The CLASP2 rocket experiment was performed on April 11, 2019. It
observed the solar chromosphere in the Mg II resonant doublet near 280 nm for three targets: the quiet Sun at the disk center, at the limb, and an active plage at midway. Slit spectra in all four Stokes parameters were obtained in the UV. The limb target revealed profile shapes for the
linear polarization caused by quantum interference in the upper term
2Po. This confirmed the theoretical prediction of Stenflo (1980),
recently thoroughly investigated by Belluzzi and Trujillo Bueno (2012).
We have developed two numerical modules for the PORTA radiative transfer code to interpret these data through forward modeling. We have solved the transfer problem numerically, including the effects of 3D geometry, partial redistribution of scattered photons, and quantum interference in a two-term atom. In this talk, I will describe the observational experiment, the theoretical method's complexity, computational demands, and approximations, show synthetic data, and discuss the importance of various effects, from spatial symmetry breaking to quantum interference.