board the SDO were used to investigate magnetic field in the source region of the jet.These magnetograms were obtained in the Fe i absorption line at 6173 Å with a spatial sampling of 0 ′′ .5 pixel −1 , a cadence of 45 seconds, and a noise level of 10 G.All the data are processed by using standard software programs in the SolarSoftWare (SSW), and differentially rotated to a reference time close to the event.In addition, to exhibit the site and appearance of the filament channel, we also use Hα line-center images obtained by the Global Oscillation Network Group (GONG; Harvey et al. 1996) at the National Solar Observatory (NSO).The pixel size of the GONG Hα images is 1 ′′ , and the time resolution is 1 minute.

OBSERVATIONAL RESULTS
On March 22, 2012, a two-sided loop solar coronal jet was observed by SDO/AIA in the eastern part of the solar disk, near the equator.The jet wasn't associated with any recorded flares or coronal mass ejections (CMEs).Upon careful examination of GONG Hα observations, we found that a big quiescent filament (QF) was presented in the same region approximately one solar rotation cycle earlier (panel (d)), and the shape of QF resembles the dark structure seen in panels (a) and (b).However, on the date when the two-sided loop jet occurred, the quiescent filament QF had already disappeared, leaving behind only some fragmented remnants, which are pointed out by the white arrows in the GONG Hα image (panel (c)).Therefore, the dark structure we see in AIA 171 and 304 Å images should be a filament channel (FC).A simultaneous zoomed-in 171 Å image is inserted in panel (a) to provide a clearer view of the jet's appearance.As depicted, the jet consists of a bright jet base and two almost antiparallel spires.However, unlike typical two-sided loop solar coronal jets, where the two branches of the jet spire are usually symmetrical, the jet presented here exhibits asymmetric branches of the jet spire, i.e., the northeastern spire is short and narrow, while the southern spire is long and wide.
Further examination of the source region of the jet (marked by white boxes in panels (b) and (c)) reveals the presence of a small filament F before the jet occurred (panels (e-g)).In the 04:40:20 UT AIA 304 Å image (panel (f)), F presented itself as a hook-like dark structure, with the length of about 47

′′
.54 or 3.447 × 10 4 km.However, it is not easy to identify this small filament in AIA 171 Å and GONG Hα images at the same time.F became more apparent a few minutes later in these wavebands (see panels (e) and (g)), and its shape and the length are differed from what was observed in the 04:40 UT AIA 304 Å image (the outline of F in this image is overlaid as black curves in panels (e) and (g)).By superimposing the contours of an almost simultaneous HMI magnetogram onto the 304 Å image (panel (f)), it was deduced that the northern foot-point of filament F rooted in the positive magnetic field region, while its southern foot-point rooted in the negative region.
Additionally, the photospheric magnetic fields beneath F's southern foot-point appeared to be highly sheared.This statement is further supported by the HMI magnetogram superposed with F's outline in panel (h).
Figure 2 presents the eruption process of the small filament F and the following two-sided loop jet.
As indicated by the black rectangle in panel (c1), the zoomed-in AIA 304 and 171 Å images of the jet's source region are presented in panels (a1-b4).The filament F is identifiable in these images.It is noticed that, at 04:40:20 UT, a patch of brightening existed beneath the south end of F already (panel (a1)).When tracking AIA and HMI observations before the time, it is inferred that the brightening had been present for about one day since 04:00 UT on March 21, and the photospheric magnetic field in the region has experienced continuous flux cancellations.Therefore, we suppose that the brightening would be a coronal bright point (CBP), which usually appeared in quiet-Sun regions and coronal holes, associated with opposite magnetic polarities in the photosphere, and had an average lifetime of about 28 hours in EUV images (Hong et al. 2014).However, the CBP did not contain any filament until 04:00 UT on March 22.In the accompanying animation of Figure 2 which starts at 04:00 UT on March 22, it is noticed that F was unrecognizable at the very beginning, and then it started to form a few minutes later.Consistent with panel (a1), at 04:40 UT F could be identified clearly at the animation.Therefore, although AIA images did not got sufficient resolution  (2015), confirms once again that two-sided loop solar jets could also be produced by the eruption of small scale filaments.It is consistent with recent observations of two-sided loop jets (Shen et al. 2019;Sterling et al. 2019;Yang et al. 2019).It is obvious that the northeastern spire of the jet is short and narrow, resembling a standard jet, while the southern spire is long and wide, resembling a blowout jet (Moore et al. 2010).In this sense, the jet is particular, since most two-sided loop jets typically have roughly symmetrical spires (Jiang et al. 2013;Sterling et al. 2019;Yang et al. 2019).
Additionally, the northeastern spire of the jet appears bright, indicating it is composed mainly of hot plasma, while the southern spire is a mixture of cold and hot mass, and exhibits noticeable untwisting motion (see the accompanying animation).AIA animation accompanying Figure 2 showed that the southern branch of the jet spire underwent a significant untwisting motion.This character was even more apparent in the time-slice plot along slit S2 (panel (d)).From the time-slice, it was evident that some complicated transverse motions took place in the jet's southern spire soon after the fast ejection of the jet.However, the width of the jet spire did not show a significant change, suggesting that the jet spire experienced only rotational motion without expansion.Untwisting motion is a common character in blowout jets (Hong et al. 2013;Moore et al. 2015;Li & Yang 2019;Chen et al. 2021), and it has been explained as the release of the twist initially stored in the twisted filament's magnetic field through magnetic reconnection with the ambient coronal magnetic field.Two episodes of the untwisting motion were traced by the dotted white lines in panel (d), with linear fitting velocities of 295.37 km s −1 and 211.46 km s −1 .These rotational velocities were relatively large compared to many previous studies (Zhang & Ji 2014;Moore et al. 2015;Yang et al. 2023b), but still within a reasonable range.The time-slice further showed that the rotational direction of the southern spire of the jet was along the forward direction of S2, from the east to the west, according to the white arrow in panel (c4) of  Yang et al.
Using the perfect observations from SDO/AIA and HMI, this paper presents an analysis of a twosided loop solar coronal jet.The jet occurred in a large filament channel located in the southeast part of the solar disk.AIA observations revealed that a pre-existing small filament erupted in the source region just a few minutes before the jet occurred, and captured the detailed processes of how the erupted small filament ejected into the adjacent filament channel and triggered the two-sided loop jet.The observations confirmed that, besides the widely accepted emerging-flux model (Shibata et al. 1994a;Yokoyama & Shibata 1995), two-sided loop solar jets can also be produced by the magnetic reconnection between the overlying magnetic field of an erupting small filament and the ambient coronal magnetic field, just like in the cases of anemone jets (Sterling et al. 2015;Hong et al. 2016;Li et al. 2017;Yang et al. 2023a).The observations further showed that two branches of the jet's spire are heavily asymmetrical, with the southern branch being wide and long while the northeastern branch is narrow and short.By investigating the coronal magnetic configuration around the jet, we suspect that this unique appearance is due to the asymmetric coronal magnetic configuration near the end of the filament channel.
The trigger mechanism of solar coronal jets is an important issue for the studies of both anemone jets and two-sided loop jets.While the event presented here maintains the picture of Sterling et al. (2015), in general, either the emerging-flux model (Shibata et al. 1994a;Yokoyama & Shibata 1995;Moreno-Insertis et al. 2008;Moreno-Insertis & Galsgaard 2013) or the mini-filament-eruption-driven picture (Sterling et al. 2015;Wyper et al. 2017Wyper et al. , 2018) ) has been supported by many observation cases respectively, and the structures and kinematic characteristics of jets produced by these two mechanisms did not show much difference from one to the other (Schmieder et al. 2022;Shen 2021).
Therefore, it seems that in a pre-existing background field, both the direct emergence of magnetic flux from the interior and the eruption of filament/mini-filament could effectively trigger reconnection between two magnetic systems and produce a solar jet.
For the jets that conform to the emerging-flux model, flux emergence is always observed and regarded as a key driver in photosphere (Shibata et al. 1994b;Zhang et al. 2000;Chen et al. 2008;Li et al. 2015).However, in the cases that jets are driven by erupting mini-filaments, flux cancella-tion is typically predominant rather than flux emergence (Sterling et al. 2015;Panesar et al. 2016;Sterling et al. 2019;Yang et al. 2019Yang et al. , 2023a)).In some cases, it seems that both flux emergence and cancellation worked together (Jiang et al. 2007;Shen et al. 2017Shen et al. , 2019)).In our event, the positive magnetic flux in the source region of the jet did not change much, but the negative flux decreased monotonously for about 7 hours before the eruptions of the small filament and the following two-sided loop jet.Therefore, although could not exclude the contribution of flux emergence completely, we still propose that the main photospheric driver of this two-sided loop jet is flux cancellation, consistent with previous studies on jets triggered by mini-filament eruptions.A study by Panesar et al. (2016) of 10 random solar jets in quiescent regions found that each one experienced flux reduction between 21% and 57% at the neutral line below a mini-filament.In another investigation of 13 random solar jets in coronal holes, the authors found that the flux reduction ranged from 21% to 73%, with a calculated average cancellation rate of 0.6 × 10 18 Mx hr −1 (Panesar et al. 2018).The value of average cancellation rate is 1.5 × 10 18 Mx hr −1 for the quiescent region jets they studied previously (Panesar et al. 2016).In comparison, the active region jets studied by Sterling et al. (2017) had an average cancellation rate of 15 × 10 18 Mx hr −1 .The two-sided loop jet in our event occurred in a quiescent region, with magnetic flux reduction of 56% and an average cancellation rate of 2.1 × 10 18 Mx hr −1 .Therefore, it seems that the flux cancellation in this event was a little more violent than the quiescent region jets studied by Panesar et al. (2016), but was still within a reasonable range.
This may be due to the small filament in this event being slightly larger than the mini-filaments in their events, thus more free magnetic energy are required to trigger the eruption of the filament.
Similar to filaments eruption of ordinary scale, the eruption of small scale filaments and the following coronal jets occasionally result in flares and narrow CMEs (Yang et al. 2020).The jets can also interact with other magnetic structures during their ejections.More recently, Joshi et al. (2023) reported two observational events of large-amplitude filament oscillations triggered by solar coronal jets.In both events, a jet formed near one end of a large filament due to flux emergence.The ejected jet hit the filament channel from the end, leading to large-amplitude oscillation in the filament, which supported the MHD numerical experiments recently conducted by Luna & Moreno-Insertis (2021).
Yang et al.
In the present event, the jet also occurred near the end of a large filament channel.However, during the jet's ejection, the filament channel contained little filament plasma, as indicated by the GONG Hα image in Figure 1.Therefore, the interaction between the jet and the filament channel was not observed in this event.
In summary, this paper presents the observations of a two-sided loop solar jet in a large filament channel with asymmetrical spires.The jet was triggered by the eruption of a small filament, and the main driver of the filament eruption and the jet was the continuous cancellation of positive and negative magnetic fluxes in the photosphere.Compared with anemone jets, there is still limited research on two-sided loop jets, and more observations with high spatial and temporal resolution are needed in the future to further understand their characteristics and underlying mechanisms.
Therefore, these numerical experiments provided an improved version of the original emerging-flux model.Accordingly, new observations also reported to support the model.Recently, Schmieder et al. (2022) reviewed the research results of surges and jets observed by the Interface Region Imaging Spectrograph (IRIS; De Pontieu et al. 2014).Several cases discussed in their review reported the jets produced by flux emergence, aligning well with the improved emerging-flux model.One of these cases was presented by Ruan et al. (2019).Utilizing IRIS spectrographic observations, they reported bi-directional flows with velocities of nearly ±200 km s −1 in an active region, and attributed these flows to the outflows of magnetic reconnection occurring between emerging flux and a long twisted loop.The magnetic reconnection produced a twisted jet as well.Joshi et al. (2020b) analyzed another twisted jet with the help of IRIS spectrographic observations.They have displayed the detailed process of how the twist transferred from a flux rope located in the reconnection site to the jet.Particularly, Joshi et al. (2020a) studied six recurrent EUV jets observed by SDO/AIA and IRIS simultaneously.They found each of these jets presented a double-chambered structure with cool and hot emissions in each vault, corresponding to the cool and hot loop regions predicted by the emerging-flux models of Moreno-Insertis et al. (2008) and Moreno-Insertis & Galsgaard (2013).
Figure 1 displays the location and the general appearance of the jet and its source region.It's worth noting that panel (d) of Figure 1 was captured on February 23, 2012, while other images in this figure were obtained on March 22.As showed by the AIA 171 and 304 Å images (panels (a) and (b)), the jet located at the northeastern end of a large curved dark structure.
Yang et al.    to reveal either the exact time or the detailed process of F's formation, we could infer that F formed between 04:00 UT and 04:40 UT on March 22.The new-formed small filament F was obvious and stable at 04:40:20 UT as showed in the AIA 304 Å image (panel (a1)), but it soon started to lift soon.When superimposing the original position of F (blue curves, obtained from panel (a1)) to other 304 Å images in the top row, it is clear that F erupted southeastwardly, approaching the big filament channel FC.The CBP also became brighter and more expanded, evolved into a small flare eventually.The subsequent movement of the erupting filament is more clearly visible in AIA 171 Å images (panels (b1)-(b4)).As pointed out by the white arrows in panels (b1) and (b2), F erupted towards the pre-existing filament channel FC, and its appearance changed a lot compared to the original form in AIA 304 Å image (blue curve).The overlying magnetic field of erupting F would reconnect with the magnetic field of FC, resulting in the launch of a jet.Dark filament plasma was observed to be ejected into FC, and some brightenings occurred at the interface of them, which were likely signatures of the magnetic reconnection (panels (b3) and (b4)).Then, bright mass flow streamed from a bright point near the flare and flowed southwardly into FC.The observed bright mass flow indicates the initiation of the jet, while the complete appearance and development of the jet were displayed in panels (c1)-(c4) of AIA 171 Å images with a larger field of view (FOV) .In these pictures, it is evident that the bright jet plasma flowed both southwardly and northeasterly, forming two roughly antiparallel jet spires, and the small flare associated with the filament eruption evolved into the jet bright point (JBP).Based on the appearance of the jet and the trajectory of F's eruption, it can be deduce that the jet was a two-sided loop solar jet triggered by the eruption of the small filament F. Therefore, this event supports the picture proposed bySterling et al.
Panel (d) of Figure 2 displays the normalized AIA 304 and 171 Å light curves measured in the white boxes in panels (a4), (b3), and (c4).The light curves show the evolution of the intensity of the CBP (or the flare, or JBP).It is noticed that the brightness of the patch began to increase sharply at around 04:45 UT, indicating the beginning of F's eruption.Both the 304 and 171 Å light curves reached their first peak simultaneously at approximately 04:49 UT, as indicated by the vertical dotted line.This peak should correspond to the maximum intensity of the small flare produced by the filament eruption.Subsequently, the intensities of the 304 and 171 Å images in the region decreased for a while, then increased again and reached the their maximum value.This second stage of increasing of intensity may indicate the launch of the jet, and the maximum of the light curve corresponds to the maxima of the JBP caused by the jet.Note that the time of maximum intensity differs when measured in different wavebands: the 171 Å light curve reached its peak at about 05:11 UT, as indicated by the vertical dashed line, while the 304 Å light curve reached its peak at about 05:07 UT, a few minutes before the maximum of the 171 Å light curve.To analyze the evolution of the jet in detail, time-slice plots were constructed from AIA 304, 171, and 131 Å images and displayed in Figure 3.In this figure, panels (a)-(c) show the time-slices constructed along slit S1, while panel (d) presents the time-slice along slit S2.Slit S1 is a curved line segment along the jet spire, pointing from the northeast to the south, as shown by the white arrows in panels (c3) and (c4) of Figure 2. Slit S2, on the other hand, is a straight line segment roughly vertical to the southern branch of the jet spire.In panel (b) of Figure 3, the horizontal dashed line represents the locations of the demarcation point of the jet spire at different times, corresponding to the blue asterisks in Figure 2. Therefore, in this image, the region above the dashed line describes the southward ejection of the jet, while the region below the line shows the northeastern ejection of the jet.The ejections along the two branches of the jet spire show notable differences.The southward Yang et al. ejections are generally more vigorous, with a larger amount of plasma being ejected.Furthermore, the southward ejections are steady and continuous, consisting of both dark and bright components, indicating the ejection of cold filament mass into the southern branch of the jet spire.All three time-slices along S1 (panels (a-c)) show that the mass ejection along the southern branch of the jet spire can be divided into two stages: a slow ejection stage with an average projected speed of about 18.09 km s −1 , as measured in 171 Å time-slice, and a fast ejection stage with an average projected speed of about 165.05 km s −1 .This behavior is similar to the eruption of a filament, which aligns with what we inferred from Figure 2 that the jet was triggered by a small filament eruption.Zhang et al.(2021) also found that the kinetic evolution of an anemone jet triggered by a mini-filament eruption can be divided into a slow rise phase and a fast rise phase, and they suggested that these two phases may correspond to the magnetic reconnections at the breakout current sheet and the flare current sheet, respectively.According to the 304 Å time-slice (panel (a)), the fast ejection of the jet started at about 05:00:29 UT, as indicated by the vertical dashed lines in the figure.On the other hand, the northeastern ejections are intermittent and primarily consist of the bright component.Three episodes of intermittent northeastern mass ejection are marked by red dotted lines in panel (b), with measured average speeds of 179.32 km s −1 , 97.53 km s −1 , and 171.52 km s −1 , respectively.Indeed, the speeds of the intermittent northeastern ejections are roughly comparable to the fast ejection of the southern branch.However, the two-stages mass ejection was indistinguishable in the northeastern branch of the jet spire, possibly because the ejection was weak there.From the time-slice plots, the presence of the cold filament mass in the northeastern ejections may not be apparent, but the accompanying AIA 171 Å animation shown in Figure2reveals that, although the majority of F's mass was ejected into the southern branch of the jet spire, a small portion of the dark filament mass was also ejected into the northeastern spire of the jet.

Figure 2 .
Figure 2.This direction is consistent with what we see in the AIA 171 Å animation.Therefore, the southern spire of the jet experienced an anticlockwise rotation when viewed from JBP.

Figure 4
Figure4and its accompanying animation displayed the situation of the photospheric magnetic field before and during the eruptions of the filament F and the following two-sided loop jet.It is revealed that, in the source region of the jet (as indicated by the black circles in the top half part of Figure4), the positive and negative photospheric magnetic polarities moved toward and canceled to each other since 21:30 UT on March 21.After 01:00 UT on March 22, the cancellation occurred mainly in the region indicated by the white ovals in the figure and the animation.When superposing the outline of the small filament F obtained from the 04:40 UT AIA 304 Å image onto the simultaneous magnetogram showed in panel (a8), it is clear that the region in the white oval is where F's southern foot-point rooted.HMI LOS magnetograms revealed that the negative polarity patch of the photospheric magnetic field (the black patch) continuously moved to northwestward, while the positive polarity patch (the white patch) moved southeastward.After several hours of movement and cancelation, the photospheric magnetic field below F's southern foot-point became highly sheared (panel (a8)), causing the southern part of the filament to raise first, followed by the entire filament's eruption.

Figure 1 .
Figure 1.General appearance of the two-sided loop solar jet and its circumstances (top row) and its source region (bottom row) showed by AIA 171 Å, 304 Å, GONG Hα line-center images, and HMI LOS magnetogram.The field of view (FOV) of the top row is 635 ′′ × 585 ′′ , while the FOV of the bottom row is 78 ′′ × 72 ′′ , which is indicated by the white boxes in panels (b) and (c).A zoomed-in image of the two-sided loop jet is inserted in panel (a) to highlight it more clearly.The outline of the axis of the small filament F obtained from panel (f) is overlaid as black curves in the other panels of the bottom row.The LOS magnetic field measured in the 04:40:16 UT HMI magnetogram are superimposed on the AIA 304 Å image (panel (f))

Figure 2 .
Figure 2. AIA 304 and 171 Å images illustrate the eruption of the filament F (panels (a1)-(b4)) and the associated two-sided loop jet (panels (c1)-(c4)).Some panels are overlaid with the outline of F's axis obtain from the 04:40:20 UT 304 Å image as blue curves.The long curved white arrow in panel (c3) and the white arrow in panel (c4) indicate the slit positions of the time slices shown in Figure 3, while the blue asterisks in panels (c2) and (c3) show the approximate position of the demarcation point of the jet's spires.The black rectangle in panel (c1) indicates the FOV of panels (a1)-(b4), which is 78 ′′ × 72 ′′ , the same as the bottom row of Figures 1.The FOV of panels (c1)-(c4) is 260 ′′ × 240 ′′ .Panel (d) presents the normalized light curves of AIA 304 and 171 Å intensities in the region indicated by the white boxes in panels (a4), (b3), and (c4).(An animation of AIA 171, 193, 304 Å direct images and GONG Hα line-center images is available.The animation lasts 15 seconds, has the same FOV as panels (c1)-(c4), and covers the time from 04:00 UT to 05:29 UT.The cadence of the AIA images is 12s, while the cadence of the GONG images is 1m.)

Figure 3 .
Figure 3.The time-space plots in AIA 304, 171, and 131 Å wavebands.Panels (a)-(c) present the timeslices along slit S1, while panel (d) presents the time-slice along slit S2.The positions of these slits are showed in Figure 2c.The vertical dashed lines indicate the start time of the jet, while the horizontal dashed line denotes the position of the demarcation point of the jet's spire.The dotted lines in panels (b) and (d) track the mass ejections in both branches of the jet spire and the transverse motion in the southern branch, respectively, with the measured projected velocities presented also.

Figure 4 .Figure 5 .
Figure 4. Evolution of the photospheric magnetic field in the source region of the two-sided loop jet.(a1-a8) HMI LOS magnetograms show the cancellation between the positive and negative magnetic polarities.The red curve in panel (a8) represents the outline of F's axis obtained from the simultaneous AIA 304 Å image, and the FOV is the same as that of the bottom row of Figure 1.(b-c) Unsigned positive and negative magnetic fluxes calculated in the areas indicated by the black circles and the white ovals in panels (a1-a8).The blue vertical lines mark the start time of F's eruption.(An animation of HMI LOS magnetograms is available.The animation lasts 27 seconds, has the same FOV as that of panels (a1)-(a8), and covers the time from 21:30 UT of March 21 to 05:29 UT of March 22.The cadence of the HMI magnetograms is 45s.)