Atomistic simulations provide the most detailed picture of grain boundary (GB) migration currently available. Nevertheless, extracting unit mechanisms from atomistic simulation data is difficult because of the zoo of competing, geometrically complex 3D atomic rearrangement processes. In this work, we introduce the displacement texture characterization framework for analyzing atomic rearrangement events during GB migration, combining ideas from slip vector analysis, bicrystallography and optimal transportation. Two types of decompositions of displacement data are described: the shear-shuffle and min-shuffle decomposition.The former is used to extract shuffling patterns from shear coupled migration trajectories and the latter is used to separate geometrically distinct, temperature dependent shuffling mechanisms. As an application of the displacement texture framework, we characterize the GB geometry dependence of shuffling mechanisms for a crystallographically diverse set of mobile GBs in FCC Ni bicrystals. Two scientific contributions from this analysis include 1) an explanation of the boundary plane dependence of shuffling patterns via metastable GB geometry and 2) a taxonomy of multimodal constrained GB migration mechanisms which may include multiple competing shuffling patterns, period doubling effects, distinct sliding and shear coupling events, and GB self diffusion.