Observational Evidence of S-web Source of the Slow Solar Wind

NOAA AR 12967 rotated onto the solar disk on 2022 March 12 (from the viewpoint of Earth). Its leading polarity contained a small, coherent positive sunspot bordered to the north and south by dispersed positive field. This configuration is shown in the magnetogram from the Helioseismic and Magnetic Imager (HMI; Scherrer et al. 2012) on board the Solar Dynamics Observatory (SDO; Pesnell et al. 2012) in Figure 1(a). By 2022 March 18, the sunspot was no longer visible in the HMI magnetograms. The following negative polarity was fully dispersed from the time it was first observed at the east limb, suggesting that the AR was already in its decay phase. At central meridian passage, the AR's total unsigned magnetic flux was ∼3.5 × 10²¹ Mx, typical for a medium-sized AR with a lifetime measured in weeks rather than months (van Driel-Gesztelyi & Green 2015).

The distribution of the photospheric magnetic field remained broadly unchanged throughout the AR's transit across the solar disk (see the online animation associated with Figure 1). Its following negative polarity field was embedded in the surrounding positive field in the southern section of the AR. In the AR's northern section, the dispersed positive field extended toward the northern polar CH.

Images from the 304 Å (log T = 4.7), 171 Å (log T = 5.8), 193 Å (log T = 6.2 and 7.3), 211 Å (log T = 6.3), and 94 Å (log T = 6.8) channels of SDO's Atmospheric Imaging Assembly (AIA; Lemen et al. 2012) are displayed in Figures 1(b)–(f). The multipanel image is taken from the online animation associated with this figure. The movie shows the chromospheric/coronal evolution of NOAA AR 12967 from 2022 March 14 12:00 UT to 2022 March 20 23:56 UT. Each image has a field of view (FOV) of 900'' × 850'' encompassing NOAA AR 12967 and the surrounding dispersed magnetic field. Bright core loops connecting the AR's main polarities are visible in the hotter AIA channels. Less bright, longer loops extend from around the positive sunspot to the negative fields of the nearby NOAA AR 12965 in the west and quiet Sun in the south. Loops rooted along the channel of positive field in the north appear to connect with the negative polarity field of NOAA AR 12967 on the eastern edge and negative field of NOAA AR 12965 on the western edge, forming a dark, narrow corridor that extends from the positive sunspot toward the north polar CH.

The dark corridor is a striking feature in the hotter 193 Å, 211 Å, and 94 Å channels (see arrows in Figure 1) but less evident in the chromospheric 304 Å and upper transition region/lower corona 174 Å channels. As the AR rotates across the solar disk, projection effects mean that some of the corridor is obscured. This is especially the case in the 171 Å channel, where there is a mixture of short, low-lying loops and long fan loops in the vicinity of the corridor. However, when the AR is close to central meridian on 2022 March 17, the full length of the corridor is clearly observed in the 193 Å, 211 Å, and 94 Å channels. In these channels, the corridor significantly widens from where it starts at the positive sunspot until it blends into the northern polar CH extension. The corridor persists and remains remarkably stable during the period of 5.5 days covered by the included movie, even after a filament eruption along the northern external polarity inversion line at 2022 March 16 ∼13:00 UT.

Hinode/EIS provided coordinated observations during the slow wind connection SOOP. EIS tracked the positive polarity of NOAA AR 12967 from 2022 March 18 to 21. Study ID 600/acronym DHB_007_v2 was employed for a total of 12 observations. The 2'' slit was rastered in 4'' steps taking 60 s exposures at each of the 62 pointing positions. A single observation took 62 minutes to build an FOV of 248'' × 512''. The north−south extent of the FOV covered the upflow region over the positive polarity and the dark corridor until the spacecraft pointing was shifted further south on 2022 March 20. Cotemporal observations from SO's Spectral Imaging of the Coronal Environment (SPICE; SPICE Consortium et al. 2020) instrument do not cover the corridor; therefore, they are not included in this analysis.

All spectroscopic data were processed with the EIS Python Analysis Code (EISPAC; https://github.com/USNavalResearchLaboratory/eispac/) software package using Python compatible level1 HDF5 files available at https://eis.nrl.navy.mil. The Fe ion emission lines were fitted with either single- or double-Gaussian functions applying the template files included in EISPAC. Single-Gaussian functions were fitted to all lines with the exception of the blended Fe xi 188.21 Å, Fe xii 195.12 Å, and Fe xv 284.16 Å lines.

The top panel of Figure 2 shows a sample of the Fe xii 195.12 Å emission-line intensity, Doppler velocity, and nonthermal velocity maps at 2022 March 19 03:29 UT, when the full length of the dark corridor is evident in the EIS observations. This time at the Sun is roughly equivalent to when the SO spacecraft first encountered the slow SW stream taking into account the magnetic connectivity and the propagation time of the SW (see Section 6). Consistent with the observed stability of the dark corridor in images from the hotter AIA channels, the Doppler and nonthermal velocities for each of the 12 observations are very similar to those shown in the top panel of Figure 2. Blueshifted upflows in the range = [−20, 0] km s⁻¹ are observed in the Doppler velocity map over the positive polarity of NOAA AR 12967, along the dark corridor of the positive field. Nonthermal velocities within the upflow regions are in the range = [0, 44] km s⁻¹.

Zoom In Zoom Out Reset image size

Table 1 contains measurements of the plasma and magnetic field parameters for subregions defined in the Fe xii 195.12 Å Doppler velocity map in the top right panel of Figure 2. Only those pixels containing Doppler upflows were considered for each parameter; pixels with downflows were masked. Values were determined using the summed spectra within a region for each emission line to improve the signal-to-noise ratio. The parameters are First Ionization Potential (FIP) bias, Doppler and nonthermal velocities, plasma density using Fe xiii ions, temperature of the peak differential emission measure (DEM), and magnetic flux density (from the corresponding HMI magnetogram). The temperatures corresponding to the peak DEMs and Si x 258.37 Å – S x 264.22 Å FIP bias were calculated using the method of Brooks & Warren (2011) and Brooks et al. (2015).

Region A is located in the relatively broad northern section of the corridor. The broader width of the region and its low magnetic flux density are more consistent with CH or polar CH extensions. As expected, it has photospheric FIP bias with low upflow and nonthermal velocities (e.g., Harra et al. 2015; Fazakerley et al. 2016). The density is slightly lower in A than in regions B and D and similar to the other two regions; however, there is little contrast in density among the five regions. The temperatures corresponding to the peak DEMs are similar across the regions with log T in the range = [6.0, 6.2]. Where the corridor is narrow in region B, the FIP bias and the upflow and nonthermal velocities are comparable to regions C–E within the AR. However, the magnetic flux there is in between CH and AR values. Overall, the plasma parameters measured in the corridor are closer to those found at the edges of ARs (e.g., Doschek et al. 2008; Harra et al. 2008; Brooks & Warren 2011, 2016) than in CHs.

The images from different AIA channels show that the dark corridor is expanding with height/temperature in the solar atmosphere. In order to verify whether this is the case, EIS intensity maps of emission lines sensitive to narrower temperature ranges are plotted for each ion from Fe viii 185.21 Å to Fe xv 284.16 Å at 2022 March 18 21:18 UT in the bottom panel of Figure 2. At this time, the corridor is close to the solar central meridian so that projection effects are minimized. The Fe emission lines represent log T in the range = [5.8, 6.3] (see the values written at the bottom of the maps in the figure). The corridor is not clearly observed in the Fe viii intensity map. In the northern section of the Fe ix map, the dark corridor becomes evident, though it is short and narrow; it continues to lengthen and widen in the intensity maps as the log T increases from 5.9 to 6.3.

A potential field source surface (PFSS) extrapolation was used to investigate the large-scale coronal magnetic configuration of NOAA AR 12967 and the dark corridor. The PFSS extrapolation of the photospheric magnetic field at 2022 March 18 18:03 UT was created using the IDL SolarSoft PFSS package (Schrijver & De Rosa 2003). With this package, PFSS extrapolations are generated using the evolving surface flux transport (SFT) model based on assimilated HMI data and known methods for evolving the observed magnetic flux. The SFT model tracks unipolar flux elements that are subject to differential rotation and meridional flows. Flux elements will coalesce and cancel. For each time step, the positions and flux amounts of the elements are output. The HMI magnetogram image resolution is then binned down by a factor of four, and all pixels within 60° of Sun center are converted to flux elements. A full-Sun flux map based on the HMI magnetogram inserted into the SFT model is used as the input into the PFSS code to create a 384 × 192 full-Sun magnetogram. Spherical decomposition is performed on the magnetogram to reconstruct the magnetic field as a function of height. The top of the PFSS model is the source surface, set at a radial distance R_ss, defined here to be at 2.5 R_⊙.

Various viewpoints of the extrapolation are shown in Figure 3. Open-field lines are overplotted on the Hinode/EIS Fe xii 195.12 Å Doppler velocity map embedded in an SDO/HMI LOS synoptic map of CR 2255 closest in time to that of the EIS raster. Open-field lines are located along the corridor of blueshifted upflows in the Doppler velocity map. The top view in panel (a) includes field lines computed along the central section of the open-field channel, and the side view in panel (b) shows the coherence along the channel with only moderate meridional expansion in the northern and southern sections. A strong lateral expansion is present as shown in panel (c), where two rows of field lines are drawn on the sides of the open-field corridor.

Zoom In Zoom Out Reset image size

Two cross sections of the PFSS extrapolation orthogonal to the corridor, one in the north and the other in the south, show that the open-field lines are bordered on either side by closed loops (Figures 4(a), (b)). The field extends radially in the center of the corridor and then diverges away from the radial direction closer to the edges/closed loops.

Zoom In Zoom Out Reset image size

The expansion properties of the open field are quantified by computing the expansion factor:

$$\begin{eqnarray}&&f(R)=\displaystyle \frac{{R}_{\odot }^{2}}{{R}^{2}}\displaystyle \frac{B({R}_{\odot })}{B(R)},\end{eqnarray} \tag{ 1 }$$

where R is the radial distance from the Sun center, R_⊙ is the solar radius, and B(R_⊙) and B(R) are the magnetic field strength at R_⊙ and R. With magnetic flux conservation, f(R) provides the flux tube cross-section expansion corrected for the spherical expansion. With R = R_ss, the radius at the source surface (set at 2.5 R_⊙), f(R_ss) is the expansion factor used in previous studies of SW sources (e.g., Wang & Sheeley 1990; Wang et al. 2009).

The expansion factor f(R) is plotted at three representative heights (1.2, 1.5, and 2.5 R_⊙) in Figures 4(c) and (d) for the northern and southern cross sections of the corridor. Cyan squares along each field line correspond to these heights. The X-axis in each plot is the distance in Mm in the photosphere of each field line from the leftmost (easternmost) one. In general, the width of the open-field corridor decreases, and the asymmetry of the closed field increases from north to south. The northern slice has an expansion that is smaller in the central part with respect to the lateral sides. At the edges, the open flux can expand rapidly when the top of the lateral closed field is reached. This magnetic field behavior is quantified in Figure 4(c). The plot of f(R) shows a smaller expansion at low heights while reaching larger values at greater heights on both sides.

Expansion properties change as the surrounding closed field and the width of the corridor vary (see abscissae at the bottom of panels (c) and (d)). An extreme in field-line asymmetry is reached along the southern slice (Figure 4(b)), where the small size of the eastern arcade field allows an expansion of the corridor field at lower heights (≲1.2 R_⊙ in Figure 4(d)). In contrast, the much larger western arcade field limits the corridor expansion at low heights, but when the maximum height of the closed field is exceeded (approx 1.5 R_⊙), the open field has a bigger volume into which it can expand until it reaches pressure balance with the nearby open field. This leads to larger f(R_ss) on the western side than on the eastern side.

Previous studies have shown that different expansion profiles, in particular f(R_ss), have direct implications for the acceleration profiles of the SW, especially its terminal speed (Wang et al. 2009; Pinto et al. 2016; Pinto & Rouillard 2017). Depending on which flux tube is crossed by SO, different SW properties, such as plasma speed, are to be expected with in situ measurements.

The open flux corridor is bordered by a drastic change in magnetic connectivity (Figure 4), with field lines changing from reaching the source surface to reaching back to the photosphere as closed field. Such a magnetic configuration is typical of the presence of a separatrix and more generally of QSLs. Initially, QSLs were defined in closed-field regions to understand the location of flare ribbons in magnetic configurations where there are no magnetic bald patches or null points present, i.e., where the magnetic topology is a priori as simple as a magnetic arcade. In fact, for ARs, typically thin coronal volumes are present where the magnetic connectivity changes are extreme. These thin regions define QSLs (Démoulin et al. 1996). In the limit where the connectivity change is discontinuous, a QSL reduces to a separatrix. A convenient way to identify strong QSLs is by using the squashing factor, Q, which quantifies the change in the magnetic connectivities at a boundary (Titov et al. 2002). More generally, Q is a geometric measure of the degree to which neighboring field lines diverge in a 3D volume.

The concept of QSLs was extended to open/closed magnetic configurations by considering the magnetic connectivity between two surfaces, e.g., the photosphere and the source surface (Titov et al. 2011), as was previously done for straightened coronal configurations (e.g., Milano et al. 1999). Since field lines extending outward from the source surface are supposed to follow a Parker spiral, there is no change in the Q-maps for a surface more distant than the source surface.

Synoptic maps of Q for each Carrington rotation (CR) are available from the SDO's Joint Science Operations Center (JSOC; see http://hmi.stanford.edu/QMap/). Figure 5(a) shows the SDO/HMI synoptic magnetogram for CR 2255, with the photospheric footpoints (r = R_⊙) of the open field in red/blue designating positive/negative radial B field components. The main ARs on the Sun near NOAA AR 12967 are labeled. Open positive field encompasses the negative polarity of NOAA AR 12967 in the configuration of a classic pseudo-streamer (Wang et al. 2007). In the northern direction, there is an extension of the northern polar CH indicated by the black arrow in the figure. It is plausible that the extension is connected to the open-field region by a thin channel (Antiochos et al. 2011) that is not detectable owing to the low spatial resolution of the connectivity computations.

Zoom In Zoom Out Reset image size

Even though magnetic synoptic maps have highly degraded spatial resolution compared to the original HMI magnetograms and with the spatial variations further smoothed with height by a potential field extrapolation, the synoptic Q-map at a height of 1.05 R_⊙ in Figure 5(b) gives a sense of the complexity of the magnetic connectivities. This is due to the fact that the photospheric magnetic fields of numerous sources are linked together. Local magnetic flux imbalances, such as with the negative polarity of NOAA AR 12967, are typically associated with closed magnetic connections to the surrounding field of the dominant opposite polarity. It results in a round flower-like Q pattern, with the "petal" boundaries defined by high Q values (see, e.g., Figure 5(b) and Figures 6(d) and (e)). The boundaries separate the central polarity connecting to surrounding opposite polarities of larger magnetic flux. This is the case around the negative polarity of NOAA AR 12967, and large-scale examples are also present in Figure 5(b) (e.g., NOAA AR 12965 and NOAA AR 12970). Boundaries of high Q are the same as the giant arcs of the S-web mentioned in the Introduction. The Q-maps in Figures 5(b) and (c) can be compared to Figure 7(a) of Antiochos et al. (2011), Figure 3 of Scott et al. (2018), and Extended Data Figure 4(c) of Chitta et al. (2023), revealing comparable patterns of Q.

Zoom In Zoom Out Reset image size

When Q-maps are computed at greater heights, the flower boundaries retract as shown in Figure 5(c) (R = 1.75 R_⊙). At such heights, all of the closed-field regions have disappeared except for a large-scale arcade that encircles the Sun (below the heliospheric current sheet). Without the presence of the ARs, this would correspond to an equatorial arcade with a streamer above it. In this active period, the arcade is warped to significant latitudes by the AR flux distribution. The rest of the sphere surface at 1.75 R_⊙ in Figure 5(c) is covered by open fields coming from different photospheric sources, so they are separated by high Q values. The photospheric linkage is indicated by labels around NOAA AR 12967. At the source surface and farther away, only these domains of different connectivities remain. In summary, Q-maps show the extension of the various sources of the SW (which come into contact at the high Q locations).

Figure 6(a) is a zoomed-in version of the synoptic magnetogram centered on the AR target and its surroundings. The open-field corridor introduced in Section 3.2 is the thin strip of positive field on the western side of the AR that broadens in its northern extension. The closed-field region surrounding the negative polarity of NOAA AR 12967, present in Figure 6(b), disappears in the map of Figure 6(c), which corresponds to a height difference of 0.1 R_⊙. For other AR polarities, the dome of closed field can extend to much greater heights such as the eastern and western ARs (NOAA AR 12971 and NOAA AR 12965), where the round QSLs are still present at r = 1.5 R_⊙ (right and left sides of Figure 6(e)). It is the retraction of the closed field with height that allows the broad lateral expansion of the open field present in between the ARs. Figure 6 provides a complementary view to the lateral expansion of the open field shown in Figure 4.

In the Q-map at r = 1.1 R_⊙ (Figure 6(c)), the negative polarity of NOAA AR 12967 is fully covered by open field forming a triangular-like shape. The open-field region continues to expand up to r = 2.5 R_⊙ (Figures 6(d)–(f)). After starting as a narrow region at r ≈ R_⊙, the corridor broadens greatly in the longitudinal direction, in contrast to only marginally increasing in the latitudinal direction. This is in agreement with the observed broadening of the corridor with increasing temperature and height in the Hinode/EIS ion intensity maps (Figure 2). The significant broadening with height is related to the large expansion factors found in Figure 4. As a consequence, a spacecraft orbiting close to the solar equatorial plane is expected to cross a wide range of longitudes with plasma originating from the corridor even if remote observations only show a thin latitudinal extent of the source region (Figures 1, 2). Indeed, the extent of the magnetic field at r = 2.5 R_⊙ is up to about 60° of longitude (Figure 6(f)), which is comparable to the typical distance between neighboring ARs in Figure 5(a).

SO tracked the decaying NOAA AR 12967 during its solar disk crossing and obtained both remote sensing and in situ observations. SO was also magnetically connected to NOAA AR 12967 and its vicinity for about 5 days. A distinguishing feature of the AR was a dark and narrow corridor observed in the EUV that expanded with height and temperature in the corona (Figures 1, 2). Plasma upflows of approximately 10 km s⁻¹ ran along the full extent of the corridor as observed by Hinode/EIS. NOAA AR 12967's magnetic field structure has a classical pseudo-streamer configuration with the following negative polarity surrounded by positive field (its leading polarity counterpart in the southwest and remnant fields elsewhere, Figures 5–6). The open positive field in the west is spatially coincident with the plasma upflows of the narrow corridor, implying that the upflows are outflows (Figures 3–4, 9).

In situ measurements at SO show that the SW stream emanating from the narrow corridor had exceptionally slow radial velocities of ∼300 km s⁻¹ with correspondingly low proton temperatures of about 5 eV and increasingly high proton density from ∼60 to 300 cm⁻³ in the time interval defined by boundaries B1–B2 (Figure 8). For approximately 1/3 days during this interval, the slow SW exhibited moderate Alfvénicity with the correlation coefficient between V and B, ∣C_vb∣, in the range [0.7, 1]. This period was preceded and followed by more variable and non-Alfvénic periods, especially closer to the B1 and B2 boundaries.

Alfvénic slow SW was first reported by Marsch et al. (1981) using Helios 2 measurements at 0.3 au. Subsequent in situ observations confirmed the presence of Alfvénic slow SW streams at ∼0.3 au (e.g., D'Amicis & Bruno 2015; Bale et al. 2019; Stansby et al. 2019; Perrone et al. 2020; Stansby et al. 2020) and at 1 au (e.g., D'Amicis et al. 2011, 2019; Wang & Ko 2019; D'Amicis et al. 2021). Alfvénicity is a key discriminator for classifying the slow SW and identifying its source regions, especially at close distances like that of SO at its perihelion (Stansby et al. 2019; D'Amicis et al. 2021). In this case, it was especially important in the absence of in situ composition data.

While proton temperatures and magnetic field flux densities are consistent with the average values measured by Helios within non-Alfvénic slow SW at a distance of 0.3–0.35 au, the velocity and density parameters detected at SO are extreme (see, e.g., Table 1 of Perrone et al. 2020). In fact, they are comparable to the values of the so-called very slow SW described by Sanchez-Diaz et al. (2016), where velocities <300 km s⁻¹ are associated with much higher and more variable densities in contrast with those reported by Perrone et al. (2020). The very slow SW was observed by Helios 1 and 2 at distances mainly <0.5 au and not above 0.7 au. Sanchez-Diaz et al. (2016) proposed that the origin of the very slow SW may be small and isolated CHs, i.e., equatorial CHs, that become more widespread during active periods of the solar cycle.

Exceedingly high densities combined with very low velocities provide crucial information for further refining the source region–spacecraft connectivity during the B1–B2 interval. SW velocities of ∼300 km s⁻¹ are more likely to originate at the southern end of the narrow corridor, where f(R_ss) is an order of magnitude higher compared to that of the north (see Figures 4(c) and (d)). Furthermore, the southern end of the corridor is adjacent to the core of NOAA AR 12967. Typically, the core loops connecting opposite polarities have the highest density (e.g., Tripathi et al. 2008) and coronal FIP bias (e.g., Del Zanna & Mason 2014; Baker et al. 2015, 2021) values within the various loop populations of ARs. The in situ plasma during B1–B2 is more characteristic of the core loops in the south compared to the CH and quiet Sun found at the northern end of the corridor. Finally, the positive polarity of NOAA AR 12967 is a natural location for reconnection to take place between core loops and nearby open field (Baker et al. 2009), so that the plasma can be transferred from the closed core loops to nearby open field.

A comparison of the plasma parameters and magnetic field within the narrow corridor and its associated SW stream with that of equatorial CHs provides insight as to where they are located on the scale of CHs. Wang & Ko (2019) compiled the basic properties of 14 equatorial CHs and their associated slow wind streams (∼350–450 km s⁻¹) at 1 au from 2014 to 2017 (during solar maximum/beginning of decay phase of cycle 24). They found the following: (1) longitudinal widths ∼ [3°, 10°], (2) relatively high Alfvénicity ∣C_vb∣ ∼ [0.6, 0.7] that tended to fall sharply approaching the edges or boundaries of the slow SW streams, (3) footpoint magnetic field strengths ∼[15, 30] G, (4) proton temperatures ∼[1, 3] eV, and (5) large expansion factors >9. These results, where applicable, correspond to those of D'Amicis et al. (2019), who identified equatorial CHs as the source regions of highly Alfvénic slow SW at 1 au.

The comparable values for key properties of the narrow corridor and its associated very slow SW are more extreme than those of equatorial CHs: the longitudinal width of the corridor in the 1.5 MK corona is up to 2 orders of magnitude smaller, the expansion factors are as much as an order of magnitude larger in the south, and the SW velocities are lower by ∼100 km s⁻¹. Alfvénicity in the B1–B2 interval possibly reaches higher levels compared to the equatorial CHs in the study of Wang & Ko (2019); however, the extent of Alfvénicity appears to be shorter. The extremely high expansion factor related to this short period of Alfvénicity is expected, as there is low expansion only in the center of the flux tube compared to the edges. Closer to the edges, the field becomes more curved so that more of the Alfvén waves are reflected back rather than traveling outward along the open field.

The narrow corridor observed by Hinode/EIS has key features predicted by the S-web model. First, at typical coronal temperatures, it is a few orders of magnitude thinner in the longitudinal direction compared to typical equatorial CHs that were confirmed source regions of the slow SW. Second, the corridor is conjectured to be topologically robust (Antiochos et al. 2007, 2011). In this case, the dark corridor was observed to be highly stable for at least 7 days. Third, the corridor expands to a latitudinal width of 60° at the source surface (=2.5 R_⊙). Fourth, the overall magnetic configuration is a pseudo-streamer formed of the embedded negative polarity of NOAA AR 12967 with the corridor formed within the surrounding positive field on the west. Open-field corridors are intrinsically associated with pseudo-streamers (Antiochos et al. 2011; Crooker et al. 2012; Scott et al. 2019).

Some of the extreme properties of the slow SW stream observed by SO are a direct consequence of the very nature of the S-web comprising open-field corridor source regions. The very slow SW recounted by Sanchez-Diaz et al. (2016) is plausibly produced by such regions. Extremely narrow corridors are characterized by super-radial expansion and therefore very slow wind velocities. Following on from this, Alfvénic slow SW is associated with overexpanded flux tubes (see D'Amicis et al. 2021, and references therein). ∣C_vb∣ was in the range [0.7, 1] for 1/3 days during the central part of the B1–B2 interval. Large variability in B_R and occasional reversal events, i.e., switchbacks, occurred while the very slow SW was moderately Alfvénic. Recent PSP and SO observations have established that switchbacks are linked to the Alfvénic slow wind from equatorial CHs (e.g., Bale et al. 2019; D'Amicis et al. 2021; Fedorov et al. 2021). It is reasonable that the induced interchange reconnection events along the S-web arcs (Higginson et al. 2017b) are a possible source of the switchbacks ever present in the inner heliosphere (D'Amicis et al. 2021, and references therein) and observed in the Alfvénic very slow SW emanating from the narrow open-field corridor.

The corridor associated with NOAA AR 12967 is not rare. On the contrary, the full synoptic HMI magnetogram shown in Figure 5(a) contains a number of possible S-web corridors identified by the yellow arrows. There is a spectrum of widths, mainly as a consequence of the B_R distribution in the photosphere. The narrowest of corridors will be beyond the spatial resolution of the magnetic models used to determine the open-field regions. Furthermore, limitations of EUV observations, e.g., instruments with limited FOV and projection effects (corridors are masked by nearby brighter regions), also contribute to the mistaken impression that they are not common features.

In conclusion, the combination of observations from SO and Hinode/EIS has made it possible to characterize the S-web narrow open-field corridor source region and its slow SW stream. The intrinsic topology of the corridors means that super-radial expansion of the open field is likely to yield extremely slow SW velocities, with moderate Alfvénic content (along the central section of the corridor), and switchbacks. Other SW properties such as composition, charge state, and density are more likely to be governed by the surroundings of the corridors as interchange reconnection at the corridor boundaries opens up pathways for closed-field plasma of the nearby quiet Sun and ARs to reach the heliosphere.