Broad line region echo from highly accreting intermediate-mass black hole candidate SDSS J144850.08+160803.1

Context. Elusive intermediate-mass black holes (IMBHs; 100 M⊙ ≤ MBH ≤ 2 × 105 M⊙) can be used as “time-squeezing” machines, enabling studies of active galactic nucleus (AGN) geometry via reverberation mapping (RM) on much shorter timescales than their supermassive siblings. Aims. Constraints on the broad line region (BLR) radius for IMBH candidates across a broad range of Eddington ratios help probe the unexplored faint end of the radius-luminosity (R − L) relation in AGNs. This development opens up the opportunity to build a more robust MBH estimator. The present study is aimed at: (a) confirming a highly accreting IMBH candidate and (b) demonstrating the feasibility of the first photometric BLR RM campaign for IMBHs with high Eddington rates. Methods. SDSS J144850.08+160803.1 (J1448+16) was identified as an IMBH candidate from a broad Hα-selected spectroscopic sample from SDSS. We carried out XMM-Newton X-ray observations to confirm its AGN status, along with narrowband Hα and broadband SDSS g′ monitoring over five months (March–July 2024) using a 60-cm telescope at the Caucasus Mountain Observatory. These time series allowed us both to probe the short-timescale variability and extract the time lag between the BLR and accretion disk (AD) continuum. Results. XMM-Newton detected J1448+16 as a bright X-ray point source with a photon index of $ \Gamma = 2.32^{+0.15}_{-0.13} $ and X-ray luminosity of $ L_{2-10\,\mathrm{keV}} = (3.3^{+0.5}_{-0.4}) \times 10^{41} $ erg s−1, confirming its AGN activity. From the SDSS optical spectrum and X-ray properties, we estimated a BH mass of ∼(0.9 − 2.4)×105 M⊙ and Eddington rate of ∼37 − 112%. We report high-amplitude ∼55% intra-night (∼1.7 h) Hα variability for this highly accreting IMBH and extract a tentative BLR RM radius estimate of ∼1 − 8 days. Conclusions. This work offers a proof of concept for further high-Eddington-rate IMBH variability studies and BLR RM campaigns, which will be essential for an efficient calibration of the R − L relation at the faint end.