The transfer equation in spherical geometry with central symmetry, which is a partial differential equation in two variables r and m, is reduced to a one-dimensional transfer equation in r parametric in m. This is done by dropping the angular derivative from the equation in conservation form without impeding the moments of the specific intensity. The justification for this reduction of the transfer equation is demonstrated analytically and numerically. The numerical demonstration is made by comparing the moments, zeroth to fourth, obtained using the analytic solutions in r of both the reduced equation and the transfer equation in conservation form using the discrete ordinates method. The agreement between the two sets of data is perfect within round-off errors.

The analytic solution for the two-streams discrete ordinates equations is carried out on the spatial domain in one-dimensional spherical geometry. The solution satisfies all the physical and the geometric conditions, namely: the natural boundary condition at the surface of the sphere, the symmetry condition at the center and the reciprocity of path. The solution obeys the inverse square law and is finite everywhere in the sphere except at the center. Graphical and numerical data obtained in cold absorbing spheres confirmed the analytical predictions. The conservation of energy is satisfied with high accuracy within round-off errors.

The discrete ordinates method with angular parameters is discussed in light of the recent extension of its underlying hypothesis. The analysis shows that the method, although it compares very well with the trajectory of photons model, it addresses a boundary value problem different from the problem of flow of energy in the interior of spheres

The angular derivative of the standard form of the transfer equation is shown to have a non-zero value for the radiation normal incidence. This is demonstrated by a rigorous proof. The transfer equation for the starter intensity turned out to be Euler first order equation its solution obeys the inverse square law.

A new set of discrete ordinates is proposed for one-dimensional radiative transfer in spheres with central symmetry. The set is structured with un-normalized circular functions. This resulted in a conservative and closed set of discrete ordinates equations that need not starter intensity: N equations with N unknowns. The angular derivative of the transfer equation is represented by a set of angular parameters that are calculated directly and without recursion from the abscissas and weights of the quadrature and without the constraint of asymptotic conditions. An analytic solution is obtained in infinite homogeneous and heterogeneous cold media for monochromatic radiation. Graphical and tabulated data validate the proposed set of discrete ordinates at all radial positions ranging from 10^(-7) to 10^7 within round-off errors

A nonsingular analytical solution for the transfer equation in a pure absorber is obtained in central symmetry and in a monochromatic radiation field. The native regular singularity of the equation is removed by applying a linear transformation to the frame of reference. Two different approaches are used to carry out the solution. In the first approach the angular derivative is interpreted in an original way that made it possible to discard this derivative from the equation for all black body media without upsetting the conservation of energy. In this approach the analytic solution is expressible in terms of exponential integrals without approximations but for practical considerations the solution is presented in the form of Gauss-Legendre quadrature for quantitative evaluation of the solutions. In the second approach the angular derivative is approximated by a new set of discrete ordinates that guarantees the closure of the set of equations and the conservation of energy. The solutions from the two approaches are numerically identical in homogeneous and non-homogeneous cold media as well as in local thermodynamic equilibrium and for all optical depths that can be managed with 16 significant digits arithmetics without numerical instabilities. The solution in either approach is made possible by splitting the original transfer equation into two equations, one for the centripetal energy flow and another one for the centrifugal energy flow. The end-points form of the solution proved to be very effective in numerical computations without spatial truncation errors.