The observed center-to-limb variation (CLV) of the scattering polarization in different lines of the Second Solar Spectrum can be used to constrain the height variation of various atmospheric parameters, in particular the magnetic fields via the Hanle effect. Here we attempt to model non-magnetic CLV observations of the $Q/I$ profiles of the Ca I 4227 A line recorded with the ZIMPOL-3 at IRSOL. For modeling, we use the polarized radiative transfer with partial frequency redistribution with a number of realistic 1-D model atmospheres. We find that all the standard FAL model atmospheres, used by us, fail to simultaneously fit the observed ($I$, $Q/I$) at all the limb distances ($\mu$). However, an attempt is made to find a single model which can provide a fit at least to the CLV of the observed $Q/I$ instead of a simultaneous fit to the ($I$, $Q/I$) at all $\mu$. To this end we construct a new 1-D model by combining two of the standard models after modifying their temperature structures in the appropriate height ranges. This new combined model closely reproduces the observed $Q/I$ at all the $\mu$, but fails to reproduce the observed rest intensity at different $\mu$. Hence we find that no single 1-D model atmosphere succeeds in providing a good representation of the real Sun. This failure of 1-D models does not however cause an impediment to the magnetic field diagnostic potential of the Ca I 4227 A line. To demonstrate this we deduce the field strength at various $\mu$ positions without invoking the use of radiative transfer.

Hanle scattering is an important diagnostic tool to study weak solar magnetic fields. Partial frequency redistribution (PRD) is necessary to interpret the linear polarization observed in strong resonance lines. Usually angle-averaged PRD functions are used to analyze linear polarization. However it is established that angle-dependent PRD functions are often necessary to interpret polarization profiles formed in the presence of weak magnetic fields. Our aim is to present an efficient decomposition technique, and the numerical method to solve the concerned angle-dependent line transfer problem. Together with the standard Stokes decomposition technique we employ Fourier expansion over the outgoing azimuth angle to express in a more convenient form, the angle-dependent PRD function for the Hanle effect. It allows the use of angle-dependent frequency domains of Bommier to solve the Hanle transfer problem. Such an approach is self-consistent and accurate compared to a recent approach where angle-averaged frequency domains were used to solve the same problem. We show that it is necessary to incorporate angle-dependent frequency domains instead of angle-averaged frequency domains to solve the Hanle transfer problem accurately, especially for the Stokes U parameter. The importance of using angle-dependent domains has been highlighted by taking the example of Hanle effect in the case of line transfer with vertical magnetic fields in a slab atmosphere. We have also studied the case of polarized line formation when micro-turbulent magnetic fields are present. The difference between angle-averaged and angle-dependent solutions is enhanced by the presence of micro-turbulent fields.