Modern geosciences are increasingly shaped by the interplay of data, physical models, and mathematical structure. Applied mathematics provides a unifying framework for modeling, analyzing, and predicting complex Earth systems, while geoscientific challenges continue to drive new mathematical developments. This session highlights recent advances at this interface, with an emphasis on how mathematical ideas, such as stochastic modeling, dynamical systems, data assimilation, and uncertainty quantification, enable deeper understanding and improved prediction of multiscale, nonlinear, and extreme phenomena. We welcome contributions that both apply mathematical methods to geoscience problems and develop new mathematical approaches motivated by geophysical applications. Topics include, but are not limited to, multiscale modeling, data-driven and machine learning methods, rare event prediction, and the integration of data with physical models. The goal is to foster sustained exchange between applied mathematics and geosciences.