Climate change and urbanization are both long-term large-scale phenomena. Like many deep-rooted socio-environmental challenges today, disastrous effects result from small yet persistent progressions accumulated over long time (e.g., greenhouse gas emissions, incremental land developments). When fully emerged, the system is already complex and difficult to change. Similar to when a person develops a disease from long-term habits, a successful cure requires both immediate treatments and long-term commitments to better habits, lasting human-environment harmony demands enduring transformations of socio-environmental trends. What socio-environmental transformations can generate long-term large-scale social and environmental wellbeing? How to implement such systematic transformations? To answer questions like these, we need a sandbox where alternative long-term futures of large-scale human-environment systems can be explored with agile computation. In this talk, I will share our research, modeling global long-term spatiotemporal dynamics of population and urban land change using statistical learning techniques, investigating large-scale interactions between urbanization and climate change for actionable resilience over the 21st century, and visions for future steps.