Morphology of ejecta features from the impact on asteroid Dimorphos

Abstract

Hypervelocity impacts play a significant role in the evolution of asteroids, causing material to be ejected and partially reaccreted. However, the dynamics and evolution of ejected material in a binary asteroid system have never been observed directly. Observations of Double Asteroid Redirection Test (DART) impact on asteroid Dimorphos have revealed features on a scale of thousands of kilometers, including curved ejecta streams and a tail bifurcation originating from the Didymos system. Here we show that these features result naturally from the dynamical interaction of the ejecta with the binary system and solar radiation pressure. These mechanisms may be used to constrain the orbit of a secondary body, or to investigate the binary nature of an asteroid. Also, they may reveal breakup or fission events in active asteroids, and help determine the asteroid's properties following an impact event. In the case of DART, our findings suggest that Dimorphos is a very weak, rubble-pile asteroid, with an ejecta mass estimated to be in the range of (1.1-5.5)×10⁷ kg.

Fig. 1. Identification of features.

a HST image at T0 + 11.86 days (08 OCT 2022 19:57:00 UTC). Didymos system and ejecta features are visible in the central part of the figure. The oblique lines (bottom-right/-left part of the image) are artifacts. b Identification of the main ejecta features: northern (blue) and southern (dark green) arms of the spiral, tail (light green) and Didymos system (yellow). For better visualization of colored regions, a zoomed box is provided to the bottom-right part of the figure. Each frame is 6480 km wide, the inset at the bottom-right of (b) is 864 km wide.

Fig. 2. Time evolution of features.

a–d Spiral feature at T0 + 1.14 days (29 SEP 2022 02:28:00 UTC) and (e–h) tail at T0 + 11.86 days (08 OCT 2022 19:57:00 UTC). The panel reports duplicated pairs of images (left unannotated, right annotated): (a, b, e, f) HST images, (c, d, g, h) synthetic images from numerical simulation (ID 02, see Supplementary Table 1). Annotations report curved streams (s1-2), cone edges (c1-4), tail edges and bifurcation line (t1-3), linear features (l1-2). Each frame is 864 km wide. See Supplementary Video 1 for full video sequence.

Fig. 3. Close up view of the inner system (20 km field of view).

a–d Spiral motion in the inner system and tail feeding mechanism involving particles that are still near the Didymos system after 14 days. Simulated trajectories are shown, with no photometric adjustments applied. e–h Argument of perihelion of orbits of ejecta with radius in the range 1–10 cm is clustered along two directions ~50 deg apart. Ejecta of cm radius returning to the system (HST view) between 8 and 12 days after the impact. See Supplementary Video 3 for full video sequence and Supplementary Videos 4–6 for larger scale views (FOV of 100 km, 1000 km, 5000 km). Figures refer to simulation ID 02 (see Supplementary Table 1).

Fig. 4. Synthetic images showing the time evolution of spiral and tail features.

a Natural bifurcations in the tail and tail curvature appear evident starting from T0 + 2.13 days (29 SEP 2022 02:28:00 UTC). b–e Evolution of the bifurcation in the subsequent 4 days. Each frame is 864 km wide. Figures refer to simulation set ID 02 (see Supplementary Table 1).

Fig. 5. Examples of rejected simulations due to unmatched dynamical conditions.

e–h Show synthetic images associated to HST reference images in (a–d) respectively. Each frame is 2160 km wide. With reference to Supplementary Table 1, e is associated with simulation set ID 16, f with simulation ID 07, g with simulation set ID 03, and h with simulation set ID 21. The mismatch between synthetic and HST images is used to constrain the properties of the ejecta.

Fig. 6. Photometric error versus mass and dSFD power-law coefficient.

Simulations forming a the tail-related feature and b the cone-related features. We identify a mass in the range 1–5 × 10⁷kg with dSFD coefficient of −2.4 to produce tail-related features, and in the range 1–5 × 10⁶kg with dSFD coefficients between −3 and −2.7 to produce the cone-related features. The colorbar is restricted to maximum values reported in each figure. The error in the yellow region is equal or larger than these maximum values. Figures refer to simulation set IDs a 19 and b 02 (see Supplementary Table 1). Source data are provided as a Source Data file.

Fig. 7. Differential SFD.

Computed combining all ejecta particles involved in cone-related and tail-related features, and its best-fit value −2.7. Data refers to simulation set IDs 19 and 02 (see Supplementary Table 1). Source data are provided as a Source Data file.

Fig. 8. Coarse identification and labeling of features.

e–h, m–p, u–x Report noise masks in light blue, cone-related masks in yellow, and tail-related masks in green. Associated HST images are reported in (a–d, i–l, q–t). Each frame is 8640 km wide.

Fig. 9. High-fidelity identification and labeling of features.

e–h, m–p, u–x Reports the identification of the main ejecta features: northern (blue) and southern (dark green) arms of the spiral, tail (light green) and Didymos system (yellow). Associated HST images are reported in (a–d, i–l, q–t). Each frame is 4752 km wide.