The Dimorphos ejecta plume properties revealed by LICIACube

LICIACube, a 6U CubeSat of the Italian Space Agency (ASI), was carried by NASA’s DART spacecraft and deployed on 11 September 2022 to image the kinetic impact with Dimorphos and its consequences. During the post-impact fly-by, LICIACube acquired and returned 426 scientific images with phase angles from 43° to 118°, using the LEIA and LUKE cameras. The science phase began 71 s before impact, at 1,466 km from Dimorphos, and continued until 320 s after impact, with closest approach at ~58 ± 2 km about 167 s after impact. The purpose was to characterize the ejecta plume’s geometry, morphology, color, and dynamics from a unique vantage point distinct from Earth- and HST-based observations, and to further constrain Dimorphos’ size and shape.

Prior to these observations, ground-based photometry showed an 8.3× brightening of the Didymos system after impact, decaying to pre-impact levels after 23.7 days (ref. 2). HST resolved the ejecta evolution from 15 minutes to 18.5 days post-impact at ~2.1 km px⁻¹, revealing complex structures typical of asteroid impact events (ref. 3). The DART momentum enhancement factor was estimated between 2.2 and 4.9 from the measured orbital period change, contingent on assumptions about Dimorphos’ mass and density (ref. 4). Comparative context comes from Deep Impact observations of comet 9P/Tempel 1, which exhibited inner-boundary ejecta velocities of ~80 m s⁻¹ and peak early ejecta speeds of ~300 m s⁻¹ (refs. 10, 11). Additional analogies are drawn to cometary plume color changes and dust fragmentation processes observed in comet 73P/Schwassmann-Wachmann 3 and studies of cometary dust dynamics (refs. 13–17).

- Spacecraft and instruments: LICIACube used two cameras—LEIA and LUKE. LEIA observed an intensity increase by ~5× in DN integrated over a fixed area comparing pre- and post-impact frames. LUKE acquired image triplets with different exposure times beginning 29 s after impact. Both instruments tracked the event through 320 s post-impact. The closest approach occurred ~167 s after impact at ~58 ± 2 km. - Geometry and cone retrieval: Assuming an axisymmetric ejecta cone, six images (five post-CA side-on, one pre-CA head-on) were analyzed. Using geometric constraints and spacecraft viewing geometry, the cone aperture and axis direction were solved. - Dimorphos shape constraint: Computer vision algorithms were applied to images with different exposure times to estimate the non-illuminated cross-sectional area of Dimorphos’ non-impacted hemisphere (~5,300 m² ± ~200 m²), validating dimensions retrieved from DART images. - Morphology tracking: High-resolution LUKE frames were used to map filamentary streams, arm-like structures, clumps, nodules, and discontinuities. Features were tracked across frames (e.g., between T+106 s and T+118 s), measuring projected displacements and applying field-of-view (FOV) projection corrections to derive velocities. - Color analysis: For a selected pre-CA LUKE triplet (0.5 ms, 4 ms, 20 ms exposures), red:blue and green:blue flux ratios were computed from RGB channels. Saturated regions in longer exposures were masked to study outer plume regions. Signal-to-noise assessments and uncertainty maps accompanied the ratio maps. - Image handling: Images were rotated and recentered on Dimorphos; Didymos was masked where necessary. Saturation in central plume regions was acknowledged; a shadow-cast dark arc between the plume and Dimorphos was identified in post-CA geometry due to an optically thick plume.

- Ejecta cone geometry: Aperture angle 140 ± 4°. Cone axis pointing toward RA ~137°, DEC ~+19° (J2000), consistent with independent analyses. - Color gradients: Inner plume appears blue, becoming progressively redder with distance. In a medium-exposure image, average red:blue flux ratio was ~0.57 in the inner part vs ~0.96 in the outer part. Green:blue ratios showed less contrast overall; inner filamentary streams showed relatively greener tones against a bluer inner background. - Morphology and structure: The plume exhibits numerous filamentary streams (18 main filaments identified at ~154 s post-impact within the inner ~250 m), arm-like structures with curving ends extending 6–8 km, discontinuities, bifurcations, bright nodules, and resolved clumps indicative of larger ejected components or aggregates/fragmenting grains. A shadowed dark arc between plume and Dimorphos indicates an optically thick plume casting a shadow. - Velocities: Pre-CA measurements show inner boundary feature speeds comparable to Deep Impact (~80 m s⁻¹). Specific examples include projected radial velocities of 67 m s⁻¹ and 47 m s⁻¹ for two filaments between T+106 s and T+118 s; a persistent clump (C10) at ~75 m s⁻¹; a bright clump (C3) at ~29 m s⁻¹. The earliest and most distant structures (T+34 s) include S1 (15.4 km from Dimorphos, 1.5 km length) with FOV-corrected 420–490 m s⁻¹, and S2 (11.7 km, 3.2 km length) with 290–400 m s⁻¹—substantially faster than HST measurements at ~2 h post-impact and consistent with early Deep Impact ejecta speeds. - Dimorphos size constraint: The non-illuminated cross-sectional area of the non-impacted hemisphere is ~5,300 m² (± ~200 m²), consistent with DART-derived dimensions. - Photometric change: LEIA images show ~5× increase in DN over a fixed area immediately after impact (local scene), aligning qualitatively with the overall post-impact brightening reported from ground-based observations (8.3×).

The LICIACube observations directly characterize the early-time ejecta plume from the DART impact, addressing the key question of ejecta geometry, composition proxies, and dynamics. The wide (140°) cone and axis orientation define the initial ejecta distribution and suggest that, if axisymmetric, directly ejected material could marginally intercept Didymos’ surface, while slower ejecta dynamics may deliver material over time. The complex, filament-rich and inhomogeneous structure indicates collimated radial outflows and heterogeneous particle populations, including larger blocks and aggregates undergoing fragmentation. The observed inner-blue to outer-red color gradient can be explained by either an abundance of sub-micron grains near the source, subsurface (less weathered) blue material ejected later than redder weathered surface material, or progressive reddening from fragmentation and dust-physics processes analogous to cometary phenomena. Velocity measurements up to ~500 m s⁻¹ for the earliest ejecta are consistent with high initial speeds seen in other impact events (e.g., Deep Impact), while inner-region speeds of tens of m s⁻¹ align with inner boundary ejecta velocities. Together, these results refine models of impact ejecta dynamics, particle size distributions, and radiative properties soon after impact, complementing HST and ground-based datasets and improving understanding of momentum transfer and plume evolution.

LICIACube provided a unique, close-up, multi-geometry view of the DART impact on Dimorphos, enabling the derivation of a wide ejecta cone (140 ± 4°), its axis orientation, detailed filamentary and clumpy plume morphology, early-time ejecta velocities from tens to ~500 m s⁻¹, and spatial color gradients (inner blue to outer red). The data also independently constrained Dimorphos’ cross-sectional area in the non-impacted hemisphere, consistent with DART results. These findings offer critical constraints for ejecta physics, particle properties, and impact momentum-transfer models. Future work should couple these observations with detailed numerical modeling of plume formation and evolution, integrate with longer-timescale observations (e.g., HST and ground-based), and investigate the mechanisms governing color changes and fragmentation to further elucidate ejecta composition and dynamics.

- Assumption of an axisymmetric ejecta cone may not capture asymmetries apparent under different viewing geometries. - Temporal coverage is limited to 71 s pre-impact through 320 s post-impact; later-time evolution relies on other facilities. - Image saturation in central plume regions and shadows from optically thick ejecta complicate photometry and color analysis. - Projection effects and FOV corrections introduce uncertainties in velocity estimates; determining true 3D velocities requires assumptions about geometry. - Instrumental factors (e.g., LEIA defocus discovered in flight) and masking of saturated/bright regions may affect measurements and feature tracking.