鉛衝突のΛ偏極から分離時に残る形状の歪みを推定する

Longitudinal hyperon polarization is generated by velocity gradients at decoupling, but its quantitative connection to the evolving collision geometry has remained unclear. We analyze 48 Pb-Pb initial states with (3+1)-dimensional ideal hydrodynamics and the isothermal local-equilibrium spin prescription. The positive kinematic-shear contribution correlates with the freeze-out eccentricity $\epsilon_{2,\mathrm{fo}}$, whereas the magnitude of the negative kinematic-vorticity contribution depends jointly on elliptic flow $v_2$ and the eccentricity survival fraction $S_\epsilon=\epsilon_{2,\mathrm{fo}}/\epsilon_{2,\mathrm{init}}$. Their competition follows $R\simeq0.84(\epsilon_{2,\mathrm{fo}}/v_2)^{0.47}$, and their sum is described by $P_{z,s2}^{\mathrm{ILE}}\simeq0.0404v_2S_\epsilon [0.850(\epsilon_{2,\mathrm{fo}}/v_2)^{1/2}-1]$. This compact relation describes $87\%$ of the variation across the 48-state scan. After its coefficients were fixed, it describes $86\%$ of the variation across 12 newly calculated initial states, including four at the previously unsampled $b_0=0.685$. A proof-of-principle inversion using ALICE polarization data then returns freeze-out eccentricities inside the model fit domain. These results show that longitudinal polarization can constrain the spatial anisotropy remaining at decoupling, complementing the momentum-space information carried by elliptic flow.