Main

The Parker Solar Probe (PSP) carries an imaging instrument, the Wide-field Imager for Solar Probe (WISPR)15. The inset in Fig. 1a shows a WISPR inner telescope (WISPR-I) image taken on 6 November 2018 at the first perihelion. The Sun is 13.5° to the left of the image and the width is about 40°. The locus of points at the apex of the contours defines the photometric axis of the F-corona. While most observations of the F-corona (or zodiacal light) have been taken from 1 astronomical unit (au) away from the Sun, two spacecraft, Helios A and B, each carrying the Zodiacal Light Experiment5, orbited the Sun from 0.3 to 1.0 au, one observing above the ecliptic plane and the other below. They measured the intensity I of the zodiacal light from varying heliocentric distances and found20 that it increases towards the Sun according to I ∝ R☉−n, where n = 2.3 ± 0.1 and R☉ is the radius of the Sun. The upper and lower limits were recorded at small and large elongations from the Sun, respectively, and were independent of the ecliptic longitude of the observer. The Sun Earth Connection Coronal and Heliospheric Investigation4 (SECCHI) heliospheric imagers HI-121, onboard the STEREO spacecraft22 orbiting the Sun at approximately 1 au, observed the corona at elongations ranging from 0.07 to 0.45 au (5°–24°) from the Sun. An analysis of intensities23 of the photometric axis of the F-corona from 2007 to 2014 found the exponent, for the entire elongation range covered by the HI-1 instrument, to be 2.31. Moreover, the analysis performed on restricted elongation ranges23 showed an identical tendency to the Helios results for the intensity gradient to increase towards the Sun (n = 2.29 ± 0.10).

Fig. 1: Intensity plots along the photometric axis of the F-corona.
figure 1

a, Observed intensities from WISPR-I for five heliocentric distances as a function of elongation (degrees) scaled to the same value at 30° elongation. The inset in a shows an image of the F-corona taken on 6 November 2018. b, Observed intensities from both telescopes for five heliocentric distances as a function of elongation (solar radii). See text for further explanation.

Full size image

Figure 1a displays a log–log plot of a sample of F-coronal intensity profiles in units of mean solar brightness (MSB) along its photometric axis as measured by WISPR-I between 15° and 50° elongation from the centre of the Sun. The sample comprises data from five different heliocentric distances of the PSP spacecraft (0.336 au to 0.166 au) obtained during the orbit inbound to the first perihelion. Colour is used to distinguish the plots. For clarity we plot only these five positions, but all the profiles during the encounter are similar. These five profiles are normalized to the maximum intensity at 30° elongation to reveal the behaviour of the profiles for the various PSP heliocentric distances. Clearly, at larger elongations, the curves have exactly the same slope, and at shorter elongations (<20°), the intensity decreases with decreasing PSP distance, with the top plot (pink) for when PSP is the furthest from the Sun and the bottom plot (dark green) the closest. Figure 1b shows the intensity profiles at the same five PSP distances for both telescopes, but plotted against elongations converted to R☉. The conversion to R☉ was performed by dividing the elongation by half the angular size of the Sun at the respective PSP distance. We note that the curves all overlie each other now, even the decreases seen in Fig. 1a. The small upward ticks are due to bright stars. The dashed blue line in Fig. 1b shows the linear fit to the F-coronal intensities for elongations between 20R☉ and 77R☉ (n = 2.31), a result identical to that obtained from both earlier observations20,23. For comparison, historical data3,24,25 have been added. The dashed green line depicts the linear fit to the LASCO-C3 data3 (light green dots) for elongations greater than 13R☉, extrapolated down to 4R☉. The exponent, n, in this case is also 2.31 (note the match between the blue and green dashed lines). The LASCO-C3 data were normalized to the WISPR value at 20R☉. For WISPR the absolute calibration was determined by analysing the intensity of stars in the field, which resulted in an error of 12%. The relative accuracy and repeatability of the WISPR are excellent, which gives us high confidence in the turnover of the intensities below 17R☉. The historical measurements represented by the black dots24 and green dots3 both have absolute errors of 20%. On the other hand, no error was given for the data represented by the red dots25.

Figure 2 shows the K-corona from both telescopes on 6 November 2018, after removal of the brighter background from the dust scattering. Supplementary Videos 1 and 2 provide background-removed videos of the images taken during the first two encounters. The grid lines for both Fig. 2 and the Supplementary Videos are in the HPLN-ARC, HPLT-ARC coordinate system26,27. The videos show the evolution of a coronal streamer during the two encounter periods of PSP observations. On the large scale, the agreement with model predictions by our team (Extended Data Fig. 1) is very good, validating the model assumptions about the configuration of the magnetic field and the mass flux of the equatorial solar wind. The representation of WISPR-I images in a latitude versus time format (Extended Data Fig. 2a) reveals that near perihelion WISPR suddenly imaged faint coronal rays that are distinct from the main streamer rays. We note that fine structure along the streamer belt has been observed before19,28. High-resolution simulations of the corona reproduce these brightness features. We interpret their displacement to higher apparent latitudes to the spacecraft motion (Extended Data Fig. 2b). This striated ‘texture’ of the background corona is caused in our model by the spatial variability of coronal magnetic flux tubes, along which the plasma is heated and accelerated to form the slow solar wind.

Fig. 2: Combined images from the inner and outer telescopes of WISPR on 6 November 2018 at 01:44 ut.
figure 2

After removal of an empirical model of the F-corona, the faint solar wind structures are revealed. A faint streamer outlining the heliospheric current sheet is visible, as are faint, radial and diffuse rays, all with apparent origin on the Sun. The image also reveals the dust trail along the orbit of the asteroid 3200 Phaethon (delineated by the white dots). The Galaxy dominates the scene in the inner part of the outer telescope accompanied by two bright objects: Jupiter (to the upper right) and the star Antares (a little below to its left) in the Scorpius constellation.

Full size image

Supplementary Video 1 shows a series of ejecta along the streamer. A particular event characterized by a big magnetic flux rope followed by several smaller ones is shown in Fig. 3. The first one (yellow arrows) has an elliptical high-density envelope surrounding a quasi-circular density depletion at its centre and a striated envelope. Although similar structures have been observed by LASCO, they could not always be resolved. The Encounter 1 images were binned 2 × 2 pixels, giving an effective 2-pixel spatial resolution15 of 60 arcsec (at 0.21 au) for WISPR-I, which is about 2× finer than the LASCO-C3 observations of this event. On the other hand, LASCO-C2 tracked this structure when it was much closer to the Sun for a few images only, but with about 2.5× better spatial resolution. The event was also recorded by WISPR-O with an effective 2-pixel spatial resolution of 96 arcsec, extending the coverage of the event. Such density features have been interpreted as the boundaries of magnetic flux ropes12,16. We have combined the spatially resolved density information with modelling to locate the structures corresponding to the internal toroidal and poloidal magnetic fields and study their interactions with the ambient plasma as the structure expands inside the streamer rays. Preliminary work demonstrates (Extended Data Fig. 3) that the structures are indeed consistent with a force-free magnetic flux rope propagating along the heliospheric current sheet, which is quite flat during this period. The heliospheric current sheet flatness may be the reason for detecting the fine-scale structure of the event. Such behaviour is extremely rare in 1 au observations29.

Fig. 3: The propagation of a CME.
figure 3

Shown are five cropped frames from Supplementary Video 1 at different times in the same coordinate system as Fig. 2. The radial range is shown at the top, and the latitudinal range is 0° ± 10° for each panel. The yellow and red arrows point to structures described in the text.

Full size image

The event shown in Fig. 3 (on 1 and 2 November 2018) also shows two additional smaller flux ropes (red arrows) following the northern boundary of the main flux rope; these are probably by-products of the interaction between the main event and the ambient corona. The quasi-circular shape and faint striations within the feature strongly suggest that it is an idealized magnetic flux rope. Smaller features, with similar morphologies, are also seen following the main ejection. The yellow arrows follow the first, main event. The red arrows follow the two following events, until they become merged with the background, although they are still visible in Supplementary Video 1. These structures were not detected by either LASCO or SECCHI, although both have observed many small ejecta14. Small dense features caused by interaction of a coronal mass ejection (CME) with its environment have proved difficult to identify positively30. Further studies will be necessary to test that idea.

Observations during the second perihelion (Supplementary Video 2) again show new dynamics in a coronal streamer. In this case, the observations capture the formation of oblong structures consistent with magnetic islands. Magnetic islands are, in two dimensions, a collection of roughly elliptical magnetic field lines that close on themselves; or, in three dimensions, helical field lines wrapping around a central (guide) field, again with a roughly elliptical cross-section. These island structures are predicted to form via the tearing-mode instability17 from magnetic reconnection in a current sheet, such as the one within this streamer, where oppositely directed magnetic fields meet. Figure 4 shows several snapshots of this streamer and the formation and evolution of one of these oblong structures. This structure first appears at the inner edge of the image around 6 April 2019 20:00 ut and propagates out, within the current sheet, as an expanding, highly elliptical shape with a high-intensity (dense) ring of emission surrounding a low-intensity core. The final panel of Fig. 4 shows a measure of the aspect ratio (ratio of minor axis to major axis) of the ellipses fitted to this structure in each of the 33 frames from time 6 April 2019 19:57 ut to 7 April 2019 02:54 ut. Each ellipse was fitted to a set of points placed by hand on the high-density ring of the oblong structure in each frame. The corresponding ellipse is shown as a green curve in the five snapshots here, with a red dot at the centre of the ellipse. The plot indicates that the structure expands with a slightly increasing aspect ratio until 23:45 ut on 6 April 2019 and then it increases more quickly until the entire structure fades into the background. This evolution, including the increase in aspect ratio, is consistent with simulations of the tearing-mode formation of islands in an expanding coronal wind18. These simulations show the un-reconnected guide field collecting at the centre, forming this low-emission core, with the reconnecting field forming the high-density ring around the core. Although such an island ejection from a coronal streamer has been reported previously31, the earlier observations were not sufficiently resolved to show this internal ring and core structure.

Fig. 4: Formation and propagation of an island-like structure within a streamer.
figure 4

The first five panels show snapshots from WISPR-I during the second perihelion. An ellipse (green line) is fitted to the high-density ring of the structure in each panel. The final panel shows the aspect ratio (minor axis/major axis) of these ellipse fits versus time. The red and blue solid lines show a linear fit to this aspect ratio from 20:00 ut to 23:45 ut on 6 April 2019 (red) and from 00:00 ut to 02:40 ut on 7 April 2019 (blue). The pairs of dashed lines on either side of these fits show the 1-sigma values.

Full size image

WISPR imaged a variety of interesting structures in the corona/solar wind during the first two PSP orbits about the Sun. The departure from linearity of the F-corona intensity profiles below about 17R☉ is opposite to that found in both Helios and STEREO data. Although this behaviour could be leading to the predicted dust-free zone close to the Sun, the intensity decrease could be due to a change in the properties of the dust scattering, or a combination of the two. WISPR has certainly not observed the dust-free zone. Theoretical analyses of the plausible existence of a dust-free zone predict8,9,10,11 the formation of circumsolar dust bands that could be observed by their thermal emission. In a compilation of the 30 observations6 made at various wavelengths from 0.8 μm to 3.6 μm during eleven solar eclipses from 1966 through to 1998, about half indicated an enhancement and the other half, including the two latest eclipses in 1991 and 1998, did not. The resolution of whether this WISPR finding represents dust depletion or something else will have to wait until PSP steps down to lower perihelia.

The near-corotation of PSP allows us to observe the radial outflow of the solar wind, without the confusing impact of solar rotation. The observations suggest that many small ejecta, commonly called ‘blobs’, may indeed be magnetic flux ropes but are usually too small to identify as such from 1 au (ref. 32). Structures larger than these are generally interpreted as CMEs, but the physical mechanism of formation may not be the same. This finding, particularly with the anticipated measurements of the same structures by PSP’s in situ payload, may finally clarify the evolution of the CME magnetic structure in the heliosphere, opening up avenues of research on internal CME dynamics. As PSP steps closer to the Sun over the next five years, these observations, together with the modelling, will certainly provide insights and opportunities to study and separate the spatial and temporal variability of the solar wind near its source and will probably increase the performance of space weather prediction schemes. This will benefit a wide range of communities from basic physics research to space situational awareness to even astrophysics through exoplanet habitability applications.

Methods

WISPR contains two telescopes and measures the intensity of the visible light corona in addition to stellar and galactic sources. The two telescopes slightly overlap and have a combined field-of-view of 13.5° to 108.5° from the Sun, corresponding to approximately 9 to 78 solar radii (R☉; 1R☉ = 696,000 km) at 35.6R☉ perihelion. The visible light corona consists of two components: light scattered by free electrons (the K-corona) and light scattered by interplanetary dust (the F-corona). The F-corona of each WISPR image is removed using a technique36 similar to that developed for the SECCHI/HI-1. The primary difference is that for the HI-1 images the initial step in the procedure analysed the horizontal lines in the image, whereas here, the initial step uses the vertical lines in the images.

All of the data presented here have been calibrated into Mean Solar Brightness (MSB) units. The calibration details will be published in a future paper, but include the removal of geometric distortion, vignetting, instrumental artefacts (stray light, and so on) and then applying the photometric calibration of the system. The vignetting is caused by two sources: the projection of the image onto the two-dimensional plane of the Advanced Pixel Sensor detector and for WISPR-I the obscuration of the objective lens of the sunward side of the image by a series of baffles (including the PSP heat shield) which are used to block the solar disk illumination and block diffraction from the edges of the preceding baffles. The absolute calibration is confirmed on-orbit by measuring the intensity of stars passing through the field. The intensity of the stars as they transit across the image is also a check on the vignetting correction.