Context-Aware Multimodal Representation Learning for Spatio-Temporally Explicit Environmental Modeling January 1, 2026 Result detail