Modeling $\Lambda$ polarization in Au$+$Au collisions at $\sqrt{s_{\rm NN}}=200$ GeV using relativistic spin hydrodynamics
We investigate spin polarization dynamics in relativistic heavy-ion collisions using ideal relativistic spin hydrodynamics, employing non-boost-invariant longitudinal solutions as the hydrodynamic background. Operating in the small-polarization regime, where spin evolves perturbatively on top of the bulk expansion, we first analyze a $(1+1)$D setup with transverse homogeneity. In this framework, symmetry-constrained initial conditions for the spin potential lead to non-trivial evolution and generate both local and global $\Lambda$ hyperon polarization consistent with qualitative experimental trends, though they fail to reproduce observed azimuthal structures. To address this limitation, we extend the framework by incorporating transverse flow and spatial anisotropy at freeze-out, constructing a $novel$ $(1+1+2)$D model that preserves the longitudinal dynamics. We demonstrate that the inclusion of a longitudinal spin acceleration component, coupled with transverse expansion, results in the emergence of a quadrupole pattern in the longitudinal polarization. The resulting momentum-dependent and integrated observables exhibit qualitative and, after fitting the spin potential initial intensity, reasonably good quantitative agreement with experimental data for Au+Au collisions at $\sqrt{s_{\rm NN}}=200$ GeV. Finally, we provide predictions for the in-plane transverse spin polarization, an observable that, to our knowledge, has not yet been experimentally measured.