A bimodal distribution of haze in Pluto's atmosphere

The study addresses whether haze particles in chemically reducing atmospheres (like those of Pluto, Titan, and Triton) have unimodal or multimodal size distributions. While Earth and Venus (oxidizing atmospheres) exhibit bimodal/multimodal aerosol distributions attributed to multiple sources and dynamics, reducing atmospheres have been commonly modeled with unimodal haze distributions due to limited constraints. Pluto’s atmosphere, similar in composition and pressure regime to parts of Titan’s and Triton’s, provides an opportunity to probe haze formation pathways. New Horizons observations indicate both blue color (implying small-particle Rayleigh-like scattering) and strong forward scattering (implying larger particles), hinting at complex particle morphology. The research aims to determine the haze size distribution and morphology in Pluto’s lower atmosphere and to test whether a unimodal population can explain observations across UV to IR and a range of phase angles.

Prior observations and interpretations suggested Pluto’s haze may be fractal aggregates of micron size comprising ~10–20 nm monomers, similar to Titan’s haze. However, purely aggregate-based models underpredict observed backscattering. On Triton, backscattering has been attributed to surface reflection and clouds, but haze particle morphology may differ. Previous works proposed various size distributions (log-normal, bimodal, power-law) and sometimes invoked surface reflection, yet degeneracies remained, particularly in backscattering. Numerical simulations for Titan’s haze indicated a transient bimodal distribution near ~0.5 Pa due to particle dimensional transition (spheres to 2D aggregates). This literature underscores the need for joint, multi-wavelength, multi-geometry analyses to resolve modality in reducing-atmosphere hazes.

The study jointly analyzes New Horizons observations of Pluto’s haze across UV to IR and multiple phase angles: UV extinction from Alice (0.185 µm); visible limb-scattered intensity from LORRI (0.608 µm) at phase angles 19.5°, 67.3°, 148.3°, 169.0°; IR forward-scattered spectra from LEISA (1.235–2.435 µm) at high phase; and visible/near-IR scattering from MVIC color filters (0.492, 0.624, 0.861, 0.883 µm) at multiple phase angles. Line-of-sight measurements (optical depth, I/F) were converted to local extinction and scattering coefficients using an Abel-transform-based linear inversion tailored for noisy data, with extrapolation above the measurement range assuming exponential decay in geopotential. The analysis focused on the lower ~50 km (pressures ~0.5–1 Pa) where all instruments provided coverage. Light scattering by particles was modeled using Mie theory for spheres and monomers, and an empirically validated fractal-aggregate scattering model (after Tomasko et al.) for aggregates, assuming tholin-like optical constants; sensitivity tests to refractive indices (including organic ices) were performed. Surface reflection and potential secondary scattering were estimated using a Hapke model with HEALPix discretization to quantify upper limits of surface contributions. A Markov-chain Monte Carlo (emcee) retrieval framework was employed to fit observations with multiple haze scenarios, varying particle morphology parameters (fractal dimension Df, monomer radius rm, monomer number per aggregate Nm, aggregate number density) and, where applicable, sphere radius and number density. Ten scenarios were tested, including monodispersed aggregates, aggregates plus surface reflection, aggregates plus spheres (bimodal), two-sphere populations, and various distributions (log-normal, power-law, exponential) for spheres or 2D aggregates. Goodness-of-fit was evaluated via the posterior probability across altitude profiles.

• A unimodal haze population (monodispersed aggregates or other single-population distributions) cannot simultaneously match UV extinction, visible backscattering, and IR forward scattering. Monodispersed ~1 µm fractal aggregates (rm ~20 nm) reproduce UV extinction and forward scattering but underestimate backscattering by ~factor 3 (LOS I/F shortfall ~5×10^-3 in lower 50 km).
• Surface reflection secondary scattering is insufficient: modeled upper-limit I/F contribution <2.5×10^-3 at relevant phase angles, at least an order of magnitude below that needed to explain the backscattering deficit.
• A bimodal haze distribution best fits all datasets: (1) large two-dimensional fractal aggregates with effective radius ~1 µm composed of ~20 nm monomers; (2) small spheres with radius ~80 nm. Aggregates dominate forward scattering; small spheres dominate backscattering at visible wavelengths. Both populations contribute comparably to UV extinction.
• Retrieved number densities are ~0.3 cm^-3 for aggregates and ~10 cm^-3 for small spheres between ~50–15 km altitude; despite differing number densities by ~two orders of magnitude, mass densities of the two populations are comparable (aggregate mass slightly greater, within a factor of ~2).
• Vertical profiles of particle parameters are nearly constant from 50 to 15 km; below ~15 km, aggregate fractal dimension increases, suggesting compaction or morphological change possibly due to cooling/condensation.
• Alternative distributions (two-sphere bimodal, and log-normal/power-law/exponential for spheres or 2D aggregates) have significantly worse fits; inclusion of IR LEISA spectra breaks degeneracies by constraining the large-aggregate size via the IR forward-scattering slope and excluding spheres-only solutions due to overly steep IR/UV scaling.
• The pressure regime of Pluto’s lower atmosphere (~0.5–1 Pa) corresponds to the particle dimensional transition predicted in Titan haze simulations (~0.5 Pa), supporting a common photochemical microphysical mechanism for bimodality in reducing atmospheres.

The findings directly address the question of haze modality in reducing atmospheres, providing observational evidence for bimodal particle populations on Pluto. The morphology and sizes align with Titan models that predict a dimensional transition from spheres to 2D aggregates near ~0.5 Pa, implying a shared microphysical pathway. For Titan’s detached haze, the Pluto result favors a photochemical origin over a purely dynamical lifting mechanism; detection of bimodality in Titan’s detached layer (e.g., via backscattering-sensitive observations from Cassini ISS) would further discriminate between scenarios. For Triton, past interpretations invoking spherical hazes and clouds plus surface reflection could be revisited to consider aggregate plus small-sphere bimodality, potentially explaining strong backscattering without excessive reliance on surface contributions. Radiatively, the newly identified small-sphere population likely alters Pluto’s energy budget: small (~80 nm) particles have short radiative relaxation but slow gas-collision heat exchange, implying they may be cooler than the ambient gas and could contribute disproportionately to mid-IR cooling signatures, possibly modifying expected mid-IR emission features relative to prior predictions. Gas condensation and composition changes (e.g., organic ices) appear insufficient to reproduce backscattering without a second particle population and have limited impact on the inferred morphology within the observed wavelength/geometry range.

This study demonstrates that Pluto’s haze possesses a bimodal particle distribution comprising ~1 µm two-dimensional fractal aggregates built from ~20 nm monomers and a second population of ~80 nm spheres. This morphology robustly reproduces New Horizons observations spanning UV extinction, visible phase functions at multiple scattering angles, and IR forward scattering. The result substantiates the occurrence of particle dimensional transitions in reducing atmospheres near ~0.5–1 Pa and supports a photochemical origin for Titan’s detached haze. It also motivates reanalysis of Titan and Triton haze observations for evidence of multimodal distributions and encourages refined radiative transfer and energy budget models that include small-particle cooling effects. Future work should: (1) analyze Titan ISS datasets focusing on backscattering of the detached layer; (2) revisit Triton’s haze with aggregate-plus-small-sphere frameworks; (3) obtain mid-IR observations (e.g., with JWST/MIRI) to test cooling signatures; and (4) pursue laboratory measurements of Pluto/Triton haze analog optical constants to reduce modeling uncertainties.

• Observational constraints are strongest in Pluto’s lower ~50 km; inversion requires extrapolation above instrument altitude limits and assumes spherical symmetry.
• MVIC panchromatic data were excluded due to calibration/coverage issues; backscattering MVIC color data have large uncertainties.
• Surface reflection estimates are simplified to provide upper limits (e.g., specular-plane assumption) and may not capture full bidirectional reflectance complexity.
• Haze optical properties adopt tholin-like refractive indices; while sensitivity tests indicate limited impact on phase functions at these sizes, true composition (including organic ices and oxygen-bearing organics) remains uncertain and affects detailed radiative budgets.
• Retrieval degeneracies are reduced but not eliminated; while monodispersed bimodal populations fit best, narrow distributions around the retrieved sizes are possible.
• Results reflect conditions during the 2015 flyby; Pluto’s seasonally varying pressure/temperature may alter haze morphology and composition over time.