The Most Extreme Temperature Records

The highest air temperature ever officially verified on Earth is 56.7°C (134.1°F), recorded at Furnace Creek in Death Valley, California, on July 10, 1913. The lowest is −89.2°C (−128.6°F), recorded at Russia's Vostok Station in Antarctica on July 21, 1983 (World Meteorological Organization). Between those two extremes sits a 145.9°C (263°F) spread that reveals how elevation, ocean currents, and basin geometry shape the planet's thermal limits, and why some famous "records" have since been thrown out.

The official record-keeper: how the WMO verifies a temperature extreme

Weather records do not become official just because a thermometer displayed a number. The World Meteorological Organization (WMO), a United Nations agency headquartered in Geneva that coordinates meteorological standards across 193 member states, maintains an Archive of Weather and Climate Extremes and runs formal evaluation committees whenever a claimed record is disputed. A panel of independent climatologists reviews the instrument type, the observer's training, the site's surroundings, and how the reading compares to nearby stations before a number is certified or struck down (WMO Bulletin, 2013).

This matters because raw thermometer readings are noisy. A weather station sitting too close to asphalt, positioned in direct sun without proper shielding, or read by an inexperienced volunteer can produce a number that looks record-breaking but reflects a measurement error rather than a real atmospheric extreme. The WMO's evaluation process exists precisely to separate genuine climate extremes from instrument artifacts, and its history includes at least one high-profile reversal that reshaped the official record books.

For 90 years, the WMO's official hottest-temperature record was 58°C (136.4°F), measured at El Azizia, Libya, on September 13, 1922. In 2012, a WMO investigative committee formally disallowed it after finding the observer likely misread an unfamiliar replacement thermometer by about 7°C (Bulletin of the American Meteorological Society, 2013). The Death Valley 1913 reading became the official record by default.

What went wrong at El Azizia in 1922

A WMO Commission for Climatology panel spent roughly two years re-examining the El Azizia case and identified five specific problems: a questionable instrument (a Bellani-Six thermometer known to be easy to misread), an inexperienced observer, an unrepresentative microclimate at the recording site, poor agreement with temperatures at nearby stations, and poor agreement with the region's subsequent 67 years of temperature data (Bulletin of the American Meteorological Society, 2013). The committee concluded the observer most likely read the wrong end of the thermometer's sliding index pin, inflating the true reading by around 7°C. The panel voted unanimously in January 2012 to strike the record, and the WMO announced the decision on September 13, 2012, exactly 90 years after the original reading.

The episode is a useful case study in how a single point of failure, one untrained volunteer with an unfamiliar instrument, can stand as an "official" world record for nearly a century before anyone checks the underlying data. It is also why the 1913 Death Valley figure, despite carrying its own controversy (discussed below), remains the WMO's certified standard rather than a newer satellite-derived estimate.

Why Death Valley gets so hot

Furnace Creek sits at 58 meters (190 feet) below sea level inside Death Valley, a narrow basin flanked by steep mountain ranges in eastern California. That geometry is not incidental to the heat record; it is the primary mechanism behind it. Air sinking into a below-sea-level basin compresses as atmospheric pressure increases with depth, and compressed air heats up through a process called adiabatic compression, the same physical principle that makes a bicycle pump warm to the touch when you use it.

The surrounding mountain walls compound the effect. Hot air rises off the valley floor during the day, drifts toward the nearby slopes, cools slightly, and sinks back down, recompressing and rewarming as it descends. That cycling keeps a layer of superheated air trapped close to the ground rather than letting it escape upward and disperse, a bit like a convection oven that keeps recirculating hot air around whatever is inside instead of venting it. Combine that trapped-air effect with a desert climate that gets under 60 millimeters (2.4 inches) of rain a year, minimal cloud cover, and dark, sun-absorbing rock and gravel on the valley floor, and the basin becomes one of the most efficient heat traps on the planet.

Death Valley's record is not a one-off fluke, either. The National Park Service and WMO both note the valley has produced many of the hottest reliably instrumented summer readings on Earth in the century since 1913, including a widely reported 54.4°C (130°F) reading in August 2020 that a WMO evaluation later treated as plausible given modern, well-maintained instrumentation (WMO Weather and Climate Extremes Archive). The 1913 reading's basin mechanism, though, is exactly why researchers keep revisiting whether the original number itself was measured correctly.

The unresolved doubt about the 1913 reading

Unlike El Azizia, the WMO has not opened a formal investigation into the 1913 Death Valley record, so it remains officially certified. But independent climate researchers have flagged reasons for skepticism. A 2025 statistical analysis published in Oxford Open Climate Change compared the 1913 reading against more than a century of subsequent July temperatures at the same location and estimated the true air temperature that day was closer to 48.9°C (120°F), roughly 7.8°C (14°F) lower than the certified figure. The researchers suggested the original instrument may have been an uncalibrated thermometer read from a ranch house veranda rather than a properly sited, ventilated weather shelter (Oxford Open Climate Change, 2025).

Because a WMO evaluation committee has never been formally convened for the 1913 case, the 56.7°C figure remains the standing record. The point for readers is not that the number is definitely wrong, but that "official record" and "verified beyond doubt" are not automatically the same thing. Extreme readings from more than a century ago, before standardized instrument shelters and trained observer networks existed, carry inherent uncertainty that modern satellite and automated station networks have mostly eliminated.

Why Vostok Station is the coldest place ever measured

Vostok Station sits at 3,488 meters (11,444 feet) of elevation on the East Antarctic Plateau, one of the highest points of Antarctica's interior ice sheet. Three factors combine there to produce the coldest air temperature ever recorded by a surface thermometer. First, elevation: air temperature drops roughly 6.5°C for every 1,000 meters of altitude gained under standard atmospheric conditions, so a station nearly 3.5 kilometers (2.2 miles) up starts from a significant thermal disadvantage before any other factor is considered.

Second, polar night: for months each winter, the sun never rises at Vostok's latitude, so the ice sheet has no incoming solar radiation to offset the heat it constantly radiates away into space. Third, and specific to the 1983 record itself, a British Antarctic Survey reanalysis identified an unusually stable atmospheric setup: a slow-moving upper-level cold air mass parked almost directly over the station for about a week, skies stayed clear of cloud and "diamond dust" ice crystals that would otherwise trap outgoing heat, and surface winds stayed calm enough that no warmer air got mixed in from lower latitudes (Turner et al., Journal of Geophysical Research: Atmospheres, 2009).

That combination let the ice sheet radiate heat into space almost unimpeded for days, the same physical principle that makes a clear desert night colder than a cloudy one, but taken to a polar extreme. The result was −89.2°C (−128.6°F) on July 21, 1983, a reading the WMO still recognizes as the official surface air temperature minimum today.

Has anything colder ever been detected?

Yes, but it does not count as the same kind of record. NASA satellite instruments have measured skin-surface temperatures over the East Antarctic Plateau dipping to roughly −93.2°C (−135.8°F) in pockets between mountain ridges during clear, calm nights in 2010 and 2013 (Scambos et al., Geophysical Research Letters, 2018). The WMO does not certify these as the official "lowest temperature" record because satellites measure the radiative temperature of the snow surface itself, not the standard 2-meter (6.6-foot) air temperature that ground-based thermometers record. The two measurement types are not directly comparable, which is why Vostok's ground-station reading remains the official benchmark despite colder numbers existing in the satellite record.

The coldest and hottest places people actually live

Extreme-temperature records at research stations are impressive, but almost nobody lives at Vostok or in the depths of Death Valley year-round. A separate, more human category tracks where communities of people have adapted to survive at the thermal edges of habitability, and the geography behind those places is worth its own look at the highest permanent human settlements, where thin, cold air rather than heat is the survival challenge.

Oymyakon, a village of around 500 people in Russia's Sakha Republic, holds the record for coldest continuously inhabited settlement. Guinness World Records lists a reading of −67.7°C (−90°F) taken there on February 6, 1933, and some regional sources cite an even colder 1924 measurement of −71.2°C (−96.2°F), though that figure has less rigorous documentation. Average winter temperatures in Oymyakon run around −50°C (−58°F), cold enough that residents keep car engines running continuously through the winter and bury bodies only after thawing the permafrost with fire, because the ground itself is frozen solid enough to resist a shovel most of the year.

The counterpart on the heat side is less a single settlement than a cluster of Gulf and Sahara cities that regularly post the hottest reliably measured air temperatures where people actually live day to day, including cities near the largest deserts on Earth. Kuwait City and several Iranian and Iraqi cities have posted verified summer readings above 53°C (127°F) in the past decade, and heat index values, which combine temperature with humidity to estimate what the body actually experiences, have exceeded 60°C (140°F) in Gulf coastal cities during peak humidity events (National Weather Service Heat Index methodology). Humid heat is more physiologically dangerous than dry desert heat at the same air temperature, because high humidity blocks the evaporative cooling that sweat depends on.

The hottest ocean surface and the coldest upper atmosphere

Sea surface temperature records work differently from land air temperature records because oceans buffer heat far more slowly than land or ice. The Persian Gulf regularly produces the warmest large-scale ocean surface temperatures on the planet, with readings exceeding 34°C (93°F) in shallow coastal areas during peak summer, driven by the Gulf's shallow average depth of about 50 meters (164 feet), which lets the sun heat the entire water column rather than just a thin surface layer, combined with limited water exchange through the narrow Strait of Hormuz (NOAA Coral Reef Watch data).

At the opposite extreme, the coldest naturally occurring temperatures on Earth are not found at the surface at all but roughly 85 kilometers (53 miles) up, in a layer of the atmosphere called the mesopause, the boundary between the mesosphere and thermosphere. Temperatures there can fall to around −143°C (−225°F) over the poles during summer, when a specific atmospheric circulation pattern pushes air upward and it cools through expansion (NASA Goddard Space Flight Center atmospheric research). These readings do not appear on most "coldest place on Earth" lists because they describe the atmosphere itself rather than a surface location, but they are colder than anywhere on the ground, including the interior of the oldest ice ever found on Earth, which preserves a frozen record of past climate rather than an active temperature extreme.

Record	Location	Temperature	Date	Source
Highest official air temperature	Furnace Creek, Death Valley, USA	56.7°C (134.1°F)	Jul 10, 1913	WMO
Lowest official air temperature	Vostok Station, Antarctica	−89.2°C (−128.6°F)	Jul 21, 1983	WMO
Formerly-official highest (invalidated)	El Azizia, Libya	58°C (136.4°F)	Sep 13, 1922	WMO (disallowed 2012)
Coldest satellite-detected surface	East Antarctic Plateau	~−93.2°C (−135.8°F)	2010 / 2013	NASA / Scambos et al., 2018
Coldest inhabited settlement	Oymyakon, Russia	−67.7°C (−90°F)	Feb 6, 1933	Guinness World Records
Warmest large ocean surface	Persian Gulf	~34°C (93°F)	Recurring, summer	NOAA Coral Reef Watch
Coldest atmospheric layer	Mesopause (~85 km altitude)	~−143°C (−225°F)	Recurring, polar summer	NASA Goddard

Every entry above traces back to a named institution's published data; none are estimates without a documented source.

What people get wrong about "hottest" and "coldest" claims

The most common mistake is treating any viral headline number as automatically WMO-certified. Local news outlets and social media routinely report un-vetted station readings, including several widely shared claims of 60°C-plus air temperatures from Middle Eastern and South Asian cities that were never submitted for, let alone passed, WMO evaluation. A genuinely record-setting reading takes months to years to certify, because the process requires cross-checking instrument calibration records, site photographs, and comparison data from surrounding stations, not just a single number appearing on a display.

A second frequent error is conflating air temperature with heat index or surface temperature. Air temperature is measured in a ventilated shelter roughly 2 meters (6.6 feet) above the ground, shielded from direct sunlight, which is the WMO standard for comparability across stations worldwide. Heat index, sometimes called "feels like" temperature, factors in humidity's effect on the body's ability to cool itself through sweat evaporation and can run 10 to 15°C (18 to 27°F) above the actual air temperature during humid heat waves. Surface temperature, meanwhile, measures the ground, pavement, or ice itself, which can run far hotter or colder than the air just above it; asphalt in direct summer sun can exceed 60°C (140°F) even when the air temperature is a comparatively mild 35°C (95°F). Confusing these three categories is how many exaggerated "hottest place on Earth" claims spread.

A third confusion involves the difference between a record and a trend. A single extreme reading, however well documented, describes one moment at one location. Whether Earth's overall climate is warming is a separate question answered by long-term global average temperature datasets, not by any individual record-breaking day. NASA and NOAA's independent global temperature analyses, which track decades of data across thousands of stations, are the appropriate tools for questions about long-term climate trends, while WMO's extremes archive answers a narrower question: what is the single most extreme reading ever reliably measured at one place and time.

The takeaway

The planet's certified temperature extremes, 56.7°C at Death Valley and −89.2°C at Vostok Station, are not arbitrary trivia numbers; each traces back to a specific, explainable physical mechanism, a below-sea-level heat trap in one case and a high-altitude polar radiative deep-freeze in the other. But both records also carry documented uncertainty: one figure was formally invalidated after 90 years, and the other has never been through the same rigorous re-examination despite comparable red flags from modern instrumentation standards. The lesson worth keeping is less about the specific numbers and more about the process: a "world record" temperature is only as trustworthy as the committee, instrument, and observer behind it, which is exactly why the WMO's evaluation archive, not a viral headline, remains the only reliable source to check.

---

Read next:

• The Coldest Laboratory Temperatures Ever Reached
• The Highest Permanent Human Settlements
• The Largest Deserts on Earth
• The Oldest Ice Ever Found on Earth

FAQ

Q: What is the hottest temperature ever officially recorded on Earth? The World Meteorological Organization recognizes 56.7°C (134.1°F), recorded at Furnace Creek in Death Valley, California, on July 10, 1913, as the official highest air temperature ever measured, though independent researchers have questioned the reading's accuracy given the instrumentation available at the time.

Q: What is the coldest temperature ever officially recorded on Earth? The official record is −89.2°C (−128.6°F), measured at Vostok Station on the Antarctic Plateau on July 21, 1983. Satellite instruments have since detected colder surface temperatures nearby, but those measure a different quantity and are not certified as the same type of record.

Q: Why was the 1922 Libya temperature record removed? A World Meteorological Organization investigative committee concluded in 2012 that the observer at El Azizia likely misread an unfamiliar thermometer by about 7°C due to inexperience and a faulty instrument, and voted unanimously to disallow the 58°C reading after a two-year review.

Q: What is the coldest permanently inhabited place on Earth? Oymyakon, a village of roughly 500 people in Russia's Sakha Republic, holds the Guinness World Records recognition for coldest continuously inhabited settlement, with a documented low of −67.7°C (−90°F) recorded on February 6, 1933.

Q: Why does Death Valley get hotter than other deserts? Its basin sits 58 meters (190 feet) below sea level and is enclosed by steep mountains, which traps descending, compressing air near the ground through adiabatic heating and prevents it from venting upward, the same effect that makes a bicycle pump warm during use.

Q: How does the WMO decide if a temperature record is official? An independent panel of climatologists reviews the instrument type, the observer's training, the observation site's surroundings, and how the reading compares to nearby stations and historical data before certifying or rejecting a claimed extreme, a process that can take months to years.

Q: What's the actual difference between an "air temperature" record and a "surface temperature" record? Air temperature is measured inside a ventilated instrument shelter positioned roughly 2 meters (6.6 feet) above the ground, deliberately shielded from direct sunlight and from the ground's own radiant heat, which is why it's the WMO's standardized measurement for comparing extremes across different locations and eras. Surface temperature, by contrast, measures the actual material at ground level, whether that's pavement, bare soil, sand, or ice, using either a contact sensor or, more often today, a satellite-based infrared radiometer reading the thermal radiation the surface emits.

The two numbers can diverge dramatically because bare ground absorbs and re-radiates solar energy far more efficiently than the open air a couple meters above it. NASA's MODIS satellite instrument has recorded land surface temperatures in Iran's Lut Desert as high as 80.8°C (177.4°F) in 2018, well beyond any certified air temperature record anywhere on the planet, because the WMO does not treat satellite-derived land surface temperature as a valid input for the official air temperature archive at all (NASA Earth Observatory, MODIS satellite observations). The same divergence explains why an Antarctic satellite reading of roughly −93.2°C, colder than Vostok's official −89.2°C ground-station mark, still hasn't replaced it as the record: one number describes the snow surface itself, the other describes the standardized air layer just above it, and the WMO's archive only certifies the latter.