Sand-swimming, a form of desert burrowing, offers interesting potential as a locomotion mode for robots operating in loose, granular environments. Unfortunately, the computational cost of modeling the relevant physics raises obstacles to a thorough exploration of the system dynamics. Geometric mechanics offers techniques for reducing the complexity of evaluating gaits, thereby offering the potential for exploring a gait design space; unfortunately, these tools have historically been restricted to systems with linear, analytical dynamics. In this paper, we present a framework for combining empirical data from nonlinear models with geometric gait evaluation methods. The resulting tools both reduce the computational costs of describing sand-swimming and reveal fundamental aspects of the motion.