Update on World Heat Record (58°C/136.48F) at Al Azizia, Libya, Sept. 13, 1922

In my blog posted on Oct. 8th, I raised questions concerning the hottest officially measured temperature in the world: the 58°C (136.4°F) reading from Al Azizia, Libya on September 13, 1922. Since that post, new information has come to my attention:

1) The director of the Libyan National Meteorological Center (LNMC), Khalid Ibrahim El Fadli, has confirmed that they officially do not accept this temperature reading.

2) Randy Cerveny, Rapporteur for Extreme Records of the WMO (World Meteorological Organization) Commission for Climatology, has advised that perhaps this record should be reevaluated in light of the Libyan rejection.

In the past month, the LNMC has also sent data for the POR (period of record) for Azizia from beginning of record in 1919 to 1984 when the station was retired. Earlier data from the site beginning in 1913 has been confirmed to have been set up by an Italian physicist, F. Eredia, and NOT by the National Geographic Society, as has been widely reported in the past (thanks to Ms. Julie Crain at NGS for investigating and clarifying this issue once and for all).
Below is a chronology of the weather observations at Al Azizia.

Important Dates in The Chronology of Al Azizia Weather Station
1913 (April): Italian Prof. F. Eredia establishes weather station at Castello Benito (a small fort on a farm located on a 46-meter/150-foot-high hill in Al Azizia, Libya), approximately 55 kilometers south-southwest of Tripoli. Local Carabinieri (police) employed to make observations. Station ceases observations in June 1915.


Figure 1. This cross-section is a schematic of the location of the weather station at Al Azizia in 1922 (on a small hill) on the Gefara Plain and near the Jabal Nafusah Mountains. (From ‘La Piu Alta Tempura del Mondo’ p. 55, by Amilcare Fantoli).

1919 (July): Italian Military takes over site and begins weather observations again. Weather station under the aegis of Amilcare Fanatoli, Chief of the Italian Meteorological Service responsible for Libya.

1922: When Chief Fantoli visits site following report of 58°C temperature reading, he finds that the usual thermometer used by the weather station had failed at some previous date (unspecified), and been replaced with a Six-Bellani thermometer (see my previous blog for explanation of this type of thermometer).

1927: Station moved off the hill and placed in civilian hands in the town of Al Azizia (unknown change in instrumentation).

1942 (December): Advent of World War II closes station.

1947 (April): Station re-established at same location. Libyan National Meteorological Center takes over operation of weather station when Libya becomes independent in 1951.

1970 (May): Station ceases observations for reasons unknown, but perhaps because of the change in government that occurred around this time.

1974 (February): Station begins reporting again.

1984 (December): Last month of observations from site.

Temperature Data by Date Sets
The new data sent to me by Director El Fadli at LNMC can be broken into five distinct periods (these are the years that complete data sets are available. September analysis:

1913-1915 (1 year): 1914 is the only complete year of data. Absolute maximum temperature for that year was 48.7°C (119.7°F) and for that September 40.9°C (105.6°F).

1920-1926 (7 years): Three months of data missing (no Septembers missing). September average absolute maximum temperature is 48.3° (118.9°F) with an absolute maximum of 58° (136.4°F) in September 1922.

1927-1942 (16 years): Two months missing (no Septembers missing). September average absolute maximum temperature is 42.2° (108.0°F) with an absolute maximum of 45.9° (114.6°) in September 1938.

1948-1968 (21 years): Three months missing (no Septembers missing). September average absolute maximum temperature is 40.3°C (104.5°F) with an absolute maximum of 44.5°C (112.1°) in September 1959.

1974-1984 (11 years): Two months missing (no Septembers missing). September average absolute maximum temperature is 41.5°C (106.7°F) with an absolute maximum of 44.0° (111.2°F) in September 1975.

For all months of the year (not just September), 50°C (122°F) or warmer has been recorded only during these months:

1919-1926: (9 months) July 1920 (50.4°C), Aug. 1921 (53.0°C), July 1922 (51.0°C), Aug. 1922 (52.0°C), Sept. 1922 (58.0°C), Aug. 1923 (56.0°C), Sept. 1923 (54.8°C), June 1926 (51.2°C), Aug. 1926 (53.4°C).

1927-1942: (2 months) June 1928 (51.9°C), Aug. 1941 (51.0°C).

1948-1968: (no temperature over 50°C recorded). Absolute maximum was 47.0°C (116.6°F) on three occasions (July 1954, July 1955, July 1965).

1974-1984: (no temperature over 50°C recorded). Absolute maximum was 48.0°C (118.4°F) in Aug. 1984.

During the September 1922 heat wave the temperature at Al Azizia reportedly exceeded 50°C for five consecutive days Sept. 11-15:

Sept. 11: 50.0°C (122.0°F)
Sept. 12: 56.0°C (132.8°F)
Sept. 13: 58.0°C (136.4°F)
Sept. 14: 53.0°C (127.4°F)
Sept. 15: 53.0°C (127.4°F)

This is an amazing feat considering no temperature over 50°C was ever measured in the 32 years of records from 1948-1984.

Diurnal fluctuations for September Periods of Record
I won’t bore you with the details of the exact data on this but what is interesting to note is that even during the 1920-1926 period the average minimum temperatures during September were not abnormally warm compared to the average maximums or compared to the average minimums in the other periods of record. So the diurnal spread (average difference between daily minimums and daily maximums) was significantly greater during the 1920-1926 period relative to all the other periods of record as this graph illustrates:


Figure 2. Average daily range of temperature (in C°) for the month of September at Al Azizia between 1919-1940. (Graph courtesy of Piotr Djakow).

Philip Eden, an eminent British weather historian, noted in an email to me recently:

”The tarred concrete (agreed, presumably near-black) would probably explain nearly all of the wide excursions during the 1922-27 period…Bear in mind also that the earliest Stevenson screens in use in the UK had no bottom, so were open to reflected radiation from the ground. These were gradually replaced here in the 1890s and 1900s, but it is conceivable that old-pattern screens were in use elsewhere beyond this time.”

Mr. Amilcare Fantoli (the Chief of the Meteorological Service in Libya in 1922 as discussed above), wrote a paper on the subject of the Al Azizia record in an Italian meteorological journal ( Rivista di Meteorologia Aeronautica, pp. 54-63 in 1958, and wrote this summary of his article La Piu Alta Temperatura del Mondo:
“The highest temperature recorded in a meteorological station next to that of Death Valley in California is the one recorded at El Azizia in Tripolitania. It should, however, be pointed out that this figure as well as the other one is the subject of doubts based particularly on the influence of the place where the meteorological screen was set up, and on the type of weather instruments used in determining the temperature. After analyzing the circumstances concomitant to the peak figure of 58°C recorded, these having occurred during a spell of ghibili (SE foehn-like wind) of an unprecedented duration and intensity, the causes are expounded [sic], which may have brought about this exceedingly high value. It is therefore proposed to reduce the recorded value to 56°C, waiting for an eventual double of the phenomenon to be ascertained.”

Conclusion
What Mr. Fantoli did not do in his article was address the disconnect between the many extreme values (not just the one 58°C reading) that were measured between 1922-1926 and those temperatures recorded thereafter in the period of record. Of course, the “eventual double of the phenomenon” has not yet occurred. Also, perhaps the most suspicious aspect of the data is the extreme diurnal excursions during the 1920-1926 period which could certainly be explained as a result of bad instrument exposure.

However, there is one question difficult to understand. If over-exposure was the problem during the 1920-1926 period of record, then why were there actually several summer months during this time that did not show excessive heat anomalies? The possible answer to this is that the Six-Bellani thermometer was badly calibrated and the other thermometer OK but occasionally failed and so was replaced from time to time by the faulty Six-Bellani. When the station was moved in 1927 a new thermometer may have been acquired and the anomalous readings ceased and, one assumes, the shelter was no longer situated on a black-tarred surface.

For weather historians, the world’s hottest measured temperature is something like the Mt. Everest of weather records. Although it matters little in the grand scale of things, it is something that sets a meteorological standard that, in the geophysical world, has some relevance as a point of reference. I hope that the reliability of this record will eventually be scrutinized and its validity is once and for all determined.

References:
Fantoli, Amilcare La Piu Alta Temperatura del Mondo, Rivista di Meteorologia Aeronautica, pp. 54-63, 1958.

Pedemonte, Roberto, Italian Meteorological Society letter to Weather magazine Vol. 56, Oct. 2001, p. 378-379.

KUDOS: Thanks to Khalid Ibrahim El Fadli at Libyan National Meteorological Center (LNMC), Julie Crain at National Geographic Society, Manola Brunet Visiting Fellow at the Climatic Research Unit, School of Environmental Sciences, University of East Anglia, UK and her colleagues in Italy for supplying the Fantoli document and other data from their files, Philip Eden, Randy Cerveny, Thomas C. Peterson, and Federico Noris for his gracious translation work.

Christopher C. Burt