A Spherical Geometric B-Spline Model for Lunar Brightness Temperature Data Approximation

Brightness temperature (TB) data from the Chinese Chang’E-2 (CE-2) microwave radiometer (MRM) are constrained by the limited quantity of the original dataset, which cannot express global TB distribution. To construct the lunar TB model with the TB data obtained by the MRM onboard CE-2, we propose a novel spherical geometric B-spline (SGB-spline) model. The model fully integrates the observed TB data with the lunar geometric features and determines optimal fitting parameters through a subdivision-based optimization process. More specifically, the establishment of the lunar TB model begins by employing spherical area coordinates (SACs) for CE-2 TB data representation across all four frequency channels, followed by applying geometric B-splines to refine the TB distribution. At the same time, it preserves the geometric integrity of the Moon. We observed that the SGB-spline model constructs the more comprehensive TB models during both lunar daytime and nighttime in the 3-D Euclidean space, providing a more detailed representation of the spherical spatial information and the effect of frequency channels. Experimental results demonstrated that the proposed SGB-spline model significantly outperforms representative interpolation approaches.