| 英文摘要: | We quantify the importance of early action to tackle urban sprawl. We focus on the long-term nature of infrastructure decisions, specifically local roadways, which can lock in greenhouse gas emissions for decades to come. The location and interconnectedness of local roadways form a near-permanent backbone for the future layout of land parcels, buildings, and transportation options. We provide new estimates of the environmental impact of low-connectivity roads, characterized by cul-de-sacs and T-intersections, which we dub street-network sprawl. We find an elasticity of vehicle ownership with respect to street connectivity of –0.15—larger than suggested by previous research. We then apply this estimate to quantify the long-term emissions implications of alternative scenarios for street-network sprawl. On current trends alone, we project vehicle travel and emissions to fall by ~3.2% over the 2015–2050 period, compared to a scenario where sprawl plateaus at its 1994 peak. Concerted policy efforts to increase street connectivity could more than triple these reductions to ~8.8% by 2050. Longer-term reductions over the 2050–2100 period are more speculative, but could be more than 50% greater than those achieved by 2050. The longer the timescale over which mitigation efforts are considered, the more important it becomes to address the physical form of the built environment. |