Discovery of a Little Red Dot candidate at $z\gtrsim10$ in COSMOS-Web based on MIRI-NIRCam selection

TL;DR

This study demonstrates the first robust identification of a z$\gtrsim$10 Little Red Dot (LRD) candidate, CW-LRD-z10, by leveraging COSMOS-Web's joint JWST NIRCam-MIRI data to exploit the characteristic V-shaped SED that distinguishes LRDs from contaminants. Through forced photometry, multi-band SED fitting (LRD and galaxy templates), and image-based morphology analyses, the object is favored as a $z_{\rm phot} \approx 10.5$ LRD with $M_{\rm UV} \approx -19.9$ and a compact size $r_e \lesssim 120$ pc. The authors derive the $z\sim10$ LRD luminosity function and estimate an LRD fraction of a few percent among UV-bright galaxies at that epoch, suggesting LRDs were relatively common in the early Universe and potentially linked to the first SMBH growth episodes. They highlight the critical role of MIRI in removing low-$z$ interlopers and outline observational paths—deeper MIRI spectroscopy, expanded filter sets, and next-generation surveys—to confirm and extend LRD demographics, with implications for seed black hole formation and early galaxy evolution.

Abstract

JWST has revealed a new high-redshift population called little red dots (LRDs). Since LRDs may be in the early phase of black hole growth, identifying them in the early universe is crucial for understanding the formation of the first supermassive black holes. However, no robust LRD candidates have been identified at $z>10$, because commonly-used NIRCam photometry covers wavelengths up to $\sim5\,{\rm μm}$ and is insufficient to capture the characteristic V-shaped spectral energy distributions (SEDs) of LRDs. In this study, we present the first search for $z\gtrsim10$ LRD candidates using both NIRCam and MIRI imaging from COSMOS-Web, which provides the largest joint NIRCam-MIRI coverage to date ($0.20\,{\rm deg^2}$). Taking advantage of MIRI/F770W to remove contaminants, we identify one robust candidate, CW-LRD-z10 at $z_{\rm phot}=10.5^{+0.7}_{-0.6}$ with $M_{\rm UV}=-19.9^{+0.1}_{-0.2}\,{\rm mag}$. CW-LRD-z10 exhibits a compact morphology, a distinct V-shaped SED, and a non-detection in F115W, all consistent with being an LRD at $z\sim10$. Based on this discovery, we place the first constraint on the number density of LRDs at $z\sim10$ with $M_{\rm UV}\sim-20$ of $1.2^{+2.7}_{-1.0}\times10^{-6}\,{\rm Mpc^{-3}\,mag^{-1}}$, suggesting that the fraction of LRDs among the overall galaxy population increases with redshift, reaching $\sim3\%$ at $z\sim10$. Although deep spectroscopy is necessary to confirm the redshift and the nature of CW-LRD-z10, our results imply that LRDs may be a common population at $z>10$, playing a key role in the first supermassive black hole formation.

Figures (12)

• Figure 1: Model SED of a $z=10$ LRD with $M_{\rm UV} = -20$ (red) and $5\sigma$ depth for point sources in COSMOS-Web (F115W, F150W, F277W, F444W, and F770W) from Casey2023 and in COSMOS-3D (F115W, F200W, F356W, and F1000W) shown in gray and gold rectangles, respectively. Model SEDs of potential contaminants of the $z\sim10$ LRD photometric selection, a $z=10$ galaxy with a stellar-origin Balmer break with $A_V=0.5$ and $M_* = 1.5 \times 10^{10}\,M_\odot$, is overlaid. The LRD model SED is from Akins2024, and the galaxy SED is generated using FSPS Conroy2009Conroy2010. The predicted model photometry is shown with diamonds for the LRD and squares for the galaxy. MIRI photometry is essential to distinguish between $z\sim10$ LRDs and $z\sim10$ galaxies with stellar-origin Balmer break.
• Figure 2: Color-color diagrams based on the LRD model SEDs from Akins2024 (red lines). The left and right panels show $m_{\rm F115W} - m_{\rm F150W}$ vs. $m_{\rm F150W} - m_{\rm F277W}$ and $m_{\rm F277W} - m_{\rm F444W}$ vs. $m_{\rm F444W} - m_{\rm F770W}$, respectively. The red filled region defined by $0 < m_{\rm F277W} - m_{\rm F444W} < 2.25$ and $1.5 < m_{\rm F444W} - m_{\rm F770W}$ can effectively distinguish $z \sim 10$ LRDs from other objects. The selected candidate in this study, CW-LRD-z10, is highlighted with a red star with error bars. Note that $m_{\rm F115W}-m_{\rm F150W}$ color for CW-LRD-z10 is a $2\sigma$ lower limit. CW-LRD-z10 also passes the color criteria for F115W dropouts Harikane2023_highz shown by the gray shaded region in the left panel. Model galaxy SEDs generated using FSPS Conroy2009Conroy2010 with various star formation histories and dust attenuation values at $z=0\,\mathchar'-\,10$ are shown in gray solid lines. Of these, we highlight two cases: (1) a stellar origin Balmer break (with an $f_\nu$ ratio of $\sim2$) and $A_V=0.5$ (green), and (2) a dusty SFG model with $A_V=5$ (cyan). Seed BH models from Inayoshi2022 at $z=8\,\mathchar'-\,14$ and LRDs with an extremely strong Balmer break at $z=9\,\mathchar'-\,15$, assuming the spectrum of The Cliff in deGraaff2025, are shown in black and orange, respectively. For these model SEDs, dots are placed at redshift intervals of $\Delta z=1$. Specifically, $z=10$ Balmer break galaxy, $z=10$ seed BH, $z=2.5$ galaxy with stellar-origin Balmer break, and $z=0.2$ dusty SFG are shown in green triangle, black triangle, green hexagon, and cyan hexagon, respectively. Photometrically-selected LRDs from Akins2024 with $S/N > 3$ are shown as orange diamonds, among them are two objects that pass our color selection (red diamonds). Dwarf stars and main sequence stars from the BT-Settl model Allard2012 are plotted in magenta crosses and pentagons, respectively.
• Figure 3: (Upper) HST + JWST $3^{\prime\prime} \times 3^{\prime\prime}$ cutout images centered on CW-LRD-z10. Each image is scaled by a noise map. If detected, the $S/N$ measured from forced photometry is shown in the top left corner. (Lower) Observed SED of CW-LRD-z10. For F606W, F814W, F115W, F1000W, and F2100W, we plot the $2\sigma$ upper limits. The red line indicates the best-fit LRD model SED from the SED fitting analysis (Section \ref{['ss:SED_fit']})
• Figure 4: Observed photometry and SED fitting results using BAGPIPES. In the left panel, black open circles, small dots, and lines represent the observed photometry, the model photometry, and the model SEDs. The estimated $z_{\rm photo}$ and corresponding $\chi^2$ value for each fit are indicated in the lower right corner. The right panel shows the $\chi^2$ distribution as a function of redshift. Red, cyan, and green lines indicate fits using the LRD model from Akins2024 over $0<z<15$, the galaxy model over $0<z<2$, and the galaxy model over $2<z<15$, respectively.
• Figure 5: Completeness as a function of redshift and UV absolute magnitude, estimated from a Monte Carlo simulation of COSMOS-Web mock observations.