Scenario-driven flood resilience assessment and critical node identification for urban road networks
This study proposes a scenario-driven framework that integrates dynamic flood simulations with a two-criterion identification method to reveal how rainfall patterns and terrain jointly shape urban road network resilience, demonstrating that prioritizing reinforcement based on both structural importance and flood exposure significantly outperforms topology-only or random strategies in mitigating connectivity loss.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
When a city faces a sudden, heavy downpour, the most critical lifeline is not the power grid or the water supply, but the road network. It is the physical stage upon which evacuation, rescue, and repair operations play out. If the roads flood, the city's ability to respond to the disaster collapses, often before the water even recedes. For decades, engineers and planners have tried to figure out which specific parts of a city's road system are most vital to protect. The traditional approach has been to look at the map and identify the busiest intersections or the longest highways, assuming that these topologically important spots are the ones that need the most reinforcement. However, this method often misses a crucial reality: a road might be a major artery on a map, but if it sits on high ground and never floods, reinforcing it does nothing to stop a flood from paralyzing the city. The real question is how to find the specific roads that are both structurally important for keeping the network connected and physically vulnerable to the water.
A team of researchers set out to solve this puzzle by simulating how a real urban road network reacts to three very different types of extreme storms. They focused on the area inside the ring expressway of Chengdu, a large city in China, mapping out nearly 65,000 intersections and roads. Instead of guessing, they fed the computer model three distinct, real-world rainfall patterns: a short, intense burst of rain; a storm with multiple heavy peaks; and a long, persistent drizzle that lasts for days. By keeping the road network exactly the same and only changing the rain, they could see how the city's resilience shifted based solely on the shape of the storm. They tracked the water depth every hour, determining which roads became impassable and how the loss of those roads caused the entire network to fragment, cutting off dry areas from the rest of the city.
The simulations revealed that the type of rain matters just as much as the amount. The shortest, most intense storm caused the network to collapse instantly, with performance dropping to its lowest point within hours, but it also rebounded quickly once the water drained. In contrast, the long, persistent rain caused a slow, sustained suppression where the network never fully collapsed but remained stuck in a weakened state for a long time. Most surprisingly, the storm that caused the deepest drop in performance was not the one that caused the greatest total loss of resilience. The multi-peak storm, which hit the city with repeated waves of heavy rain, caused the most cumulative damage because the network stayed in a damaged state for longer, preventing recovery between the peaks. This finding suggests that looking only at the lowest point of a flood or the total rainfall is misleading; the entire timeline of the event, from the first drop to the final drain, determines the true cost.
The study also uncovered a hidden mechanism in how local flooding destroys city-wide connectivity. When a few roads flood, they do not just disappear; they act like severed arteries, isolating large blocks of dry, usable roads from the main network. The researchers found that for every one percent of the road network that failed, the overall ability of the city to move around dropped by about three percent. This "fragmentation" meant that the damage was far greater than the sum of the flooded roads. It also revealed a stark mismatch between what traditional planning considers important and what actually matters during a flood. The most structurally important roads, those with the highest traffic flow on a normal day, were often located on high ground and stayed dry during the storms. Conversely, the roads that stayed underwater for the longest time were often minor, peripheral streets that were not considered critical by standard maps.
To fix this, the researchers proposed a new way to choose which roads to reinforce. They developed a method that looks for roads that satisfy two conditions at once: they must be structurally important enough to hold the network together, and they must be located in areas that actually get flooded. They tested this by simulating the reinforcement of 1,000 specific nodes, which represented about 1.6 percent of the entire network. When they reinforced these dual-critical nodes, the total resilience loss dropped by between 22 and 28 percent across all three storm scenarios. This was more than five times better than reinforcing only the structurally important roads and more than 23 times better than picking roads at random. The study concludes that the most effective flood protection strategy is not to fortify the busiest roads or the wettest roads alone, but to target the specific, often overlooked intersections where high structural value meets high flood risk. This approach provides a clear, quantitative basis for city planners to spend limited budgets where they will actually save the city's lifeline.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.