2025ApJ...990..149K

The rates and properties of tidal disruption events (TDEs) provide valuable insights into their host galaxy central stellar densities and the demographics of their central supermassive black holes. TDEs have been observed only at low redshifts (z ≲ 1), due to the difficulty in conducting deep time-domain surveys. In this work, we present the discovery of a high-redshift TDE candidate, HZTDE-1, in the COSMOS-Web survey with JWST's NIRCam, using a novel selection technique based on color and morphology. We outline a methodology for identifying high-z TDEs in deep infrared imaging surveys, leveraging the unique spectral energy distributions and morphologies of these transients. While focused on TDEs, this methodology could also be applied to find other UV-bright transients, such as superluminous supernovae (SLSNe). We apply this technique to COSMOS-Web in filters F115W, F150W, F277W, and F444W, and identify HZTDE-1, a transient point source relative to archival UltraVISTA infrared observations. If we assume it is a TDE, we estimate a photometric redshift of $z\,=\,5.0{2}_{-1.11}^{+1.32}$. HZTDE-1 cannot be explained by reasonable supernova or active galactic nuclei models. However, an SLSN at z ≳ 3 can also plausibly explain this transient and would be the highest-redshift SLSN yet known. If confirmed with follow-up observations, HZTDE-1 would represent the highest-redshift TDE discovery to date, and suggest an enhancement of the TDE rate in the high-redshift Universe. Our method, which can be applied to future deep surveys with the JWST and Nancy Grace Roman Space Telescope, offers a pathway to identify TDEs at z > 4 and probe black hole demographics at early cosmic times.