2511.15359v1
Changing-look Active Galactic Nuclei from the Dark Energy Spectroscopic Instrument. V. Dramatic Variability in High-Ionization Broad Emission Lines
First listed 2025-11-19 | Last updated 2025-11-19
Abstract
We present a systematic search for changing-look (CL) quasars at high redshift z > 0.9 by cross-matching the spectroscopic datasets from the Dark Energy Spectroscopic Instrument Data Release 1 and Sloan Digital Sky Survey Data Release 18. We identify 97 CL quasars showing significant variability in high-ionization broad emission lines, including 45 turn-on and 52 turn-off events, corresponding to a detection rate of approximately 0.042%. This rate is lower than that found for low-ionization CL quasars, likely due to both selection effects and physical differences in high-ionization lines. Based on the CL quasar sample, we find that CL quasars generally exhibit lower accretion rates compared to typical quasars, with average Eddington ratios of log lambda_Edd approximately -1.14 in the bright state and approximately -1.39 in the dim state, compared to approximately -0.65 for typical quasars. While high-ionization lines in CL quasars follow the Baldwin effect on a population level, some individual sources show inverse Baldwin trends. We also find a positive correlation between the variability of high-ionization lines such as Mg II and C III] and the change in bolometric luminosity. In addition, we estimate a characteristic rest-frame timescale of approximately 3 years for CL transitions, with no significant difference between turn-on and turn-off cases. Taken together, these results support an accretion-driven origin for the CL phenomenon and provide new insights into the variability of high-ionization emission lines.
Short digest
Cross-matching DESI DR1 and SDSS DR18 spectra, the authors systematically uncover 97 z > 0.9 changing-look quasars via dramatic variability in high-ionization broad emission lines (45 turn-on, 52 turn-off; detection rate ~0.042%). They infer a characteristic rest-frame transition timescale of ~3 years with no on/off asymmetry, and find CLQs have systematically lower accretion rates than typical quasars (log λ_Edd ≈ −1.14 bright, −1.39 dim vs. ≈ −0.65). On population scales the high-ionization lines follow the Baldwin effect, though some objects show inverse trends; variability in lines such as Mg II and C III] correlates with ΔLbol. The low detection rate relative to low-ionization CLQs likely reflects both selection limitations and physical differences in high-ionization lines, reinforcing an accretion-driven origin for the CL phenomenon.
Key figures to inspect
- Figure 1: Check the pseudo-magnitude vs. contemporaneous survey photometry at the SDSS and DESI epochs to validate spectrophotometric consistency; objects straying beyond the ±0.5 mag band flag potential calibration/fiber-loss issues affecting CL classification.
- Figure 2: Inspect the luminosity–redshift distribution of the CLQs relative to the SDSS DR14 quasar population to see where these high-z CL events reside in L–z space and to gauge selection biases; compare means marked by dashed lines.
- Figure 3: For five exemplars, read long-baseline light curves (CRTS/PS1/ZTF) against the SDSS/DESI spectroscopic epochs and pseudo-magnitudes; in the difference spectra, identify which high-ionization lines (e.g., C IV, Si IV, C III], Mg II) flip or fade and how associated absorption or S/N impacts the detection of tu…
- Figure 4: Examine a borderline candidate to understand failure modes—how modest continuum changes, absorption features, or low S/N can mimic or obscure BEL transitions—and how their pipeline handles such edge cases.
Discussion
Log in to view the paper discussion, see votes, and leave your own feedback.