Distinguishing Tidal Disruption Events and Changing-look Active Galactic Nuclei via Variation of Mid-infrared Color

At present, there is a lack of effective probes to distinguish between mid-infrared (MIR) outbursts induced by tidal disruption events (TDEs) and changing-look active galactic nuclei (CLAGNs) based on only MIR data. Here, we propose that the time variation of MIR color (K-corrected W1-W2 after subtracting the quiescent fluxes) is a promising probe. With an optically selected sample containing TDEs, ambiguous nuclear transients (ANTs), and CLAGNs, we studied the MIR color variation of their MIR counterparts using NEOWISE-R data. We found that the MIR color of TDEs and ANTs turns red faster than CLAGNs during the rising phase, and TDEs have a redder color than ANTs at the earliest phase. The former may be caused by the difference between the ultraviolet light curves of TDEs/ANTs and CLAGNs, or be related to no or relatively weak underlying AGN in TDEs/ANTs, while the latter may be related to the difference in the dust geometry. Based on color variation rate, we selected high-probability TDE, ANT, and CLAGN candidates from MIR outbursts in samples of Jiang et al. (2021) and Masterson et al. (2024). We found that both samples are mixtures of TDEs/ANTs and CLAGNs. For MIR outbursts whose hosts are not Seyfert galaxies, we estimated that $\sim50\%-80\%$ are TDEs and inferred a rate of infrared TDEs of $1.5-2.8\times10^{-5}$ galaxy$^{-1}$ yr$^{-1}$, comparable with that of optical TDEs. The rest are CLAGNs, suggesting the presence of weak AGNs that cannot be identified using common diagnoses. We predicted that with our method, a large amount of dust-obscured TDEs could be selected from future infrared surveys with higher data quality and cadence.