Re-accretion of Giant Impact Ejecta Can Drive Significant Atmospheric Erosion on Terrestrial Planets

TL;DR

This work demonstrates that the long-term re-accretion of giant-impact ejecta can drive significant atmospheric erosion on terrestrial planets, potentially removing substantial portions of an atmosphere over tens of millions of years. By parameterizing debris ejection with $f_{ m esc}$ and $D_{ m max}$ and modeling collisional grinding and re-accretion, the authors show that even modest debris masses (as low as ~1% of $m_p$) can erode Earth's atmosphere by factors of several over 10–100 Myr depending on the debris size distribution and initial atmospheric properties. The Moon-forming analog is used as a reference case, revealing that re-accretion can erase a present-day Earth-like atmosphere within ~30 Myr for plausible values of $f_{ m esc}$ and $D_{ m max}$, and that such erosion is robust across a broad parameter space up to ~2 au. The findings imply that giant impacts have a substantial, previously underappreciated role in shaping the long-term atmospheric evolution and geochemical evolution of Earth-like worlds, and that ignoring debris re-accretion could underpredict atmospheric loss in planetary formation histories.

Abstract

Giant impacts, the collisions between planetary embryos, play a crucial role in sculpting the planets and their orbital architectures. Numerical simulations have advanced our understanding of these events, enabling estimations of mass and atmospheric loss during the primary impacts. However, high computational costs have restricted investigations to the immediate aftermath, limiting our understanding of the longer-term consequences. In this study, we investigate the effect of re-accretion of giant impact debris, a process previously overlooked, on the atmospheres of terrestrial planets. Following the collisional and dynamical evolution of the debris ejected during the primary impacts, we quantify the amount of debris that would be re-accreted by the progenitor. We find that $\sim 0.003\ M_{\oplus}$ would be re-accreted over a wide range of Earth-like planet properties, assuming $1\%$ of their mass is ejected as non-vaporised debris. Over a prolonged period, the secondary impacts during re-accretion drive enhanced atmospheric loss. Notably, the impacts from the debris of the canonical Moon-forming impact would have gradually eroded an atmosphere similar to present-day Earth's in $\sim 30$ Myr. More generally, any planet growing via giant impacts within $2$ au is likely to experience significant post-impact atmospheric erosion unless the initial atmosphere was at least $5$ times more massive than Earth's. Our results highlight the crucial role secondary impacts from giant-impact ejecta could have in driving the long-term atmospheric evolution of Earth-like planets, and demonstrate that giant impacts can be significantly more effective at eroding such atmospheres than previously thought, when re-accretion of debris is considered.

Figures (8)

• Figure 1: Top: Evolution of the ejected debris mass over time following a hypothetical Moon-forming impact with ${D_{\rm{max}}}=100\ {\rm{km}}$, according to \ref{['eq:mdisk_sol']}. The colors correspond to different values of ${f_{\rm{esc}}}$ (see legend), which determines the initial disk mass, ${M_{\rm{d0}}} = {f_{\rm{esc}}} {m_{\rm{p}}}$. Vertical dashed lines indicate the collisional depletion timescale (${\tau_{\rm{c}}}$, \ref{['eq:tau_c']}) for each ${f_{\rm{esc}}}$. Middle: Mass of debris reaccreted onto the planet as a function of time (\ref{['eq:m_reacc']}). The black dashed vertical line indicates the re-accretion timescale (${\tau_{\rm{acc}}}$, \ref{['eq:tau_acc']}), which does not depend on ${f_{\rm{esc}}}$. Bottom: Atmospheric mass loss over time due to the re-accretion of debris (\ref{['eq:dm_atm']}). For reference, the horizontal dashed line represents the atmospheric loss expected from a canonical Moon-forming impact 2020_Kegerreis.
• Figure 2: Top: Evolution of the debris mass over time (\ref{['eq:mdisk_sol']}). The colors correspond to varying values of ${D_{\rm{max}}}$ (see legend), assuming ${f_{\rm{esc}}}=0.01$. Vertical dashed lines indicate the collisional depletion timescale (${\tau_{\rm{c}}}$) for each ${D_{\rm{max}}}$. Middle: Mass of debris reaccreted onto the planet as a function of time (\ref{['eq:m_reacc']}). The black dashed vertical line indicates the re-accretion timescale (${\tau_{\rm{acc}}}$). Bottom: Atmospheric mass loss over time due to the re-accretion of debris (\ref{['eq:dm_atm']}). The horizontal dashed line indicates the atmospheric loss expected from a canonical Moon-forming impact 2020_Kegerreis.
• Figure 3: Fraction of atmosphere retained $1$ Gyr after a hypothetical Moon-forming impact with ${f_{\rm{esc}}} = 0.01$ and ${D_{\rm{max}}}=100\ {\rm{km}}$, as a function of the initial atmospheric mass (${m_{\rm{atm0}}}$). Filled circles represent an Earth-like, volatile-rich atmosphere with $\mu = 29$, and the filled diamonds represent an atmosphere with a primordial (solar) composition with $\mu = 2.35$. For completely depleted atmospheres (${m_{\rm{atm}}}/{m_{\rm{atm0}}} < 10^{-4}$), the time of depletion ($t_{\rm{atm,lost}}$) is indicated by the fill colors. Dotted lines connecting markers of the same $\mu$ values are included as visual guides.
• Figure 4: Amount of debris re-accreted over $10^{9}$ years as a function of planet mass (${m_{\rm{p}}}$) and semi-major axis (${a_{\rm{p}}}$). The planet is assumed to orbit a Sun-like star with an impact debris disk characterized by ${f_{\rm{esc}}}=0.01$ with ${D_{\rm{max}}} = 100\ {\rm{km}}$.
• Figure 5: Post-giant-impact atmospheric erosion as a function of planet mass (${m_{\rm{p}}}$) and semi-major axis (${a_{\rm{p}}}$) for varying initial atmospheric masses (${m_{\rm{atm0}}}$). Top panels: The colours represent the fraction of the initial atmospheric mass lost within $1$ Gyr following the primary giant impact for a planet with ${m_{\rm{atm0}}} = 10\ {m_{\rm{atm\earth}}}$ (left) and for a planet with ${f_{\rm{atm}}} = 10\ {f_{\rm{atm\earth}}}$ (right), with the black lines indicating $25\%$, $50\%$ and $75\%$ loss in each case. Bottom panels: The lines represent $50\%$ atmospheric erosion contours for different initial atmospheric masses (see contour labels). Note that the spacing between contours representing different atmospheric erosion percentages (not explicitly shown) also varies with ${m_{\rm{atm0}}}$, typically exhibiting narrower spacing at lower ${m_{\rm{atm0}}}$ and wider spacing at higher ${m_{\rm{atm0}}}$. Left panels: ${m_{\rm{atm0}}}$ is varied independent of ${m_{\rm{p}}}$. Right panels: ${m_{\rm{atm0}}}$ scales with ${m_{\rm{p}}}$ as ${m_{\rm{atm0}}} = {f_{\rm{atm}}} {m_{\rm{p}}}$. All scenarios assume a Sun-like host star and an impact-generated debris disk with ${f_{\rm{esc}}}=0.01$ with ${D_{\rm{max}}} = 100\ {\rm{km}}$.