D‐Rex++: A New Tool to Simulate Olivine's Crystallographic Preferred Orientation in the Upper Mantle With Grain Size‐Coupled Rheology

Understanding the co‐evolution of olivine's crystallographic preferred orientation (CPO) and grain size is essential for capturing the dynamics of Earth's upper mantle. Existing CPO evolution models, such as D‐Rex, successfully predict CPO development but overpredict CPO strength and do not provide a physically meaningful grain size, limiting their ability to capture realistic seismic anisotropy strength and the feedback between microstructure and rheology. In this work, we introduce D‐Rex++, a CPO evolution algorithm based on the kinematic D‐Rex code, capable of simulating realistic mantle microstructure evolution and tying it to a larger‐scale geodynamic flow regime. D‐Rex++ uses dimensionally resolved grain size evolution that accounts for both grain growth and dynamic recrystallization within polycrystalline olivine aggregates. By predicting a physically dimensional grain‐size distribution and its evolution during deformation, D‐Rex++ enables direct coupling between the microstructural state and the dominant deformation mechanisms. To demonstrate this capability, we integrated D‐Rex++ with the mantle convection code Aspect, passing the evolving mean grain size to update the viscosity field in real time. D‐Rex++ reproduces first‐order trends in dynamic recrystallization and texture development consistent with observations from experiments and natural samples. In particular, the new algorithm reproduces realistic CPO strength at high strains, comparable to experimental and natural observations. Although discrepancies remain due to incomplete physics and limited calibration data sets, the algorithm's dimensionalization enables future refinement as additional microphysical constraints become available. D‐Rex++ thus represents a key step toward more predictive geodynamic models that incorporate realistic microphysical processes.