Remote sensing has been used in Antarctic studies as an earth observation technique to study the polar region. A remote sensing forward model is an important tool in polar research to study and understand scattering mechanisms and sensitivity of physical parameters of snow and sea ice. In this paper, a reliable theoretical model to study sea ice is developed. The theoretical model in a prior work was improved by including multiple-surface scattering, based on an existing integral equation model and additional second-order surface-volume scattering. This model is applied to a desalinated ice layer above thick saline ice and analyzed using different frequencies, bottom surface roughness and sea-ice layer thickness. Improvement in calculation of the backscattering coefficient of the sea-ice layer is investigated for both co-polarized and cross-polarized returns. The effect on each scattering mechanism is also investigated, to understand in more detail the effect of surface multiple scattering and second-order surface-volume scattering. Comparisons are also made with field measurement results, to validate the theoretical model. Results show improvement in the total backscattering coefficient for cross-polarized return in the studied range, suggesting that multiple-surface scattering and surface-volume scattering up to second order are important scattering mechanisms in the sea-ice layer and should not be ignored in polar research.