Intercity Connectivity Enhances Urban Mobility Resilience to Extreme Rainfall
Mobility data from 1.1 billion devices across 366 Chinese cities shows that intercity connectivity — not just local infrastructure — drives urban resilience to extreme rainfall.
Urban climate resilience has long been framed as a local problem — a matter of better drainage, sturdier infrastructure, smarter early-warning systems. But cities do not exist in isolation. They are embedded in webs of economic, cultural, and logistical ties that quietly shape how well they absorb and recover from shocks.
Haiyan Liu and Xuemei Bai set out to quantify something that has been theoretically appealing but empirically thin: whether the quality of those intercity connections actually changes how cities perform during extreme rainfall events. Drawing on daily mobility data from more than 1.1 billion mobile devices across 366 Chinese cities — spanning 2013 to 2023 — they construct a picture of intracity mobility disruption and recovery for 450 extreme rainfall city-days in 2023 alone, and then trace how that picture changes with the structure of the intercity network surrounding each city. The result is a striking empirical case for rethinking resilience as a property not just of cities, but of the networks they inhabit.
Key Findings
Mobility disruption is brief but spatially extensive
During extreme rainfall events in China in 2023, intracity mobility dropped by an average of 9.1% on the day of the event (interquartile range: 2.3–13.0%), recovering rapidly in the days that follow. Although most individual city exposures lasted only a single day, 53.9% of events affected more than one city concurrently, making simultaneous regional disruptions — not local incidents — the dominant form of exposure.
Intercity connectivity improves recovery, not initial resistance
Across the three components of resilience measured — maximum impact (resistance), recovery time, and recovery shape — intercity connectivity shows its most consistent effects on recovery dynamics rather than on how hard a city is hit. Higher eigenvector centrality (a measure of how well-connected a city is to other well-connected cities) was associated with a 7.6% reduction in total performance loss (TPL), driven primarily through faster and smoother recovery rather than reduced initial impact. This distinction matters: connectivity appears to accelerate the post-shock rebound rather than buffer the shock itself.
Ties to economically dominant cities are particularly protective
Elite-city linkage — the strength of a city’s connections to China’s most economically powerful urban centers — was associated with an 8.5% reduction in TPL. The interpretation offered is that high-capacity hub cities can mobilize resources, personnel, and logistical support rapidly, as documented empirically during the 2021 Henan floods, when over 1,800 professional rescue personnel and 18,500 pieces of equipment were dispatched from surrounding provinces within days.
Constrained networks sharply increase vulnerability
Counterfactual simulations substituting 2023 connectivity configurations with those from the COVID-19 lockdown period (2021, when mobility between cities was heavily restricted) and the earlier socioeconomic rollback period (2013–2014) yield striking results. Relative to observed 2023 connectivity, the lockdown-era configuration would increase exposure-weighted national TPL by 54.7%; the rollback configuration would increase it by 527.1%. More integrated networks, by contrast, are associated with reduced TPL.
Projected railway development could avoid billions in losses by 2050
By 2050, global climate models project annual extreme rainfall frequency to increase by 8.7% under SSP245 and 12.2% under SSP585 relative to 2023, with multicity events accounting for over 77% of all affected city-days. Under a railway expansion scenario (full connectivity by 2035), median city-level TPL is projected to decline by 72.5% relative to 2023; under a railway equity scenario (prioritizing underdeveloped cities), the reduction reaches 76.1%. These pathways are estimated to avoid 12.8–14.2 billion Chinese Yuan in indirect annual economic losses by 2050 — losses that remain modest and statistically insignificant under socioeconomic development alone without strategic network expansion.
Smaller cities gain the most in relative terms
Although large cities benefit most in absolute economic terms from improved connectivity, the largest relative gains accrue to smaller cities (population under 1 million), where avoided indirect economic losses under the railway equity pathway could reach 1.9% of local GDP by 2050 — compared to 0.3‰ for medium-sized cities and 0.7‰ for megacities. Prioritizing network enhancement in poorer, less-connected cities therefore delivers the highest marginal resilience return.
The Network Is the Infrastructure
The conventional playbook for climate-proofing cities — harden local systems, strengthen early warning, upgrade drainage — is necessary but incomplete. What Liu and Bai demonstrate is that resilience is not just built within city limits; it is woven into the intercity fabric that surrounds them. The implication for climate adaptation planning is concrete: investing in connectivity is not merely an economic or mobility policy — it is a disaster risk reduction strategy, one whose returns compound as extreme rainfall events become more frequent and spatially extensive. As Liu & Bai show, cities with nowhere to turn when the rain comes are cities that recover more slowly, lose more, and have the least margin for error in a warming world.
Sign up for our newsletter or connect with us on social media to stay up-to-date with our latest posts and permaculture inspiration.
Explore our inspiring series and posts:
Love the post? Share it with your circle, inspire your people.