A false sense of security can be a dangerous thing. The "Safe Development Paradox" is a phenomenon where anti-flooding infrastructure makes people feel safe, so communities subconsciously start building closer and closer to historically vulnerable lowlands. We can study this effect by looking at where people settled before and after infrastructure installation. Researchers at the International Research Institute of Disaster Science (IRIDeS), Tohoku University, and Meijo University compared Meiji era natural landscapes and built-up area distribution to the present day, with flood simulations and analyses that paint a fascinating picture of what it really takes to keep civilians safe from danger.

Focusing on the historically flood-prone Marumori Town in Miyagi Prefecture, Japan, the researchers looked at how construction over the past century could have altered the severity or exposure of floods. However, evaluating how these physical changes affected flooding has historically been challenging, because rainfall patterns, river structures, and urban growth have all changed simultaneously. To measure the exact impact of these man-made structural changes, researchers had to use highly detailed computer simulations and old maps to recreate and study the original, untouched natural landscape.

"By looking back at conditions in the Meiji era as a natural baseline, we can use models to determine whether the flood prevention infrastructure we installed is helping," says Nilo Lemuel J. Dolojan (IRIDeS). "Surprisingly, we discovered that these 'upgrades' might have made things worse in some cases."

It may sound counterintuitive, but installing infrastructure to reduce flooding can sometimes lead to more floods. Manmade walls along rivers called levees are designed to keep swollen rivers from spilling over during heavy rain and extreme weather events. However, during lighter rain, the very same walls meant to protect us can trap rainwater inside neighborhoods, causing worse inland flooding today than there was over a century ago.

During heavy storms with widespread flooding, the simulations suggested that flood volumes decreased thanks to levees; but during weak rain where flooding is scarce, flood volumes increased. Further analysis revealed that there is a tipping point beyond which modern terrain modifications transition from mitigating to exacerbating flood depths. Assessing where that tipping point lies for each body of water may be key when planning whether flood prevention infrastructure is necessary or not.

"In general, the levees are still effective during heavy rainfall. But even then, we still found exceptions," says Shuji Moriguchi (IRIDeS). "During the unexpectedly strong 2019 Typhoon Hagibis, modern flood-control structures were completely overwhelmed."

In other words, levees cannot keep us completely safe from floods under every weather condition. Although levees are not failproof, the researchers noted that nearby communities acted as if they were. The "Safe Development Paradox" was in full effect, evidenced by 110% growth in settlement area exposed to flooding.

These findings are not just helpful for Marumori Town - they can be used by researchers to assess other areas as well. The computer models in this study separate the effects of man-made modifications from the effects of climate change and urban growth, allowing for quantitative verification of the "Safe Development Paradox." The researchers hope to continue improving this model, with the hope that it sparks conversation about urban planning that reduces hazard exposure and keeps people safe.

The research was completed as a comprehensive partnership agreement between Tohoku University and Meijo University, published in the Journal of Hydrology: Regional Studies on September 14, 2026.