The GFS is an operational global numerical weather prediction model run by the National Oceanic and Atmospheric Administration (NOAA), USA. The ARW model is a research mesoscale numerical weather prediction model built and maintained by the National Center for Atmospheric Research, USA, and can better resolve mesoscale systems such as Low B.
Snow and gales in eastern England from a North Sea polar low: 6/7 January 2010
Version of Record online: 24 DEC 2010
Copyright © 2011 Royal Meteorological Society
Special Issue: Severe winter weather
Volume 66, Issue 1, pages 10–13, January 2011
How to Cite
Lawson, J. (2011), Snow and gales in eastern England from a North Sea polar low: 6/7 January 2010. Weather, 66: 10–13. doi: 10.1002/wea.740
- Issue online: 24 DEC 2010
- Version of Record online: 24 DEC 2010
While many people across the British Isles were digging out cars and gritting driveways on the morning of 7 January 2010, a mesoscale low was bringing gale force winds and snow showers to the Norfolk and Lincolnshire coasts. The precipitation at this time was illustrated by the 0630 utc radar image from the Met Office network (Figure 1). The mesoscale low was one of many wintry weather events in the winter of 2009/2010, and, at the time it formed, 22 people had already died in the UK due to car accidents, avalanches, and ice-related accidents (Times Online, 2010). Impacts of the severe winter weather included shortages of road grit, widespread workplace and school closures, and severe disruption to travel (BBC, 2010). Because attention was directed elsewhere, the media barely mentioned this mesoscale low. Although the snowfall and winds that affected the Norfolk coast added to the disruption, they were not explicitly addressed in meteorological reports (for example, Brugge, 2010).
Mesoscale lows can have a large impact in small regions and over short periods of time. Due to their small diameter, which numerical models often fail to resolve, they are often more difficult to forecast than synoptic-scale low-pressure systems. Climatology of such mesoscale vortices (Forbes and Lottes, 1985) indicates that they are most frequent in polar regions during the winter, with the term ‘polar low’ used to designate stronger systems. Rasmussen (1981) discovered that certain warm-core variants of these polar lows are different in genesis and structure from synoptic-scale low-pressure systems. The latter are typically associated with strong cyclonic vorticity advection aloft, producing a comma-shaped cloud pattern in satellite imagery. In contrast, many polar lows are driven by convection (so-called ‘Arctic hurricanes’) and produce a spiral-shaped cloud pattern.
At 0600 utc on 4 January, northwest Europe lay under a diffluent northerly flow at 500mbar, being located downstream of a zonal shortwave trough axis across the Norwegian Sea. As the trough started to move south and amplify, it advected cold air southwards. By 0600 utc on 6 January (Figures 2 and 3), a 500mbar closed low was analysed over the English Channel, collocated with a filling surface mesoscale low (Low A). A 999mbar pressure minimum lay to the southwest of Denmark (Low B, the subject of this article), and a trough extended to the east of it (analysed as an occluded front). With a 1019mbar surface high over Norway, a strong eastnortheasterly surface flow extended from the Skagerrak across the North Sea, as seen in the low-level cloud streets on the infrared satellite image (not reproduced). Close to the Danish and British coasts, deeper convection was breaking out in the vicinity of the occluded front.
The surface temperatures in western Sweden were −12 to −14°C, with dew-point depressions of 2 to 3 degC. The temperature gradient was orientated east–west, hence cold air was being advected from the cold land over the warmer waters of the Skagerrak and the North Sea. The strong temperature gradient perpendicular to the Norwegian coast is analogous to the sea-ice margin often seen as a prerequisite for certain types of polar low formation (Rasmussen and Turner, 2003), which require a shallow baroclinic region.
Is Low B a polar low?
Rasmussen and Turner (2003, p. 12) define a polar low as asmall, but fairly intense, maritime cyclone that forms poleward of the main baroclinic zone…approximately between 200 and 1000km [in diameter] and surface winds near or above gale force [standard Beaufort scale]. For the purpose of this article, a polar low is defined as a small meso-α (by definition, 200–1000km in diameter) vortex containing near gale-force 10-minute average winds (14ms–1/27kn) or greater, whose genetic region must lie poleward of the main baroclinic zone. A secondary criterion, from results in previous spiraliform polar low studies (Rasmussen, 1981), is a warm core generated by strong convection and latent heat release. Warm-core low-pressure systems, such as tropical cyclones, are associated with winds decreasing with height (i.e. the strongest winds are observed close to the surface).
Due to the paucity of surface observations in the North Sea, models were used to supplement existing data. The Global Forecast System (GFS) and the high-resolution Advanced Research Weather Research and Forecasting (WRF-ARW) run resolved Low B's structure (Figure 4) and moved the polar low into southeast England, but did not resolve the heavy snow and strong winds on the Norfolk coast.1
Is Low B poleward of the polar front?
The polar front lay over northern Spain, southern France and eastward into central Europe on 6/7 January, hence Low B was located well poleward (500–1000km) of the main baroclinic zone. Although Low B formed at the relatively southern latitude of 55°N, other polar-low-like systems have been documented as far south as 40°N, for instance by Homar et al. (2003). In that case, a quasi-tropical cyclone formed in the Mediterranean poleward of an intense cold front, and involved processes characteristic of both cold-low polar lows and tropical cyclones.
Are 10-minute average wind speeds near or at gale force (greater than 14ms−1)?
A 10-minute average wind speed of 23ms−1 (45kn) was observed in the North Sea (at 2000 utc on 6 January at Tyra Oest), and shortly after landfall at Weybourne, Norfolk, where observations indicated a 10-minute wind speed of 16ms−1 (31kn) at 1300 utc on 7 January (Figures 5 and 6). Due to the sparse observations in the North Sea, it is difficult to ascertain for how long these strong wind speeds were present, but the model output shows gale force winds associated with Low B for a 12-hour period on the morning of the 7th.
Is Low B between 200km and 1000km across?
An infrared satellite image from 0522 utc on 7 January can be used to estimate the size of Low B, which is about 10° longitude at 55°N (Figure 7). This gives Low B a diameter of roughly 700km, i.e. meso-α scale.
Is there conclusive evidence of a warm core?
The logical starting point of a warm-core diagnosis is the temperature distribution. GFS initialisations on the 6th and 7th both indicate a region of warmer air in the centre of Low B at 500mbar and 850mbar, most likely a signal of latent heat release from convection. However, the system is almost vertically stacked (i.e. the central geopotential height minima are collocated throughout the troposphere), whereas a typical warm-core system would contain an anticyclone above a cyclone. Nevertheless, the model shows small anticyclonic vortices around the low centre, one of which is collocated with strong convection in the boundary layer, suggesting an outflow from cumulonimbus as seen in previous spiraliform cases (Rasmussen, 1981). Interestingly, the ARW model output profile and a sounding from Ekofisk at 0000 utc on 7 January (Figure 8) compare favourably with a typical vertical wind profile associated with a tropical cyclone. The skew T–log p diagram shows decreasing winds with height (a barotropic signature) and almost calm conditions near 500mbar. This evidence argues for the storm having a warm core.
A warm-core low contains its highest wind speeds close to the surface, with speeds decreasing with height. This results in a thermal wind vector that may oppose the low's motion, as seen within Low B. Wilhelmsen (1985) classified polar lows into broad categories. One of these categories, reverse-shear polar lows, was defined as polar lows that contained a thermal wind vector that opposed propagation of the thermal wave and movement of the low-pressure centre. This is opposite to the typical orientation where all three vectors are unidirectional. In fact, there are similarities between this case and another reverse-shear polar low case observed in the Hudson Bay (Albright and Reed, 1995).
Low B met all the criteria for, and should therefore be considered as, a reverse-shear polar low. It formed in the eastern North Sea when an upper trough moved into a downstream surface baroclinic zone, advecting cyclonic vorticity aloft. It then moved west over warmer sea and became more organised. After differential cyclonic vorticity advection was minimised by the collocation of the surface low and upper trough, destabilisation as a consequence of cold air advection over the warm sea surface probably became dominant in aiding intensification, while condensational heating prompted Low B to develop a warm core as seen in model output. Strong convection was seen in the northwest quadrant of Low B, as is often the case in polar lows. This convection was connected with destabilisation from cold-air advection over the warmer waters, and collocated with strong surface wind – and therefore increased evaporation and low-level wind shear – due to the increased pressure gradient force between the Norwegian high and the European low. Upon reaching the coast of England, Low B filled, probably as a result of coming into contact with the colder land and losing the energy source of the relatively warm North Sea waters. Further study of this interesting case in the form of sensitivity experiments with ARW simulations may reveal more evidence for the importance of surface fluxes of heat and moisture and latent heat release, all potentially responsible for the formation of Low B.
Many thanks are due to David John Gagne II at the University of Oklahoma for his help with Python scripts and creating imagery from the ARW run, Kent Knopfmeier at the University of Oklahoma for assistance with the model runs, the Met Office for satellite imagery and data, Neil Lonie at the University of Dundee Receiving Station for high-resolution satellite data, David Smart at University College, London, for many useful images, Dr David Schultz at the Universities of Manchester and Helsinki and Finnish Meteorological Institute for providing comments that strengthened this article, and Dr David Stensrud at the NOAA/National Severe Storms Laboratory for initiating and advising the report.
- 1995. Origin and structure of a numerically simulated polar low over Hudson Bay. Tellus 47A: 834–848. , .
- BBC. 2010. Frozen Britain. Retrieved May 2010, from BBC News. http://news.bbc.co.uk/2/hi/in_depth/uk/2010/frozen_britain/default.stm [accessed May 2010].
- 2010. Roger Brugge's Weather Diary. http://www.met.rdg.ac.uk/∼brugge/diary2010.html [accessed May 2010]. .
- 1985. Classification of mesoscale vortices in polar airstreams and the influence of the large-scale environment on their evolutions. Tellus 37A: 132–155. , .
- 2003. Numerical diagnosis of a small, quasi-tropical cyclone over the western Mediterranean: Dynamical vs. boundary factors. Q. J. R. Meteorol. Soc. 129: 1469–1490. , , , , .
- 1981. An investigation of a polar low with a spiral cloud structure. J. Atmos. Sci. 38: 1785–1792. .
- 2003. Polar Lows: Mesoscale Weather Systems in the Polar Regions. Cambridge University Press: Cambridge, MA. , .
- Times Online. 2010. Deaths during Britain's Big Freeze. http://www.timesonline.co.uk/tol/news/weather/article6979895.ece [accessed May 2010].
- 1985. Climatological study of gale-producing polar lows near Norway. Tellus 37A: 451–459. .