Water absorption confirms cool atmospheres in two little red dots

TL;DR

The paper uses JWST/NIRSpec Prism spectra to distinguish intrinsic red continua from dust-reddened hot disks in Little Red Dots by detecting a rest-frame near-infrared water absorption feature at ~$1.4\,\mu$m in two $z\sim2$ LRDs, revealing a cool, dense gas component with $T\lesssim 3000$ K contributing ~20–30% of the near-infrared flux. Through two-temperature photosphere modeling and low-density atmosphere calculations, the authors demonstrate that a single-temperature blackbody cannot explain both the optical continuum and the molecular absorption, whereas a composite model with a $T\sim2000$ K cold component and a hotter $T\sim4000$ K component fits the data, and suggests very low metallicity. The findings imply intrinsically red thermal emission and substantially lower bolometric luminosities and SMBH masses than in dust-reddened AGN interpretations, reshaping our understanding of rapid black hole growth at cosmic noon and establishing molecular absorption as a diagnostic tool for LRDs. The work motivates further high-resolution spectroscopy and denser atmosphere grids to test geometry (spherical vs disk) and metallicity across the LRD population.

Abstract

Little red dots (LRDs) are an abundant population of compact high-redshift sources with red rest-frame optical continua, discovered by the James Webb Space Telescope (JWST). Their red colors and power sources have been attributed either to dust reddening of standard hot accretion disks or to intrinsically cool thermal emission from dense hydrogen envelopes, in both cases surrounding accreting supermassive black holes. These scenarios predict order-of-magnitude differences in emission temperature but have lacked decisive temperature diagnostics. Here we report a prominent absorption feature at rest-frame $\sim 1.4 \, μ\mathrm{m}$ in two out of four LRDs at $z \sim 2$ with high signal-to-noise JWST spectra, among the coolest from a large LRD sample. The feature matches the shape and wavelength of the water absorption band seen in cool stars. Atmosphere models require $T \lesssim 3000\, \mathrm{K}$ to reproduce it, confirming unambiguously the presence of a cool, dense gas component contributing $20-30\%$ to the emergent continuum. A composite model reproduces both the absorption and the rest-frame optical-to-infrared continuum shape and suggests a temperature range ($\sim2000\, \mathrm{K} - 4000 \, \mathrm{K}$) rather than a single blackbody predicted by some gas envelope models. Molecular absorption demonstrates that the red continua of some LRDs are intrinsic rather than dust-reddened, implying order-of-magnitude lower bolometric luminosities and black-hole masses, and providing a new diagnostic of the emitting gas.

Figures (10)

• Figure 1: Detection of water absorption in two LRD spectra at $2<z<3$.a) The two LRD JWST/Prism spectra with detected water absorption are plotted in blue and cyan, respectively. Black dashed lines show power-law fits to the continuum (i.e., no water absorption) to guide the eyes. Observed M dwarf spectra with varying water absorption strength are shown in yellow and brown for comparison Filippazzo2015Pineda2021. We emphasize that these M dwarf spectra are not fits to the LRD data, but are included solely for illustrative purposes. The absorption features in LRDs are consistent with the characteristic water bands seen in these stellar spectra. To quantify the absorption in a model independent way, we define a simple continuum-normalized water absorption index based on the ratio of flux densities in two narrow rest-frame windows bracketing the break at $1.310\,\mu\mathrm{m}$ and $1.363\,\mu\mathrm{m}$ (gray vertical bands): $r_{\mathrm{H_2O}} = f\lambda,1.310/f\lambda,1.363 - 1$ (i.e., more negative values indicating stronger absorption). b) Measured water absorption strength of the LRDs (blue and cyan, shaded area indicate $1\sigma$ uncertainties) compared to $0.1\,{\rm Z}_{\odot}$Phoenix stellar models and low-density non-stellar models, across 7 orders of magnitude in densities spanning a range of temperatures. Models above $\sim 3{,}000\,\mathrm{K}$ fail to produce the observed water absorption, establishing firm evidence for the existence of low temperature, $\sim 2{,}000\,\mathrm{K}$, gas in LRDs.
• Figure 2: Models fits to the optical to near-infrared continua of two LRDs with detected water absorption requires cold ($T<3{,}000\,\mathrm{K}$) gas. The observed JWST/NIRSpec Prism spectra are shown in dark gray, with wavelength regions excluded from the fit indicated in light gray. Single–temperature blackbody models are shown in black. Best-fit single-temperature blackbody models (green) broadly reproduce the continuum shape but fail to account for the depth of the water absorption feature and yield temperatures too high for the survival of water vapor. The best-fit two-component photospheric models are shown in orange, with the warm and cool components overplotted in blue and red, respectively. Fit residuals are shown below each panel. This minimal model reproduces both the overall continuum shape and the observed water absorption band. The inferred cool component has an effective temperature of $T\sim2{,}000\,\mathrm{K}$ and contributes approximately $20\%$ and $30\%$ of the total flux at rest-frame $1.4\,\mu\mathrm{m}$ in the two objects, respectively, while a very low metallicity ($Z<10^{-3}\,Z_\odot$) warmer component with $T\sim4{,}000$ K is required to match the optical continuum.
• Figure 3: Bolometric luminosities and temperatures of the two cold LRDs in context of the LRD population.a) The bolometric luminosities of the sources studied here, inferred using the two-temperature stellar atmosphere model, are shown as blue and cyan stars. For context, we include the bolometric luminosities of a larger sample of LRDs derived from single temperature modified black body model fits deGraaff:sample. For our sample blackbody fits produce practically identical bolometric luminosities (within 5-10%). For comparison, the bolometric luminosities assuming a dust-reddened AGN interpretation are shown as unfilled stars of the same colors and would be at least an order magnitude higher. b) The bolometric luminosities versus wavelength at which the rest optical continuum peak for LRDs, showing only sources at $z<4.5$ where the blackbody model fits are well constrained. Longer peak wavelengths correspond to lower effective temperatures, and an approximate effective temperature conversion is shown on top. The LRDs of this paper, marked as stars, have bolometric luminosities within the range of the larger LRD population but occupy the colder side of the temperature distribution.
• Figure E.1: Sample of this paper. The observed JWST/Prism spectra are plotted in black, whereas the uncertainties are shown in blue. 2D spectra are shown on the top. Composite images on 1$"$$\times$ 1$"$ cutouts are included to the right.
• Figure E.2: Cutouts from the available JWST/NIRCam imaging for (a) WIDE-EGS-2974 and (b) UNCOVER-A2744-20698. Each panel is 1.3$"$$\times$ 1.3$"$. The NIRSpec slitlets are overlaid in cyan.