The GlimmIr: Spectroscopic Variability in a z~7 LRD Indicates Rapid Changes in Both the Narrow and Broad Line Regions

Erini Lambrides, Taylor A. Hutchison, Rebecca L. Larson, Pablo Arrabal Haro, Casey Papovich, Weida Hu, Nikko J. Cleri, Steven L. Finkelstein, Jonathan R. Trump, Pablo G. Perez-Gonzalez, Bingjie Wang, Dale D. Kocevski, John Chisholm, Amy Secunda, Sarah E. I. Bosman, Hollis Akins, Mitchell Karmen, Mark Dickinson, Volker Bromm, Bren E. Backhaus, Marco Chiaberge, Olivia R. Cooper, Yukta Ajay, Guillermo Barro, Danielle A. Berg, Jenna Cann, M. C. Cooper, Norman A. Grogin, Michaela Hirschmann, Marc Huertas-Company, Jeyhan S. Kartaltepe, Anton M. Koekemoer, Ray A. Lucas, Arianna S. Long, Roberto Gilli, Colin Norman, Andrew F. Ptak, Chris T. Richardson, Jane R. Rigby, Brittany N. Vanderhoof, L. Y. Aaron Yung, Jorge A. Zavala

First listed 2026-04-28 | Last updated 2026-04-28

Abstract

The enigmatic population of ``Little Red Dots'' (LRDs) sit at the center of some of the largest debates in extragalactic astronomy today. The source(s) of ionizing emission and the physical scale over which it governs is still largely unknown. We show for the first time spectroscopic variability in a z ~ 7 LRD. Comparing a recently obtained 10.2 hr JWST/NIRSpec F290LP/G395M spectrum via the C3PO survey to an 8.4 hr F290LP/G395M spectrum taken 99 days earlier (~13 rest-days) via the THRILS survey, we find a ~30% $ difference in the continuum and broad-line flux, and a 42% difference between [OIII]5008 flux in the two epochs. Through rigorous testing, we confirm that such differences are not the result of differing MSA slit placements on source nor merely flux calibration offsets. These results are further corroborated by both a similar continuum and [OIII]5008 flux differences found in NIRSpec prism/clear observations of the source at an epoch taken approximately a year earlier than the THRILS observations via RUBIES and an additional observation fortuitously taken during the THRILS epoch (within a rest-day) via the CAPERS survey. Assuming LRDs are a type of accreting black hole system, this implies direct sight-lines must exist from the accretion disk to the surrounding nebular gas on scales beyond the broad-line region, and thus any high-density gas interpretations must allow for covering fractions < 100%. Furthermore, these results show the [OIII] line emission is likely not galaxy process-dominated, with a significant population of the narrow-line emitting gas closest to the broad-line region being directly ionized by the LRD. Finally, these results highlight the need for new approaches in inferring black hole properties of these systems, accounting for the lack of significant ionization via star formation, and/or exploring more exotic host-galaxy conditions at these early epochs.

Short digest

First detection of spectroscopic variability in a z≈7 Little Red Dot (the GlimmIr): two deep JWST/NIRSpec F290LP/G395M epochs 99 days apart (~13 rest-days) show ~30% changes in the continuum and broad-line flux and a 42% drop in [O III]5008. Careful tests rule out slit placement and calibration systematics, and NIRSpec prism data (RUBIES ~1 year earlier; CAPERS within a rest-day of THRILS) reproduce similar continuum/[O III] differences. The fast [O III] response ties part of the narrow-line gas to the variable engine, implying direct sight-lines from the accretion disk beyond the BLR and covering fractions <100% for any high-density screen. Assuming an accreting BH origin, the result argues that [O III] is not galaxy-process dominated and black-hole property inferences must be rethought for LRDs.

Key figures to inspect

• Figure 1: Compare the two G395M/F290LP epochs (THRILS vs C3PO) in 2D/1D and the subtraction—verify ~30% continuum/broad-line changes and the pronounced [O III]5008 shift; use the per-element ratio panel and the CEERS-10444 MSA overlays to see why slit placement cannot explain the differences.
• Figure 2: Inspect CAPERS vs RUBIES prism/clear spectra—confirm that continuum and [O III] differences persist in low‑resolution data taken ~1 year earlier and within a rest-day of THRILS, reinforcing true variability across modes and timescales.
• Figure 3: Follow the synthetic F444W fluxes across all epochs and slit orientations—check that variability trends track across configurations, supporting that orientation/throughput is not the driver.
• Figure 4: Look at brightest-nebular-line flux ratios for comparison sources—the GlimmIr stands out beyond the assumed 30% systematic band, strengthening the intrinsic-variability interpretation.