A Forecast Is Not Protection: Governing Extreme Weather Before It Becomes Disaster

In 2024, the most intense floods to hit Assam in more than a decade triggered about 2.5 million movements within India. Across Africa in 2025, floods accounted for more than half of reported extreme weather events, killing more than 200 people in Nigeria and 160 in the Democratic Republic of Congo, according to the World Meteorological Organization. Different places, but the same question: why do increasingly monitored and forecast hazards continue to cause such predictable harm?

The familiar answer is that extreme weather is becoming more severe. That is true. The World Meteorological Organization reports that 2015 to 2025 were the hottest eleven years on record, with the global average temperature in 2025 about 1.43°C above the 1850 to 1900 average. But temperature explains only part of the harm. A flood encounters drainage, housing, roads, hospitals and electricity networks. A heatwave encounters homes that trap heat, outdoor workers, strained power supplies and unprepared clinics.

Extreme weather is therefore testing entire systems of government. The weakest point is often the space between knowing that danger is coming and making protection possible. A forecast becomes protective only when it changes what public institutions do before the event arrives.

When one hazard creates several failures

Governments divide responsibility by sector. One department manages water, another health, another transport and another housing. Extreme weather does not observe those divisions. Heavy rainfall can flood roads, block access to clinics, interrupt food supplies and displace families at the same time. A power failure during extreme heat can affect cooling, water pumping and health services. Yet preparedness is still often planned and financed within separate institutions, allowing each agency to fulfil a narrow responsibility while the overall response breaks down.

Africa illustrates this tension. The continent has recorded a 72 per cent improvement in the strength of its early-warning systems since 2015, yet it still has the lowest overall capability of any region. Progress in forecasting is real, but it does not automatically produce safe housing, reliable transport or capable local administration.

The cost is substantial. The United Nations Office for Disaster Risk Reduction estimates direct disaster losses at about US$202 billion annually. Once damage to health, livelihoods, supply chains and ecosystems is counted, the cost exceeds US$2.3 trillion. This reflects the force of hazards, but also the price of development that ignores known risks.

A warning is where responsibility begins

There has been genuine progress. By 2025, 119 countries, representing 60 per cent of all countries, reported having systems capable of warning people about several hazards. Disaster mortality is nearly six times lower in countries with stronger warning systems than in those with limited capacity.

But sending an alert is not the same as enabling action. A warning should trigger a decision, identify a responsible authority, release money and change how relevant public services operate. If any link is missing, the warning remains information rather than protection.

Who decides when schools close, or shelters open? Who arranges evacuation transport? When do hospitals secure backup power, and when must employers suspend unsafe outdoor work? If these questions are answered only after an emergency begins, the problem is no longer climate information. It is an administrative failure.

Weather agencies cannot solve this alone. They can explain what the weather may do, but they cannot clear drainage, provide evacuation transport or stock a clinic. Those outcomes require weather services, local government, health departments, transport agencies, employers and community organisations to work together.

India offers practical examples. Heat action plans in cities such as Ahmedabad link forecasts to public health advice, hospital preparedness and reduced exposure. Flood forecasting, improved drainage, cool roofs, shaded spaces and reliable transport similarly show that forecasts matter when they change how a city operates.

There is also a political difficulty. Rescue teams, relief supplies and official visits are visible. Prevention is quieter. Drainage maintenance, building enforcement and tested evacuation plans rarely create dramatic images, but they determine how many people later need rescuing. Response should not be mistaken for preparedness.

Can people actually act on the warning?

The final test lies outside government offices. Advice to remain indoors offers little protection in an overcrowded, poorly ventilated room. Avoiding outdoor work means something different to a salaried employee and to a street trader choosing between heat and lost income. Evacuation assumes access to transport and a safe destination. Migrants may fear questions about documents, while older people and persons with disabilities may need practical assistance.

These groups are not passive recipients. They often know local risks and practical ways to reach neighbours. Authorities must use that knowledge and remove barriers to action. A warning that ignores working hours, language, mobility, income or trust is not inclusive simply because it was widely broadcast.

Resilience cannot be downloaded onto a mobile phone. Success should be measured by who received the warning, who understood it, who could act and what support reached those who needed help.

Extreme weather also forces people to move. Storms, floods and other hazards triggered 29.9 million movements within countries in 2025. Mobility is therefore part of preparedness, not a separate migration issue. Plans must cover evacuation, temporary accommodation, healthcare, schooling, property protection and safe return. An evacuation that leaves people without income, medicine or education is only a partial success.

Turning forecasts into protection

The starting point is to treat a warning as a public service, not a message. Every alert level should trigger agreed actions, assigned to named institutions and supported by money available immediately. Plans must be tested with the agencies and communities expected to use them. A plan that fails in a crowded settlement is not functioning.

Governments must map more than the physical hazard. They need to know who lives and works in exposed areas, which services may fail and what prevents people from responding. Local knowledge cannot replace forecasting, but it shows how danger will be experienced.

Public investment is equally important. A 2025 World Bank assessment of 24 Indian cities estimated that timely adaptation could prevent US$5 billion in annual rain-related flood losses by 2030 and save more than 130,000 lives from extreme heat by 2050. Drainage, cool roofs, green spaces, prepared health services and reliable transport are basic requirements for safe cities. Across Africa, investment must reflect different realities while giving local authorities the money and capacity to act.

Finally, every major warning should be reviewed. Did it reach the right people? Could they act? Which services changed their operations? Was funding available before the damage? Were displaced residents able to access essential services and return safely? The answers should shape the next budget, infrastructure decision and preparedness plan.

Not every loss can be prevented as weather extremes become more severe. But intensity cannot explain every institutional failure. The next time an alert is issued, policymakers should ask more than whether the forecast was accurate. What changed because of it? Which resources were released, which services acted and who became safer? If those answers are uncertain, the forecast has not yet become protection.

Disclaimer: The views expressed in this article are solely those of the author and do not necessarily represent the views, policies, or positions of the organisation.