Interaction of a coronal mass ejection and a stream interaction region: A case study

We investigated the interaction of a coronal mass ejection (CME) and a coronal hole (CH) in its vicinity using remote-sensing and 1AU in situ data. We used extreme-ultraviolet images and magnetograms to identify coronal structures and coronagraph images to analyze the early CME propagation. The Wind spacecraft and the Advanced Composition Explorer (ACE) provide plasma and magnetic ﬁeld data of near-Earth interplanetary space. We applied various diagnostic tools to the images and to the time-series data. We ﬁnd that the CME erupts under a streamer and causes the evacuation of material at its far end, which is observable as dimming and subsequent CH formation. The CME is likely deﬂected in its early propagation and travels southwest of the Sun-Earth line. In situ data lack signatures of a large magnetic cloud, but show a small ﬂux rope at the trailing edge of the interplanetary CME (ICME), followed by an Alfvénic wave. This wave is identiﬁed as exhaust from a Petschek-type reconnection region following the successful application of a Walén test. We infer that the two spacecraft at 1AU most likely traverse the ICME leg that is in the process of reconnection along the heliospheric current sheet that separates the ICME and the high-speed stream outﬂowing from the CH


Introduction
Coronal mass ejections (CMEs) in the vicinity of coronal holes (CHs) are known to be deflected from their expected trajectory in the undisturbed case (e.g., Jiang et al. 2007;Gopalswamy et al. 2009a).The magnitude of the deflection can vary depending on the configuration and properties of large-scale magnetic structures in the solar corona, that is, CHs and active regions, ARs (e.g., Kay et al. 2017;Sahade et al. 2021).Coronal dimmings (e.g., Dissauer et al. 2018) can be used as a diagnostics for the erupting coronal magnetic field.The eruption of a CME under the footpoint of streamer loops can trigger the expulsion of coronal material at its opposite footpoint (e.g., Moore & Sterling 2007), leading to the generation of dimmings, also known as transient CHs (e.g., Gutiérrez et al. 2017).Deflection of CMEs away from the Sun-Earth line can cause the CME to partially or even entirely miss Earth, resulting in different IP signatures (e.g., Möstl et al. 2015).When a spacecraft encounters a shock driving an interplanetary CME (ICME), there is usually an interval of characteristic plasma and magnetic field features after the shock and the compressed solar wind (SW) in front of the ICME (i.e., the sheath) have passed.Most importantly, a nose encounter of the ICME often leads to signatures of magnetic clouds (MCs; Burlaga et al. 1981) or flux ropes (FRs), which are characteristic of weakly fluctuating magnetic field components, one of which shows a smooth rotation, as well as decreased temperature, low plasma β, and bidirectional suprathermal electron fluxes (Kilpua et al. 2017).How pronounced these signatures are depends on the crossing distance of the spacecraft to the FR axis (i.e., the impact parameter), among others."Driverless" hence refers to the complete lack of a characteristic shock like this that drives ICME features.A spacecraft crossing the flank or leg of an ICME observes different properties than a spacecraft that traverses the nose (e.g., Richardson & Cane 2010;Rodríguez-García et al. 2022).In addition to a change in propagation direction of the CME close to the Sun, morphological changes in interplanetary space can occur.This includes FR erosion triggered by magnetic reconnection along a current sheet (e.g., Ruffenach et al. 2015) or the compression of the magnetic structure due to the surrounding SW conditions (e.g., Kilpua et al. 2017;Manchester et al. 2017).Interplanetary magnetic reconnection of Petschek-type is a phenomenon in the SW identified via Alfvénic exhaust jets (i.e., strong, sudden increases of plasma speed) and current sheet crossings along which the correlation or anticorrelation of velocity and magnetic field vector components reverses (e.g., Gosling et al. 2005a).The most prominent current sheet in interplanetary space is the heliospheric current sheet (HCS; e.g., Gosling et al. 2005b).Most importantly, the B x component changes polarity at the HCS and the strahl of the electron pitchangle distribution (PAD) changes from outward to inward oriented or vice versa.The HCS is embedded in the heliospheric plasma sheet (HPS; e.g., Winterhalter et al. 1994) that is identified by enhanced density and reduced magnetic field strength following and/or preceding the crossing of the HCS.Petschek-type reconnection is often found in ICME sheath regions (e.g., Feng & Wang 2013) or at the interface between two successive ICMEs (e.g., Maričić et al. 2014).In the unperturbed case, ICMEs expand due to higher pressure inside than outside of the structure (e.g., Klein & Burlaga 1982).Assuming self-similar expansion in the radial direction, the expansion at the front is added to the propagation speed, while at the trailing edge, it is subtracted from the propagation speed.Depending on the relative locations of CME source region and source of HSSs (i.e., CHs), different parts of the ICME can become accelerated and thus change its expansion behavior (e.g., Lugaz et al. 2022).Occasionally, compression of the ICME can give rise to a trailing edge that is faster than the leading one (i.e., a positive speed gradient), as noted by Dumbović et al. (2022).We here present the interaction of a CME with a nearby CH, followed by the interaction of the ICME and the trailing HSS in interplanetary space.This interaction results in the observation of a magnetic reconnection region located between the trailing edge of the ICME and the HSS.
The paper is structured as follows.First, we provide information about the remote-sensing and in situ data used for this study, together with the methods of analysis, in Sect. 2. We then give the results and accompanying discussion of this analysis in Sect.3, followed by a brief summary and our conclusion based on these findings in Sect. 4.

Remote sensing
For the remote-sensing analysis of the CME source region, its low coronal signatures, and the nearby CH, we used extremeultraviolet (EUV) images from the 193 and 211 Å channels of the Atmospheric Imaging Assembly (AIA; Lemen et al. 2012) and magnetograms from the Helioseismic and Magnetic Imager (HMI; Scherrer et al. 2012) on board the Solar Dynamics Observatory (SDO; Pesnell et al. 2012).To analyze the coronal dimming associated with this event, we used the method developed by Dissauer et al. (2018), who used logarithmic base-ratio (LBR) images to detect the dimmings, LBR = log Here, I n represents the image at a given time step n, and I 0 is the reference image before the onset of the eruption and associated dimming.
We extracted the properties of the CHs that are visible on disk using the code called collection of analysis tools for coronal holes (CATCH; Heinemann et al. 2019), which is based on intensity threshold calculations.CATCH uses AIA 193 Å images and the HMI line-of-sight photospheric magnetograms as input to determine the area, intensity, center of mass, and extent, as well as the underlying photospheric magnetic flux.In addition, we used a magnetic field extrapolation based on magnetograms to analyze the global coronal magnetic field.This extrapolation was done via a potential field source surface model (PFSS; Schrijver & De Rosa 2003;Schrijver 2001) that is implemented in SSW-IDL.In general, the PFSS model reveals largescale regions of open and closed magnetic structures at different heights in the solar corona.It thus helps interpreting processes visible in remote-sensing data before and after the eruption of the CME.
To study the propagation of the erupted structure in the upper corona, we used white-light images from the Large Angle and Spectrometric COronagraph (LASCO; Brueckner et al. 1995) on board the SOlar and Heliospheric Observatory (SOHO; Domingo et al. 1995) as well as COR 2 images from the Sun Earth Connection Coronal and Heliospheric Investigation (SECCHI; Howard et al. 2008) on board the Solar TErrestrial RElations Observatory (STEREO; Kaiser et al. 2008) twin spacecraft.
Early CME appearance close to the Sun was modeled by the application of the graduated cylindrical shell (GCS; Thernisien et al. 2006;Thernisien 2011) reconstruction.The main assumption of this model is that the CME can be approximated as a toroidal structure with two footpoints anchored on the solar surface and a broad front at its apex.It uses contrastenhanced coronagraphic images of SOHO/LASCO (C2 or C3) and STEREO/SECCHI (COR1 or COR2) as input.Reconstructing the erupted structure with GCS has multiple degrees of freedom, such as width, height, and tilt of the curved cylinder-like shape.

In situ measurements
To link the remote-sensing data to in situ measurements, we searched the list of Richardson/Cane ICMEs1 (RC-list; Richardson & Cane 2010), which occasionally provides a reference to the LASCO CME catalog2 (Gopalswamy et al. 2009b), for the best eruption candidate causing the near-Earth disturbance.To further ensure the correct association of CME and ICME, we used the drag-based interplanetary propagation model (DBM; see Čalogović et al. 2021;Dumbović et al. 2021) elaborated by Vršnak et al. (2013) to estimate the arrival time at Earth's orbit3 .DBM efficiently calculates an estimate of the ICME arrival time assuming a change in propagation velocity due to interaction with the ambient SW, similarly to the aerodynamic drag.Therefore, input into DBM includes not only CME and ambient SW speed, but also an estimate for the drag parameter.We took γ = 1 × 10 −7 km −1 as input for the drag parameter (because the structure is moderately dim; see Vršnak et al. 2013) and an ambient SW speed of w = 360 km s −1 (the average SW speed measured at 1 AU one day prior to the shock arrival in Wind plasma data).The output of the DBM tool is presented in Sect.3.1.
We analyzed near-Earth SW conditions using in situ data from Wind (Ogilvie & Desch 1997) and the Advanced Composition Explorer (ACE; Stone et al. 1998) spacecraft.From Wind, we used data from the Three-Dimensional Plasma and Energetic Particle Investigation (3DP; Lin et al. 1995), the Magnetic Field Investigation (MFI; Lepping et al. 1995), and Solar Wind Experiment (SWE;Ogilvie et al. 1995).From ACE, we used ion compositional data from the Solar Wind Ion Composition Spectrometer (SWICS; Gloeckler et al. 1998).
Finally, we applied the Walén test (e.g., Phan et al. 2006) to in situ data to confirm the occurrence of Alfvénic waves close to the region of interplanetary magnetic reconnection.The test gives an estimate for the SW velocity components v pred across the reconnection region, where v ref is the vector components of the measured velocity at the edges of the region, and ρ ref is the mass densities at the same point.Eq. (2) also uses the magnetic field components at the same points, B ref , as well as the magnetic field strength and mass density inside the region, B and ρ.The pressure anisotropy factor, α ref , is assumed to be zero because no measurements are available for this (as noted in Ruffenach et al. 2015).The plus sign signifies Alfvénic waves propagating antiparallel to the magnetic field, producing correlated changes in velocity and magnetic field strength components.The minus sign hence signifies parallel propagation and anticorrelated changes (e.g., Gosling et al. 2005a).

Remote sensing
The left panel of Fig. 1 shows a 193 Å AIA image at the time of the eruption.Around the central meridian lies a large AR complex that produced two M-class flares in the course of 2.5 h.The first flare (ID 098250 in the Hinode Flare Catalog, see Watanabe et al. 2012) took place on 2014 February 4 at 01:23 UT at latitude S09 and longitude W13 4 .Two hours prior to the flare, running-difference images show a slowly expanding bright loop that originated at the exact same location.At 04:00 UT on the same day, a second flare (ID 098260) can be observed closer to the central meridian (at S14 W07) that has higher X-ray emission (M5.2 vs. M3.8 in the case of the first flare).A more detailed view of the AR complex is given in the central panel of Fig. 1, showing a composite image of AIA 211 Å and HMI magnetogram.The magnetic connection between regions of opposite magnetic field strength sign where the flares occur is clear.This connection can be seen as bright arcs that span from positive to negative values of the magnetic field strength.From this, an estimate of the location and orientation of the polarity inversion lines and thus the original orientation of a possible CME can be inferred (e.g., Marubashi et al. 2017).East of the central meridian lies a CH with a large elongation in the N-S direction, stretching from slightly south of the equator (S10 as obtained via CATCH) to N48 in the field of view.Its center of mass is located at (26.2 ± 0.1) • E and (22.9 ± 0.1) • N. The 4 Obtained from solarmonitor.orgunsigned mean magnetic field strength of the underlying photospheric region is (8.67 ± 0.05) G, and the signed mean field strength is (−3.27 ± 0.05) G. Thus, the CH has an overall negative polarity at the photospheric level.Remarkably, south of the CH, the emission coincident with the expanding loops and the first flare drops.The right panel of Fig. 1 shows the logarithmic base-ratio image (calculated according to Eq. ( 1)) of this region 45 min after the first flare.We find that the most prominent dimmings occur in two separate areas: a horseshoe shaped region at the bottom of the image, and a narrow area located in between the AR and the CH.Given the rather large distance to the flaring region, we consider these dimmings to be remote (see, e.g., Jin et al. 2022, especially Fig. 9 there).Furthermore, one day after the flares, the dimming developed into a narrow CH extending to high southern latitudes.This small CH lasted for several days and is still visible in AIA 193 Å images after it moved out of the line of sight, as shown in Fig. 2.
Figure 3 then shows an EUV image when this newly formed CH passes the central meridian (left panel) as well as the extraction of it using CATCH (right panel).This CH extended from (19.5 ± 0.1) • W to (10.4 ± 0.01) • E, and from (24.5 ± 0.6) • S to (69.6 ± 0.4) • S. The category factor of the extracted CH is 1.07, thus making it an extraction with medium stability, as defined in Heinemann et al. (2019).The unsigned mean magnetic field strength is (7.2 ± 0.1) G, and the signed magnetic field strength is (−2.3 ± 0.1) G. Thus, it has the same (negative) underlying magnetic polarity at photospheric level overall as the larger CH to the north.
Figure 4 shows an HMI magnetogram after the eruption on February 4 at 12:04 UT, together with the extrapolated PFSS magnetic field lines at a starting height of 1.2R 0 .The large-scale magnetic connection between the flaring region at the west and the coronal dimmings in the southeast is clear: these are coronal loops.The open field lines of opposite polarities confining the closed loop structure longitudinally suggest a streamer above the AR.The location of the first flare is located beneath one footpoint, and the dimming region is located beneath the other footpoint.
A168, page 3 of 13   The first C2 appearance of the CME is on February 4 at 01:25 UT, as stated in the LASCO CME catalog.The speed of the ejected structure at 20R 0 is given as 500 km s −1 in the LASCO CME catalog.Figure 5 shows the GCS reconstruction of the CME that appears as a rather faint structure in all of the coronagraphs.In STEREO A (right panel of Fig. 5), it appears as an elliptic feature that erupts toward the southeast.From the SOHO point of view, the CME appears in the southwest, thus implying likely no nose-hit at Earth.We note that the latitude (36 • S) and longitude (24 • W) of the GCS fit suggest a deviation from radial propagation to the southwest given the location of the flaring region (S09 W13).Still, the initial orientation is likely conserved during early propagation, as a comparison of the reconstructed CME in SOHO with the western polarity inversion line in the central panel of Fig. 1 suggests.
Figure 6 clearly shows that the CME propagates away from both STEREO A and B. Furthermore, the Sun lies in the line of sight between STEREO B and the CME, emphasizing the difficulty of reconstructing the structure in STEREO B images.Remote coronal dimmings connected to so-called streamer-puff CMEs have been described by Moore & Sterling (2007) based on a case study by Bemporad et al. (2005).There, the eruption of a CME below the loops of a streamer leads to a remote secondary ejection that is visible as dimming.Similarly, in the present case, the flare located close to one footpoint of the streamer appears to be triggering an ejection of material at the other footpoint.This leads to the secondary structure that is visible southeast of the main ejecta in Fig. 5 (as indicated by the red arrows).Additionally to streamer-puff CMEs, remote coronal dimming can also occur as a consequence of two adjacent loop systems that reconnect via magnetic component reconnection along quasi-separatrix layers (QSLs; e.g., van Driel-Gesztelyi et al. 2012).Figure 4 shows that at the site of the dimming, the field lines of two loop systems meet above photospheric magnetic fields of the same polarity.In both cases, the remote dimming is a consequence of magnetic reconnection, either above the streamer (in the case of streamer-puff CMEs) or above the adjacent loop systems (in case of component reconnection).
Global reconfiguration of the coronal magnetic field can also be seen in the EUV images of STEREO A and B given in Fig. 7.In the STEREO B image taken 10 days prior to the event, we observe the same coherent CH as in SDO images at the time of CME eruption (e.g., left panel of Fig. 1).In the STEREO A image taken 13 days after the event, however, the previously cross-equatorial single CH has split into two high-latitude CHs in both hemispheres.This dynamics indicates a transport of open flux toward the south pole of the Sun.The underlying photospheric polarity of the CHs (negative) is equal to the polarity of the south polar CH that is formed until the end of SC 24.This could be an example of an observation of the long-term development of the global magnetic field of the Sun through the mechanism of open flux transport described by Golubeva & Mordvinov (2016).
The application of the DBM tool results in an arrival time of 02:50 UT on February 8 of the CME at 1 AU.This is less than 3 h after the onset of the ICME interval that is visible in Wind data at 00:30 UT on February 8 (as presented in Sect.3.2).

Shock and sheath region
Figure 8 shows Wind measurements of the IP disturbance beginning on DOY 38.68 with the arrival of a fast forward shock5 (solid vertical blue line), leading to an increase in all parameters.Prior to the shock, all plasma parameters show values characteristic of slow SW (e.g., Schwenn 1990).The PAD reveals predominant outward (hence field-aligned) electron strahl, but there is a slight asymmetry 12 h before the shock that is indicative of magnetic mirroring effects (Carcaboso et al. 2020).
In the sheath region behind the shock, the highest values of the total magnetic field strength, proton temperature, and speed of the whole disturbance interval are reached.The variability in the magnetic field components is strong, with the Bx-component and Bz-component changing polarity multiple times.This strong variability is reflected by the fluctuations of the total magnetic field strength as well as in the large short-lived peaks of the plasma β (third to last panel).Furthermore, electron pitch angles show a smeared-out distribution due to particle scattering.The charge state of iron is not yet enhanced, which further indicates that there is no erupted material in this subinterval (e.g., Lepri et al. 2001).The end of the sheath region shows a peak of the temperature, which is characteristic of an ICME leading edge (Temmer & Bothmer 2022).

Front region
At DOY 39 (first dashed vertical blue line in Fig. 8) the plasma temperature rapidly drops to values comparable to the expected temperature.The expected temperature, first introduced in Lopez (1987), gives an estimate of the proton temperature using the proton speed alone.A drop in the actual proton temperature below the expected temperature is a strong indicator of cool plasma in the SW, that is, ICME material.At the same time, there is a significant enhancement of plasma density from ∼5 to ∼30 cm −3 .Fluctuations in the magnetic field strength remain at a high level, which is also the case for plasma β.The PAD shows equally strong contributions from both 0 and 180 • , indicating an interval of bidirectionality.Iron charge-state data now rise to values above 12, which indicates ICME material in this subinterval (Goryaev et al. 2020).High density values after the sheath region are indicative of the front region of the ICME (Kilpua et al. 2013), but unlike in the present study, they are usually lower than in the sheath region.This peculiarity could arise from the propagation geometry: because the ICME travels off the Sun-Earth line, the pile-up of material toward Earth is not as strong as in a direct hit, for instance.Kilpua et al. (2013) only analyzed ICMEs with clear MC signatures that are, above all, observed in direct hits.Thus, their study, esp.throughout the subinterval.After measuring a dominant outward population of electrons, the main contribution of electrons shifts to 180 • at the change of polarity of the Bx-component.Toward the end of the subinterval, the distribution becomes isotropic, but is faint overall.The iron charge state decreases toward the end of the subinterval, but still remains above 12.

Back region and reconnection exhaust
Starting with DOY 40.25 (second dashed vertical red line), another interval of very low magnetic field fluctuations lasts until DOY 40.45.Except for a strong local disturbance, the Bx component is rather constant at −8 nT, while By has overall positive polarity.Bz rotates from −4 to 4 nT.The temperature still varies strongly, but at lower values overall than before.Accordingly, plasma β 1 again.The SW speed remains at values of ∼400 km s −1 .Pitch-angle data show a strong contribution of electrons moving outward along the magnetic field lines.Still, similar to the part after DOY 38, the contribution from 180 • is not negligible, again indicating mirroring effects inside this subinterval.The iron charge state gradually decreases throughout and drops below 12 at the end of the subinterval.
At DOY 40.45 (second dashed vertical blue line) the speed, temperature, and density suddently increase, coincident with a drop in the magnetic field strength.This interval lasts for less than 1 h.Furthermore, Bx changes from negative to positive values again at this point, and simultaneously, the pitchangle electrons shift to a clear inward strahl.These features are indicative of a HCS crossing, followed by a traverse through the HPS.After these characteristic measurements, we observe a global rise in plasma speed and temperature (now significantly above the expected temperature), to a maximum of ∼540 km s −1 and ∼3 × 10 5 K. Compositional data drop to the lowest values in the whole data range plotted.We ascribe these changes to the occurrence of an HSS (e.g. , Schwenn 1990) that peaks in speed at DOY 41.05 (February 10, 01:15 UT).This HSS is Figure 9 provides a closer look at the back region of the ICME disturbance, that is, the part after the second dashed red line in Fig. 8. FR-like magnetic field changes are shown in the first panel, with a rotation of the z-component and a nearly constant x-component.The y-component also follows the expected behavior for an SEN-oriented FR (i.e., left-handed, and main axis oriented toward the east; see, e.g., Palmerio et al. 2018), except for a drop in the middle, where the x-and z-components also show perturbed behavior.This middle part also includes sudden rises in plasma beta and speed, and it is connected to a drop in total suprathermal electron counts.After this part, the magnetic field components return to the course expected of an SEN-type FR.At the dashed blue line, the reconnection exhaust is visible as a drop in magnetic field magnitude and increases in plasma density, speed, and β.
Figure 10 shows a Walén test (Eq.( 2)) applied to the region in between the small FR-like magnetic obstacle and the onset of the HSS.On the front side of the interval, magnetic field and velocity components are anticorrelated, whereas toward the trailing edge, velocity and magnetic field components are clearly correlated along the x-and y-directions.
The successful application of the Walén test is a strong indicator of Petschek-like reconnection.It confirms Alfvén waves propagating parallel or antiparallel on different sides of the reconnection exhaust, as expected from the simple theoretical 2D reconnection model suggested by Gosling et al. (2005a).We obtain a duration of the exhaust region of ∼16 min, similar to the reconnection exhaust reported by Gosling et al. (2005a).

Interpretation
The disturbance at 1 AU features clear sheath and ICME front region in the entire time interval, but it lacks signatures of a MC.Due to the inhomogeneity of the whole structure, it rather A168, page 7 of 13 resembles a compound stream or complex ejecta (e.g., Lugaz et al. 2017), with occasional small-scale field rotations.The interaction of ICME with the trailing HSS shows a small magnetic structure similar to what has been described as small FR (SFR; e.g., Moldwin et al. 2000).Qualitatively, a small FR resembles an MC, but it is much smaller and hence far shorter (>8 h vs. minutes to hours; e.g., Cartwright & Moldwin 2010).We propose two explanations for the SFR at the end of the disturbance: 1. Local (magnetic island formation): The proximity of ICME back region, HSS, and current-sheet crossing indicates strong turbulence immediately at the transition from ICME disturbance to HSS.It is known (e.g., Moldwin et al. 2000, and A168, page 8 of 13 Fig. 9. Selected plasma and magnetic field measurements for the back region of the ICME disturbance.The first panel shows the GSE-components of the magnetic field with same color-coding as in Fig. 8, and the absolute value of the magnetic field strength is plotted in black, with negative and positive sign.Again, the horizontal dashed line marks 0 nT.The second panel displays the proton density (black) and the plasma β (red).The third panel shows the proton speed.The fourth panel shows the PAD.Like in Fig. 8, the dashed blue line indicates the onset of the slow-mode wave.The magnetic field behaves similarly as in MCs.Feng et al. 2011) that in the proximity of current sheets, small FRs can form as remnants of magnetic reconnection.The observation of reconnection exhausts is also the strongest evidence that reconnection still occurs at 1 AU and therefore requires other physical mechanisms than small FRs formed in vicinity of the Sun (as given in, e.g., Rouillard & Sheeley 2011).Therefore, we do not observe parts of the ICME FR, but a newly formed magnetic island (MI) in the vicinity of a strong reconnecting current sheet.In support of this explanation, we point out i) the similarities of PADs simulated by Khabarova et al. (2020) to measurements of electron pitch angles by the Solar Wind Electron Proton Alpha Monitor (SWEPAM; McComas et al. 1998) on board ACE (Fig. 11) and ii) the generation of MIs at the flanks of modeled ICMEs as they interact with ambient SW.Hosteaux et al. (2018) observed the formation of MIs following the eruption of a breakout CME, similarly to swirls in a river flow due to obstacles.
2. Global (ICME leg): No clear MC-connected magnetic field changes are visible inside the disturbance interval.The orientation of the erupted CME implies that the front of the ICME structure passes below the Sun-Earth line and is thus not measured by the L1 spacecraft.However, the eastern flank of the GCS reconstructed ICME (assuming unaltered self-similar expansion) crosses the Sun-Earth line.This together with the strong radial component of the SFR (see Fig. 9) indicate a traverse from above through the trailing part of the ICME.Kilpua et al. (2013) used the radial magnetic field strength as diagnostics to determine whether a spacecraft passes the leg of A168, page 9 of 13 Fig. 10.Walén test of the slow-mode wave at the interface of ICME and HSS.From top to bottom, the x, y, and z components and the absolute value of the proton velocity and magnetic field strength are given.In each panel, the black curve shows the measured value of the proton velocity, and the blue diamonds give the predicted value, as determined according to Eq. ( 2).The solid red curve shows the magnetic field strength, and the dashed red curve marks a field strength of 0. The vertical dashed black lines indicate the start and end time of the reconnection exhaust interval.The jump in the predicted velocity marks the center of the exhaust region, where the anticorrelation of the velocity and magnetic field components changes to correlation.A168, page 10 of 13 an ICME.If the radial component accounts for more than 70% of the total magnetic field magnitude, as is the case for the present event, the spacecraft likely crosses the ICME leg.We note that the local changes in the magnetic field components inside the SFR at DOY 40.38 are similar to the global changes at the reconnecting current sheet: the polarities of the magnetic field components are equal and their strengths are similar, the SW speed also shows an enhancement similar to that after the current sheet crossing, and the density also takes the same values as before and after the Alfvén wave.
We thus picture a situation as given in Fig. 12.It shows the spacecraft path through the ICME flank, the HCS/HPS, and finally the HSS.In this picture, the FR is gradually removed by the interaction with the ambient SW (e.g., Lavraud et al. 2014, andRuffenach et al. 2015), implying an originally larger FR.The HCS, nearly perpendicular to the ecliptic plane due to the N-S orientation of the streamer (Fig. 4), is thought of as vertical wall separating the ICME and the HSS.Because the HPS envelopes the eastern side of the flank, plasma sheet material is additionally inserted into the FR (explaining the similarities of the local changes inside the SFR with the global changes afterward).This situation was described in Winslow et al. (2016) for the nose of an ICME.As the turbulent HPS is crossed, there is a short-term isotropy and overall low count of PAD, probably arising due to field line disconnection associated with the reconnecting HCS (Gosling et al. 2005b).Suprathermal electron data and the x-component of the magnetic field then indicate that the spacecraft has entered the trailing HSS, which is located on the other side of HCS/HPS.
To summarize the above discussion, the SFR is either in the process of being formed near the HCS or is eroded by it.In the first case, this indicates local formation mechanisms as suggested by Khabarova et al. (2020).In the second case, the erosion occurs globally to the ICME FR (e.g., Lavraud et al. 2014).By tracking in situ properties of similar ICMEs throughout the heliosphere by multispacecraft observations, the question of whether we observe global or local structures could be answered more confidently.

Summary and conclusion
In conclusion, the CME that erupted on February 4, 2014, at 01:30 UT was likely deflected away from the Sun-Earth line.This can be inferred both from the GCS reconstruction of the CME as well as from the lack of MC signatures in situ.The latter is indicative of missing the main body of the ICME and just crossing the leg.We ascribe the deflection of the ICME to a CH east of the CME source region.The CME erupted below the footpoint of a streamer and caused evacuation of plasma at the other footpoint.This evacuated site developed into a small and long-lasting CH over one day with the same predominant (negative) photospheric polarity as the already present one.in situ data at 1 AU show that the non-MC ICME drives a shock and well-visible sheath region and features a small FR at its trailing A168, page 11 of 13 edge, followed by the reconnection exhaust at the interface of the SFR and the trailing HSS.This study therefore demonstrates that apart from compression of a CME due to a trailing HSS, reconnection can also occur at the interface of the two structures.In a future study, the frequency of the occurrence of interplanetary reconnection in between CME and HSS will be analyzed.Furthermore, recurring properties of reconnection, such as magnetic field orientation or the location of the reconnecting region with respect to the CME, will be evaluated statistically.This will help us to understand the importance and relevance of reconnection in interplanetary space.

Fig. 1 .
Fig. 1.Analysis of the solar corona during eruption.Left: AIA 193 Å image of the CME source region west of the disk center and the CH east of the center taken on 2014 February 4 at 01:13:45 UT. (1) denotes the location of the first flare (at 01:23 UT), and (2) indicates the location of the second flare (at 04:00 UT).The box marks the cutout used for the dimming analysis (Fig. 1).The arc inside this rectangle outlines the region that undergoes the strongest dimming in emission shortly after eruption.Center: AIA 211 Å and HMI magnetogram composite image of the region inside the box given in the left panel.In the magnetogram, red refers to positive values of the radial magnetic field strength, and blue shows negative values.The black lines give the approximate location of the polarity inversion line, located below bright arcs connecting regions of opposite polarity of the magnetic field strength.Right: logarithmic base-ratio image of the dimming region in 211 Å at 02:15 UT.The base reference image is taken on February 3 at 22:00 UT.Dark blue corresponds to a relative change of ≥100%, thus marking no change or increase in emission.Dark red is associated with the regions of strongest emission decrease, with a relative change of ≤−81%.

Fig. 2 .
Fig. 2. Sequence of AIA 193 Å full-disk images showing the AR complex and the development of the coronal dimming into a small CH in the southern hemisphere.The ellipse in the top left frame marks the extent of the AR prior to the eruption.In the top right frame, the well-visible CH is marked by an arrow.The first frame in the top left corner is taken on February 3 at 22:29 UT, and the last frame in the bottom right corner is taken on February 7 at 04:59 UT.

Fig. 3 .
Fig. 3. New CH in the southern hemisphere.Left: AIA 193 Å full-disk image taken on February 5 at 10:59 UT.A newly formed CH is visible along the central meridian south of the equator, marked by the box.Right: extraction of the CH inside the box using CATCH.The red line gives the CH boundaries as obtained with a threshold of 35% of full-disk median intensity.The yellow cross marks the center of mass of the surface covered by the CH.

Fig. 4 .
Fig. 4. PFSS extrapolation obtained on February 4 at 12:04 UT.In the PFSS extrapolation, purple and green indicate open field lines of negative and positive polarity, respectively.White is used for closed field lines.In the magnetogram, regions of strong positive (negative) magnetic field are shown in white (black).The magnetic connection of the eruption region, given by the cross, to the location of the dimming, given as red shaded area, is illustrated by an arc tracking a field line.The starting height of the extrapolation is 1.2R 0 .

Fig. 5 .
Fig. 5. Early interplanetary evolution of the CME.Top panels: coronagraph base-difference images of COR2/STEREO B (left), C3/SOHO (center), and COR2/STEREO A (right) taken on February 4 at 03:30 UT.The arrow points to the secondary structure southeast of the main ejecta.Bottom panels: GCS reconstruction of the erupted structure plotted over the corresponding images in the top panels.The green grid envelops the CME body from its front to its footpoints at the Sun.
Fig. 5 there, is indicative of CMEs travelling very close to the Sun-Earth line.Still, Kilpua et al. (2013) provide a good description of ICME front regions and a meaningful explanation of what we observe after the sheath region.
3.2.3.Central regionAt DOY 39.28 (first dashed vertical red line) the fluctuations of the total magnetic field strength suddenly decrease, and a drop of β to values 1 is measured.The density also declines to values of ∼10 cm −3 .The components of the magnetic field do not show clear FR-like changes.Another substructure shows a B x rotation within this subinterval that lasts until DOY 40.25.Close to the rear boundary of this subinterval, the components are again highly turbulent and oscillate around 0. The plasma

Fig. 6 .
Fig. 6.Orbital configuration of Earth (green) and STEREO A (red) and B (blue) on February 4. The arrow points to the propagation direction of the CME nose as inferred from the GCS reconstruction.
temperature stays at values similar to the expected temperature, showing additional small-scale fluctuations as well as an overall increase coincident with the late enhanced fluctuations of the magnetic field components.The proton speed overall slightly decreases inside this subinterval, notably showing strong local increases in the back part.This is simultaneous with temperature and magnetic field measurements.Bidirectionality in the PAD ceases and does not show a clear directional preference A168, page 6 of 13

Fig. 7 .
Fig. 7. Long-term evolution of the large CH close to the eruption site.Left: STEREO B/EUVI 195 Å image taken on January 25 at 02:36 UT showing the coherent CH stretching from slightly south of the solar equator to high northern latitudes.Right: STEREO A/EUVI 195 Å image taken on February 17 at 05:05, demonstrating the bifurcation of the CH into a northern and southern part, with the division occurring at lower latitudes.
included in the catalog by Besliu-Ionescu et al. (2022), while Heinemann et al. (2020) ascribed this HSS to the northern CH described in Sect.3.1.

Fig. 8 .
Fig. 8. Plasma and magnetic field measurements for the interval of ICME disturbance.The first panel shows the absolute magnetic field strength in black and its fluctuation in gray.The second panel shows the geocentric solar ecliptic (GSE) components of the magnetic field, and in red, we show the x-component, in blue the y-component, and in green the z-component.The horizontal dashed line marks 0. The third panel shows the plasma temperature in black and the expected temperature (Lopez 1987) in red.The fourth panel displays the proton density (black) and the plasma β (red).The fifth panel shows the proton speed in black and the iron charge state in gray, and the dashed gray line marks a charge state of 12.The last panel shows the PAD of electrons with energies of 116.1 eV, in which the lowest counts are depicted in dark blue and the highest counts in dark red.In all the panels, the solid vertical blue line marks the occurrence of a fast forward shock on February 8 at 16:15 UT.Then, the dashed vertical blue line indicates the end of the sheath region and the beginning of ICME frontal region on February 9 at 00:30 UT.The dashed vertical red line that follows is indicative of the main ICME part starting at 05:30 UT.The second dashed red line marks the beginning of small FR-like field developments on February 10 at 06:00 UT.Finally, the second dashed blue line indicates a slow-mode wave that is simultaneous with the end of the ICME disturbance and onset of the HSS stream, at 10:45 UT.

Fig. 11 .
Fig. 11.Comparison of measured and simulated PAD data.Left: pitch-angle data from ACE/SWEPAM for the back region of the ICME.All the measured energy channels are plotted, with the highest energy in the top panel and the lowest energy in the bottom panel.Right: Fig. 39 from Khabarova et al. (2020) showing the simulated PAD for a merging MI close to a reconnecting current sheet.The top panel shows the trajectory of a virtual spacecraft through the merging MI.In the middle panel, normalized high-energy PAD measurements of this virtual spacecraft are plotted.The bottom panel shows the measurements for low-energy channels.The counterstreaming in the highest-energy channels on the left is focused at 07-09 UT, similar to the simulated PAD on the right.Figure 39 from Khabarova et al. (2020) reproduced with permission from SNCSC.

Fig. 12 .
Fig. 12. Schematic representation of the ICME-HSS border.The left and right panel show the traverse of the spacecraft (dashed line) in the plane of ecliptic and above the plane of ecliptic, respectively.The red shaded structure symbolizes the small FR at the end of the ICME interval, and the light blue shaded area represents the HCS and the HPS enveloping it.The light green region indicates the HSS.Blue stars are indicators of magnetic reconnection and blue arrows indicate exhaust jets.The black arrows indicate magnetic field lines.