Diversity of Cold Worlds: Predicted Near- to Mid-infrared Spectral Signatures of a Cold Brown Dwarf with Potential Auroral Heating

TL;DR

This study investigates W1935, a cold brown dwarf with a prominent methane emission at $3.326\ rac{\mu}{m}$, to test whether upper-atmosphere heating (likely auroral) via a ~300 K thermal inversion shapes its spectrum. By extending retrieved spectra to 1–20 μm and comparing inversion vs non-inversion scenarios using HITRAN/ExoMol cross sections and the SEDA framework, the authors quantify the inversion's impact and contrast it with self-consistent atmosphere models (e.g., Sonora Elf Owl, LB23, ATMO 2020). They find atmospheric heating contributes about $15\%$ to the bolometric luminosity, mainly from wavelengths longer than ~5 μm, and that the inversion suppresses CH$_4$, H$_2$O, and NH$_3$ features at low pressures (~$0.002$–$0.04$ bar) while leaving CO/CO$_2$ from deeper layers largely unchanged; importantly, the inversion predicts a new methane emission at ~7.7 μm and possible NH$_3$ features, offering testable JWST predictions. Despite these spectral implications, W1935 appears as an outlier only in CMDs that include the Ch2 band (4.4 μm, CO/CO$_2$ sensitivity), and the observed CMD dispersion is not fully explained by heating alone, with binarity likely contributing to its overluminous position. Overall, the work highlights the spectral diversity of cold brown dwarfs, clarifies the role of aurora-driven heating in shaping mid-IR spectra, and provides concrete predictions for future observations to validate the inversion scenario.

Abstract

Recent JWST/NIRSpec observations have revealed strong methane emission at 3.326 microns in the $\approx$482 K brown dwarf CWISEP J193518.59$-$154620.3 (W1935). Atmospheric modeling suggests the presence of a $\approx$300 K thermal inversion in its upper atmosphere, potentially driven by auroral activity. We present an extension of the retrieved spectra of W1935 with and without inversion spanning 1--20 microns, to identify thermal inversion-sensitive spectral features and explore the origin of the object's peculiar characteristics. Our analysis indicates that atmospheric heating contributes approximately 15% to the bolometric luminosity. The model with inversion predicts an additional similar-strength methane emission feature at 7.7 microns and tentative ammonia emission features in the mid-infrared. Wavelengths beyond $\sim$2 microns are significantly influenced by the inversion, except for the 4.1--5.0 microns CO$_2$ and CO features that originate from atmospheric layers deeper than the region where the inversion occurs. W1935 appears as an outlier in Spitzer/IRAC mid-infrared color-magnitude diagrams (CMDs) based on the $m_{\rm Ch1}-m_{\rm Ch2}$ (IRAC 3.6 microns $-$ 4.5 microns) color, but exhibits average behavior in all other combinations that trace clear sequences. This anomaly is likely due to the Ch2 filter probing vertical mixing-sensitive CO$_2$ and CO features that do not correlate with temperature or spectral type. We find that the thermal inversion tends to produce bluer $m_{\rm Ch1}-m_{\rm Ch2}$ colors, so the overluminous and/or redder position of W1935 in diagrams involving this color cannot be explained by the thermal inversion. This analysis provides insights into the intriguing dispersion of cold brown dwarfs in mid-infrared CMDs and sheds light on their spectral diversity.

Figures (6)

• Figure 1: Observed and predicted SEDs of W1935. The observations include the NIRSpec G395H spectrum (original $R\sim2700$ resolution in light green and $R\sim250$ convolved spectrum in dark green) and NIRI, IRAC, WISE, and JWST/MIRI magnitudes (Section \ref{['sec:observations']}), as indicated in the bottom-right legend. The best retrieved spectra in Faherty_etal2024 using either a thermal inversion (black line) or no thermal inversion (brown line) in the upper atmosphere are shown. Synthetic fluxes from both retrieved spectra for the filters with observations, along with IRAC Ch3 (5.6 $\mu$m) and Ch4 (7.6 $\mu$m) are plotted using the same color as their corresponding spectra. Additional synthetic IRAC Ch1 and Ch2 fluxes from the NIRSpec spectrum are shown in green. The main spectral features are indicated (see Figure \ref{['fig:SED_CF']}), including the principal methane emission regions.
• Figure 2: Retrieved spectra for W1935 (with and without thermal inversion in black and brown, respectively) and W2220 (orange curve; no temperature inversion). The NIRSpec G395H spectra for W1935 and W2220 are shown in green and blue, respectively, with a light color for the original resolution ($R\sim2700$) spectra and a darker color for convolved ($R\sim100$) spectra. To facilitate comparison between the two objects, we plotted absolute fluxes, accounting for their respective distances (14.43$\pm$0.79 pc and 10.47$\pm$0.23 pc for W1935 and W2220, respectively; Kirkpatrick_etal2021).
• Figure 3: Top panel: SED for W1935 in Figure \ref{['fig:SED']} along with the absorption cross sections for key molecular species at a representative temperature of 475 K and a pressure of 1 bar. Middle panel: Contribution function for the retrieved spectrum with thermal inversion. The horizontal dashed line indicates the approximate maximum pressure at which the thermal inversion occurs. Bottom panel: Contribution function for the retrieved spectrum without thermal inversion. The horizontal dashed line indicates is the same as in the middle panel.
• Figure 4: Best retrieved spectra for W1935 with (top panel) and without (bottom panel) thermal inversion. The color axis in each panel shows the corresponding contribution function, which indicates the pressure contributing the most to the fluxes emerging at different wavelengths. The maximum pressure probed by both retrieved spectra is similar ($\approx$60 bar). However, due to the thermal inversion, the minimum pressure probed by the retrieved spectrum with inversion ($\approx$0.002 bar) is much lower that the one from the model without inversion ($\approx$0.1 bar, with a few wavelength points reaching as low as 0.03 bar). The W1935 2.9--5.1 $\mu$m G395H spectrum probed pressures $\approx$0.002--10 bar. The main spectral features are indicated, including the principal methane emission regions.
• Figure 5: CMDs for all combinations of IRAC Ch1, Ch2, Ch3, and Ch4 magnitudes using observed data for $\ge$T5 dwarfs in the UltracoolSheet catalog Best_etal2024 and from Patten_etal2006, as well as synthetic values for objects with NIRSpec PRISM spectra in Beiler_etal2024. The positions of W1935 and W2220 considering synthetic values derived using the retrieved spectra in Faherty_etal2024 are indicated with a red star (from model with inversion) and a blue star (from model without inversion) for W1935 and a green star for W2220. Labeled objects in (a) and (b) panels indicate the targets in the GO Cycle 1 2124 program.