Is Extreme Weather Getting Worse? What the Data Shows

by peterpfhst | Oct 1, 2026

Stencil illustration of a weather station beneath storm clouds with muted purple accents.

Extreme weather trends depend on the hazard, location, and period. Heat extremes have increased, while cold extremes have declined. Heavy rainfall has intensified across many well-observed regions. Floods, droughts, hurricanes, and fires show more varied patterns. No single global trend accurately describes every type of extreme weather. [1]

My argument about extreme weather has a weakness: the phrase hides several different questions. Some extremes have increased, others have decreased, and several trends remain unclear.

The strongest objection concerns storms. Historical records do not show every storm type increasing everywhere.

Consider hurricanes reaching the United States. Records show no clear increase across the past century. That directly challenges claims that warming means more hurricanes hitting America. [2]

Other evidence also challenges a simple story of worsening conditions. NASA research found global burned area fell 24% between 1998 and 2015. [3]

A reasonable person could therefore reject a sweeping “everything is worsening” conclusion. I would reject it too.

Imagine a skeptical reader who remembers earlier floods and brutal summers. They want measurements, clear dates, and fair comparisons.

They suspect someone chose a starting year that makes the trend look alarming. They wonder whether better reporting explains rising disaster counts.

They want old storms, falling death tolls, and uncertain findings acknowledged. They also distrust anyone who turns scientific findings into automatic policy demands.

Those concerns deserve answers, not labels.

This article has limits too. It cannot cover every region, weather record, or competing study.

The sources cover different periods, which appear alongside the relevant findings. A finding ending in 2015 cannot describe every year afterward.

Nor can admitting weaknesses make an argument impossible to challenge. You should still check the evidence and question the reasoning.

Here is the conclusion I can defend: several important weather hazards have increased. However, the direction depends on the hazard, location, period, and measurement.

That answer sounds less dramatic than either side might prefer. It also gives us something useful to examine.

We need to separate heat from storms, rainfall from flooding, and damage from weather. Otherwise, we risk debating different questions while using the same words.

How Do Scientists Measure Extreme Weather Trends?

Start with a simple example: two towns experience the same heavy downpour.

One town has open fields and drains that handle the water. The other has paved surfaces and drains that overflow.

Their rainfall matches, but their flood damage differs.

That difference matters when someone uses damage figures to describe weather trends. The figures may reflect more buildings, poor drainage, or more people facing floods.

Imagine a coastline with ten houses several decades ago. Today, that same coastline holds hotels, shops, and thousands of homes.

An identical storm could now cause much greater losses. Higher losses alone would not prove stronger winds.

The reverse also matters. Better buildings could reduce damage despite stronger winds.

We therefore need separate questions:

Frequency: How often does the event happen?
Intensity: How strong does it become?
Duration: How long does it last?
Area: How much land does it affect?
Impact: What happens to people, buildings, crops, and wildlife?

These measures can move in different directions. Fewer storms could still bring heavier rainfall.

We also need a clear meaning for “extreme.”

A temperature might count as unusual in northern Sweden but common elsewhere. Local seasons and past conditions shape that judgment.

The comparison period matters just as much. Comparing today with one unusual year can produce a misleading picture.

A fair comparison checks several decades when reliable records allow it. It also explains gaps and changes in measuring methods.

For hurricanes, earlier observers missed some storms far from shipping routes. Satellites later improved coverage, complicating comparisons with older storm counts. [2]

That does not make historical records worthless. It means researchers must account for their limits.

Most importantly, observed trends and future projections answer different questions. A forecast for 2100 cannot prove what has already happened.

Before accepting any broad claim, ask what the author actually measured. A graph showing costs cannot answer a question about wind speeds.

Likewise, one flooded street cannot establish a worldwide rainfall trend. It can show real suffering without proving every proposed explanation.

Are Heatwaves and Heavy Rainfall Becoming More Common?

Heat provides some of the strongest evidence of changing extremes.

The IPCC finds that hot extremes have increased worldwide since 1950. Cold extremes have moved in the opposite direction. Human influence drives the broad temperature changes. [1]

This does not mean every summer beats the previous summer. It describes a longer shift across many observations.

Think about a town that rarely crossed a particular heat threshold. A warming climate can push more days beyond that threshold.

The town can still experience cool weeks. Those weeks do not erase the longer pattern.

Likewise, a harsh winter does not settle the global question. Local weather and worldwide trends operate across different areas and periods. [4]

American city records offer a concrete example. EPA’s 2024 assessment reports rising heatwave frequency, duration, and intensity across major cities. Its comparison covers records beginning in the 1960s. [5]

That starting point deserves attention. It answers a specific question about changes since that period.

It does not establish that every earlier decade lacked severe heat. Nor does it make every city’s history identical.

Good evidence can support a clear trend without supporting an unlimited claim.

Cold extremes deserve equal honesty. Their decline counts as a real change in extreme weather.

We should not quietly exclude them because they complicate the headline. “More heat extremes” describes the evidence better than “everything grows more extreme.”

Heavy rainfall also shows a broad increase across well-observed land regions. However, the evidence does not support identical changes in every location. [1]

The physical explanation helps. Warmer air can carry more water vapor when moisture is available.

Storms can then release more water through intense rainfall. However, winds and available moisture still shape where rain falls. [6]

A warmer world therefore does not mean constant rain everywhere. It can produce heavier downpours while some places still face dry spells.

The important distinction involves rainfall strength versus rainfall regularity. A month’s rain arriving suddenly creates different problems from gentle, repeated showers.

Are Floods, Droughts, and Hurricanes Increasing?

More intense rain does not automatically mean every river floods more often.

River flooding also depends on soil moisture, snowmelt, and land conditions. The IPCC finds mixed regional trends in rivers’ highest flows. Researchers lack confidence in a single worldwide trend for those peaks. [1]

This uncertainty deserves plain language. We cannot confidently describe all river floods with one rising global line.

Drought requires similar care because the word covers different shortages.

A farmer might worry about dry soil around crop roots. A water supplier might worry about low reservoirs.

Those problems overlap, but they do not measure exactly the same thing.

Heat can worsen drought by increasing moisture loss. Evaporation and water loss through plant leaves generally increase with temperature. [7]

That means rainfall totals alone can miss part of the problem. The same rainfall may leave less available water under hotter conditions.

Research covering the American West illustrates this point. A NOAA summary describes heat-driven evaporation dominating recent drought intensity. Its analysis compares 1948–1999 with 2000–2022. [8]

That finding concerns a specific region and measure. We should not turn it into “everywhere experiences more drought.”

Hurricanes require another distinction: storm numbers versus storm strength and rainfall.

Climate scientist Tom Knutson reviews hurricane evidence on NOAA’s research website. His review reports no clear century-long rise in American hurricane landfalls.

Future models generally project fewer or similar total tropical cyclone numbers globally. However, they also project stronger average intensity and heavier storm rainfall. Those forecasts do not establish increases in every historical record. [2]

A season’s storm count therefore cannot capture every changing risk. Yet neither can one powerful hurricane prove a long-term trend.

Tornadoes add another warning against simple counting. NOAA researchers examined American records covering 1954–2013, excluding the weakest tornadoes.

They found fewer tornado days, but more tornadoes on those days. That finding describes clustering, rather than a simple rise across all measures. [9]

These examples explain why careful answers can sound uneven. The evidence itself varies between hazards.

A clear heat trend does not remove uncertainty about tornadoes. Uncertain tornado trends do not erase the evidence about heat.

Why Have Global Burned Area and Disaster Deaths Declined?

Fire offers one of the strongest challenges to careless climate claims.

NASA research found a substantial global burned-area decline during 1998–2015. Changes in farming and land use helped drive that decline. [3]

Anyone claiming that all fire measures increased should address this evidence directly.

However, total burned area combines very different landscapes. Grasslands, savannas, forests, and farmland do not tell the same story.

NASA describes declining burning in some grasslands alongside worsening fire activity in northern forests. A worldwide total can therefore hide opposing regional changes. [10]

Fire also needs more than hot weather. Fuel, ignition, and land management help determine what actually burns.

NASA’s broader fire overview describes how farming changes reduced burning across several landscape types. It also connects warming with growing fire risks in other settings. [11]

We should therefore separate fire-friendly weather from actual fires and burned acreage. Those measures answer related questions, but they are not interchangeable.

Death tolls deserve the same care.

WMO’s assessment covering 1970–2019 found that recorded disaster deaths fell almost threefold. Better warnings and disaster management helped save lives. [12]

That is meaningful progress. We should acknowledge it without adding a dismissive “but” immediately afterward.

People can reduce their exposure and improve their chances of survival.

WMO’s later update, covering 1970–2021, also emphasizes improved warnings and disaster management. These gains matter when evaluating society’s ability to handle dangerous weather. [13]

Still, survival rates cannot measure wind speed or rainfall.

Imagine two equally strong storms striking decades apart. During the later storm, forecasts help families reach safe shelters.

Fewer deaths would show successful protection. They would not show that the storm weakened.

The same logic applies when hazards grow stronger. Better protection can partly offset a worsening threat.

Conversely, poor preparation can turn familiar weather into a major disaster.

We should welcome falling deaths while measuring weather separately. Otherwise, we might mistake progress in safety for proof about atmospheric conditions.

Both questions deserve attention because both affect real lives.

How Do Scientists Link Climate Change and Extreme Weather?

After a major flood, people naturally ask what caused it.

The question sounds simple, but several causes can operate together. Weather patterns deliver the rain, while local conditions shape the flooding.

Climate change may also alter the rainfall’s likelihood or strength.

Researchers study that contribution through event attribution. They compare possible conditions with and without human-caused warming.

The Met Office explains that these studies examine changes in probability and intensity. They do not simply stamp every disaster with the same explanation. [14]

Consider a purely hypothetical example. Researchers estimate that warming doubled the chance of a particular downpour.

That would not mean warming caused half the resulting property damage. Drainage, building locations, and flood defenses would still matter.

Nor would it mean the downpour could never happen naturally. A change in probability differs from creating an entirely new possibility.

The strength of any result depends on the evidence and methods.

Researchers need observations and models suited to the event. They must also define its location, duration, and severity clearly. [15]

For example, one hour of rain differs from a wet winter. A study answering one question cannot automatically answer the other.

The Met Office highlights this problem for short-duration British rainfall. Natural swings can hide long-term trends, while researchers examine warming’s role in individual events. [16]

This is why uncertainty should appear beside the conclusion. Readers need to know whether the evidence is strong, limited, or mixed.

They also need room to distinguish scientific findings from policy choices. Evidence can describe changing risks without settling every response.

Different people may weigh costs, fairness, and practical options differently.

That disagreement does not require pretending the physical evidence says nothing. Nor does accepting the evidence require accepting every proposed solution.

The goal should remain understanding rather than winning trust through confident wording.

An honest article cannot promise to destroy your skepticism. It should help you direct that skepticism toward precise, answerable questions.

Ask what changed, where it changed, and how researchers know. Then ask what evidence could alter their conclusion.

How Can You Check Extreme Weather Claims?

The evidence supports a mixed answer with several clear directions.

Heat extremes have increased, while cold extremes have declined. Heavy rainfall has intensified across many regions with sufficient observations. [1]

Other hazards require more local detail and careful measures. Hurricane counts, river floods, drought, and fires resist one universal answer.

That does not place every claim on equal footing. Strong evidence deserves more weight than a striking photograph or isolated example.

It also does not justify hiding evidence that complicates the picture. Falling burned area during a documented period belongs in the discussion.

So do uncertain storm trends and improvements in disaster survival.

The most useful answer depends partly on your actual concern. A homeowner near a river needs different information from a farmer.

Someone caring for older relatives may focus on prolonged summer heat. A coastal community may need rainfall and storm-surge evidence alongside wind records.

A worldwide headline cannot replace those local questions.

When you encounter the next alarming claim, try this checklist:

Which type of extreme weather does the claim describe?
Does it measure frequency, strength, duration, or damage?
Which location and time period does it cover?
Have reporting methods changed during that period?
Does the source describe observations or future projections?
What uncertainty or contrary evidence does the author acknowledge?

These questions also apply to reassuring claims. “Nothing has changed” deserves the same scrutiny as “everything is getting worse.”

You do not need to choose between alarm and dismissal. You can accept clear findings while challenging claims that outrun them.

That approach leaves room for new evidence. It also protects against choosing facts merely because they feel comfortable.

Extreme weather deserves better than a headline that outruns its evidence. Neither a frightening image nor a reassuring slogan answers your local risk question. Check your national weather service’s long-term record for one hazard near you.