Water absorption confirms cool atmospheres in two little red dots

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.

💡 Research Summary

**
The authors present a decisive temperature diagnostic for the enigmatic “little red dots” (LRDs) – compact, high‑redshift sources identified by JWST that show red rest‑frame optical continua. Two competing interpretations have existed: (1) a standard hot accretion‑disk spectrum that appears red because of heavy dust extinction, implying very high bolometric corrections, extreme luminosities, and massive super‑massive black holes (SMBHs); and (2) intrinsically red thermal emission from a dense, quasi‑spherical hydrogen envelope surrounding the SMBH, which would be much cooler (a few thousand kelvin) and thus have far lower luminosities and SMBH masses. The key difference between these scenarios is the characteristic temperature of the emitting surface, but until now no direct temperature probe existed.

To resolve this, the team examined JWST/NIRSpec Prism spectra of LRDs at 2 < z < 3 with signal‑to‑noise > 5 per pixel. From a parent sample of ~17,000 low‑resolution spectra, four LRDs satisfied strict compactness and “v‑shape” SED criteria. Two objects, WIDE‑EGS‑2974 and UNCOVER‑A2744‑20698, display a broad depression centered at rest‑frame ≈ 1.4 µm. The feature’s wavelength and shape match the water (H₂O) absorption band that dominates the spectra of cool M‑type stars (1.33–1.50 µm). The authors ruled out reduction artifacts, emission‑line contamination, or systematic effects.

Using a suite of stellar atmosphere models (PHOENIX) extended to very low gravities (log g < 0) appropriate for a gaseous envelope, they explored the dependence of the water band on temperature, density, and metallicity. The absorption strength is primarily a temperature indicator: models with T ≳ 3000 K cannot reproduce the observed depth, while those with T ≈ 2000–3000 K match both the band shape and its equivalent width. The band’s presence therefore proves that a substantial fraction of the emitting surface is cool enough for water molecules to survive.

To quantify the contribution of this cool component, the authors fitted the full rest‑frame optical‑to‑NIR continuum (excluding emission lines) with two approaches. Single‑temperature blackbody fits (≈ 3800 K) reproduce the overall SED shape but leave large residuals around the water band and imply temperatures too high for water to exist. A minimal two‑temperature model, consisting of a warm component (T ≈ 4000 K, very low metallicity) and a cool component (T ≈ 2000 K), simultaneously matches the continuum slope, the Balmer‑limit region, and the depth of the water absorption. In the best fits, the cool gas contributes ~20 % (WIDE‑EGS‑2974) and ~30 % (UNCOVER‑A2744‑20698) of the flux at 1.4 µm.

These results have several important implications:

1. Intrinsic Redness – The red optical continua of at least some LRDs are intrinsic, not the product of dust reddening. Consequently, bolometric luminosities derived under a dust‑obscured AGN (AGN) scenario are overestimated by roughly an order of magnitude. The revised luminosities place these objects comfortably within the luminosity distribution of the broader LRD population.

2. SMBH Masses – Lower bolometric luminosities imply proportionally lower SMBH masses (via standard Eddington arguments), alleviating the tension that LRDs would otherwise over‑produce the high‑z black‑hole mass density.

3. Physical Structure – The need for a temperature range (≈ 2000–4000 K) rather than a single photospheric temperature suggests a more complex envelope structure than simple hydrostatic, single‑temperature photospheres. Possible explanations include multi‑layered radiative‑convective zones, clumpy or anisotropic flows, or a combination of a quasi‑static envelope with a hotter inner accretion flow.

4. Population Context – Within the high‑S/N sample, 2 of 4 objects show water absorption, while the other two have similar SEDs but lack the feature, perhaps because they are slightly hotter or have lower water column densities. Across the full LRD catalog (z ≈ 2–7), the two water‑bearing sources sit at the redder, cooler tail of the color‑temperature distribution, consistent with the expectation that the reddest LRDs are the coolest.

5. Future Diagnostics – Water absorption provides a powerful, direct probe of temperature, density, and metallicity in LRDs. At z < 3 the JWST/NIRSpec prism can access the band; at higher redshifts, JWST/MIRI or deeper NIRSpec integrations will be required. Systematic surveys of molecular features will test whether cool thermal emission is a universal property of LRDs and will constrain the physics of rapid black‑hole growth in the early universe.

In summary, the detection of a water absorption band in two LRDs furnishes the first unequivocal evidence for a cool (≤ 3000 K), dense gas component contributing significantly to their near‑infrared output. This finding decisively favors the intrinsically red, thermally emitting envelope scenario over the dust‑reddened hot‑disk picture, reshapes estimates of their luminosities and black‑hole masses, and opens a new spectroscopic window onto the physical conditions governing early, rapid SMBH accretion.