Meteorological forecasts for the approaching weekend (January 26-27, 2013) show the presence of a very deep sea level pressure minimum in the North Atlantic ocean, at high latitudes, located between Iceland and UK. Without pretending to be exhaustive, we will mention the minima forecasted by some of the most used models: less than 940 hPa (on 27th at 00 UTC) by ACCESS, 939 by ECMWF (on 27th at 00 UTC) and GEM (26th at 18 UTC), 932 by JMA (on 27th at 12 UTC), 930 by NOGAPS (27th at 00 UTC), less than 930 hPa (26th at 12 UTC) by GFS, and less than 928 hPa (26th at 12 UTC) by UKMO.
The pressure values are quite low and resemble those measured near or inside tropical cyclones. However, these storms have a different genesis and structure and are classified in the category of extratropical cyclones. A question that one might ask is whether these values can be considered unusual, or even exceptional.
In this blog, I will try to give a brief survey of world barometric pressure records, with the help of this blog of Christofer C. Burt and, for the maps, the very useful database stored in wetterzentrale.de. Most of the information and scientific results, as well as one figure reported in this blog, came from the original papers published by S.D. Burt, listed in the references.
As one can easily imagine, excluding the small scale phenomena such as tornadoes, the lowest pressures observed on the Earth have occurred during tropical cyclones.
The following table reports the absolute minima observed for each region.
The second minimum in the list is questioned by the Australian Bureau of Meteorology, that places a minimum pressure for the storm at 905 hPa on the same date and location. For North Atlantic, the minimum value is 882 hPa: for sake of comparison, the lowest value observed during the track of hurricane Sandy in October 2012 was 940 hPa, and that of hurricane Katrina in August 2005 was 902 hPa.
When we move to extratropical cyclones, the situation is different. The minima are a little bit higher, but not so much as one could imagine.
For Southern hemisphere, the record belongs to a storm developed near Antarctica, with 919 hPa observed at Casey station on the Windmill Islands (just outside the Antarctic Circle) on Vincennes Bay (66°17’S 110° 31’ E) on August 8-9, 1976.
In the northern hemisphere, the most probable record of lowest pressure is held by the storm of January 10th, 1993, which showed a central pressure of 912-915 hPa between Iceland and Scotland near 62°N 15°W. The map from wetterzentrale.de, shows, at 00 UTC of January 11st, a deep minimum lower than 930 hPa, as a part of a large depression covering the whole northern Atlantic. I am remembering here that, in the site, the maps are available every 24 hours, at 00 UTC.
Also the storm occurred on December 15th-16th, 1986, deepened to at least 916 hPa south-east of Greenland, near 62°N 32°W, a value assessed by the British Meteorological Office, while the West German meteorological service proposed a pressure possibly as low as 912-913 hPa. The map below, at 00 UTC of December 15th, shows a minimum lower than 925 hPa, and a situation quite different from the previous one, in which the depression size was smaller and supposedly with most intense winds. In both cases, the evolution of the system was quite rapid: the low intensified by 25-30 hPa in only 24 hours, and filled by 25-30 hPa in a similar time.
On December 15, 1986, the ship Uyir, sailing southeast of Greenland, recorded the value of 920.2 hPa. According with the British Met. Office, the central pressure of the storm, which was centered some distance southeast of the ship, was 916 hPa. The map below, at 00 UTC of December 15th, shows a minimum lower than 925 hPa, very close to Greenland. The almost explosive deepening of the depression was followed by a fast filling on the subsequent day.
It is interesting also to mention some record values observed on the territory of individual nations. The Iceland one is 923.6 hPa, observed on December 2nd, 1929, at Storhofoi. The map on 3rd at 00 UTC shows a minimum of 945 hPa close to the Iceland; the depression has undergone a rapid deepening and a subsequent rapid filling.
Other two relevant minima studied by the British weather historian Stephen Burt are:
921.1 hPa on Feb. 5, 1870 measured by the ship Neier at 49°N 26°W;
924 hPa on Feb. 4, 1824, measured at Reykjavik, Iceland (the lowest on land measured pressure in the North Atlantic). These two minima are too old in order to draw a map.
Also the value of 925.5 hPa, recorded on Dec. 4, 1929 by the SS Westpool somewhere in the Atlantic (the exact location is unknown) is noticeable: in this case, the map on 5th at 00 UTC shows a minimum of less than 940 hPa off the Ireland, with very strong winds on the Irish coasts.
In UK, the minimum value of sea level pressure recorded over the territory was 925.6 hPa, on January 26th, 1884, at Ochtertyre, Pershire. In this case, the map from wetterzentrale.de on 27th at 00 UTC reveals a minimum of 945 hPa located between England and Norway.
For this event, a surface chart of the cyclone drawn 6 hours before (Burt, 2006) reveals the minimum of 925.6 hPa at Ochtertyre, in Scotland.
Even if not in the north Atlantic, also the most powerful storm observed in Alaska in modern history in October 25-26, 1977, at Dutch Harbor on the Aleutian Island of Unalaska, is remarkable. The minimum pressure observed was 926 hPa on the evening of October 25th. Winds gusted to 130 mph at Adak, and gusted for 12 consecutive hours exceeding 110mph. The map below reveals as, even in this case, the size of the depression, the proximity of the isobars, leaving deduce a very strong wind speed associated with the cyclone.
Worthy of mention is also the minimum recorded in Ireland, at Belfast, on December 8th, 1886: 927.2 hPa. In that occasion, the map on 7th at 00 UTC reveals a narrow depression elongated from North Pole to Mediterranean sea and centered on Scotland, already filled to less than 955 hPa.
Finally, we want to mention also the Finland record of 939.7 hPa, recorded on March 1st, 1990. The map of February 28th at 00 UTC clearly shows the minimum (lower than 950 hPa) over the southern Finland as the center of a large depression covering most of northwestern Europe. Differently on previous cases, the evolution of this minimum was slower, and the cyclone, formed on February 26th near Iceland, moved northwards, gradually filling.
This short overview of meteorological situations favourable to produce very low cyclones in the northern hemisphere can be sufficient to say that the vale predicted for the next weekend in the north Atlantic can be considered not exceptional and not unusual, even if rare. Normally these storms are confined, in both hemispheres, at latitudes not lower than 50°, with very few exceptions.
The range of pressures predicted by the different models for the storm of the January 26-27 weekend is of about 12 hPa (according with the forecasts available at 10 p.m. of January 22nd, the minimum will range between 928 and 940 hPa), and is in agreement with the standard deviation of 15 hPa of the ensemble prediction of ECMWF model (see figure below).
To get a more quantitative picture, it is possible to use the data founded by Von Ahn et al. in this study of 2005, that will be shorthly summarized here. The authors examined all extratropical or mid-latitude storm systems showing a wind speed larger than the value defined as “hurricane force” (HF, equal to or larger than 64 kts). They used near-surface winds from the National Aeronautics and Space Administration (NASA) QuikSCAT scatterometer. In this way, they observed, from October 2001 to April 2004, a total of 120 HF cyclones. The distribution in terms of minimum pressure was the one reported in the figure below.
The authors, examining the evolution of these cyclones, have noticed that, for most cyclones, HF conditions were observed to occur at or near the time of minimum central pressure (the mature phase of the cyclone), lasting on average less than 24 hours, a relatively short-life compared to the average life span of 5 days for ocean storms. The typical evolution of an ocean storm follows the scheme depicted by Shapiro and Keyser (1990) and well summarized in the paper of Von Ahn et al., that I summarizes here. The cyclone begins as an open frontal wave with a warm front and cold front (I). As the cyclone intensifies, also the frontal wave amplifies. The cold front pushes eastward (South of the low) and the temperature gradient tightens to the West of the low center (II). The front associated with this tightening temperature gradient west of the low is referred to as the bent back front or occluded front. The wave continues to amplify (III) and the bent back (occluded) front and associated temperature gradient swings eastward to the southwest of the low center. The strongest temperature gradient in phase III is associated with the continuous warm to bent back front and not in association with the cold front to the south. Phase III is referred to as the frontal T-bone. Phase IV shows the mature cyclone or warm core frontal seclusion. At this point, the very strong temperature gradient (or front) has encircled the surface low center. A shallow pocket of relatively warm air has migrated to the low center and become cut off or secluded (thus the term warm seclusion). Within the warm seclusion the air is very unstable and convection may occur. An arc of very strong temperature gradient surrounds this pocket of warmer air with cold air found to the exterior of this temperature gradient. A very strong pressure gradient exists on the cold side of the temperature gradient (south of the low). It is in this area of strong pressure gradient that HF conditions are often observed.
Using the winds derived by QuikSCAT images to create composites of the maximum winds for 17 open ocean HF cyclones (11 in the North Pacific and 6 in the North Atlantic) near maturity or close to minimum central pressure, Von Ahn et al. have determined where HF conditions occur most frequently, deriving a conceptual model of cyclone development shown in the figure below. According with this model, the area of HF winds recorded by QuikSCAT, indicated in the figure with red hatching, is located from southeast to nearly west of the low center during the mature phase of an ocean cyclone. Figure 5 illustrates where to anticipate HF winds in a mature cyclone.
It is needless to say that, despite the short duration of HF conditions, HF cyclones can indeed be very dangerous, especially if the area of strongest winds will impact on land area. A well know example is the sequence of storms that hit France and central Europe during the Christmas holidays of 1999, at unusually low latitudes.
In the case of landfall, contrary to what happens for tropical cyclones, in which the “engine” is constituted by the presence of water vapor, the system does not dissipate too much during the landfall, but follows its natural dynamics.
Cristopher Burt for his two posts on weatherunderground;
Stephen Burt for list of North Atlantic pressure records and the list of his papers;
Steve Gregory for the map of the Alaskan storm of 1977 and related information.
Burt, Christopher C. (2011) Weather extremes. Super Extra-tropical Storms; Alaska and Extra-tropical Record Low Barometric Pressures, available on http://www.wunderground.com/blog/weatherhistorian/comment.html?entrynum=49
Burt, Christopher C. (2011) Weather extremes. World and U.S. Lowest Barometric Pressure Records, available on http://www.wunderground.com/blog/weatherhistorian/comment.html?entrynum=50
Burt, Stephen D. (1983) New UK 20th Century low pressure extreme. Weather, 38, pp. 209-13
~ (1985) Remarkable pressure fall at Valentia, 17 October 1984. Weather, 40, pp. 48-51
~ (1987) A new North Atlantic low pressure record. Weather, 42, pp. 53-56
~ (1987) A new North Atlantic low pressure record. The Marine Observer, 57, No. 297 (July 1987), pp. 122-125
~ (1987) Deep depressions. Letters to the Editor, The Times, London: REPRINTED IN Letters to the Editor, J Meteorol, 12, pp 348-8
~ (1989) London’s lowest barometric pressure in 167 years. Weather, 44, pp 221-5
~ (1993) Another new North Atlantic low pressure record. Weather, 48, pp 98-103
~ (2006) Barometric pressure during the Irish storm of 6-7 January 1839. Weather, 61, pp 22-27
~ (2006b) Britain’s highest barometric pressure on record is incorrect. Weather, 61, pp 210-1
~ (2007) The Lowest of the Lows … Extremes of barometric pressure in the British Isles, Part 1 – the deepest depressions. Weather, 62 (1), pp 4-14
~ (2007) The Highest of the Highs … Extremes of barometric pressure in the British Isles, Part 2 – the most intense anticyclones. Weather, 62 (2), pp 31-41
~(2008) The intense anticyclone over NW Russia, early January 2008. Weather, 63, pp 174-76
~ (2009) Long-term variations in extremes of barometric pressure in the British Isles. Weather, 64, pp 187-189
~ (2011) Barometric pressure during the Irish storm of 6-7 January 1839. Published online at http://www.irishmetsociety.org/jan-1839-storm