Earth’s atmosphere protects the biosphere from nearby supernovae

The study investigates how Earth’s atmosphere and climate respond to radiation increases from nearby supernovae, which are expected to occur within ~100 parsecs roughly once per million years. Prior hypotheses proposed that nearby supernovae could deplete stratospheric ozone, elevate surface ultraviolet radiation, increase aerosols and cloud cover, and trigger climatic cooling with possible links to mass extinctions. The authors evaluate these potential impacts using a comprehensive Earth system modeling framework that includes atmospheric chemistry and aerosol-cloud interactions, aiming to quantify ozone, aerosol, cloud, radiative, and potential biospheric effects under realistic geomagnetic and heliospheric conditions. They also examine implications for a low-oxygen atmosphere representative of early terrestrial life to assess evolutionary constraints.

Early conceptual and low-dimensional modeling studies (e.g., Ruderman 1974; Whitten et al. 1976; Aikin et al. 1980) suggested significant ozone losses from supernova-driven ionization. Later two-dimensional chemistry-climate studies (Crutzen & Brühl 1996; Gehrels et al. 2003) found less severe depletion, indicating catastrophic effects would require extremely close (≤8 pc) events, for which geological evidence is lacking. Thomas (2018) argued increased UV from supernova-induced ozone loss may affect species abundances but likely not cause mass extinctions, whereas some studies (e.g., Eide et al.) proposed that ionizing radiation from ~20 pc events could increase extinction risks. Separately, links between galactic cosmic rays (GCRs) and clouds/climate have been debated (Kirkby 2007; Mironova et al. 2015). Svensmark and colleagues suggested strong GCR influences on cloud condensation nuclei (CCN), climate, and biodiversity across the Phanerozoic, including large potential effects from nearby supernovae. The present work builds on this literature with a comprehensive Earth system model including state-of-the-science aerosol nucleation (CLOUD-based) and detailed ionization schemes to reassess atmospheric sensitivity to a 100-fold GCR increase.

The authors use a comprehensive Earth System Model with Atmospheric Chemistry (EACM/EMAC ECHAM/MESSy framework, MESSy v2.54.0) to simulate atmospheric responses to a nearby supernova. Key components include: (1) Detailed gas-phase and heterogeneous chemistry from the surface to the mesosphere, including catalytic ozone cycles (NOx, HOx) and polar stratospheric cloud processes via the MSBM submodel to represent heterogeneous reactions of ozone-depleting compounds. (2) Aerosol microphysics: an aerosol nucleation submodel parameterized with CERN CLOUD measurements to compute neutral and ion-induced new particle formation (NPF) rates; growth to CCN sizes is simulated additionally with the University of Helsinki multicomponent aerosol model for neutral and charged particles (ion-UHMA; 60 size sections from 1.8–100 nm), including ion dynamics, ion-aerosol interactions, condensation, coagulation, and deposition. Two sets of box-model simulations bracket sulfuric acid vapor conditions representative of the lower troposphere (0.1–5 × 10^-5 molecules cm^-3). (3) Cosmic ray ionization: a GCR ionization scheme as a function of atmospheric depth and geomagnetic cut-off rigidity, with heliospheric modulation and geomagnetic coefficients (IGRF 2010) and cut-off rigidity calculated per established methods; values between tabulated points are interpolated. Table 1 provides monthly heliospheric modulation and geomagnetic coefficients used. (4) Supernova forcing experiments: the ion-pair production rate is increased by a factor of 100 relative to control simulations for 1–2 years to represent periods of elevated GCR intensity from a nearby supernova; ionization is applied as a function of atmospheric column depth accounting for zenith angle up to 85°. (5) Process analysis: Ion lifetimes, recombination, and competition between nucleation and growth are assessed. High ionization shortens ion lifetimes (controlled by ion-ion recombination), leaving only a small fraction of ions available to form HSO4− and nucleate H2SO4–HSO4− clusters, limiting the efficiency of ion-induced nucleation. (6) Scenarios: Present-day atmospheric composition without anthropogenic ozone-hole drivers and a paleo-sensitivity case with 2% atmospheric O2 (early Cambrian-like) to assess biospheric implications in a low-oxygen atmosphere. Outputs include NOx, HOx, O3 profiles and columns, H2SO4, NPF rates (1–17 nm), aerosol number, CCN at cloud base (0.2% supersaturation), cloud cover, and radiative fluxes/forcing at TOA and surface.

• Gamma-ray bursts from nearby supernovae are strongly attenuated before reaching the lower stratosphere, implying small direct impacts on atmospheric chemistry.
• A 100-fold increase in GCR: NOx increases throughout the troposphere and stratosphere; HOx decreases in the stratosphere and moderately increases in the troposphere. Due to compensating catalytic cycles, net ozone depletion is moderate.
• Ozone: Seasonal local maximum stratospheric ozone depletion is about 30% over polar regions, comparable to present-day polar ozone loss from anthropogenic emissions. Global mean depletion is much smaller (order ~10%), with ozone increases in the tropics; most losses occur at high latitudes due to weaker geomagnetic shielding and dynamical isolation. Present-day Antarctic ozone-hole losses (60–70%) exceed the simulated supernova-induced depletion.
• Ozone column metrics: Present-day global mean ~300 DU; in a 2% O2 atmosphere (Cambrian-like), the nominal ozone column is ~100 DU, which already absorbs sufficient UV for terrestrial life; supernova-driven changes remain moderate due to vertical redistribution of ozone formation to lower altitudes.
• Aerosols and CCN: Global mean CCN at cloud base increase by ~20% (0.2% supersaturation), with regional increases up to ~100% over remote oceans with low background CCN. Despite a 100× ionization, modeled changes in H2SO4, NPF rates (1–17 nm), and total aerosol number are modest owing to short ion lifetimes and recombination limiting ion-induced nucleation.
• Clouds and radiation: Increased CCN leads to enhanced cloud reflectivity; estimated top-of-atmosphere radiative forcing is comparable in magnitude but opposite in sign to current anthropogenic forcing, implying a transient cooling of similar order. Changes in surface solar radiation are only a few percent globally, largest at high latitudes.
• Biospheric exposure: Background cosmic-ray effective dose ~0.35 mSv/year; a 100× increase implies ~35 mSv/year. While elevated, epidemiological constraints on long-term chronic exposure at this level are limited; direct health impact assessment is outside the study scope.
• Overall, impacts on global ozone, UV at the surface, and climate are modest; Earth’s atmosphere effectively shields the biosphere from nearby supernovae.

The simulations show that even a large (100-fold) increase in cosmic-ray ionization yields only moderate stratospheric ozone depletion, largely constrained to polar regions due to geomagnetic shielding and stratospheric circulation. Catalytic chemical feedbacks involving NOx and HOx both destroy and regenerate ozone, reducing net losses. In the tropics, ozone can increase, further limiting global mean depletion and surface UV increases to small values (a few percent on average). Aerosol and CCN increases are buffered by ion recombination and competition between nucleation and growth, leading to modest changes in aerosol burdens. The associated radiative forcing from cloud adjustments is negative and of a magnitude comparable to, but opposite in sign from, current anthropogenic forcing, implying cooling rather than catastrophic climate perturbation. In a low-oxygen Cambrian-like atmosphere, baseline ozone columns are smaller (~100 DU) but still sufficient to absorb harmful UV; supernova-induced ionization causes only moderate additional changes due to vertical redistribution of ozone production. Together, these results indicate that Earth’s atmosphere and geomagnetic field substantially mitigate the atmospheric and climatic impacts of nearby supernovae, making mass-extinction-level consequences unlikely under the modeled conditions.

Using an Earth system model with comprehensive chemistry and ion-induced nucleation parameterized from CLOUD, the study finds that radiation from a nearby supernova—modeled as a 100-fold increase in cosmic rays—produces moderate, regionally confined ozone depletion (seasonal polar maxima ~30%), small global changes in surface UV, and modest aerosol and CCN increases that yield a negative radiative forcing comparable in magnitude to current anthropogenic forcing. Gamma-ray bursts are strongly attenuated and have small atmospheric effects. A low-oxygen (2% O2) Cambrian-like atmosphere exhibits similarly moderate responses owing to compensating chemical and dynamical effects. Overall, Earth’s atmosphere and magnetic shielding effectively protect the biosphere from severe supernova impacts. Future work should quantify event-specific scenarios (e.g., different distances, spectra, and interstellar magnetic configurations), explore longer timescales and coupled ocean–biosphere feedbacks, and assess biological and health impacts of elevated chronic radiation exposure with improved dosimetry and epidemiological constraints.

• The simulations idealize supernova forcing as a uniform 100× increase in GCR ionization for 1–2 years; real events may differ in duration, spectrum, anisotropy, and timing.
• The study does not comprehensively assess direct health risks; radiation dose implications are mentioned but not modeled for biological outcomes.
• The low-oxygen scenario focuses on atmospheric composition and ozone but does not represent Cambrian climate or full paleoenvironmental conditions.
• Model uncertainties include parameterizations of ion-induced nucleation, sulfuric acid availability, ion-ion recombination, and aerosol–cloud interactions.
• Potential extremes (e.g., aligned interstellar magnetic fields increasing ground-level flux) are noted but not explicitly simulated.
• Results may be sensitive to geomagnetic field strength/configuration and heliospheric modulation choices; only one set of coefficients (IGRF 2010, Table 1) is used.