It is 24.8°C (76.6°F) at 02:30 local on Wednesday 4 November.
Commentary: The grill has been uncovered.
Our synoptic decoder reads air temperature only, so this station reports no dew point, wind or pressure here.
Between its coldest night and its hottest day, this place has swung 18.4°C.
At 14:30 it was 29.8°C. By 02:30 it was 24.8°C. The heat did not hold.
New facts every day. Report an error.
3.3°C short of the record for 7 November
Today’s high 29.8°C against 33.1°C in 1997. A reading that beats it is flagged here automatically, once a second cycle confirms it.
The highest in available records since 1976. Source: GHCN-Daily, 1976 to 2026.
The lowest in available records since 1976. Source: GHCN-Daily, 1976 to 2026.
This station lies about 1,100 km south of the equator. It lies within the tropics.






























Land surface temperature for 2026-11-06, from NASA GIBS. This is the temperature of the ground seen from orbit, not the air, and it is never ranked against air readings. Cloud hides the ground, so gaps are unread rather than guessed.
View more on Google Maps9.8000°S, 139.0333°W
| Month | High | Low | Mean | Days | Readings |
|---|---|---|---|---|---|
| November 2026 | 29.8°C | 24.8°C | - | 2 | 4 |
| October 2026 | 33.1°C | 23.2°C | 28.1°C | 7 | 15 |
Means are of the readings we stored, weighted by how often this station reported, so they are not daily averages and not climate normals.
The nearest other station is 833.4 km away. Quality control compares a reading against stations within 150 km, so there is nothing close enough to cross-check this one against.
Atuona (WMO 91925) is a SYNOP weather station in French Polynesia, 51 m above sea level.
We have kept a daily record here for 9 of the 23 days since 16 October 2026, from 19 observations.
The same record at three widths, picking up where the last 24 hours above leaves off. Each period is summarised from the grain that suits it, so nothing here is one reading repeated thousands of times into the page.
Over the last week we have readings for 2 days here, running from 24.8°C to 29.8°C, a spread of 5.0°C.
Entered the hundred hottest reporting stations on Earth, reaching 29th of 8,794, on each of 3 days running.
Its high of 33.1°C was the highest of the 10 stations we hold in French Polynesia, on each of 3 days running.
These are not official records. They are the highest and lowest readings this site has stored for Atuona since October 2026, across 9 days we hold a reading for. Official climatological records are adjudicated by national meteorological services and by the WMO.
Highest 33.1°C, lowest 23.2°C, mean 27.9°C. 9 days recorded from 19 readings.
Readings on 9 of the last 30 days. Typically 2 readings a day.
| Day | High | At | Low | At |
|---|---|---|---|---|
| 2026-11-04 | 29.8°C | 00:00 UTC | 24.8°C | 12:00 UTC |
| 2026-11-03 | 27.2°C | 18:00 UTC | 27.2°C | 18:00 UTC |
| 2026-10-27 | 31.0°C | 00:00 UTC | 25.0°C | 12:00 UTC |
| 2026-10-26 | 32.0°C | 00:00 UTC |
We have readings for 9 of the 23 days since we started storing this station on 16 October 2026.
Its last day with a reading was Wed 4 Nov 2026. Its longest unbroken run is 4 days.
The warmest month on record here is December, when the highest reading has reached 37.6°C (99.7°F) on Monday, 3 December 1990.
NOAA GHCN-Daily, highest reading recorded in each month
The coldest month on record here is October, when the lowest reading has fallen to 19.2°C (66.6°F) on Sunday, 23 October 1988.
NOAA GHCN-Daily, lowest reading recorded in each month
37.6°C (99.7°F) on Monday, 3 December 1990. That is the highest in the daily record we hold for this station, which runs since 1976. It is not an officially certified record unless a meteorological authority has recognised it.
NOAA GHCN-Daily
19.2°C (66.6°F) on Sunday, 23 October 1988. That is the lowest in the daily record we hold, which runs since 1976.
NOAA GHCN-Daily
24.8°C (76.6°F), measured at 02:30 local on Wednesday 4 November.
Live reading from the station, after our quality checks
Readings here arrive through NOAA WMO SYNOP bulletins. SYNOP is the surface synoptic report that national weather services file to the WMO, typically at the main hours of 00, 06, 12 and 18 UTC and often three hourly between them. It covers places with no airport, which is why the coldest readings on Earth arrive this way.
| Figure | Source | Type |
|---|---|---|
| Latest reading, today so far | NOAA Aviation Weather Center (METAR) and NOAA NWS WMO bulletins (SYNOP) | Measured, station |
| All-time, monthly and on-this-day records | NOAA NCEI Global Historical Climatology Network daily (GHCN-Daily), 1976 to 2026 at this site and its predecessors, and our own observations, 9 days | Measured, station |
| Nearby stations and distances | Computed from station positions in our own database | Calculated |
Every reading on this page has passed the checks below. A reading that fails any of them is stored as refused and can never be ranked, shown as a record, or counted in a daily high or low.
No reading from this station has been refused in the last 14 days.
A reading with no usable latitude and longitude is discarded. It cannot be ranked, mapped, or compared with its neighbours, and a temperature with no place is not an answer to which place is hottest.
Some feeds fill an unknown position with a stand-in value such as 99.99 degrees or an elevation of 9999 m. Those are not coordinates. They are caught separately from a missing position because a station placed at a fake point has no neighbours, so the neighbour check below can never see anything wrong with it.
SYNOP arrives as files that are only re-read when they change. A file that has stopped being updated is set aside whole, before any reading inside it is considered.
Different networks get different windows, because they report at different rates: 2 hours for METAR, which transmits continuously, and 9 hours for SYNOP, which reports at 00, 06, 12 and 18 UTC. A single rule would have discarded the entire synoptic network, including every station in Antarctica.
Across 11 Novembers, 1979 to 2025, this station's November average is level. A fit through its own record, not a climate finding.
Warmest November here: 1987, averaging 31.9°C. Coldest: 1989, averaging 30.8°C.
Source: NOAA GHCN-Daily. The upper figure is the highest ever recorded here in that month and the lower one is the lowest, each with the day it happened. The current month is outlined.
| Station | Away | Height | High | Low | vs here |
|---|---|---|---|---|---|
| TakaroaFrench Polynesia | 833.4 km | none | none | ||
| HaoFrench Polynesia | 937.7 km | 28.2°C | 28.2°C | −1.6°C | |
| Hao/Tuamoto French Polynesia | 938.5 km | 28.0°C | 26.0°C | −1.8°C | |
| Tahiti Island/Faaa IntlFrench Polynesia | 1429 km | 28.0°C | 27.0°C | −1.8°C | |
| Tahiti-FaaaFrench Polynesia | 1429.5 km | none | none | ||
| RikiteaFrench Polynesia | 1537.8 km | 22.1°C | 22.1°C | −7.7°C |
The last 7 days we recorded: 7 days recorded, 16 readings. From the daily highs and lows we recorded.
The last 30 days we recorded: 9 days recorded, 19 readings. From the daily highs and lows we recorded.
The last 12 months: 9 days recorded, 19 readings. From our monthly totals.
| 27.9°C |
| 12:00 UTC |
| 2026-10-25 | 33.1°C | 18:00 UTC | 23.9°C | 12:00 UTC |
| 2026-10-19 | 31.9°C | 18:00 UTC | 31.9°C | 18:00 UTC |
| 2026-10-18 | 24.7°C | 06:00 UTC | 24.7°C | 06:00 UTC |
| 2026-10-17 | 31.9°C | 18:00 UTC | 31.9°C | 18:00 UTC |
| 2026-10-16 | 25.1°C | 06:00 UTC | 23.2°C | 12:00 UTC |
Observations stored by HottestAndColdest. A day we were not yet tracking is shown as an empty square, and says nothing about whether the station reported.
Anything above 56.8 °C or below -95 °C is refused. The ceiling sits just above the highest reading WMO accepts, so a genuine new world record would be refused here too: a record needs adjudicating, not a live tile on a home page. The floor is deliberately further from its record, because the Antarctic plateau is thinly instrumented and a real extreme there has far less around it to corroborate it.
A change of more than 15 °C from this station's previous reading within 60 minutes is treated as a failing sensor. The limit is set from measured data: real weather reaches about 12 °C in an hour at desert sunrise or under a thunderstorm outflow, and the faults found were 20 °C and more.
A reading more than 12 °C from the median of other stations within 150 km is refused. It needs at least 3 neighbours to run at all, so it protects dense networks and can say nothing about an isolated station, which is why the other checks above have to stand on their own.