Extreme Heat Is Spreading Beyond Traditional Heat Seasons: Why Monitoring Calendars Need to Change
Extreme heat is becoming more frequent.
But a new global study asks a different question:
Is extreme heat also arriving at different times of the year?
The answer appears to be yes.
Researchers examining 45 years of climate data found that extreme dry- and humid-heat seasons have expanded across approximately half of the world’s land area.
More importantly, that expansion is often uneven.
Some regions are experiencing a greater increase in unusually hot days before their historical heat season.
Others are experiencing more extreme heat after the season would traditionally have ended.
That matters because environmental and occupational preparedness is often organised around an expected calendar.
If extreme heat increasingly occurs outside that calendar, monitoring and risk-management programmes may also need to change.
What did the researchers investigate?
The study was conducted by Catherine C. Ivanovich, Benjamin Cook and Sonali McDermid of the NASA Goddard Institute for Space Studies and collaborating institutions.
It was published in AGU Advances on September 10, 2026.
Rather than studying only whether annual temperatures are increasing, the researchers examined when during the year extreme heat occurs.
They analysed global conditions from 1980 through 2024, covering the six inhabited continents.
Two types of heat were examined separately:
Dry heat, using daily maximum air temperature.
Humid heat, using wet-bulb globe temperature, or WBGT.
This distinction is important because high air temperature and high physiological heat stress are related but are not identical environmental conditions.
What does “extreme heat” mean in this study?
The word “extreme” requires careful interpretation.
The researchers did not use one universal temperature—such as 35°C or 40°C—to define an extreme event everywhere in the world.
Instead, they established a location-specific historical threshold.
For each grid cell, the researchers calculated the 95th percentile of daily maximum temperatures during their baseline period.
Days reaching or exceeding that threshold were classified as extreme dry-heat events.
They repeated the process using the 95th percentile of WBGT to identify extreme humid-heat events.
This means an extreme day is defined relative to what was historically unusual at that particular location.
That is scientifically useful for analysing climate change.
But it also means the results should not be interpreted as a global occupational or medical danger threshold.
A 95th-percentile event identifies an unusually hot condition.
Whether that condition constitutes an occupational heat-stress exceedance depends on the applicable exposure metric, work intensity, clothing, radiant heat, acclimatisation, applicable standard and other site-specific factors.
How did the researchers define a heat season?
The team began with the earliest decade available in the primary dataset: 1980–1989.
At each location, they identified the three consecutive months containing the greatest proportion of extreme heat events.
Those three months became the location’s historical extreme heat season.
The researchers then identified two-month periods immediately before and after it.
They called these the pre-season and post-season, or collectively the shoulder seasons.
This approach is important because the hottest period is not the same everywhere.
It also avoids assuming that “summer” has the same environmental meaning across the globe.
The method worked even across much of the tropics, where annual temperature variability is smaller than in temperate regions.
What changed over 45 years?
The researchers compared the historical baseline decade of 1980–1989 with the latest decade in the dataset, 2015–2024.
They found that extreme dry-heat seasons had expanded significantly across approximately 50% of global land area.
Extreme humid-heat seasons expanded across approximately 48%.
But the expansion was usually not symmetrical.
In several regions—including the western United States, eastern China, northern Africa and eastern Europe—extreme heat increased more strongly after the historical heat season.
In other regions—including western Europe, southern Africa and northwestern India—the greater increase occurred before the historical heat season.
The implication is that climate change is not simply making the traditional hottest months hotter.
In many places, it is altering when unusually hot conditions occur.
Dry heat and humid heat can behave differently
Another important result is that dry and humid heat do not necessarily shift in the same direction.
Phoenix, Arizona provides an instructive example.
During the historical baseline, virtually all of the city’s extreme dry- and humid-heat days occurred within their locally defined heat seasons.
By 2015–2024, however, the timing had changed.
The median date of Phoenix’s dry-heat extremes shifted approximately 10 days later, while the median date of humid-heat extremes shifted approximately 5.5 days earlier.
That difference illustrates why air temperature alone does not describe every dimension of heat exposure.
Humidity, solar radiation and wind conditions influence the heat load experienced by the human body.
What is WBGT?
Wet-bulb globe temperature is a heat-stress metric that incorporates several environmental factors rather than temperature alone.
The study calculated WBGT using air temperature together with factors representing humidity, radiant energy and wind conditions.
WBGT is widely used in occupational, military and sports heat-stress applications because it more closely represents the environmental heat load affecting human thermoregulation than air temperature alone.
But another distinction is essential.
The researchers defined an extreme WBGT event as the local historical 95th percentile.
That is not the same thing as saying a particular occupational WBGT exposure limit was exceeded.
The study’s WBGT metric is being used primarily to examine the changing seasonality of humid heat.
Occupational exposure assessment requires the appropriate workplace standard and field-monitoring protocol.
Why unexpected heat can create additional risk
Heat outside its expected season may create problems even when the absolute temperature is not unprecedented.
Early-season heat can arrive before people have acclimatised or before organisations have activated seasonal cooling and heat-management measures.
Late-season heat can extend cumulative exposure after workers and communities have already experienced months of elevated temperatures.
The Columbia Climate School also notes that cooling centres, warning systems and other heat-response measures are often operated according to expected seasonal calendars.
This creates an important adaptation question:
Should preparedness end because the calendar says the heat season is over, or because monitored environmental conditions show that the hazard has actually declined?
Rising average temperature does not explain everything
The researchers expected that a warmer climate might simply shift more days above historical extreme thresholds.
They tested that hypothesis.
At each location, they calculated the increase in annual average temperature between the historical and recent periods.
They then added that amount of warming to the historical temperature distribution to create a synthetic climate in which the existing seasonal pattern simply became warmer.
This reproduced some of the increase in extreme heat during the core heat season.
But it generally failed to reproduce the observed asymmetry between the pre- and post-seasons.
That suggests the timing changes involve more than uniform warming.
Regional processes such as atmospheric circulation, rainfall, soil moisture, land-surface conditions and natural variability may influence when extremes occur.
This is not a complete climate-attribution study
The distinction above is important.
The researchers did not conclude that every observed change in heat-season timing was caused directly by anthropogenic climate change.
Their experiment isolated whether annual-mean warming alone could explain the changes.
It often could not.
But anthropogenic climate change can affect far more than annual average temperature.
It can also influence atmospheric circulation, humidity, persistence, rainfall and land-atmosphere interactions.
The authors therefore state explicitly that their analysis does not directly attribute the observed seasonal changes to climate change.
Further climate-model experiments are needed to separate the effects of human forcing from internal climate variability and other processes.
That qualification should remain central to any responsible interpretation of the paper.
How reliable is the global pattern?
The researchers primarily used NASA’s MERRA-2 atmospheric reanalysis dataset.
Reanalysis combines observations with weather-model calculations to produce a continuous representation of atmospheric conditions through space and time.
Because any reanalysis dataset can contain biases, the researchers repeated their major analyses using the ERA5 reanalysis produced by the European Centre for Medium-Range Weather Forecasts.
They report that the principal conclusions were broadly robust across both datasets.
They also repeated several analyses using different time windows and extreme-event thresholds.
The overall patterns remained broadly consistent.
Rare events create statistical uncertainty
There is nevertheless an important limitation.
Extreme heat is rare by definition.
Extreme heat occurring outside the normal heat season is rarer still.
At many locations, the historical baseline therefore contained very few pre- or post-season events.
The researchers used bootstrapping and sensitivity analyses to address this limited sample size.
But the authors still describe the findings as an important first line of evidence, rather than the final word on changing global heat seasonality.
Climate-model simulations can provide many additional representations of the climate system and may help determine whether the observed patterns persist when larger samples of extreme events are available.
What does this mean for Trinidad and Tobago?
Trinidad and Tobago already has a locally defined heat season.
The Trinidad and Tobago Meteorological Service defines a hot day as one on which maximum temperature reaches or exceeds 34.0°C.
According to TTMS climatological information, the local heat season generally extends from March through October.
The greatest concentrations of excessive heat occur during two periods:
April through May, and August through October.
This is an important starting point for local heat-risk planning.
But the TTMS definition and the new global study use different methodologies.
The Columbia/NASA study defines a three-month historical extreme-heat season based on each grid cell’s 95th-percentile events.
TTMS uses its own local climatological definition and threshold.
The numerical findings should therefore not be combined as though they describe exactly the same metric.
Does the study show Trinidad and Tobago’s heat season has expanded?
Not from the evidence reviewed here.
The global study includes tropical regions and demonstrates widespread changes in extreme-heat seasonality.
But the paper does not provide a Trinidad and Tobago-specific trend estimate in the main results.
It would therefore be inappropriate to claim that the country’s March–October heat season has already expanded by a particular number of days or months based on this study alone.
That question requires analysis of Trinidad and Tobago observations or appropriately resolved climate data.
The international study instead identifies a monitoring question for the Caribbean:
Is the frequency of locally extreme heat changing before or after the historically recognised heat season?
That is a testable question.
Heat exposure in Trinidad and Tobago is already a management issue
The Ministry of Health advises residents to take heat precautions during the dry season, including avoiding strenuous activity during the hottest part of the day, maintaining hydration and taking frequent breaks during outdoor work.
Heat, however, is not confined to the dry season.
TTMS climatology places one of the country’s main excessive-heat peaks during August through October, well within the wet season.
That is a useful reminder that rainfall season and heat season are not the same environmental concept.
Organisations therefore need to base heat-risk decisions on environmental conditions rather than assuming that rain automatically means lower heat exposure.
Why this matters for workplaces
Outdoor and industrial workers may experience heat generated from several sources simultaneously.
These can include:
- High ambient temperature
- High humidity
- Direct solar radiation
- Heat-producing machinery or processes
- Limited air movement
- Protective clothing
- Physical workload
The relevant risk is therefore not always captured by the outdoor air temperature alone.
A workplace can also have a different thermal environment from the nearest meteorological station.
Industrial processes, roofing materials, enclosed spaces, ventilation, paved surfaces and equipment can all alter worker exposure.
This is why site-specific measurement matters.
A calendar is not a monitoring instrument
Many organisational controls are seasonal.
Equipment is serviced before an expected hot period.
Hydration programmes are intensified.
Work-rest procedures may be adjusted.
Supervisors become more vigilant.
Heat awareness campaigns are launched.
Those practices remain valuable.
But the new research suggests that the environmental trigger for heat controls should not necessarily be a fixed calendar date.
If unusual heat occurs earlier or persists later, a system that automatically stands down because “the season is over” may miss genuine exposure.
Monitoring provides an alternative.
Controls can be linked to actual conditions.
What should heat monitoring measure?
The correct monitoring method depends on the question being asked.
For broad climate monitoring, air temperature and humidity may provide valuable information.
For occupational heat-stress assessment, additional variables may be required depending on the applicable method or standard.
These can include radiant heat, air movement, humidity, temperature, worker metabolic workload, clothing and exposure duration.
The important principle is that the measurement method should match the decision being made.
A weather forecast is not automatically an occupational exposure assessment.
Likewise, indoor thermal comfort measurements are not necessarily interchangeable with a formal occupational heat-stress evaluation.
Where Ecotox can contribute
Ecotox Environmental Services identifies occupational monitoring as one of its established service areas.
Its current service information specifically includes heat-stress and thermal-comfort assessment and identifies measurable conditions including:
- Temperature
- Humidity
- Air velocity
- Ventilation
- Noise
- Vibration
- Lighting
Ecotox can therefore support organisations seeking to understand workplace environmental conditions and develop evidence-based occupational monitoring programmes.
However, the current verified Ecotox web material does not explicitly state that WBGT measurement is offered.
This distinction should be preserved.
If a project specifically requires WBGT-based occupational heat-stress assessment, the appropriate instrumentation, measurement method and applicable exposure standard should be confirmed before that capability is advertised or proposed.
From seasonal monitoring to condition-based monitoring
The new research suggests a useful change in thinking.
Traditional approach:
“Heat controls operate during these months.”
More resilient approach:
“Heat controls operate whenever measured and forecast conditions indicate relevant exposure.”
That shift becomes increasingly valuable when historical seasonality is changing.
It also allows organisations to retain seasonal planning while adding environmental triggers capable of responding to unusual early- or late-season events.
Heat may increasingly overlap with other hazards
Changing timing also matters because environmental hazards do not occur independently.
The study specifically identifies potential overlap between extended heat seasons and wildfire or hurricane seasons.
This is particularly relevant to climate-risk planning.
Extreme heat occurring alongside another emergency can increase electricity demand, complicate evacuation, affect outdoor workers, reduce recovery capacity and place additional pressure on infrastructure.
The research does not quantify those compound risks for Trinidad and Tobago.
But the monitoring principle is highly relevant to Caribbean disaster planning:
Risk should be evaluated as a system rather than one hazard at a time.
Long-term environmental data become more valuable as seasons shift
One of the broader lessons from this study is the value of historical monitoring.
Researchers could identify changing seasonality because they had decades of comparable climate information.
The same principle applies at the facility and project level.
A few days of environmental measurements can establish conditions during an assessment.
Repeated measurements can reveal seasonal patterns.
Multi-year monitoring can begin to show whether those patterns themselves are changing.
That difference becomes increasingly important under climate change.
What the research does not establish
The study does not provide a universal temperature at which conditions become dangerous.
It does not establish an occupational exposure limit.
It does not show that every 95th-percentile WBGT event constitutes unsafe working conditions.
It does not calculate a Trinidad and Tobago-specific expansion of the heat season.
It does not show that annual-average warming alone explains changing heat seasonality.
And it does not formally quantify what proportion of the observed seasonal changes is attributable to anthropogenic climate change.
Those limitations are not weaknesses to hide.
They define what the study can legitimately tell us.
Conclusion
The new global analysis changes the way we should think about extreme heat.
The issue is no longer only:
How hot are the hottest days becoming?
We also need to ask:
When are those days occurring?
Across 45 years of climate observations, extreme dry- and humid-heat seasons expanded significantly across approximately half of the world’s land area.
In many places, the expansion was asymmetric, moving more strongly into the months before or after the traditional heat season.
Average warming alone could not fully explain those patterns.
For Trinidad and Tobago, this study does not establish that the national heat season has expanded.
But it creates an important local monitoring question.
TTMS already recognises a lengthy heat season from March through October.
If locally extreme heat begins occurring more frequently outside expected periods, occupational monitoring, public-health preparedness and climate-risk management may also need to remain active for longer.
The practical lesson is straightforward:
Heat risk should ultimately be managed according to measured conditions—not simply according to the calendar.
Linked Sources
Ivanovich, C.C., Cook, B. & McDermid, S. (2026). Extreme Dry and Humid Heat Seasons Are Changing Asymmetrically. AGU Advances, 7(6), e2026AV002516. DOI: 10.1029/2026AV002516. Primary peer-reviewed study
NASA Goddard Institute for Space Studies — publication record. NASA GISS study record
Columbia Climate School — Dangerous Hot Days Are Spreading Beyond Summer. Columbia Climate School article
Trinidad and Tobago Meteorological Service — Climate Averages and Extremes. TTMS heat-season information
Trinidad and Tobago Ministry of Health — Dry Season Health Guidance. Ministry of Health guidance
Ecotox Environmental Services — Occupational Monitoring Services. Ecotox occupational monitoring