The Docket | Emerging Science

Fire After Dark

The End of the Overnight

By Maya Jiménez— June 26, 2026

Feature image for article: The End of the Overnight

Firefighting shifts, evacuation plans, and the basic act of going to bed in a fire-prone community all rest on the same assumption: fire slows down at night. A 2026 study in Science Advances traces the erosion of that nighttime pause across fifty years, documenting a 36 percent rise in potential burning hours since 1975. On January 8, 2025, the Eaton Fire grew from 2,227 to 10,600 acres between 1 a.m. and sunrise. The clock that safety systems were built around is losing time. The systems haven't been rebuilt.

Fire After Dark

The End of the Overnight

By Maya Jiménez— June 26, 2026

Firefighting shifts, evacuation plans, and the basic act of going to bed in a fire-prone community all rest on the same assumption: fire slows down at night. A 2026 study in Science Advances traces the erosion of that nighttime pause across fifty years, documenting a 36 percent rise in potential burning hours since 1975. On January 8, 2025, the Eaton Fire grew from 2,227 to 10,600 acres between 1 a.m. and sunrise. The clock that safety systems were built around is losing time. The systems haven't been rebuilt.

Fire's Expanding Reach

Five peer-reviewed studies published this year describe a fire regime that has outgrown the systems designed to contain it. Crews plan around daylight. Evacuation orders assume fires quiet after dark. Air quality networks monitor one pollutant while fires produce others that harm through entirely different pathways. Death certificates have no field for smoke.

The research spans fire behavior and public health. What connects it is less dramatic than any single finding: the baseline assumptions in firefighting, air monitoring, and mortality tracking were set for a version of wildfire that is already gone. Below is what the science says, and what it means for the people living inside these numbers.

Fire's Expanding Reach

Five peer-reviewed studies published this year describe a fire regime that has outgrown the systems designed to contain it. Crews plan around daylight. Evacuation orders assume fires quiet after dark. Air quality networks monitor one pollutant while fires produce others that harm through entirely different pathways. Death certificates have no field for smoke.

The research spans fire behavior and public health. What connects it is less dramatic than any single finding: the baseline assumptions in firefighting, air monitoring, and mortality tracking were set for a version of wildfire that is already gone. Below is what the science says, and what it means for the people living inside these numbers.

Burning Hours

North American Fires Burn a Third Longer Now

Luo et al. in Science Advances found annual potential burning hours across North America rose 36% between 1975 and 2024, with a third of active fire days exceeding twelve hours. Fires now routinely outlast the crew rotations, evacuation windows, and overnight rest that suppression strategy was built around.

Nighttime Peaks

Fourteen Percent of Fire Days Peak After Dark

From the same Luo et al. study: 14% of active fire days reached peak burning intensity at night rather than afternoon. For communities and fire crews, the old reliable assumption that darkness brings a reprieve is becoming a planning liability for evacuations and deployment alike.

Smoke Deaths

Wildfire Smoke Linked to 24,000 U.S. Deaths Yearly

Zhang et al. in Science Advances estimated wildfire-smoke PM2.5 caused roughly 24,100 U.S. deaths per year from 2006 to 2020, identifying no safe exposure threshold. Smoke kills at concentrations most people walk through unbothered, and death certificates almost never record it as a cause.

Stroke Risk

Years of Low Smoke Build Silent Stroke Risk

Hao et al. in the European Heart Journal tracked roughly 25 million Medicare beneficiaries and found each 1 μg/m³ rise in three-year average wildfire-smoke PM2.5 corresponded to a 1.3% increase in stroke risk. The cardiovascular damage accumulates across seasons, invisible in any single summer's air quality readings.

Ozone Reversal

Wildfires Have Erased Four Years of Ozone Progress

Deng et al. in Science found wildfire emissions flipped U.S. ozone trends from −0.65 ppb/year improvement (2003–2015) to +0.13 ppb/year worsening (2015–2024), wiping out 3.9 years of clean-air gains. Ozone triggers different health effects than the PM2.5 most smoke coverage focuses on, and requires different monitoring.

The Research Base

A weakened diurnal weather constraint leads to longer burning hours in North America

Luo et al., Science Advances, 2026. Tracking roughly 9,000 large fires through hourly satellite observations and reconstructing weather patterns back to 1975, researchers measured how overnight cooling and humidity recovery have historically dampened fire activity. Those natural brakes are failing. Potential burning hours climbed 36% over five decades as the atmospheric conditions that once slowed fires after dark grew weaker and less reliable.

Who loses sleep over this, literally?

Fourteen percent of active fire days now peak at night, collapsing the overnight lull that firefighting crews and evacuation plans depend on.

When did fire season stop having edges?

Extreme burning days of twelve-plus hours surged 81 to 233% across fire-prone regions, with the steepest gains in spring and fall shoulder seasons.

Wildfire smoke PM2.5 and mortality rate in the contiguous United States

Zhang et al., Science Advances, 2026. Using causal inference methods across all 3,068 U.S. counties from 2006 to 2020, this study isolated wildfire-specific fine particulate matter from other pollution sources and linked chronic exposure to roughly 24,000 excess deaths annually. The dose-response curve never flattened. No concentration was low enough to register as safe.

What does "no safe threshold" actually mean?

The health damage from smoke does not level off at low concentrations, undermining any regulatory assumption of a harmless baseline.

Why don't death certificates reflect this?

Coroners record the disease that killed someone, not the years of smoke exposure that accelerated it. The toll stays invisible.

Long-term exposure to wildfire smoke particulate matter and incident stroke

Hao et al., European Heart Journal, 2026. Rather than measuring acute smoke events, this study tracked three-year rolling averages of wildfire PM2.5 exposure across the Medicare population, linking sustained low-level smoke inhalation to elevated stroke incidence. The finding opens a cardiovascular pathway that is distinct from the respiratory harms most smoke research has focused on.

Why does the three-year window matter?

It captures the cumulative burden of living through repeated smoke seasons, not just surviving a single bad week.

Whose risk profile changes?

Medicare enrollees face stroke risk from chronic smoke at concentrations well below what triggers air quality alerts or keeps people indoors.

Fires reverse progress toward ozone air quality standards in the United States

Deng et al., Science, 2026. Ozone is a different beast from particulate matter, formed when wildfire carbon monoxide drifts hundreds of miles and reacts with sunlight. Using deep learning to build a gapless one-kilometer ozone dataset, researchers showed that wildfire emissions have reversed a decades-long decline in U.S. surface ozone since 2015. The existing monitoring network was too sparse to catch it happening.

How far does the damage travel?

Wildfire-driven ozone exceeded federal standards for 148 million Americans in 2023, many of them nowhere near a fire.

Can regulators keep calling these "exceptional events"?

When wildfire ozone exceedances recur annually, the regulatory framework built around rare spikes stops making sense.

The Research Base

A weakened diurnal weather constraint leads to longer burning hours in North America

Luo et al., Science Advances, 2026. Tracking roughly 9,000 large fires through hourly satellite observations and reconstructing weather patterns back to 1975, researchers measured how overnight cooling and humidity recovery have historically dampened fire activity. Those natural brakes are failing. Potential burning hours climbed 36% over five decades as the atmospheric conditions that once slowed fires after dark grew weaker and less reliable.

Who loses sleep over this, literally?

Fourteen percent of active fire days now peak at night, collapsing the overnight lull that firefighting crews and evacuation plans depend on.

When did fire season stop having edges?

Extreme burning days of twelve-plus hours surged 81 to 233% across fire-prone regions, with the steepest gains in spring and fall shoulder seasons.

The Research Base

Wildfire smoke PM2.5 and mortality rate in the contiguous United States

Zhang et al., Science Advances, 2026. Using causal inference methods across all 3,068 U.S. counties from 2006 to 2020, this study isolated wildfire-specific fine particulate matter from other pollution sources and linked chronic exposure to roughly 24,000 excess deaths annually. The dose-response curve never flattened. No concentration was low enough to register as safe.

What does "no safe threshold" actually mean?

The health damage from smoke does not level off at low concentrations, undermining any regulatory assumption of a harmless baseline.

Why don't death certificates reflect this?

Coroners record the disease that killed someone, not the years of smoke exposure that accelerated it. The toll stays invisible.

The Research Base

Long-term exposure to wildfire smoke particulate matter and incident stroke

Hao et al., European Heart Journal, 2026. Rather than measuring acute smoke events, this study tracked three-year rolling averages of wildfire PM2.5 exposure across the Medicare population, linking sustained low-level smoke inhalation to elevated stroke incidence. The finding opens a cardiovascular pathway that is distinct from the respiratory harms most smoke research has focused on.

Why does the three-year window matter?

It captures the cumulative burden of living through repeated smoke seasons, not just surviving a single bad week.

Whose risk profile changes?

Medicare enrollees face stroke risk from chronic smoke at concentrations well below what triggers air quality alerts or keeps people indoors.

The Research Base

Fires reverse progress toward ozone air quality standards in the United States

Deng et al., Science, 2026. Ozone is a different beast from particulate matter, formed when wildfire carbon monoxide drifts hundreds of miles and reacts with sunlight. Using deep learning to build a gapless one-kilometer ozone dataset, researchers showed that wildfire emissions have reversed a decades-long decline in U.S. surface ozone since 2015. The existing monitoring network was too sparse to catch it happening.

How far does the damage travel?

Wildfire-driven ozone exceeded federal standards for 148 million Americans in 2023, many of them nowhere near a fire.

Can regulators keep calling these "exceptional events"?

When wildfire ozone exceedances recur annually, the regulatory framework built around rare spikes stops making sense.

Who Controls the Air Inside

Shelter in Whose Place

The advice during smoke events is always the same: stay indoors, close your windows. Good advice if you control the building and can afford a filter. Most people breathing wildfire smoke have neither.

Han et al. (2025) found over a billion people globally face substantial indoor wildfire-PM2.5 days each year. Air purifiers reduce concentrations. Meeting WHO guidelines would cost an estimated $69 billion annually. But in rental units, aging schools, and care facilities, the person breathing the air almost never decides what happens to it. A teacher can smell something wrong in the classroom. Whether the filters get changed is somebody else's call, in some other office, on some other budget line.

The engineering exists. The question of who pays, and who decides, doesn't show up in any filtration study.

Who Controls the Air Inside

Shelter in Whose Place

The advice during smoke events is always the same: stay indoors, close your windows. Good advice if you control the building and can afford a filter. Most people breathing wildfire smoke have neither.

Han et al. (2025) found over a billion people globally face substantial indoor wildfire-PM2.5 days each year. Air purifiers reduce concentrations. Meeting WHO guidelines would cost an estimated $69 billion annually. But in rental units, aging schools, and care facilities, the person breathing the air almost never decides what happens to it. A teacher can smell something wrong in the classroom. Whether the filters get changed is somebody else's call, in some other office, on some other budget line.

The engineering exists. The question of who pays, and who decides, doesn't show up in any filtration study.

TAKE NOTE

Closed windows aren't enough: Indoor PM2.5 stays at 55–60% of outdoor levels during smoke events with doors and windows shut. Older buildings leak more.

36,000 U.S. schools need HVAC updates nationally. Funding depends largely on local property taxes.

Evidence gap: Most filtration research covers portable air cleaners. Evidence on HVAC performance during smoke events is nearly nonexistent. Institutional buildings rely on HVAC.

No mandate: Wildfire PM2.5 isn't regulated like industrial emissions. ASHRAE's 2024 building smoke guideline is voluntary.

24,100 deaths per year in the U.S. linked to wildfire smoke PM2.5 (2006–2020 average). None listed as wildfire deaths on any death certificate.