Skip to main content
Institute for Social Vision Design

Disaster

4 items

Insights & Analysis

For a Distant Tsunami That Brings No Shaking, Only Trust in the Warning Drives Evacuation — the Dilemma of False Alarms and Misses

In June 2026, an earthquake near the Philippines put a tsunami advisory on Japan's coast, and when some municipalities issued wide-area evacuation orders the government warned this could undermine trust in the warning. Evacuation rates vary — Noto 59.3%, Tohoku's immediate 57%, the distant Kuril 19.9%. For a distant tsunami with no felt shaking, trust in the warning is the only trigger, yet false alarms lower evacuation and issuing warnings freely for fear of a miss has limited value. This reads the design of trust, not accuracy.

Insights & Analysis

Ten Riot Police, Twenty Minutes, a House That Had Not Collapsed — False Rescue Requests Went from 1 in 573 to 104 in 1,091

After the 2026 Kumamoto earthquake of 28 July, false rescue requests naming addresses that do not exist, and AI-generated videos purporting to show the damage, spread on social media. Measured by the same system, false posts among rescue requests numbered 1 out of 573 in the 2016 Kumamoto earthquake and 104 out of 1,091 in the 2024 Noto Peninsula earthquake. What changed is not the volume of malice but who pays for the correction: no longer the attention of readers, but the deployment of fire and police.

Practice Guides

Practical Guide to Social Vulnerability Mapping During Disasters — Lessons from the Noto Peninsula Earthquake

Disaster damage concentrates on socially vulnerable groups. A guide to vulnerability mapping, inclusive shelter management, and community resilience building.

Insights & Analysis

15 Years After 3/11, 2 Years After Noto — The Structural Limits of Japan's ¥41 Trillion Recovery

Fifteen years after the Great East Japan Earthquake and two years after the Noto Peninsula earthquake, Japan's ¥41 trillion recovery budget rebuilt infrastructure but failed to bring residents back. Population declined in 90% of 42 affected municipalities. A structural analysis of hardware-biased recovery and the absence of a recovery model for depopulating areas.