A Population of Little Red Dot-like Quasars in SDSS

Quinn O. Casey, Ryan C. Hickox, Nikko J. Cleri, Jonathan H. Cohn, David M. Alexander, Emmanuel Durodola, Kelly E. Whalen, Raphael E. Hviding, Tonima Tasnim Ananna

First listed 2026-06-25 | Last updated 2026-06-24

Abstract

Compact and red sources in the high redshift ($z\sim5$) Universe, known as "Little Red Dots" (LRDs), are among JWST's most intriguing discoveries. These sources have broad Balmer emission lines, weak X-ray emission, and unique spectral energy distributions (SEDs) poorly fit by either stellar or AGN templates. Local analogs of LRDs allow for detailed studies of the underlying physical processes with archival multi-wavelength datasets unavailable in the high-$z$ Universe. We show that the SDSS $ugriz$ filters at $z\approx0.4, 0.8$ overlap well with the JWST filters used to select LRDs at $z\sim5$. We use SDSS quasars to define a sample of $\sim1300$ Local Red Dots (LoRDs) which share the same photometric colors of LRDs. A subset of the LoRD sample selected to have V-shaped continua ($N=244$) show prominent higher-order Balmer absorption features and [NeV]$λ$3426 emission, both of which would likely be missed in JWST/PRISM observations given the low spectral resolution. A composite SED of the LoRDs differs from a typical quasar SED in the rest-frame UV/optical, but the two agree with each other in the NIR. The LoRD SED matches well with a stack of LRDs and can be modeled either as a reddened AGN combined with a host galaxy, or as a reddened AGN combined with a host galaxy and a cool blackbody. Interestingly, the LoRDs are X-ray detected at a rate comparable to typical quasars. However, the probability that LoRDs and typical quasars would go undetected, if subject to the LRD X-ray upper limits, is $>50\%$.

Short digest

Using the overlap between SDSS ugriz filters at z≈0.4 and 0.8 and the JWST bands used to select z∼5 little red dots, this paper builds a large sample of about 1300 SDSS quasar analogs dubbed Local Red Dots, or LoRDs. A V-shaped subset of 244 objects shows higher-order Balmer absorption and [Ne V] λ3426 emission, highlighting spectral structure that would likely be missed at JWST/PRISM resolution while preserving the broad-Balmer, red-continuum phenomenology that defines LRDs. Their composite SED departs from a typical quasar in the rest-frame UV and optical but agrees in the NIR, and it matches an LRD stack with fits consistent with a reddened AGN plus host galaxy, with or without an added cool blackbody. Just as importantly, the LoRDs are X-ray detected at rates comparable to ordinary quasars, and the paper argues that current LRD X-ray upper limits are loose enough that similar sources would often remain undetected, tempering claims that LRD-like colors necessarily imply intrinsic X-ray weakness.

Key figures to inspect

• Figure 1. Use this to show the selection logic. It demonstrates the rest-frame filter overlap between JWST and SDSS and then maps where the LoRDs sit in color space relative to the control quasars, making clear how the local analog sample is defined rather than assumed.
• Figure 3. This is the key spectral-evidence figure for the V-shaped subset. The stacked spectra isolate the Balmer-limit-centered continuum shape and show the higher-order Balmer absorption plus [Ne V] emission that strengthen the case that LoRDs reproduce distinctive LRD-like spectroscopic behavior.
• Figure 5. This figure is the cleanest synthesis of the SED argument because it directly compares the LoRD and LRD composite SEDs and shows the exploratory component fits. It matters because the paper’s bottom-line physical interpretation rests on the close match between the local analog stack and the observed LRD stack.
• Figure 6. Include this for the X-ray result, which is one of the paper’s most consequential takeaways. The comparison of Hβ-based expectations, LoRD detections and limits, and published LRD upper limits makes the paper’s point that present LRD non-detections do not by themselves prove anomalous X-ray weakness.
• Figure 7. This figure captures the observational-resolution caveat in the most concrete way by degrading SDSS LoRD spectra to JWST/PRISM resolution and comparing them directly with observed LRD spectra. It shows why narrow absorption and line-diagnostic structure visible locally could be washed out in current high-redshift data while leaving the overall LRD-like spectral shape intact.