The baryonic Tully-Fisher relation in the Simba simulation

Abstract: We investigate the Baryonic Tully-Fisher Relation (BTFR) in the $(100\,h^{-1}{\rm Mpc})^3$ Simba hydrodynamical galaxy formation simulation together with a higher-resolution $(25\,h^{-1}{\rm Mpc})^3$ Simba run, for over $10,000$ disk-dominated, HI-rich galaxies. We generate simulated galaxy rotation curves from the mass distribution, which we show yields similar results to using the gas rotational velocities. From this we measure the galaxy rotation velocity $V_{\rm circ}$ using four metrics: $V_{\rm max}, V_{\rm flat}, V_{2R_e},$ and $V_{\rm polyex}$. We compare the predicted BTFR to the SPARC observational sample and find broad agreement. In detail, however, Simba is biased towards higher $V_{\rm circ}$ by up to 0.1 dex. We find evidence for the flattening of the BTFR in $V_{\rm circ}>300$ km s$^{-1}$ galaxies, in agreement with recent observational findings. Simba's rotation curves are more peaked for lower mass galaxies, in contrast with observations, suggesting overly bulge-dominated dwarf galaxies in our sample. We investigate for residuals around the BTFR versus HI mass, stellar mass, gas fraction, and specific star formation rate, which provide testable predictions for upcoming BTFR surveys. Simba's BTFR shows sub-optimal resolution convergence, with the higher-resolution run lowering $V$ in better agreement with data.