Under the natural and human-induced global change, the intensification of multiple drivers has triggered nonlinear changes in many ecosystems. Research on aquatic systems is also undergoing a transition from “changes in single processes” to “systematic changes in multi-process coupling relationships”. The impacts of global change are often transmitted within lake ecosystems by altering habitats and regulating internal processes, manifested as the reorganization of multiple processes and their coupling relationships in lakes, which further affects system structure and function. Over more than a century of development, traditional limnology has formed subdisciplines such as physical, chemical, and biological limnology. In response to emerging global challenges, this study proposes and defines “global change limnology” as a framework for comprehensively understanding how lake ecosystems respond to global change. It follows the conceptual sequence of “external drivers—changes in multiple processes and their coupling—functional changes and feedbacks”, focusing on how key physical, chemical, and biological processes within lakes and their coupling change under non-stationary stress, and how these changes affect ecosystem functioning and environmental outcomes and generate feedbacks. Centered on the core scientific questions of “how do multiple drivers reshape the coupling among lake processes, what mechanisms underlie these changes, and how can this mechanistic understanding improve our ability to predict lake ecosystem responses”, our study constructs a “driver–process–function–feedback” research framework. It shifts the research perspective from a focus on species and states to one centered on the coordinated changes of processes, and proposes a pathway that integrates multi-source observation, controlled experiments, mechanistic inference through modeling and theoretical analysis. At present, global change limnology urgently needs to develop an indicator system that is not dependent on states, such as specific species composition or nutrient concentrations, but that has both process-indicative capacity and cross-lake comparability. It also needs to overcome the bottleneck of multi-scale integration, so as to promote the elevation of empirical understanding into generalizable theoretical frameworks and risk-governance tools.