Green urbanism, eco-cities and innovative green infrastructures are being promoted and adopted worldwide to improve urban well-being. However, all these efforts are guided by aesthetic, conceptual, and qualitative principles rather than a quantitative understanding of the underlining biophysical processes, scales and feedbacks. To fill this knowledge gap, the following hypotheses are framed: (i) urbanization generates critical transitions in urban biophysical states; (ii) general scaling laws and bulk relations linking metrics of urbanization, vegetation and ecosystem services exist and can be used to identify critical thresholds and early warning signals; (iii) different vegetation types and patterns can enhance resilience to climate fluctuations by re-coupling cities to their local biosphere. The overarching goal is to identify efficient urban vegetation strategies and introduce quantitative tools (i.e. scaling laws, dynamical system theories, eco-hydrological modeling) in the design of future cities.