Western European Snow of 1-2 February 2009 By Richard H. Grumm National Weather Service Office State College PA 16803. 1. INTRODUCTOIN A rare snow event struck Western Europe on 1-2 February 2009. This was reported, by newspapers and on the United Kingdom Meteorological Office 1 (UKMO ) to be largest snow fall over London and southern England in 18 years. The last event of this magnitude was observed on 7-9 February 1991. The cold episode of February 1991 was documented by Bruge (1991). The snowfall was observed in the United Kingdom and in portions of France. New accounts implied that many business activities, schools, and transportation systems had to be close. New accounts reported “transportation” nightmares as trains and tubes slowed or came to a halt. Snow being a relatively a rare phenomena, trains are not equipped to remove snow from the tracks. Additionally, with significant snowfall being so uncommon, snow removal equipment is normally not purchased or on hand to remove snow. The snow therefore was estimated to have cost about 1.3 billion British Pounds in losses due to the closing of so many business and government activities. The new report that the snow produced what was in essence a National snow day (credit Washington Post ). Snow fall amounts were nearly 1 foot in and around London. On the roads massive delays and traffic jams resulted. The M25 reportedly had a traffic jam, due to snow; about 50 miles long 1 The story was found at the UKMO news release site: http://www.metoffice.gov.uk/corporate/pressoffice/ 2009/pr20090203.html (credit The Guardian ). Extra fees were put in place on the roads to in an attempt to keep the traffic down. The snow was forecast in advance by the UKMO so the event was not a surprise. Note the term snow storm has not been used here. The term storm is usually associated with a surface cyclone of considerable magnitude. It will be shown that this and the February 1991 event were associated with easterly flow around a massive anticyclone to the north and east. This note will serve to document the pattern and conditions associated with the snow event of February 2009 and compare it to conditions associated with the snow event of 7-9 February 1991. 2. METHODS Data for this study include re-analysis climatological data from the NCEP/NCAR global re-analysis project (GR: Kalnay et al 1996). These data were used to reconstruct the pattern associated with previously documented snow event of 7-9 February 1991. The means and standard deviations were used to compute standardized anomalies, displayed in standard deviations from normal (SDs). Forecasts from the NCEP FS were used here to show the conditions associated with the snow event and to show forecasts of these conditions. Anomalies of key fields were produced from these data similar to the GR data. All data was displayed using GrADS. Anomalies were computed as described Hart and Grumm (2001) and Grumm and Hart
Transcript
Western European Snow of 1-2 February 2009
By
Richard H. Grumm
National Weather Service Office State College PA 16803.
1. INTRODUCTOIN
A rare snow event struck Western Europe on 1-2 February 2009. This was reported, by newspapers and on the United Kingdom Meteorological Office1 (UKMO) to be largest snow fall over London and southern England in 18 years. The last event of this magnitude was observed on 7-9 February 1991. The cold episode of February 1991 was documented by Bruge (1991).
The snowfall was observed in the United Kingdom and in portions of France. New accounts implied that many business activities, schools, and transportation systems had to be close. New accounts reported “transportation” nightmares as trains and tubes slowed or came to a halt. Snow being a relatively a rare phenomena, trains are not equipped to remove snow from the tracks. Additionally, with significant snowfall being so uncommon, snow removal equipment is normally not purchased or on hand to remove snow.
The snow therefore was estimated to have cost about 1.3 billion British Pounds in losses due to the closing of so many business and government activities. The new report that the snow produced what was in essence a National snow day (credit Washington Post). Snow fall amounts were nearly 1 foot in and around London.
On the roads massive delays and traffic jams resulted. The M25 reportedly had a traffic jam, due to snow; about 50 miles long
1The story was found at the UKMO news release site: http://www.metoffice.gov.uk/corporate/pressoffice/2009/pr20090203.html
(credit The Guardian). Extra fees were put in place on the roads to in an attempt to keep the traffic down.
The snow was forecast in advance by the UKMO so the event was not a surprise. Note the term snow storm has not been used here. The term storm is usually associated with a surface cyclone of considerable magnitude. It will be shown that this and the February 1991 event were associated with easterly flow around a massive anticyclone to the north and east.
This note will serve to document the pattern and conditions associated with the snow event of February 2009 and compare it to conditions associated with the snow event of 7-9 February 1991.
2. METHODS
Data for this study include re-analysis climatological data from the NCEP/NCAR global re-analysis project (GR: Kalnay et al 1996). These data were used to reconstruct the pattern associated with previously documented snow event of 7-9 February 1991. The means and standard deviations were used to compute standardized anomalies, displayed in standard deviations from normal (SDs). Forecasts from the NCEP FS were used here to show the conditions associated with the snow event and to show forecasts of these conditions. Anomalies of key fields were produced from these data similar to the GR data. All data was displayed using GrADS. Anomalies were computed as described Hart and Grumm (2001) and Grumm and Hart
(2001). Shaded values show the standardized anomalies computed as:
Figure 1. NCEP GFS 00-hour forecasts valid at 0000 UTC 01 February 2009 showing a) mean sea level pressure (hPa) and pressure anomalies, b) precipitable water (mm) and precipitable water anomalies, c) 850 hPa temperatures and anomalies, and d) 850 hPa winds (kts) and u-wind anomalies.
SD = (F – M)/σ)
Where F is the value from the reanalysis data at each grid point, M is the mean for the specified date and time at each grid point and σ is the value of 1 standard deviation at each grid point. For brevity not all forecast systems data or all levels and parameters are show. Furthermore, all times will be referred to as 02/0000 UTC for 0000 UTC 02 February 2009. Other years and dates when used will be spelled out in order to distinguish them from this event.
3. RESULTS
i. The pattern associated with the 2009 event
The large scale pattern at 01/0000 UTC over Western Europe is shown in Figure 1. The key features include a strong cyclone west of Spain with a -4 to -5SD pressure anomaly and a strong anticyclone over Scandinavia with anomalously high pressure anomalies. A strong gradient in the isobaric pattern implies easterly flow, which is depicted in Figure 1d which shows the 850 hPa wind and u-wind anomalies. The strong u-wind anomalies were northwest of the cyclone off the Iberian Peninsula.
Figure 2. As in Figure 1 except valid at 1200 UTC 1 February 2009 showing a) 500 hPa heights 9m) and height anomalies, b) mean sea level pressure (hPa) and pressure anomalies, c) c) 850 hPa temperatures and anomalies, and d) 850 hPa winds (kts) and u-wind anomalies.
The 850hPa temperatures (Fig. 1c) and precipitable water (PW) (Fig. 1b) showed a frontal zone pushing eastward across the Britain with dry, cold air over Europe moving toward Britain. There was a positive PW anomaly over Ireland, ahead of the advancing cold front.
By 01/1200 UTC the front had pushed west of Ireland (Fig. 2) and the anomalous cold air at 850 hPa had moved over southeastern Britain. The u-wind anomalies were -2 to -3 SDs below normal over much of Britain (Fig. 2d). At 500 hPa an upper-level low was present over Germany with negative height anomaly while a large positive height anomaly was present over Scandinavia. This pattern is often associated with the negative phase of the North Atlantic
Oscillation (NAO which brings generally cold weather to Western Europe.
At 01/1800 UTC (Fig. 3) and 02/0000 UTC (Fig. 4) the surge of anomalous cold air at 850 hPa had pushed into and through the region, reading as far west as Ireland. The deep 500 hPa low with -1 to .1.5SD height anomalies moved over Britain around 02/0000 UTC (Fig. 4a), accompanied by cold air and strong easterly flow (Figs. b-d).
The 500 hPa low was squarely over southern England at 02/0600 UTC (Fig. 5) then slid off southwestern coast. At the surface, a deep cyclone was present over France with an inverted trough extending northward over
southern England (Fig. 5a & 6b). Inverted troughs of this nature often are associated with precipitation along the East Coast of the United States. It is also interesting to note that with the cold air to the west, the easterly flow at 850 hPa implied strong warm advection over much of southern England.
A surface cyclone developed in the trough over the English Channel was captured in the GFS analysis at 02/1800 UTC (not shown) and then moved over England as shown in Figure 6b. The upper-level low at 500 hPa and the surface cyclone suggest that in a snow storm did develop.
ii. The pattern associated with the 1991 event
The snow event of 7-9 February 1991 produced over 10 cm of snow over England, Wales, and portions of southern Scotland. The UKMO report also suggests that about 15 cm of snow affected London during the morning of 9 February 1991. An as much as 51 cm of snow was observed Bingley in Yorkshire (credit UKMO).
Figure 3. As in Figure 2 except valid at 1800 UTC 1 February 2009.
There are many similarities in the patterns between the snow events of February 2009 and February 1991. Key features included the anomalously cold air moving over the region from the east, the strong anticyclone and upper-level ridge to the north and east and a strong low-level easterly, as depicted by the 850 hPa winds and wind anomalies.
Conditions from 0000 UTC 7 to 0000 UTC 9 February 1991 are shown in Figure 8-12. These data show the upper-level (500 hPa) trough and height anomalies over the region. For brevity 0000 UTC data is shown with the exception of Figure 11 where 1200 UTC data is shown. In writing this section data from 0000 UTC through 0000 UTC 11 February was examined in 6-hourly increments.
Though not shown, the PW and 850 hPa temperature fields showed a frontal zone pushing westward on 6-7 February 1991. At 0000 UTC 6 February a massive 1048 hPa cyclone over Scandinavia and the implied easterly flow would push extremely cold air over Germany westward over Britain by 7 February.
Thus, at the onset there was a push off cold continental air westward.
Figure 4. As in Figure 3 except valid at 0000 UTC 2 February 2009. Return to results.
In the 7-9 February 1991 event, the initial surge of cold air and the 500 hPa low retrograded over Britain (moved westward) but on the 9th the system became progressive and moved westward over Britain and back over Western Europe. Without good snowfall and radar data it is unclear when the heaviest snow fell, during the retrogressive or progressive phase of the event. However, UKMO reports of 15 cm of snow fall in and around London on the early on 9 February suggest heaviest snowfall may have been in that region may have been during the more progressive stage of the event.
Figure 5. As in Figure 3 except valid at 0600 UTC 2 February 2009.
iii. Forecasts
The overall pattern, to include the 500 hPa ridge and positive height anomalies, the large anticyclone, and the anomalous cold air were well predicted by the NCEP GFS and GEFS. The details on the snow are not available thought the GFS and GEFS both showed a potential for enhanced precipitation over much of eastern Britain during the period of interest.
For brevity a few forecasts are shown to demonstrate the value of Ensemble Prediction System (EPS) data and anomalies.
Figure 13 shows NCEP GEFS plume diagrams for a point near London, UK from forecasts initialized at 27/0000 and 29/0000 UTC. The
longer range forecasts have more uncertainty with some warmer members; however both plumes show the trend toward colder temperatures at 2m, 925, and 850 hPa. A sharper and more consistent change to cold is shown in the forecasts initialized at 29/0000 UTC. The sharpness of the cold front on the 1 February is quite evident in these forecasts. The precipitation type plumes from the GEFS suggest that snow was a possible outcome over London are shown in Figure 14. The ensemble mean snowfall was around 0.15 in the forecasts shown implying, to a first guess, 1-2 inches of snow (2.5 to 5cm) with one member from 29/0000 UTC suggesting over near 0.8 inches or ~8 inches (20cm) of snow.
The pattern associated with the cold episode was probably the more successful outcome of the
event for the GEFS and it can be verified and directly compared to data presented here. The 500 hPa heights and MSLP fields valid at 02/0000 UTC from the 30/0000 UTC GEFS is shown in Figure 15. These data show the key features including the 500 hPa ridge over Scandinavia and the 500 hPa low over southern Britain. The MSLP fields show the strong surface anticyclone and the gradient supporting the strong easterly flow. The GEFS also predicted the anomalously cold air at 850 hPa. Figure 16 shows the forecasts initialized at 31/0000 UTC. These shorter range forecasts show less uncertainty with the cold air, easterly flow, and 500 hPa cyclone over southern Britain.These data compare remarkably well with the data shown in Figure 4, the GFS 00-
hour forecasts which is a good proxy for ground truth.
Figure 6. As in Figure 2 except valid at 1200 UTC 2 February 2009.
4. CONCLUSIONS
A cold continental air mass pushed westward over Western Europe and Britain on 31 January and 1 February 2008. The strong easterly flow and an upper-air low brought snow to most of southern England. The snow was a high impact event in that it caused the transportation network to grind to a halt. The cost of this event, in lost work, was estimated to be in excess of 1.3 billion British Pounds. This was the largest snowfall across much of England since 7-9 February 1991. With such large intervals between significant snow events it clear that
most government agencies do not invest in snow removal equipment.
The overall pattern was similar to the pattern observed during the snow event o 7-9 February 1991. These two snow events included a strong ridge at 500 hPa and surface anticyclone over Scandinavia and strong easterly flow. Each event included the passage of a strong cold front from the east with abnormally cold air behind the frontal boundary. Both events were characterized by strong low-level easterly flow and the evolution of an upper-level closed cyclone. Finally, in each event, a surface cyclone developed.
Walsh et al (2001) showed the patterns of surface anomalies associated with North American and western European cold outbreaks. The positive pressure anomalies over Scandinavia shown for this event were consistent with the precursor to the cold episodes documented by Walsh et al (2001). The anomalies shown here were in the same location as those shown in their Figure 10.
Bruge (1991) documented the cold episode of February 1991 over western Europe and Britain. During this event, record snow fell over portions of Britian. This event, as shown from the GFS relative to the NCEP/NCAR reanalysis data shared many of the characteristics of the event 1991 as shown by Bruge (1991) and the MSLP anomalies were aligned well with those presented by Walsh et al (2001).
This case and the case of February 1991 show the value of anomalies in identifying potentially significant events. In this case a cold outbreak with locally heavy snow fall in Britain. Walsh et al (2001) clearly documented the value of mean sea level pressure anomalies to anticipate cold episodes in North America and Western Europe. Their results and the examples shown here seem of reinforce the value of anomalies and the value of anomalies when applied to model and ensemble forecasts.
The success of the GEFS may explain why news accounts of the cold and snow said the event was
well predicted. The UKMO has a model of comparable skill and it too likely predicted this event quite well. Additionally, the UKMO has access to the European Center (EC) Model2 and the 50-member EC ensembles which likely performed well in this event.
The data shown here show that the pattern and timing were relatively well forecast. The lack of data leaves many open questions. Does the relatively warm North Sea with cold air and easterly flow set up a lake effect or sea effect snow potential such as observed over the Great Lakes of the United States or the Sea of Japan? If so, this pattern and local forecaster experience probably are quite valuable.
5. Acknowledgements 6. 7. References
Bruge,R. 1991:The cold snap of February 1991. Weather, 46, 222-231.
Walsh, J. E, A.S Philips, D.H. Portis,and W.L Chapman,2001: Extreme Cold outbreaks in the United States and Europe 1948-1999.Jour. of Climate,14,2642-2658.
2 Model skill scores suggest the UKMO, NCEP,and EC models are of comparable skill though overall the EC model is the most skillful model available. No UK or EC data is available to display here.
Figure 7. As in Figure 3 except valid at 0000 UTC 3 February 2009.
Figure 8. As in Figure 2 except NCEP/NCAR global re-analysis valid at 0000 UTC 6 February 1991.
Figure 9. As in Figure 8 except valid at 0000 UTC 8 February 1991.
Figure 10. As in Figure 7 except valid at 0000 UTC 7 February 1991.
Figure 31. As in Figure 8 except valid at 1200 UTC 8 February 1991.
Figure 22. As in Figure 7 except valid at 0000 UTC 9 February 1991.
Figure 13. NCEP GEFS forecast from a point near London, UK, from forecasts initialized at 0000 UTC 27 and 29 January 2009. Data shown include temperatures for 2m, 925 and 850 hPa. Each ensemble member forecast is a thin line of variable color. The thick yellow line is the median and the thick black line is the mean of all forecasts. Units are degrees Fahrenheit for2m temperature and Celsius for 925 and 850 hPa data.
Return to text.
Figure 14. As in Figure 13 except GEFS forecasts initialized at 0000 UTC 29 January and 01 February 2009 showing precipitation, and accumulated precipitation by type. The 4 types are shown to the left. Return to text.
Figure 15. NCEP GEFS forecasts initialized at 0000 UTC 30 January 2009 valid at 0000 UTC 2 February 2009. From left to right are 5hPa forecasts, mean sea level pressure forecasts, and 850 hPa temperatures. Upper panels show spaghetti plots of select contours from each ensemble member and the thick black line shows the ensemble mean. Shading shows the spread about the mean. Lower panels show the ensemble mean and the departure of this field from normal in standardized anomalies.
00
Return to text.
Figure 16. As in Figure 15 except forecasts initialized at 0000 UTC 31 January 2009 and zoomed in closer to the United Kingdom. Return to text.
