Synthetic spectral lines from radiative magnetohydrodynamic models are essential in order to interpret the observations of the solar chromosphere, but only a few computations in the literature take 3D effects in the synthesis into account. The lines form in the middle to upper chromosphere and are well suited to studying the structure of the chromosphere. However, the details of their formation in the solar chromosphere are not fully understood. We studied the effects of 3D on the rt line properties and tested the known correlations between the underlying atmosphere and spectral line features in a new model of the chromosphere. We forward modeled the with partial frequency redistribution (PRD) in a self-consistent 3D radiative magnetohydrodynamic (rMHD) simulation with non-local thermodynamic equilibrium (NLTE) energy transport and non-equilibrium (NE) hydrogen ionization of an region simulated with the . We compared the spectra computed with 3D RT to those computed with 1.5D RT and to observations from the Interface Region Imaging Spectrograph (IRIS). Furthermore, we tested the correlations between the spectral line properties and underlying atmospheric properties such as temperature and velocity structure. lines in 3D rt en muramche The spatially averaged model, the qualitative difference between 1.5D and 3D RT results is even more pronounced than in the public Bifrost snapshot, as given in the literature. We found that this large discrepancy might be partly attributed to the horizontal velocities that are naturally included in the full 3D synthesis, but not in typical 1.5D computations. We confirmed that the correlations between spectral line properties and the underlying atmosphere from the MURaM-ChE simulation are similar to those obtained from Bifrost, but show more scatter due to the more dynamic atmosphere. In addition to already known correlations, we found that the ̨twov (blue) peak of the spectral lines computed with 3D RT match approximately a typical IRIS observation, which includes quiet sun and network elements. However, the peak separation is still slightly lower in the simulation. In contrast, the 1.5D RT spectra tend to overestimate the peak intensities and the central minimum significantly. In the muramche rt rt line forms preferably in upflows, whereas the ̨twor (red) peak forms preferably in downflows. The lines computed with 3D RT match the observations better in the core intensities and their distribution on the Sun compared to 1.5D computations. This underlines the importance of 3D RT in the forward modeling of
Ondratschek et al. (2026) studied this question.