From educational accessibility to high-fidelity forward modeling: Development and validation of a lightweight python radiative transfer model

Built upon the Gauss–Seidel iterative method–based gsit model, liteMRT demonstrates that sophisticated radiative transfer modeling can be achieved through a concise, well-structured codebase, making it highly accessible for educational purposes and algorithm development.

In addition to its pedagogical clarity, this research also provides three benchmark cases for scalar (non-polarized) radiative transfer, covering homogeneous and non-homogeneous atmospheric profiles over Lambertian gray and urfaces. Validation against the benchmark Linearized Discrete Ordinate Radiative Transfer (LIDORT) model shows excellent numerical agreement, with maximum relative errors below 0.04%. The code maintains computational efficiency through Numba JIT acceleration and requires no external dependencies beyond standard scientific Python libraries, ensuring reproducibility across different computing environments.