One-dimensional profile of the rms velocity v_rms ≡ (σ_los² + v_los²)^1/2, where v_los is the line-of-sight mean velocity after subtracting the mean systemic velocity of 213 km s⁻¹. Left panel: One-dimensional v_rms profile for JWST-NIRSpec measurements from this paper. Middle panel: One-dimensional v_rms profile for the Keck-KCWI measurements from Shajib et al. (2023). The vertical error bar represents the total uncertainties, including both statistical and systematic contributions, and the horizontal error bars denote the annulus widths. The dashed lines represent the best-fit model predictions from JWST-only data (red), KCWI-only data (blue), and from a joint fit (black), and the associated shaded regions illustrate the 1σ extent of the predictions from the model posterior. The horizontal arrows represent the FWHMs of the corresponding PSF in each panel. Right panel: One-dimensional light profile from double Sérsic fits performed on the NIRSpec white-light image (red line) as part of the lens modeling in this paper, and from that, on the HST imaging in the F814W filter (blue line) from Shajib et al. (2023). The vertical dashed line marks the $1\stackrel{″}{.}3$ transition radius, where we stitch the light profiles with IR in the inner region and the visible one in the outer region that is used in the dynamical model of the NIRSpec 1D profile. From a likelihood-ratio test, we find that the JWST and KCWI measurements are consistent at ∼1.2σ (i.e., the likelihood-ratio test statistic’s p-value is ∼0.22).