Bar properties as a function of wavelength in TNG50: analysis of mock images

Abstract

Recent studies used TNG50 galaxies to study bar formation, evolution, and properties like length, strength, and pattern speed. In simulations, these are typically derived from particle positions and mass distribution, neglecting stellar light and extinction effects. However, observational studies indicate that bar appearance depends on wavelength. To test whether this dependence exists in TNG50 at $z \sim 0$, we analysed 50 strongly barred galaxies using mock images from SKIRT radiative transfer simulations covering infrared to ultraviolet filters (Spitzer 3.6 um, SDSS i, r, g, S-PLUS J0378, GALEX NUV, and GALEX FUV). Bar ellipticity and length were measured via ellipse fitting. Ellipticity generally increases by 6 percent from 3.6 um to g band, and by 9 percent to J0378 band. On average, TNG50 bars cannot be said to be longer in bluer filters when the entire sample is used. However, the trend is detected when only star-forming galaxies are considered. In this star-forming subsample, bar length increases by 10 percent from 3.6 um to g band, and by 17 percent to J0378 band; moreover, the bar can appear up to 20${-}$30 percent longer in bluer mocks than in the mass map. Over 90 percent of bars vanish in UV due to minimal emission by dominant stellar populations. We reproduced the bar properties morphological dependence phenomenon using age-filtered mass maps, where older stars form shorter, rounder bars, and younger stars generate longer, more elliptical ones. TNG50 bars exhibit a wavelength-dependent trend similar to observations: bars appear more elliptical and longer in bluer filters, with this effect being stronger in star-forming galaxies.