The Earth is a complex system comprising many interacting spatial and temporal scales. We developed a transdisciplinary data-model integration (TDMI) approach to understand, predict, and manage for these complex dynamics that focuses on spatiotemporal modeling and cross-scale interactions. Our approach employs human-centered machine-learning strategies supported by a data science integration system (DSIS). Applied to ecological problems, our approach integrates knowledge and data on (a) biological processes, (b) spatial heterogeneity in the land surface template, and (c) variability in environmental drivers using data and knowledge drawn from multiple lines of evidence (i.e., observations, experimental manipulations, analytical and numerical models, products from imagery, conceptual model reasoning, and theory). We apply this transdisciplinary approach to a suite of increasingly complex ecologically relevant problems and then discuss how information management systems will need to evolve into DSIS to allow other transdisciplinary questions to be addressed in the future.