Sonora Diamondback Atmosphere Models

• The model couples a one-dimensional radiative-convective atmosphere to thermal evolution, incorporating refractory clouds and non-solar metallicity to yield self-consistent emergent spectra and cooling tracks.
• It defines a cloudy, warm, chemical-equilibrium regime for L and early T dwarfs, demonstrating how cloud opacity and composition alter temperature profiles and spectral features.
• The grid spans T_eff=900–2400 K, log g=3.5–5.5, and varying f_sed values, providing a comprehensive benchmark for spectral fitting and atmospheric retrieval studies.

Sonora Diamondback denotes the cloudy atmosphere-and-evolution tranche of the Sonora substellar model program, designed for warm substellar objects in which condensate clouds remain astrophysically important, including roughly L dwarfs, early T dwarfs, and directly imaged giant planets in the same temperature regime. Its defining feature is the self-consistent coupling of a one-dimensional radiative-convective atmosphere model to thermal evolution calculations while explicitly incorporating refractory clouds and non-solar metallicity, so that atmospheric structure, emergent spectra, colors, and cooling tracks are computed within a single framework rather than as disconnected products (Morley et al., 2024).

1. Place within the Sonora model sequence

Within the Sonora series, Diamondback occupies the cloudy, warm, chemical-equilibrium branch. Sonora Bobcat established the coupled cloudless atmosphere-and-evolution architecture in rainout chemical equilibrium across a broad ultracool parameter space (Marley et al., 2021). Sonora Cholla then added cloud-free, solar-metallicity chemical disequilibrium driven by vertical mixing (Karalidi et al., 2021), while Sonora Elf Owl extended self-consistent disequilibrium modeling to varying metallicity and C/O ratio, again in cloud-free atmospheres (Mukherjee et al., 2024). Diamondback is the corresponding warm cloudy grid: it targets the regime where silicate, iron, and alumina condensates materially affect the pressure-temperature structure, spectra, and inferred cooling history (Morley et al., 2024).

This placement matters because Diamondback is not a generic brown-dwarf grid. It was built specifically for objects whose observable properties are strongly shaped by refractory clouds, and it therefore addresses the domain in which cloud-free grids are known to become inadequate, particularly across the L-dwarf sequence and into the early T regime (Morley et al., 2024). A later extension, Red Diamondback, carried the same atmosphere-evolution logic to hotter and younger objects by stitching Diamondback to SPHINX cloud-free M-type atmospheres across $2000$–$2400$ K, thereby extending the usable boundary conditions for early brown-dwarf evolution (Davis et al., 9 Oct 2025).

2. Atmospheric and evolutionary formalism

Diamondback atmospheres are computed in radiative-convective equilibrium and chemical equilibrium. Thermal radiative transfer follows the Sonora/Marley framework, using the source-function technique with correlated-$k$ opacities. The opacity set includes updated molecular and atomic line lists relevant to warm brown dwarfs, including $\mathrm{H_2O}$, $\mathrm{CH_4}$, CO, NH$_3$, TiO, VO, FeH, alkalis, H$^-$, and CIA opacities (Morley et al., 2024). In this formulation, the atmospheric structure and the chemistry are solved self-consistently, but the chemistry is assumed to remain at equilibrium throughout the atmosphere.

Clouds are treated with the Ackerman & Marley mass-balance framework, in which vertical mixing competes with sedimentation. Diamondback writes the cloud transport balance as

$$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$

where $K_{zz}$ is the eddy diffusion coefficient, $w_*$ the convective velocity scale, $2400$0 the total vapor-plus-condensate mixing ratio, $2400$1 the condensate mixing ratio, and $2400$2 the sedimentation efficiency (Morley et al., 2024). Smaller $2400$3 yields loftier, thicker, more vertically extended clouds; larger $2400$4 yields more efficient settling and thinner clouds. Cloud formation is triggered by the condensation condition

$2400$5

and the refractory condensates emphasized in Diamondback are Mg$2400$6SiO$2400$7, MgSiO$2400$8, Fe, and Al$2400$9O$k$0 (Morley et al., 2024). The grains are assumed to follow a wide log-normal size distribution with $k$1, and cloud opacity is computed with Mie scattering for spherical particles.

The atmosphere grid is also used as the outer boundary condition for thermal evolution. The evolution equation is written as

$k$2

with a fully convective interior and the atmospheric temperature at 10 bar, $k$3, providing the atmosphere-interior connection (Morley et al., 2024). Nuclear terms include deuterium burning and the pp chain. This coupling is central to Diamondback’s scientific role: clouds and metallicity modify not only the emergent spectrum but also the cooling trajectory itself.

3. Grid design and released products

The published Diamondback grid spans warm cloudy substellar atmospheres over a parameter range intended to cover a broad mass and age interval (Morley et al., 2024).

Quantity	Coverage
$k$4	$k$5–$k$6 K
$k$7	$k$8–$k$9
$\mathrm{H_2O}$0	$\mathrm{H_2O}$1
$\mathrm{H_2O}$2	$\mathrm{H_2O}$3, plus cloud-free
Approximate masses	$\mathrm{H_2O}$4–$\mathrm{H_2O}$5

These dimensions are the operational basis for most later applications of Diamondback. In the SPHEREx benchmarking study, the model-summary table additionally describes Diamondback as a cloudy, chemically equilibrated, coupled atmosphere-interior grid with $\mathrm{H_2O}$6–2400 K, $\mathrm{H_2O}$7–5.5, free parameters $\mathrm{H_2O}$8, $\mathrm{H_2O}$9, $\mathrm{CH_4}$0, and $\mathrm{CH_4}$1, native resolving power $\mathrm{CH_4}$2, and 1440 models (Rustamkulov et al., 1 Jul 2026).

The release is explicitly open-access. The spectra, evolution tracks, and other data products are published online, and updated opacity products are distributed through Zenodo, including correlated-$\mathrm{CH_4}$3 coefficients at Zenodo 10.5281/zenodo.7542068 and high-resolution opacities at Zenodo 10.5281/zenodo.6600976 (Morley et al., 2024). This open distribution has made Diamondback usable both as a forward-model grid for direct spectral fitting and as a benchmark target for independent atmospheric codes.

4. Predicted atmospheric, spectral, and evolutionary behavior

A major Diamondback result is that clouds substantially warm the atmosphere at fixed $\mathrm{CH_4}$4. Compared with cloud-free models, cloudy profiles are hotter at a given pressure because cloud opacity blankets the atmosphere and pushes the radiative-convective boundary deeper. The warming can reach hundreds of kelvin at a given pressure in the upper atmosphere, and cloudy models can develop detached convective zones that are absent in cloud-free atmospheres (Morley et al., 2024). In one $\mathrm{CH_4}$5 example, the deep radiative-convective boundary shifts from about 0.6 bar at 2400 K to about 40 bar at 900 K.

Cloud structure varies systematically across the grid. Lower $\mathrm{CH_4}$6 makes clouds more optically thick; lower gravity shifts cloud bases upward to lower pressures; higher metallicity also thickens clouds and moves the cloud base upward; and lower $\mathrm{CH_4}$7 produces more vertically extended clouds (Morley et al., 2024). These trends are central because they generate coupled changes in thermal structure and photospheric opacity rather than merely rescaling an otherwise fixed spectrum.

The spectral consequences are correspondingly broad. Clouds redden the near-infrared by adding relatively gray opacity and suppressing flux from windows such as $\mathrm{CH_4}$8–$\mathrm{CH_4}$9. They also weaken atomic and molecular features because the continuum forms higher in the atmosphere. In the mid-infrared, the cloud-warmed pressure-temperature structure shifts the onset of methane so that $_3$0 becomes spectroscopically important at effective temperatures roughly $_3$1–$_3$2 K higher than in cloud-free equilibrium models (Morley et al., 2024). Metallicity alters the spectra as well, especially in the K band: higher metallicity strengthens the CO feature near $_3$3, whereas lower metallicity produces smoother spectra more strongly influenced by $_3$4 CIA.

The evolutionary effects are comparably large. Diamondback quantifies that clouds and metallicity can change the inferred temperature at a given age by up to about $_3$5–$_3$6 K, depending on mass and composition (Morley et al., 2024). The hybrid cloudy-to-clear cooling tracks differ from cloud-free tracks and show cooling slowed or stalled around the cloud-clearing transition near the L/T boundary. The paper considers both a fixed transition at 1300 K and a gravity-dependent cloudy-to-clear transition in which lower-gravity objects clear at lower temperatures.

5. Diagnostic uses in spectroscopy and variability studies

Diamondback has been used as a controlled laboratory for isolating the effects of cloud opacity, temperature, and gravity in specific spectral features. In the analysis of the $_3$7 doublet at $_3$8, 325 cloudy Sonora Diamondback models at solar metallicity were convolved to $_3$9 and fit with pseudo-Voigt profiles. Two observables were extracted: the average full width at half maximum and the average maximum depth. At $^-$0 K, the maximum depth correlated strongly with cloudiness, with $^-$1 against $^-$2, while showing only weak correlation with gravity, $^-$3; conversely, the FWHM correlated very strongly with gravity, $^-$4, and was uncorrelated with cloudiness, $^-$5 (Baldelli et al., 18 Apr 2025). Over $^-$6 K, FWHM remained strongly and linearly correlated with gravity with $^-$7, while the cleanest cloud-gravity separation occurred for $^-$8 K.

Diamondback has also been used as the principal cloudy-theory comparison for mid-infrared JWST/MIRI photometric diagnostics of silicate clouds. Synthetic MIRI magnitudes derived from Spitzer spectra of 113 M5–T9 ultracool dwarfs showed that the F770W/F1000W combination best separates cloudy from cloud-free L dwarfs, and that L dwarfs with $^-$9 mag are about seven times more likely to be cloudy than cloud-free (LHeureux et al., 13 Apr 2026). In those comparisons, cloudy Sonora Diamondback sequences improved over cloud-free models in reproducing the L-dwarf locus, especially in $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$0, but they still did not reproduce the observed red $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$1 colors. The interpretation given is that Diamondback may capture indirect cloud effects, particularly through the $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$2 water band, while still underestimating the direct $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$3 silicate absorption (LHeureux et al., 13 Apr 2026).

Time-domain studies have used Diamondback spectra as heterogeneous surface components. For JWST/NIRISS spectroscopy of SIMP J01365662+093347, no single Diamondback model fit the time-averaged spectrum; a Bayesian Information Criterion analysis favored a 6-model admixture, while a 3-model mixture was only slightly worse and consistent with the principal-component requirement of at least three distinct regions (Akhmetshyn et al., 29 Aug 2025). In HST/WFC3 variability work on L/T transition dwarfs, the observed disk-integrated spectrum was modeled as

$$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$4

with $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$5 and $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$6 taken from Diamondback. Those fits linked variability to temperature contrasts, cloud opacity, and cloud coverage, with 2MASS J2139 behaving like a classic patchy-cloud object, whereas at least two early-T dwarfs appeared to require additional physics beyond a simple patchy-cloud interpretation (Chapleski et al., 20 Nov 2025).

6. Benchmarks, limitations, and later extensions

Diamondback has been benchmarked both as a physical model and as a practical fitting grid. PICASO 4.0 used Sonora Diamondback as its main brown-dwarf cloud benchmark, reproducing self-consistent cloudy models at $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$7, 1300, and 2400 K with $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$8 and $$-K_{zz}\frac{\partial q_t}{\partial z}-f_{\rm sed}w_*q_c=0,$$9. The agreement in pressure-temperature structure was described as good, and the largest discrepancy was a cloud optical-depth difference of about 5% in the 2400 K case, attributed to a slight temperature offset near the condensation region (Mang et al., 25 Feb 2026). This established Diamondback as a validation target for independent cloudy radiative-convective solvers.

Forward-model benchmarks show a more mixed picture. A comparison of MCMC interpolation and Random Forest Retrieval on 11 benchmark L/T companions found Diamondback to be the preferred grid for early- and mid-type L dwarfs and, in some cases, early T dwarfs; the study also reported that MCMC gave better fit quality and tighter parameters, whereas RFR was orders of magnitude faster after training (Lueber et al., 26 May 2025). A uniform analysis of 142 age-benchmark late-M, L, and T objects found that Sonora Diamondback and SM08 yielded broadly consistent $K_{zz}$0 and radius estimates, but Diamondback gave statistically better spectral fits for only 40 of 111 objects in the main sample. That work also identified a significant age dependence of $K_{zz}$1 in L4–L9 dwarfs, with young benchmarks ($K_{zz}$2 Myr) showing systematically lower $K_{zz}$3 than older ones, and emphasized residual mismatches in TiO-, FeH-, CH$K_{zz}$4-, and K I-sensitive regions (Mader et al., 26 Jan 2026).

Large-sample spectral tests likewise show that Diamondback is informative but not exhaustive. In SPHEREx spectra of 37 nearby brown dwarfs, Diamondback served as the cloudy-equilibrium benchmark: it could help suppress overly prominent J, H, and $K_{zz}$5 peaks in some late-L cases, but clouds alone did not resolve the major mismatches in the $K_{zz}$6 opacity window or in the CO$K_{zz}$7 and CO features (Rustamkulov et al., 1 Jul 2026). In K-band forward modeling of young companions and isolated brown dwarfs, Sonora was limited by its equilibrium-chemistry and solar-metallicity assumptions in that study’s implementation, and no single grid dominated the full sample (Palma-Bifani et al., 4 Jul 2025). In multi-modal fitting of $K_{zz}$8 Pictoris b, Sonora achieved the highest Bayesian evidence for the GRAVITY-only dataset, but its metallicity and cloud posteriors ran to the grid edges, so the paper treated the resulting high-metallicity, low-$K_{zz}$9 solution cautiously (Ravet et al., 29 Sep 2025).

Several limitations recur across these studies. The original Diamondback paper explicitly states that the Ackerman & Marley cloud framework in this form does not naturally reproduce the strong $w_*$0–$w_*$1 silicate feature seen in some L dwarfs, because the particle-size and vertical-distribution solution cannot place enough small grains at the relevant altitudes without making the near-infrared unrealistic (Morley et al., 2024). This limitation is echoed by the MIRI synthetic-photometry analysis (LHeureux et al., 13 Apr 2026). A broader practical limitation is that atmospheric fits can yield unstable $w_*$2 and mass estimates when residual opacity errors, rainout uncertainties, and incomplete cloud physics are absorbed into the parameter posteriors (Mader et al., 26 Jan 2026).

Red Diamondback is the principal published extension of the framework. By blending Diamondback with SPHINX atmospheres through the $w_*$3 boundary condition across $w_*$4–$w_*$5 K, it extends the model coverage to $w_*$6–$w_*$7 K, $w_*$8–5.5, and ages $w_*$9 Myr, thereby enabling early-time Hayashi-phase evolution for high-mass brown dwarfs that lay outside the original $2400$00 K ceiling (Davis et al., 9 Oct 2025). This suggests that Diamondback should be understood not as a fixed terminal grid, but as the cloudy warm branch of an evolving Sonora infrastructure whose central scientific role remains the same: self-consistent prediction of how refractory clouds and composition reshape substellar atmospheres, spectra, and cooling histories.