Introduction

On the night of 12 November 1970, the Bhola cyclone drove a storm surge across the low islands and coastline of what was then East Pakistan. Around 500,000 people died — still one of the deadliest natural disasters ever recorded.1 There were almost no shelters, and almost no warning reached the people in its path.2

Fifty years later, Bangladesh is held up worldwide as a model of disaster preparedness. Its coast is no less exposed: the Bay of Bengal’s shallow, funnel-shaped shoreline still produces some of the highest storm surges on Earth. What changed was the country’s ability to see storms coming and get people out of their way.

>100×

Fall in deaths from comparable severe cyclones, from 500,300 in 1970 to 4,234 in 2007 — even though Cyclone Sidr’s winds were stronger.

Similar storms, very different outcomes

The clearest evidence comes from comparing storms of similar strength. A review in the Bulletin of the World Health Organization compiled every recorded cyclone in Bangladesh from 1960 to 2009.1 Four storms had wind speeds between 223 and 250 km/h — all category 6 on the scale the authors used.

Storms of similar strength, very different death tolls
Deaths from the four severe cyclones (223–250 km/h winds) recorded 1960–2009: 1970 Bhola 223 km/h, 1991 225 km/h, 1997 225 km/h, 2007 Sidr 250 km/h · logarithmic scale
Source: Haque et al., “Reduced death rates from cyclones in Bangladesh: what more needs to be done?”, Bulletin of the World Health Organization 90(2), 2012, Table 1

The 1970 storm (223 km/h) killed 500,300 people. The 1991 cyclone (225 km/h) killed 138,958. Just six years later, a storm of the same strength in 1997 killed 410. And Cyclone Sidr in 2007 — the strongest of the four, at 250 km/h — killed 4,234.1

Because the numbers span three orders of magnitude, the chart uses a logarithmic scale: each gridline is ten times the one below it.

The decline across decades

Totalling every recorded cyclone by decade tells the same story. The 1970s and the 1990s were dominated by single catastrophic storms; by the 2000s, total cyclone deaths for the whole decade were under 4,500.1

Cyclone deaths by decade
Total recorded cyclone deaths in Bangladesh, 1960–2009 · logarithmic scale
Source: Haque et al., Bulletin of the World Health Organization 90(2), 2012, Table 1 (sum of recorded cyclones per decade)

The trend has continued since. Cyclone Amphan, one of the strongest storms ever recorded in the Bay of Bengal, killed 26 people in Bangladesh in 2020.3 Before Cyclone Remal in 2024, nearly 800,000 people were evacuated; the official death toll was at least 10.4

What made the difference

No single measure explains the decline. The WHO review credits a combination of early warning systems, shelters, evacuation plans, coastal embankments, reforestation and community awareness.1

  • Warnings that reach people. The Bangladesh Meteorological Department tracks storms with radar stations at Dhaka, Khepupara and Cox’s Bazar and with international satellite data, and the Disaster Management Bureau broadcasts alerts as soon as a cyclone is detected.1
  • Volunteers who carry the warning the last mile. The Cyclone Preparedness Programme, launched by the Bangladesh Red Crescent Society after the 1970 disaster, trains coastal volunteers to spread warning signals by megaphone, flag and door-to-door visits.1 It now has around 76,000 volunteers.2
  • Somewhere safe to go. Only 44 cyclone shelters existed in 1970.2 Before 2007 there were about 1,500, each able to hold up to 5,000 people, and after Sidr the government began building 2,000 more across 15 coastal districts.1
  • Physical defences. Around 4,000 km of coastal embankments and about 120,000 hectares of replanted mangrove forest blunt the force of storm surges; mangroves are credited with shielding the south-west during Sidr.1

What still needs to be done

Saving lives is not the same as protecting livelihoods. Sidr still destroyed homes, crops and fishing boats across the south-west, and each storm pushes coastal families back into poverty.2

Evacuation also remains incomplete. Some people stay behind to guard livestock and property, distrust warnings after previous false alarms, or avoid shelters that are in poor condition. During the 1991 cyclone, only two in five shelters were usable because of flooding.1

The WHO review recommends a dense network of smaller, sturdy shelters within a 2 km walk of every household, better maintenance, and warnings designed for people without radio or television.1 With rising seas making storm surges reach further inland, that unfinished work matters more each year.

Endnotes

  1. M. A. Haque et al., “Reduced death rates from cyclones in Bangladesh: what more needs to be done?”, Bulletin of the World Health Organization, vol. 90, no. 2 (2012), pp. 150–156, doi:10.2471/BLT.11.088302. Death tolls and wind speeds from Table 1. ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9 ↩10 ↩11

  2. International Centre for Climate Change and Development (ICCCAD), “50 years of cyclone preparedness: success in saving lives, but not livelihood,” The Business Standard (2022). ↩ ↩2 ↩3 ↩4

  3. “50 years of cyclone preparedness,” The Business Standard, citing official figures for Cyclone Amphan (2020). ↩

  4. Associated Press, “Cyclone floods coastal villages and cuts power in Bangladesh, where 800,000 had evacuated” (27 May 2024); Ministry of Disaster Management and Relief figures as reported by the Bangladesh press. ↩