From "The Cliff" to "Virgil": Mapping the Spectral Diversity of Little Red Dots with JWST/NIRSpec

TL;DR

This work characterizes a large population of Little Red Dots (LRDs) at $z ≥ 3$ by combining six-field JWST/NIRCam photometry with publicly released NIRSpec/PRISM spectra. It shows a wide range of rest-frame UV slopes and optical colors, with a bimodal tendency toward breakers ( Balmer-break–driven curvature) at red colors and power-law continua at blue colors, and it documents robust correlations between continuum shape, emission-line properties, and color. The authors propose a semi-empirical two-component model (galaxy + BH*) that reproduces the observed continuum diversity by varying BH* luminosity, attenuation, and the host-to-BH luminosity ratio, providing a unified view of LRDs and their evolution. The study highlights the significant role of host galaxies in shaping LRD SEDs and identifies key observational avenues (e.g., MIRI follow-up, higher-resolution spectroscopy) to further disentangle AGN and host contributions and to test the dense-gas BH* scenarios. The findings have important implications for understanding early black hole growth, AGN–host coevolution, and the diversity of high-redshift compact sources in JWST surveys.

Abstract

One of JWST's most unexpected discoveries is the emergence of "Little Red Dots'' (LRDs): compact sources at $z \gtrsim 3$ with blue rest-frame UV continua, red optical slopes, and broad Balmer emission lines that challenge standard models and suggest a population of early, unusual active galactic nuclei (AGNs). Using a comprehensive photometric selection and public NIRSpec/PRISM spectroscopy across six JWST deep fields, we identify a large sample of 118 LRDs with high-S/N spectra, enabling a population-wide analysis of their UV-optical continuum and emission lines. We find clear correlations between rest-frame color ([0.3-0.9\,$μ$m]) and slopes: bluer LRDs have blue UV slopes ($β_{ν,\mathrm{UV}} \sim 0.3$) and red optical slopes, while redder LRDs exhibit redder UV slopes ($β_{ν,\mathrm{UV}} \sim 1.1$). The continuum shape shows a similar trend: redder LRDs display prominent Balmer breaks and curvature, while bluer LRDs follow power-law-like optical SEDs. From literature compilations, $\sim$60% of known broad-line AGNs satisfy our LRD criteria, and up to 90% of LRDs show broad Balmer lines. Emission-line diagnostics reveal a shift from high H$_α$/H$_β$ and low [OIII]$\lambda5007$/H$_β$ in redder LRDs to the opposite in bluer ones, along with stronger narrow-line equivalent widths, suggesting a transition from AGN- to host-dominated emission. We fit the spectra with a two-component model combining a gas-enshrouded black hole (BH) and a galaxy host. Redder LRDs require higher-luminosity, unreddened BHs and modestly reddened hosts; bluer LRDs require lower-luminosity, reddened BHs and dust-free galaxies. This framework reproduces the diversity in colors and spectral shape by varying BH luminosity, obscuration, and host-to-BH luminosity ratio.

Figures (13)

• Figure 1: Redshift distribution of the photometric LRD sample (gray) and the NIRSpec/PRISM spectroscopic subset (black, scaled by a factor of 3). Horizontal lines show the median redshift and percentiles for each sample. The full sample spans $z\sim3–9$ with peaks at $z\sim5$ and $z\sim7$; the spectroscopic subset is skewed slightly lower, likely due to selection biases in spectroscopic follow-up favoring brighter sources.
• Figure 2: Color–color selection diagram of LRDs based on the barro25 criterion (solid lines; gray area to reject brown dwarf contaminants) for the six JWST deep fields analyzed in this work. Small gray dots show the full photometric catalog, while the orange density map marks the distribution of sources selected as photometric LRDs. Gray circles indicate LRDs with high-quality NIRSpec/PRISM spectra from DJA. Colored symbols highlight literature LRDs within this sample; when an object appears in multiple studies, only the original reference is shown. Black crosses denote sources classified as BLAGNs in previous studies. Yellow circles indicate compact spectroscopic LRDs that were missed by the photometric selection and were added manually for completeness. The diagram demonstrates that the vast majority of previously identified spectroscopic LRDs fall within the photometric selection region.
• Figure 3: Example of an LRD with a prominent Balmer break illustrating the difference between the optical slope and the rest-frame UV-to-optical color. The black line shows the NIRSpec/PRISM continuum spectrum, with a green line indicating the best-fit power law to the optical continuum redward of the Balmer limit (H$_\infty$). The shaded gray regions mark the wavelength ranges used to compute the [0.3$-$0.9 $\mu$m] color. In LRDs with prominent continuum breaks, the rest-frame color is often significantly larger than the optical slope, as the latter fails to capture the presence of a sharp discontinuity.
• Figure 4: Rest-frame UV slope versus optical color for LRDs with high-S/N (S/N$_{\rm UV}$$>$2) NIRSpec/Prism continuum detections. Left: Distribution for 118 spectroscopic LRDs. Colored markers highlight sources from the literature, including those reported to show strong continuum breaks near the Balmer limit wang24_lrdfurtak24labbe24bdegraaff25naidu25taylor25. Dashed and dashed-dotted lines indicate selection thresholds from kocevski24 and barro24. Orange density contours and histograms show the full photometric LRD sample using SED-fit values; gray histograms show the same for the spectroscopic subset. Medians and 68% percentiles for both samples are indicated. The LRD population spans nearly 4 magnitudes in [0.3$-$0.9 $\mu$m] color, reflecting substantial diversity in their SEDs. Most LRDs have blue UV slopes ($\beta_{\nu, \mathrm{UV}} = 0.3^{+0.5}_{-0.5}$), but there is a clear trend toward redder UV slopes at redder colors, reaching $\beta_{\nu, \mathrm{UV}} \sim 1$ for the reddest 7% with [0.3$-$0.9 $\mu$m] $>$3.5. Right: Same color–slope diagram, now color-coded by the curvature of the rest-optical continuum, computed from second-degree polynomial fits to the log-log SEDs. Red squares mark LRDs best fit by power laws. Strongly curved SEDs (yellow) are concentrated among the reddest LRDs, while bluer LRDs typically exhibit power-law–like continua. Some overlap remains around the median [0.3$-$0.9 $\mu$m] color (see discussion in § \ref{['ss:breakers_vs_powerlaws']}).
• Figure 5: Left: Same color–slope diagram as in Figure \ref{['fig:rest_slopes']}, now divided into five bins of [0.3$-$0.9 $\mu$m] color from reddest to bluest. Right: Top to bottom, the five panels show individual spectra of representative LRDs with high continuum S/N in each color bin, highlighting the diversity in SED shapes. Black lines show stacked LRD spectra from Perez-Gonzalez (in prep.) for the same bins. Gray lines show the stacked spectra from redder bins, carried forward to each panel to highlight differences across the sequence. The reddest LRDs (top two panels; [0.3$-$0.9 $\mu$m] $>$ 3, $\sim$20% of the sample) exhibit strong Balmer breaks, pronounced curvature, and redder UV slopes. Toward bluer LRDs (bottom three panels), the breaks and curvature diminish, giving way to smoother, power-law–like SEDs. The UV slope also becomes markedly bluer, with median values near $\beta \sim 0.3$ across all three bluest bins.