Title: From mobility patterns to scaling in cities

Abstract: The recent availability of data for cities has allowed scientists to exhibit scalings which present themselves in the form of a power-law dependence with population of various socio-economical (GDP, gasoline consumption, number of patents, etc.) and structural indicators (urbanized land area, lane miles of the road network, etc.). Despite numerous attempts, the gap between empirical regularities and the theoretical understanding of cities has not yet been bridged. We propose here a dynamical, stochastic theory of urban growth which accounts for the observed scalings and we confirm these results on US and OECD empirical data. Our results suggest that the key needed to decipher most of cities' complexity is the spatial structure of the mobility pattern, controlled by congestion. In particular, we show that the dependence with population size of the total number of miles driven daily, the total length of the road network, the total traffic delay, the total consumption of gasoline, the quantity of $CO_2$ emitted and the relation between area and population of cities, are all governed by a single parameter which characterizes the sensitivity to congestion. Interestingly, we also find that Zipf's law for cities can be interpreted in terms of cost minimisation. Finally, our results suggest that the diseconomies associated with congestion scale superlinearly with population size, implying that, despite polycentrism, cities whose transportation infrastructure rely heavily on traffic sensitive modes are unsustainable.