Coronal mass ejections (CMEs) are sudden releases of large amounts of plasma dragging magnetic fields throughout the heliosphere (Webb and Howard, 2012). CMEs have been imaged in remote-sensing data since the 1970s, with missions such as OSO-7 (Tousey et al., 1973), Skylab (MacQueen et al., 1974), the Solar Maximum Mission (MacQueen et al., 1980), and P78-1 (Solwind; Sheeley et al., 1980).

Shortly after the first observation, a relationship was discovered between CMEs at the Sun, and in situ signatures recorded near the Earth (Burlaga et al., 1982) further strengthened in the following years (Richardson et al., 2000). The remote-sensing observations were improved with the continuous, near-Earth imaging of the Large Angle and Spectrometric COronagraph Experiment (LASCO, Brueckner et al., 1995) aboard the SOlar and Heliospheric Observatory (SOHO, Domingo et al., 1995), beginning in 1996 and still operating today. Moreover, since 2006, after the launch of the Solar TErrestrial RElations Observatory (STEREO, Kaiser, 2005), multiple viewpoints and imaging from the solar surface to beyond 1~au are obtained with the Sun Earth Connection Coronal and Heliospheric Investigation instrument suite (SECCHI, Howard et al., 2008).

This improved observational coverage allowed a direct link between the imaging data and in situ data (Rouillard, 2011; DeForest et al., 2013), followed by statistical studies comparing properties of CMEs throughout the heliosphere to the in situ properties (Hess and Zhang, 2017; Wood et al., 2017) of their counterparts, the interplanetary coronal mass ejections (ICMEs). From a comprehensive analysis of 310 ICMEs, Zurbuchen et al. (2016) found that not only ICMEs exhibit systematic element abundance enhancements compared to quasi-stationary solar wind but also that the in situ composition measurements provide a unique way to determine the location and dynamics of their source region. Furthermore, it was found that more than 70% of CMEs are associated with prominence eruptions (Munro et al., 1979; Gopalswamy et al., 2003), which are related to the settlement of cool and dense plasma in the highly sheared, mainly horizontal, magnetic field above polarity-inversion lines (Priest et al., 1989; Gibson, 2018). Moreover, inside ICMEs, the composition has been found to be elevated in He²⁺/H⁺ (Borovsky, 2008; Yermolaev et al., 2020) particularly with larger values when the ICMEs are related to prominences (Hirshberg et al., 1972; Borrini et al., 1982; Lepri and Rivera, 2021).

Prominence eruptions are routinely observed, near the Sun, with visible (e.g., Hα) and extreme ultraviolet (EUV) telescopes, in the middle corona with visible light coronagraphs and with heliospheric imagers at larger distances from the Sun (a review on prominence observations can be found in Parenti, 2014, and references therein). In favorable circumstances, these remote-sensing observations may be supplemented by in situ measurements, which provide highly localized information about the speed, composition, and magnetic field.

While propagating from the Sun into the heliosphere, prominences and associated CMEs must be observed in widely disparate domains as no single instrument can provide an overview of their evolution (Thernisien et al., 2009; Luhmann et al., 2020). For the most part, gaps between these measurements and the challenges of processing data in a unified way have inhibited complete analyses of eruption initiation and propagation to in situ instruments.

In contrast, one example of this kind of study that does achieve the rare unification of observations that span from the Sun to the heliosphere is described in a series of papers by Howard and DeForest (2012) and DeForest et al. (2013), using data from STEREO-A, the Advance Composition Explorer (ACE, Stone et al., 1998), the Wind spacecraft (Ogilvie and Desch, 1997), and extensive image processing to track a small CME from its genesis at the Sun through the heliosphere to Earth.

More recently, studies have demonstrated the feasibility of closing some important observational gaps in the early evolution of erupting prominences using extended EUV observations. Reva et al. (2016) used the TESIS instrument assembly on the Russian CORONAS-PHOTON mission, along with images from LASCO, to observe the onset and early evolution of an erupting prominence, producing a complete trajectory from initiation to about 25 R_⊙. Seaton et al. (2021) used a special campaign of the GOES Solar Ultraviolet Imager to track several small eruptions continuously from the solar surface to heights that are well-observed by coronagraphs. Other studies have leveraged the large field of view and off-pointing capabilities of the SWAP EUV imager on PROBA2 to characterize the early evolution of eruptions (O’Hara et al., 2019) and, when combined with STEREO-A observations, to develop three-dimensional understanding of erupting features (Mierla et al., 2013).

The advent of the Parker Solar Probe (PSP, Fox et al., 2016) and Solar Orbiter (SolO, Müller et al., 2020) missions has provided a new opportunity to study the origins and evolution of eruptions as they propagate from the Sun into the heliosphere. With a highly elliptical orbit that, using a series of gravity assists from Venus, will continue to get closer to the Sun, PSP enables an unprecedented ability to observe at heliospheric distances as close to the Sun as 0.04 au (less than 9 R_⊙). While not getting as close to the Sun as PSP, SolO still takes science observations down to 0.28 au (60 R_⊙) and further improves observational coverage with remote-sensing capabilities extending from the EUV to visible light, including coronagraphy and, at a later stage of the mission, by leaving the ecliptic plane.

Given the combined ability to observe both near the Sun and off of the Sun–Earth line, these spacecrafts significantly extend our ability to track eruptions continuously, with multiple perspectives, and the capability to directly sample localized plasma and magnetic conditions within an eruption and, when combined with additional observations, three-dimensional characterization of these features.

Unlike the past 1 au missions, PSP and SolO, due to their rapid changes in radial distance from the Sun, are not intended to take consistent data at regular intervals but are instead designed to focus science observations at specific periods and locations. Given the need to inform the in situ observations at one satellite with the context of remote-sensing observations at another, exploiting the uniquely advantageous orbital configurations is vital to achieving deep understanding, significantly augmenting the science results that could be achieved by any individual mission (Velli et al., 2020).

A favorable configuration occurred in late 2021 April, while PSP was approaching its eighth perihelion and was in near quadrature with SolO. Rodriguez et al. (2023) reported in detail an eruptive event that happened on 2021 April 22. With multiple independent lines of evidence (from different instruments and spacecraft), they showed, for the first time, the direct connection between the eruption and continuous magnetic reconnection and heating processes. The related ICME, which arrived on the Earth late on 2021 April 24, causing a minor geomagnetic storm, highlights the importance to establish a connection between solar eruptions observed in the solar corona and their in situ signatures measured by spacecraft in interplanetary space.

Two days later, during the late hours of 2021 April 23, another unrelated eruption event took place, reaching PSP in the early hours of 2021 April 25. In this investigation, we identify the eruption of prominence material at the surface of the Sun as the source of a helium-enriched plasma structure measured in situ by the PSP spacecraft. We use multi-spacecraft remote-sensing observations (STEREO-A and SolO) to track the propagation of the associated CME from the Sun to the PSP location.

This work is organized as follows. In Section 2, we describe the models, locations, and datasets of the different spacecraft used to track the CME from the Sun to PSP. In Section 3, we present the PSP in situ observations and how we use them to predict solar wind conditions at 10–20 R_⊙. Then, we present an overview of the remote-sensing observations (Section 4): from the EUV images (Section 4.1), coronagraphs (Section 4.2), and heliospheric images (Section 4.3). The results of the reconstruction using the Graduated Cylindrical Shell model (GCS Thernisien et al., 2006; Thernisien et al., 2009) are shown in Section 4.4. Finally, we summarize, discuss our results in a height–time plot (distances between 1.5 and 20 R_⊙) combining all observations, and present our conclusions in Section 5.

To study the full evolution of the event, we combine the in situ detectors on board Parker Solar Probe (PSP, Fox et al., 2016), including the plasma parameter measurements from the Solar Wind Electrons Alphas and Protons (SWEAP, Kasper et al., 2016) and the Electromagnetic Fields Investigation (FIELDS, Bale et al., 2016) instruments, with remote-sensing observations from the Solar TErrestrial RElations Observatory (STEREO, Kaiser, 2005), Solar Orbiter (SolO, Müller et al., 2020), and PSP. Due to the location of the prominence on the far side of the Sun from the Earth’s perspective (see Figure 1), there were no obvious signatures in any Earth-orbiting in situ or remote-sensing observatories.

On PSP, the in situ plasma conditions are obtained from the data registered by three electrostatic analyzers, called the Solar Probe ANalyzers (SPAN, Whittlesey et al., 2020), and the Solar Probe Cup (SPC, Case et al., 2020), which are part of SWEAP suite, while the magnetic field measurements are from FIELDS instrument. On SolO, the plasma is measured by the Solar Wind Analyzer (SWA, Owen et al., 2020) and the magnetic fields with the magnetometer (MAG, Horbury et al., 2020).

For remote sensing, SolO carries the Extreme Ultraviolet Imager (EUI, Rochus et al., 2020), the Metis coronagraph (Antonucci et al., 2020), and the SolO Heliospheric Imager (SoloHI, Howard et al., 2020). EUI includes the Full Sun Imager (FSI), a wide field of view EUV imager with 174 and 304 Å channels which overlaps with the fields of view of Metis, that images the corona in the Visible and Ly-α. PSP, in addition to extensive in situ instrumentation, and carries its own Wide-field Imager for Solar Probe (WISPR, Vourlidas et al., 2016). STEREO images are obtained with the Sun-Earth Connection Coronal and Heliospheric Investigation instrument suite (SECCHI, Howard et al., 2008) consisting of five telescopes: an EUV imager (EUVI), two white light coronagraphs (COR1 and COR2), and two wide-field imagers (HI-1 and HI-2).

To image the prominence near the Sun, in the inner and middle corona, we use FSI from SolO and SECCHI from STEREO-A spacecraft. FSI was operating alternatively in its 174 and 304 Å channels, resulting in over eight images per hour of each type (Mampaey et al., 2022). SolO was at 0.87 au distance from the Sun, resulting in an FSI field of view of 12.4 R _⊙.

In the middle and outer corona, we utilize the COR2 and Metis coronagraphs. The Metis white light (WL) and Lyman-α channel observations were obtained during a synoptic program, with a cadence of one dataset per hour. Each dataset consists of four polarized (WL) images, which, combined together, produce one (pB) image and three UV images. The acquired data were processed and calibrated following the procedure described in Romoli et al. (2021) and subsequently updated by Andretta et al. (2021). Small variations of the baseline level of data in the UV channel were corrected with a procedure devised and later implemented in the standard data processing pipeline.

Imaging at even greater heights is provided by SoloHI and WISPR. SoloHI observes off the east limb of the Sun with a 40° field of view, centered near 25° from the center of the solar disk. SoloHI consists of four separate 2k×2k APS detectors, arranged in a square pattern and sharing a single optical system for a total 4k×4k field of view. The boundaries between these tiles form a visible, stable cross-pattern in the images when the four tiles are correctly assembled into a single mosaic. WISPR consists of two 2k×2k APS detectors. The WISPR-Inner telescope (WISPR-I) has a 40° field of view, centered approximately 13.5° off the western limb of the Sun, relative to the spacecraft. The WISPR-Outer telescope (WISPR-O) has a 60° field of view centered at approximately 73°.

The locations of Earth, PSP, STEREO-A, and SolO are shown in Figure 1 in the Heliocentric Earth Ecliptic (HEE) coordinate system. PSP and SolO orbits are shown in orange and purple solid lines, respectively. PSP and SolO were in nearly perfect quadrature when the CME was launched to space from the Sun (its direction is marked with a blue arrow in the figure), allowing us the remote-sensing observation of the eruption from SolO’s point of view. In Table 1, we summarize the radial distance (R), longitude (lon), and latitude (lat) of all spacecraft. In both Figure 1 and Table 1, the locations of PSP are given for both the approximate onset time of the prominence eruption as well as for 24 h later, when the in situ signatures are first observed. In this time span, PSP moves 5.7° in longitude and 7.2 R _⊙ toward the Sun.

We model the background magnetic field using the Magnetohydrodynamic (MHD) Algorithm outside a Sphere (MAS) model from Predictive Science Inc. (PSI). We use the semi-empirical thermodynamic approach described in Riley et al. (2021), which includes conductive and radiative losses, and an empirical term for the coronal heating term (see Lionello et al., 2001; 2009; Reeves et al., 2010, for details involved in solving the MHD equations). The model consists of coronal solution (1–30 R_⊙) and heliospheric solution (30 R_⊙—1 au), with the heliospheric solution being driven by the results of the coronal solution. The boundary conditions for the coronal solution are derived from the Helioseismic and Magnetic Imager (HMI Schou et al., 2012) aboard Solar Dynamics Observatory (SDO Pesnell et al., 2012), and the boundary conditions for the heliospheric solution are based on the Distance from the Coronal Hole Boundary (DCHB) method (Riley et al., 2001; 2015). This model has been found to have very good agreement with magnetic field measurements taken in situ at the Earth, STEREO-A, and PSP for the first four PSP perihelia (Riley et al., 2021). Figure 2 shows the magnetic field obtained from the model and its relationship to the Sun and the location of PSP during the eruption.

In order to model the kinematics of the CME, we use the Graduated Cylindrical Shell model (GCS Thernisien et al., 2006; Thernisien et al., 2009), which is an empirical geometric model used to represent the flux-rope configuration of CMEs. This model was used to reconstruct many CMEs from LASCO and SECCHI data (e.g., Hess and Zhang, 2017; Kilpua et al., 2019; Zhao et al., 2019; Scolini et al., 2020; Palmerio et al., 2021; Nieves-Chinchilla et al., 2022). The same approach can be adapted to include WISPR and SoloHI data while still including the older datasets (Braga et al., 2022).

To reconstruct the solar wind conditions from PSP data (speed and density) back to 10–20 R_⊙, we solve Bernoulli’s equation (Landau and Lifshitz, 1987) assuming the adiabatic expansion of a polytropic flow and neglecting the contribution of gravity force. Backward reconstruction methods have been extensively used (Parker, 1958; Weber and Davis, 1967; Tasnim et al., 2018; Biondo et al., 2021) with their limitations extensively discussed (Schatten, 1971; Pizzo, 1981; Ness and Burlaga, 2001).

In Figure 3, from top to bottom, we show the 15-min time series of the PSP in situ measurements of (first panel) the normalized differential energy flux versus azimuth angle SPAN − ϕ computed from SPAN instrument (Livi et al., 2022); (second panel) the PSP distance to the Sun (R, in black) with both flow angles SPC − θ (in peach) and SPC − ϕ (in teal) measured from SPC (Case et al., 2020) ¹ ; (third panel) the magnitude of the proton speed by both SPC (SPC-|V| in black) and SPAN (SPAN-|V| in gray) and the magnetic field strength (|B|, shown in navy) overplot. In the fourth panel, we show the radial (B _R , in blue), tangential (B _T , in green), and normal (B _N , in red) components (measured by FIELDS). Then, in the fifth panel, the electron pitch angle (PA) for an energy of 314.45 eV with the color scale represents the logarithm of the energy flux vs. PA (from SPAN instrument, Whittlesey et al., 2020). In the sixth panel, we show the velocity distribution functions (VDFs) measured by SPC. Then, we present the SPC-β parameter and proton temperature (SPC-T) in the seventh panel while SPC-N and SPAN-N densities in the eighth panel.

For this period, the majority of the solar wind flow was out of the SPAN field of view (SPAN − ϕ > 160°, top panel) resulting in large and systematic underestimates in the proton density, SPAN-N (bottom in Figure 3), from that instrument. The solar wind is fully within the SPC sensor field of view; however, the flow SPC-ϕ direction (second panel) is only roughly estimated due to the crosstalk anomaly described in the SPC data quality flags. For this analysis, we approximate that |V|∼|V _R |, which is roughly consistent with both sets of observations. We note that the SPAN spectrogram, capturing the tail of the proton distribution, does not show significant signs of flow deflection for at least the first half of the period.

PSP crossed over the heliospheric current sheet (HCS) on 2021 April 24 16:15 UT. Before this time, the magnetic field components and plasma parameters show quiet solar wind conditions, the flow was propagating at a speed between 230 and 260 km s⁻¹ and positive B _R value. The electron pitch angle corroborates the change of magnetic field sector which we marked with a black vertical dash line. After the HCS crossing, the spacecraft passed through a high-density region (reaching a maximum value of 692 cm⁻³) traveling at a very low speed of 250 km s⁻¹.

The structure of interest arrived at the spacecraft on 2021 April 25 at 01:15 UT (marked with purple dashed vertical line) with no shock signatures and characterized by SPC-N = 214 cm⁻³, SPC-T = 96,423 K, β = 0.38, and |B| = 73 nT. The low density, beta parameter, and temperature values have been found as characteristic in situ signatures of magnetic clouds (Zurbuchen and Richardson, 2006). Moreover, in several investigations, it has been suggested that magnetic clouds are related to the eruption of prominence material (Bothmer and Schwenn, 1994; Burlaga et al., 1998; Wang et al., 2018).

We also observed a clear and continuous He²⁺/H⁺ > 8% ratio, that is, the presence of alpha-particles. The alpha-particle population is identified, in this case, from the clear secondary peak in the 1D reduced distribution functions (see sixth panel from Figure 3) from the SPC instrument (VDF, Case et al., 2020). The primary peak is, usually, associated with the core protons in the plasma, and it is well-measured throughout this period. The secondary peak is observed at roughly twice the kinetic-energy-per-charge of the proton core peak. This is often expressed in terms of the proton-equivalent speed, V = Z / m V P , where Z and m denote the charge number and mass number of the species. The second peak is most likely alpha-particles (Z = 2, m = 4) that are co-moving with protons, hence the apparent magnitude V = 2 V P .

Alternatively, a secondary peak might be associated with a proton beam component in the solar wind (Marsch et al., 1982; Alterman et al., 2018). In Figure 4, we present the 82-s time series of the alpha speed (first panel) and alpha ratio (second) estimated by SPAN’s partial field of view. In the third panel, we show the alpha ratio computed with SPC. In the fourth panel, we present the magnetic field direction where ϕ _B is shown in black and θ _B in blue. In the fifth panel, we show B _R /|B| (shown in blue) and SPC-|V| for the VDF first (V₁) and second (V₂) peaks (in black) with the VDF maxima marked with black +. The VDF maxima were located computing the first and second derivatives of the VDF function.

In this case, we find the proton beam interpretation to be highly unlikely, as such secondary proton beams typically drift along the local magnetic field direction, relative to the core, with a drift speed that is at most on the order of the local Alfvén speed. Thus, the radial component of a proton distribution with a beam component would be expected to exhibit roughly Alfvénic or sub-Alfvénic drift speeds that modulate with the magnetic field angle. For the periods under consideration, the local Alfvén speed is roughly 100–200 km s⁻¹, which is comparable to the observed difference in the proton-equivalent speed and, coincidentally, to 2 V P . However, the apparent drift speed is not correlated with the magnetic field angle. We estimated a Pearson correlation coefficient of 0.46 between V ₂−V ₁ and B _R /|B|, where V ₁ and V ₂ are the speeds of the first and second peaks on the VDF. We furthermore note that although the proton and alpha distributions are only partially measured by the SPAN instrument at this time, this experiment offers corroboration that the alpha component was significantly enhanced during this period (see second and third panels of Figure 4).

The remote-sensing observations (from both spacecraft SolO and STEREO-A) show that the eruption occurred in two phases. The first one, a smaller outburst, began in the more southerly part of the prominence around 2021 April 23 at 21:25 UT, and the second, characterized by a substantial eruption, originated from the more northerly part of the prominence around 2021 April 24 at 02:32 UT. It is impossible to say from our observations whether these individual features seen in the EUV and in coronograph observations are really two distinct events or part of a single, more complex one.

For the sake of clarity in the text, we refer to the earlier, smaller outburst as the first eruption and the latter, substantial one, as the second eruption. As these features propagated into the field of view of the several heliospheric imagers we used in this study, they began to become less distinct. Moreover, by the time they reach the PSP spacecraft, they can no longer be distinguished as completely separate events.

During the event, the eruptions propagated to the backside of the Sun from the SDO point of view. In Figure 5, we show the modeled background magnetic field (see Section 2) in different views combined with FSI 304 Å images of the Sun. We note that the model is based on SDO/HMI data from the previous visibility of this region, so there is some uncertainty in the boundary conditions. Assuming that the general magnetic configuration is preserved, the model gives insight into the background magnetic field configuration on 2021 April 23–24. The top panels show the magnetic field configuration from the solar North Pole. The bottom panels are from PSP (left) and SolO (right) points of view.

In Figure 5, we also show several open (gray), closed (red), and short, low-lying (yellow) magnetic field lines obtained with the MAS model. Purple, green, and blue colored lines are related to the different structures tracked from the Sun to PSP following Figures 2, 3. These lines originate close to the boundary between open and closed field lines (see Figure 5 bottom left panel). The top right and bottom left panels of Figure 5 show these field lines bending forward from the solar back-side toward the front-side and also toward the solar equator. The low-lying lines barely stick out behind the limb in the bottom right perspective and are roughly at the same solar latitude as the erupting filaments (−25 to −45°).

The magnetic field model results predict a scenario in which both eruptions strongly interact with their surroundings. It also predicts the possible filament source region close to the boundaries between open and closed field lines. At the source region, the eruptions are expected to encounter a domain where the magnetic field is complicated, while at farther distances from the Sun, the structures are embedded in regions dominated by open magnetic field lines. In this case, the results also suggest the eruptions being deflected toward the ecliptic plane.

To relate features observed in FSI and Metis, COR2, and PSP, we constructed a height–time plot (Figure 13) that combines the FSI and COR2 data, with overlays from measured or extrapolated positions of the eruptions as observed in other instruments. The plot shows the average over a radial cut through the images along the primary propagation direction of the eruption, between about 108° and 122° in the heliocentric radial coordinate system (where 0° is at the Sun’s north pole and 90° points due east).

Above 4 R_⊙, we use COR2 observations to track the motion of the eruption, interpolated to match the spatial and temporal resolution of the FSI observations, and corrected to match the line of sight based on the known spacecraft locations and our 3D reconstruction (see Section 4.1). Note that the timing error introduced by SolO and STEREO-A’s different distances from the Sun is about 45 s, which is negligible compared to the uncertainty introduced by the observation cadence of COR2 and the assumptions underpinning the de-projection of the STEREO-A data to match SolO’s perspective, and are, therefore, neglected in this plot.

To relate the multiple observations, reconstructions, and extrapolations of the eruption from different instruments to one another and to validate that each tracking method has indeed produced self-consistent results, in a simple, unified view, we overlay our measurements of the position of the eruption on the height–time plot, including FSI (red +), COR2 GCS reconstructions (light blue *), Metis (orange □), and backward extrapolations from PSP’s position using the velocities measured in situ (first transient in magenta +, constant speed patch in green +, and the second transient in indigo +; these colors are selected to generally correspond to the colors of specific field lines in Figure 2 but must be lightened somewhat to be visible on the dark height–time image).

From the in situ observations, we analytically reconstruct the speed values from the PSP location back to 10–20 R_⊙. We track back three different patches of plasma: the first structure (vertical magenta dash line in Figure 3) that crossed PSP at 412 km s⁻¹, the second at a constant speed of 300 km s⁻¹ (delimit between the two vertical green dash lines), and the third structure at 310 km s⁻¹ (vertical indigo dash line).

We assume steady flows of adiabatic gas in each patch to solve the 1D spherical symmetric Bernoulli’s equation (integrated momentum equation) and obtain their speeds at ten different distances between 10 and 20 R_⊙. We neglect gravity forces, assume a polytropic gas (Shi et al., 2022, in our case, γ = 5/3), and include the molecular weight μ to consider the He²⁺/H⁺ density ratio measured by SWEAP, to get the following: v 2 = 2 γ γ − 1 k T μ m H 1 − v 0 r 0 2 v r 2 γ − 1 , where k is the Boltzmann constant, m _H is the proton mass, v ₀ and r ₀ are the speed and radial distance near the Sun (in this case between 10 and 20 R_⊙), and v and r are the speed and radial distance at the PSP location, respectively. We reconstruct analytically the information of the structures back to the Sun by assuming constant speed motion (see Figure 13).

In this case, we did not pursue running MHD numerical simulations (e.g., ENLIL (Odstrcil and Pizzo, 2009), PLUTO (Mignone et al., 2007; Mignone et al., 2012), and EUHFORIA (Pomoell and Poedts, 2018) numerical codes) by extrapolating the photospheric fields due to the difficulty of analyzing the event at its source region, observational limitations, and possible prediction of transit times to follow the eruption from the Sun to the PSP location with an analytical approach.

The analytical equations predict between 10 and 20 R_⊙: the passage of the first structure on 2021 April 24, 08:40 UT, propagating away from the Sun with a speed of 389 km s⁻¹, the constant speed structure on 2021 April 24, 10:00 to 18:00 UT, at 290 km s⁻¹, and the third transient on 2021 April 24, 23:00 UT, at 302 km s⁻¹.

Fitting the slopes of the two eruptions observed in the height–time plot, corresponding to the feature observed near the first sets of orange Metis points, we obtain a velocity of 335 ± 12 km s⁻¹. The second major eruption, around the second set of orange Metis points, has a velocity of 322 ± 12 km s⁻¹. A dark intermediate feature, corresponding to the third set of red points, is probably the trailing edge of the first eruption and has a slower speed, 230 ± 8 km s⁻¹.

The strong agreement between the different sets of observations suggests we are tracking the same features in all of our data. The velocities we measure from these height–time plots also agree well with the PSP-derived velocities, indicating that the propagation of the CME is not strongly influenced by interactions in interplanetary space.

Early in the first eruption, the prominence appears to follow the local magnetic field in the low corona, which is highly non-radial, and thus the radially plotted height–time diagram does not fully capture the prominence’s trajectory or velocity. Outside of the FSI field of view, the eruption moves primarily radially, with essentially constant velocity.

By combining the analytical model with the connectivity derived from the MAS model, it is clear that, for much of the propagation of the eruption, the CME is propagating constantly along the open field lines connected to PSP because of the close proximity of the initial prominence to these field lines. While the magnetograms used in the MAS model for the area around the eruption are necessarily somewhat out of date because the eruption occurred on the backside of the Sun with respect to Earth, all of the remote-sensing data supports the notion of a simple propagation originating in the south and following open field lines northward toward the equator.

We reported the propagation of a complex prominence eruption that reached PSP on 2021 April 25 at 01:00 UT when the spacecraft was located at 46 R_⊙. To study the full evolution of the event, we combined multi-spacecraft remote-sensing observations with the in situ measurements onboard PSP. The structure, as sampled by SPC, was characterized as a low temperature and low-density transient with complex magnetic field configuration and a He²⁺/H⁺ > 8% ratio indicating the presence of alpha particles identified from the clear secondary peak in the 1D reduced distribution functions moving at a velocity ranging between 550 and 650 km s⁻¹. Although the structure lacks a coherent magnetic field configuration, the rest of the characteristics observed are signatures common in magnetic clouds, particularly when related to prominences. We identified the complex prominence eruption as the source by tracking from the Sun to PSP location the propagation of the associated CME, which was only possible due to the nearly perfect quadrature of PSP and SolO.

We analyzed FSI, Metis, and COR2 images, tracked the structure evolution up to 20 R_⊙, and identified that the eruption occurred in two phases: a smaller outburst beginning in the more southerly part of a prominence and the substantial eruption originated from the more northerly part of the structure. Below 4 R _⊙, the set of remote-sensing observations showed that the structure is complex.

Above 20 R_⊙, the CME kinematics was modeled using the GCS reconstruction method over WISPR and COR2 coronagraph images and by modeling the background magnetic field using the MHD PSI/MAS model and backward analytic reconstruction starting from PSP in situ data. The strong agreement between the different sets of observations and models showed that the ICME propagated radially at a constant speed and that it was not strongly influenced by interactions in interplanetary space.

This work highlights the importance of studying the propagation of transients from a multi-spacecraft point of view, as their combination enables a better understanding of the phenomena by closing gaps between the sets, e.g., the distance ranges covered by the different instruments, the image dependence with the spacecraft location which also reverberates on the possibility to model the events from remote-sensing observations, and the limits of the current reconstruction models back to the Sun from in situ data. Particularly in this case, we were able to follow a complex structure that, during the first stages of propagation, it seems to be evolving non-radially before propagating at a constant rate. Moreover, the eruption conserves its plasma and magnetic properties up to 46 R_⊙.