ARE IRIS BOMBS CONNECTED TO ELLERMAN BOMBS?

We identify IBs based mainly on the FUV spectra. We first select all substantial and compact brightenings in the Si iv 1393.755 Å intensity image. Line profiles in these brightenings are then inspected. IBs are defined as those brightenings with chromospheric absorption lines superimposed on greatly broadened and enhanced non-Gaussian profiles of transition region lines (e.g., Si iv and C ii). The absorption feature needs to be obvious for at least the Ni ii 1393.330 and 1335.203 Å lines. Based on this criterion, we have identified 10 IBs sampled by the slit and their locations are indicated in Figure 1(E).

Figure 3 presents images of the Si iv 1393.755 Å intensity and line width, Mg ii k core and wing, SJI 1400, 1330 and 2796 Å, ${{\rm{H}}}_{\alpha }$ core and wings, line of sight component of the photospheric magnetic field, and AIA 1700, 171 and 193 Å in a small region enclosing each identified IB. We can see that all IBs are associated with an enhanced line width of the Si iv line. IBs are also clearly observed in both the SJI 1400 and 1330 Å images, which can be understood since the Si iv and C ii lines are usually greatly enhanced in IBs. Signatures of IBs are less obvious in the SJI 2796 Å images, suggesting that the Mg ii k line does not always have a relevant response.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The four IBs identified by Peter et al. (2014) are all associated with mixed magnetic field polarities and at least one of them shows a clear signature of flux cancellation during the observation. Vissers et al. (2015) also found association of a few IBs with strong opposite-polarity fluxes. The HMI data presented in Figure 3 reveal a similar pattern: most IBs appear to be sitting at the magnetic field polarity inversion lines. This likely suggests energization of these IBs through interaction between strong fluxes with opposite polarities, likely anti-parallel magnetic reconnection. The association with strong opposite-polarity fluxes is not obvious for IBs 5, 8, and 9, which might be consistent with the third of the three scenarios proposed by Georgoulis et al. (2002). This scenario involves interaction between the emerging vertical fields and pre-existing horizontal fields. In this case the line of sight components of the interacting magnetic fluxes are not necessarily opposite in polarity. However, it might also be related to the line of sight effect as the observed region is close to the limb.

We also try to identify possible coronal signatures of IBs from both the IRIS spectra and AIA images. We find no obvious emission of the IBs in both the Fe xii 1349.38 Å and Fe xxi 1354.08 Å lines (not shown here). We realize that both lines are forbidden lines, so that the absence of these lines might be caused by the large density in the source regions of IBs. However, AIA observations in the 171 Å (dominated by emission from Fe ix/Fe x) and 193 Å (dominated by emission from Fe xii) passbands also reveal no obvious brightening at locations of most IBs. A possible exception is IB 3, where a very intense loop-like brightening can be identified from both the 171 and 193 Å images. We thus conclude that IBs are generally not heated to coronal temperatures.

Typical line profiles of the 10 IBs are shown in Figures 4–9, where the line profiles averaged over a quiet plage region (Solar-X = [−7885 : −7756], Solar-Y = [−2228 : −1852]) are also overplotted for reference. The reference line profiles are also used to perform absolute wavelength calibration. For the Si iv 1393.755 Å spectral window, the chromospheric Fe ii 1392.817 Å and Ni ii 1393.330 Å lines in the reference spectrum are assumed to have zero Doppler shifts. These cold lines are known to show negligible average velocities in quiet regions. For the Si iv 1402.770 Å window, in principle we can assume a zero shift of the S i 1401.514 Å line. However, this line appears to be weak and also too close to the O iv 1401.156 Å line. Instead, we calibrate the wavelength for this window by forcing the Si iv 1402.770 and 1393.755 Å lines in the reference spectra to have the same Doppler shift. For the C ii window, the Ni ii 1335.203 Å line is assumed to have zero shift. The wavelength calibration has been confirmed by the similar Doppler shifts of the Ni ii 1335.203 and 1393.330 Å absorption lines in the IBs. Absolute wavelength calibration in the Mg ii window is much easier since many strong neutral absorption lines are present in the reference spectrum. These neutral lines can be safely assumed to have zero shifts.

As expected, all the IBs show greatly enhanced and broadened profiles of the Si iv, C ii, and Mg ii lines, although the Mg ii signature appears to be less obvious for some IBs. Obvious enhancement in one or both wings of the Si iv lines, which are usually believed to be associated with reconnection outflows (bidirectional jets or unidirectional jets, Innes et al. 1997), can be clearly identified for most bombs. These line profiles are generally similar to those of transition region explosive events (EEs), which are also believed to result from reconnection (e.g., Dere et al. 1989; Innes et al. 1997; Chae et al. 1998; Madjarska et al. 2004; Ning et al. 2004; Teriaca et al. 2004; Zhang et al. 2010; Huang et al. 2014; Gupta & Tripathi 2015). The most distinct difference between EEs and IBs may be the formation height: EEs are formed in the transition region, and IBs are formed lower down as suggested by the chromospheric absorption lines. It is also likely that some previously identified EEs are actually IBs. As demonstrated by Yan et al. (2015a), the narrow absorption features in the FUV spectra of IRIS may not be unambiguously resolved by previous moderate-resolution instruments such as the Solar Ultraviolet Measurements of Emitted Radiation instrument (Wilhelm et al. 1995) on board the Solar and Heliospheric Observatory. So from spectra taken with these instruments, one cannot distinguish between EEs and IBs. We also see obvious absorption features at the cores of the two Si iv lines for IBs 4 and 6. The intensity ratio of the two Si iv lines is ∼1.55 for these two line profiles. This ratio is much smaller than the ratio derived from the reference line profiles, which is 1.93 and close to the expected value 2 in optically thin cases. The dip appears to be stronger in the 1393.755 Å line. These results suggest that the central absorption feature likely results from self-absorption of the Si iv lines rather than bidirectional jets (Yan et al. 2015a). It is unclear why the opacity effects become prominent in these IBs.

Some singly ionized absorption lines are clearly superimposed on the enhanced line profiles of Si iv and C ii. These lines include Ni ii 1335.203 Å, Fe ii 1392.817 Å, Ni ii 1393.330 Å, Fe ii 1403.101 Å, and Fe ii 1403.255 Å, with the Ni ii lines being the strongest. They are clearly emission lines in the quiet plage region. Under ionization equilibrium these lines are typically formed at a temperature of log (T/K) ≈ 4.15 in the upper chromosphere. In IBs they generally reveal a blueshift less than 10 km s⁻¹, although a large blueshift of ∼20 km s⁻¹ is found for IB 9. It is believed that these absorption lines result from the largely undisturbed upper chromosphere and they suggest the presence of hotter gas (up to the Si iv formation temperature ∼8 × 10⁴ K) below the upper chromosphere (Peter et al. 2014; Vissers et al. 2015).

The Mg ii 2798.809 Å line is a self blend of two lines at 2798.754 and 2798.822 Å. In most observations they appear as absorption lines, but they come into emission above the limb and in energetic phenomena such as flares (e.g., Tian et al. 2015). Pereira et al. (2015) undertook a forward modeling study of this line using three-dimensional radiative magnetohydrodynamic models, and found that it changes from absorption to emission when strong heating occurs in the lower chromosphere. Interestingly, this line shows enhanced wings in all IBs except IBs 1 and 9. The shape of this line is similar to the Mg ii k line at 2796.347 Å and Mg ii h line at 2803.523 Å for most IBs. This feature was also noted by Vissers et al. (2015).

We now examine the relationship between IBs and EBs. Since EBs are usually defined from ${{\rm{H}}}_{\alpha }$ images, we first examine possible signatures of IBs in the ${{\rm{H}}}_{\alpha }$ core and wing images taken with NVST. It has already been demonstrated that the NVST ${{\rm{H}}}_{\alpha }$ data can be very useful when studying filaments and dynamic events in the lower solar atmosphere (e.g., Bi et al. 2015; Yan et al. 2015b, 2015c; Yang et al. 2015a, 2015b; Xue et al. 2016). The ${{\rm{H}}}_{\alpha }$ data acquired during our joint observation are good enough for the identification of EBs. The contrast on the NVST images appears not constant over time due to the varying seeing conditions, which has little effect on our identification of EBs since the ${{\rm{H}}}_{\alpha }$ wing brightenings discussed below are mostly very strong.

From Figure 3 we find that some IBs appear to be associated with EBs and others are not. IBs 1–4 are clearly not connected to EBs. No obvious brightening can be identified from either the ${{\rm{H}}}_{\alpha }$ wing or core images for IBs 1, 2, and 4, while for IB 3 significant brightening can only be identified from the ${{\rm{H}}}_{\alpha }$ core image. These signatures are not EB signatures. IBs 5–7 reveal typical signatures of EBs, which include substantial brightening in both wings of ${{\rm{H}}}_{\alpha }$ and no obvious enhancement in the ${{\rm{H}}}_{\alpha }$ line core. The latter suggests that these IBs lie below the chromospheric fibril canopy visible in the ${{\rm{H}}}_{\alpha }$ core images. The wing brightenings clearly exceed those from the much more ubiquitous facular bright points (pseudo-EBs or network bright points, Rutten et al. 2013). We notice that IB 7 reveals as a long bright structure in the Si iv intensity image. Its northern part coincides with significant enhancement in both wings of ${{\rm{H}}}_{\alpha }$. However, only the red wing of ${{\rm{H}}}_{\alpha }$ shows a less prominent enhancement in the southern part. The SJI 1400 Å images in the online movies suggest that these two parts might be related to two different brightenings. Despite this possibility, we still regard these two neighboring parts as one IB since they could not be separated in the Si iv intensity image. In the following we select one position at each of the two parts in IB 7 for line profile analysis (positions a and b marked in Figure 3). IBs 8–10 are possibly EBs. These three IBs all show no obvious core enhancement and clear red wing enhancement. However, the blue wing enhancement is less obvious for IB 8 and is not present for IB 9, which may result from the obscuration by spicules. From the online movie associated with Figure 2, we see that this is indeed the case for IB 9. The blue wing shows no obvious signature when a spicule is launched nearby. However, the blue wing reveals a significant enhancement after the spicule fades away. No significant blue wing enhancement can be identified for IB 10. We notice that IB 10 is an anemone jet (Shibata et al. 2007) revealing an obvious inverted-"Y" morphology (Figure 2). The jet reveals as a bright collimated structure in the SJI 1400 Å images and a dark one in the ${{\rm{H}}}_{\alpha }$ blue wing images, similar to the quiet-Sun network jets or rapid blueshifted excursions (Tian et al. 2014a; Rouppe van der Voort et al. 2015). We select three positions in IB 10 for detailed analysis of the line profiles: a at the upward jet, and b and c at the two legs of the inverted-"Y" structure. The line profiles at positions b and c discussed below suggest that IB 10 is possibly an EB.

Peter et al. (2014) found that the O iv 1401.156 and 1399.774 Å lines are absent in their four identified IBs. In our data we find that the O iv lines are absent in some IBs but clearly present in other IBs. We have calculated the intensity ratio of the Si iv 1402.770 Å and O iv 1401.156 Å lines for each line profile presented in Figures 4–8, and found that the ratio is larger than 60 for IBs 5–10 (positions b and c for IB 10). The absence of the O iv line emission has been previously found in sub-arcsecond bright dots in the transition region above sunspots (Tian et al. 2014b) and small-scale brightenings at the footpoints of hot loops (Testa et al. 2014). Olluri et al. (2013) found that non-equilibrium ionization can lead to the absence of the O iv lines. The O iv lines can also be greatly suppressed in the presence of non-Maxwellian electron distributions (Dudik et al. 2014). A third explanation for the absence of these forbidden lines is the dominance of collisional de-excitation from the meta-stable level over radiative decay in a high density environment (Feldman & Doschek 1978; Young 2015). Our result appears to support the third scenario because the absent or weak O iv lines are all found in EB-related IBs or possible EB-related IBs. Since EBs are a pure photospheric phenomenon (e.g., Watanabe et al. 2011), the density should be very high and likely high enough to suppress the radiative decay. On the other hand, IBs 1–4 are not EBs and the strong O iv lines probably suggest a relatively lower density at the formation height. Since the Ni ii 1335.203 and 1393.330 Å absorption lines are also present, these IBs should be located below the upper chromosphere and thus are likely in the lower or middle chromosphere. Using the intensity ratio of the O iv 1401.156 and 1399.774 Å lines, we have derived the electron densities of these IBs under the assumption of ionization equilibrium (CHIANTI v7.1, Landi et al. 2013). The derived densities for IBs 1–4 are in the range of log (N_e/cm⁻³) = 11.2–11.9, which also suggests the formation of these IBs in the chromosphere. Thus, our results are not necessarily inconsistent with Judge (2015), who placed the origin of IBs in the low-middle chromosphere or above.

The Mg ii k and h lines also show distinctly different behavior for the EB-related IBs and other IBs. From Figures 2 and 3 we can conclude that these Mg ii lines may also be used to identify EBs. For IBs 5–10, we see significant brightening in the Mg ii k wing (–1.33 and +1.33 Å images are similar and thus summed) but no obvious brightening in the Mg ii k line core. These IBs are exactly the ones connected or possibly connected to EBs. While for IBs 1, 2, and 4, we could not identify any significant brightening from either the wing or core images of Mg ii k. At the edge of IB 3 brightening is seen from both the Mg ii k wing and core images. If we replace Mg ii k by ${{\rm{H}}}_{\alpha }$, these properties would indicate that these events are not EBs, which is exactly what we concluded above. The different Mg ii k and h line profiles for these two types of IBs can also be seen from Figures 4–9, where we generally see significant enhancement of the Mg ii wings and no dramatic change of the line cores for IBs 5–10. The NUV continuum between the Mg ii k and h lines is also enhanced for IBs 5–10, with the only exception at position b of IB 7. While for IBs 1–4, there is no substantial enhancement of the Mg ii wings beyond ∼–1.33 Å/+1.33 Å from the line cores. In addition, the NUV continuum is even suppressed or only slightly enhanced for these IBs. Since the Mg ii k and h cores sample the chromospheric fibrils and wings are formed lower down, the different behavior mentioned above confirms that IBs 5–10 are likely also EBs formed in the photosphere and that IBs 1–4 are not. The enhanced NUV continuum in IBs 5–10 also supports our argument that these IBs are likely generated in the photosphere. Our finding suggests that the Mg ii k and h lines may be used similarly to the ${{\rm{H}}}_{\alpha }$ line for the identification and investigation of EBs. This would open a very promising new window for EB studies since the Mg ii k and h data are routinely acquired in the seeing-free IRIS observations. To elaborate this we plan to perform a more detailed analysis using more coordinated observations between IRIS and NVST in the near future.

We also notice that the chromospheric absorption lines such as Ni ii 1393.330 and 1335.203 Å, which are superimposed on the broadened and enhanced wings of the Si iv and C ii lines, are usually very strong in EB-related IBs. These absorption features are generally much weaker (shallower) for other IBs. This difference may also indicate that the EB-related IBs are formed deeper in the atmosphere, thus experiencing stronger absorption at the wavelengths of these chromospheric lines.

Another interesting feature unique to the EB-related IBs is the superposition of the Mn i 2795.640 Å absorption line on the greatly enhanced blue wing of the Mg ii k line. This feature appears to be very obvious for IBs 5, 6, 7 (position a), 8, and 10 (positions b and c). Similarly, we also see the Mn i 2799.093 Å absorption feature superimposed on the enhanced red wing of the Mg ii 2798.809 Å line for IBs 5, 7, and 10 (position b). All these IBs are connected or possibly connected to EBs, as we mentioned above. Since the Mn i lines are formed in the upper photosphere, their absorption lines superimposed on the greatly enhanced Mg ii lines suggest the formation of these hot IBs below the cooler upper photosphere. This again supports our argument that these IBs are also EBs. Vissers et al. (2015) also found this feature for a few EBs. Similar to us, they also attributed these absorption lines to the foreground upper-photosphere gas above the EBs. These NUV absorption lines generally have no obvious Doppler shift, although a small blueshift of Mn i 2795.640 Å appears to be present for IBs 7 (position a) and 8. This confirms that the upper-photosphere gas is generally not impacted by these EB-related IBs, which are a pure photospheric phenomenon.

From Figures 6, 7, and 9 we see that the shape of the S i 1401.514 Å line profile is similar to those of Mg ii k, Mg ii h, and Mg ii 2798.809 Å for IBs 5, 6, 7 (position a) and 10 (position b), revealing enhanced wings bridged by a dip at the core. In the quiet plage region, this line is simply a weak emission line and its shape is close to Gaussian. We find that the C i 1354.288 and 1355.844 Å lines reveal a similar behavior for these IBs, while the optically thin O i 1355.598 Å line is still very narrow. Figure 10 shows the profiles of these lines for IBs 5, 6, 7, (position a) and 10 (position b). For wavelength calibration the O i 1355.598 Å line in the quiet reference spectrum is assumed to have a zero shift. The dramatic change of the S i and C i line profiles in these EB-related IBs is likely caused by the opacity effect. A recent two-cloud model of Hong et al. (2014) reveals an increase of the optical depths when EBs occur, which may result from direct heating in the lower cloud or illumination by enhanced radiation on the upper cloud. The S i and C i line profiles we report here might be related to these processes.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Finally, from Figure 3 we notice that the EB-related IBs mostly show strong and compact brightening in AIA 1700 Å images, while this is less obvious for other IBs. It seems that IB 8 is not associated with any compact brightening in the AIA 1700 Å image shown in Figure 3. However, an inspection of the associated online movie suggests that this is because IB 8 was scanned during its decaying phase (at 03:02 UT). In its early phase, at around 02:57 UT, this bomb shows a strong and compact brightening in AIA 1700 Å. Since the AIA 1700 Å passband samples emission mainly from the upper photosphere, it is not surprising that the chromospheric IBs have little response in the AIA 1700 Å passband. We notice that strong EBs have been found to show obvious brightenings in both the 1700 Å (Vissers et al. 2013) and 1600 Å passbands (Qiu et al. 2000), which is consistent with our finding. Rutten (2016) mentioned that the AIA 1700 and 1600 Å channels are also good EB diagnostics because at high temperature they are dominated by the Balmer continuum which shares the ${{\rm{H}}}_{\alpha }$ properties.

We summarize the characteristics of the 10 IBs discussed above in Table 1. These observational signatures indicate that IBs 1–4 are independent of EBs and that IBs 5–10 are likely connected to EBs. For IB 7 the characteristics at position a are shown in the table. As we mentioned above, the northern (around position a) and southern (around position b) parts of IB 7 might be related to two different brightenings in AIA 1700 Å images. We find that position a shows all characteristics typical of EB-related IBs, while position b reveals some differences: the NUV continuum is not enhanced and the S i 1401.514 Å line does not reveal a central reversal. However, the O iv lines are weak and the Mn i 2799.093 Å absorption feature is superimposed on the enhanced wing of Mg ii 2798.809 Å. In addition, the Mg ii k and h wings are enhanced, although not as significant as the enhancement at position a. These characteristics suggest that the southern part is also possibly connected to an EB. IB 10 appears to be an anemone jet (Shibata et al. 2007). The characteristics at the footpoints (positions b and c) of the jet, including the absence of the O iv lines, the superposition of the Mn i absorption lines on the enhanced Mg ii wings, the centrally reversed S i line, and the intense AIA 1700 Å brightening, suggest that it is likely produced in the photosphere and thus is connected to an EB. These features are absent at the tip of the jet (around position a), consistent with the upward flow location higher up in the atmosphere. In Table 1 we only show the characteristics at position b for IB 10.

Table 1.  Characteristics of the 10 Identified IBs

IB	${{\rm{H}}}_{\alpha }$ Enhancement	Si iv /O iv	Mg ii k and h Enhancement	NUV Continuum Enhancement?	Mn i Absorption on Mg ii k Wing?	Ni ii Absorption	AIA 1700 Å Brightening	S i Broadened with Reversal?
1	no	7	no	no	no	moderate	diffuse, weak	no
2	no	49	no	no	no	weak	diffuse, weak	no
3	core	7	wings, core	no	no	weak	diffuse, weak	no
4	no	32	no	no	no	weak	diffuse, weak	no
5	wings	154	wings	slightly	yes	strong	compact	yes
6	wings	69	wings	yes	yes	moderate	compact, strong	yes
7	wings	498	wings	yes	yes	strong	compact, strong	yes
8	red wing	129	wings	yes	yes	strong	diffuse, weak	no
9	red wing	143	wings	yes	no	strong	compact, strong	no
10	red wing	239	wings	slightly	yes	moderate	compact, strong	yes

Note. Columns are as follows: ${{\rm{H}}}_{\alpha }$ enhancement, intensity ratio of the Si iv 1402.770 Å and O iv 1401.156 Å lines, Mg ii k and h enhancement, enhancement of the NUV continuum between the Mg ii k and h lines, superposition of the Mn i 2795.640 Å absorption line on the greatly enhanced blue wing of Mg ii k, chromospheric absorption lines such as Ni ii 1393.330 Å and 1335.203 Å, brightening in the AIA 1700 Å passband, and broadened S i 1401.514 Å line with a central reversal. For IBs 7 and 10 characteristics at positions a and b are shown, respectively. These observational signatures indicate that IBs 1–4 are independent of EBs and that IBs 5–10 are likely connected to EBs.

Download table as:  ASCIITypeset image

Our finding that some IBs are also EBs greatly challenges previous modelings of EBs, which almost unexceptionally predict a temperature increase by only a few hundred to ∼3000 K in the photosphere or lower chromosphere (e.g., Fang et al. 2006; Isobe et al. 2007; Bello González et al. 2013; Berlicki & Heinzel 2014; Hong et al. 2014). Our observations suggest the need for models which can produce a much stronger temperature increase in the photosphere. A recent numerical simulation by Ni et al. (2015) predicted heating of the chromospheric plasma to ∼8 × 10⁴ K, which might explain our IBs 1–4. It would be interesting to move the reconnection sites down to the photosphere and investigate the associated heating process. By assuming local thermodynamic equilibrium (LTE) for line extinctions during the hot and dense EB onsets, recently Rutten (2016) put the formation of the Si iv lines in a temperature environment of 1–2 × 10⁴ K. This temperature, although lower than ∼8 × 10⁴ K under the assumption of ionization equilibrium, is still much hotter than that predicted by all non-LTE modeling of EBs.

So far we have concluded that some IBs are connected to EBs and other are not. One may then ask another question: does every EB correspond to an IB? From our observation it is clear that not all EBs are connected to IBs. It is impossible to examine the IRIS spectra for all EBs since many EBs were not sampled by the IRIS slit. In addition, sometimes the distinction between EBs and magnetic concentrations (network bright points) is not easy due to the varying seeing condition. Nevertheless, a rough examination of the ${{\rm{H}}}_{\alpha }$ wing images suggests that the IRIS slit crossed ∼30 possible EBs, among which only six show typical IB-type line profiles. So it seems that only a small fraction of the EBs are heated to IB temperatures. The other EBs appear not to be efficiently heated, and thus do not show IB signatures in the Si iv lines. We notice that these events usually also display wing enhancement in the Mg ii k and h lines, although for many of them the enhancement appears to be weaker compared to that for IB-related EBs. In the near future we plan to perform more coordinated observations between IRIS and NVST, and perform a detailed comparison between the unbinned Mg ii and ${{\rm{H}}}_{\alpha }$ data for EB detection.