Storm Catalogue The Netherlands
The goal of this website is to visualize the storms, chosen for reanalysis with the high resolution atmospheric model HARMONIE to determine in later stage the time and spatially varying extreme wind fields needed as input for the next safety assessment of the primary water defences.
This website is an extension of the Deltares report "Selecton of historical storms for atmospheric model validation" (2011). The full analysis of these 17 events is expected to fulfil the aim of the reanalysis project: ‘To assess the quality of the model and improve where needed, in order that in 2017 and onwards more accurate HBC can be determined.’
Embedding of wind modelling within SBW program
To comply with Dutch legal rules for the assessment of the safety level of water defences, the government funded program WTI (“Wettelijk Toets Instrumentarium”: legal assessment instruments) produces the required safety assessment instruments. This program is supported by the SBW (“Sterkte en Belastingen Waterkeringen”: strengths and loads of water defences) program, also funded by the government, which addresses relevant knowledge gaps. More details on the framework of SBW and the link with WTI, as well as motivation and background, are given in the report.
Within the SBW program, there is a project dealing with knowledge gaps in the determination of hydraulic loads that water defences should be able to withstand. This project, in turn, contains a wind modelling component dealing with the way in which extreme wind fields, used in the computations of extreme hydraulic load conditions, are determined. The first step of this recently started “SBW hydraulic loads wind modelling” project is the selection of a number of storms. These storms will be modelled subsequently and the modelling results will be validated using observations, so that choices can be made concerning appropriate models and methods of analysis. This report is devoted to the first step, the storm selection.
Accordingly the following criteria have been defined to select the storms to
be simulated using the high-resolution numerical weather prediction model HARMONIE. The
entire set of selected storms should:
1. Be relevant in terms of hydraulic loads (waves and water level at each water system).
2. Contain high wind speeds from different sectors.
3. Contain a range of stability characteristics above land and water.
4. Consist of a range of small to large storms.
5. Consist of a range of slowly to rapidly moving storms.
6. Consist of a range of paths followed by storms.
7. Contain variety in season.
8. Contain variety in the role/presence of fronts in the weather system.
9. Have available measurements to validate the model results.
10. Have suitable lateral boundaries from a global atmospheric model.
In applying the criteria above we have restricted ourselves to the period from 1979 until 2010. For this period ERA-interim data is available to force the HARMONIE model. For historical reasons the 1953 storm was added to the selection. ERA-40 data is, for the same historical reasons, available for this storm and will be used to drive the HARMONIE model for this period.
The stormlist on this website contains animations of weanthercharts and links to relevant documentation from various sources. The weathermaps are based on NCEP-reanalysis (being available on short term) and KNMI-analysis based on Hirlam-output (since 2003). The weathermaps contain the pressure pattern at surface, the main flow at 500 hPa and the temperature distribution at 850 hPa. NCEP-analysis are presented in steps of 24 hours, Hirlam-analysis in steps of 6 hours. The windseries are based on information from the Hydra-project 1998-2005 (Verkaik). Indications of dimensions of the storm are based on visual inspection of the NEP-charts around the main event as an impression of f.i. the size and speed of the storm.
Example of a reanalysis ofthe storm of 09 November 2007, the white lines represent the surface pressure, the coloured lines the geopotential height (in decametres, see bar of the right. The centre pressure of the stormlow near Skagerak is close to 875 hPa, the high pressure area on the Atlantic is apprx. 1038 hPa, the stormfield is at ist maximum along the Norwegian coast. The main upperflow over the Northsea is northwest.
Example of the temperature distribution at 850 hPa (aboit 1,5 km), indicating the distribution of airmasses with cold air over the Northsea and warm air over the central Atlantic and southern Europe.
Example of a surface chart with isobars and fronts. The pressure differences at 12 UTC on 09 November 2007 are at its maximum over the northern coast og the Netherlands and the German Bight, just northwest of the occluded front (purple front)t. The blue front from Poland to Pyrenees represents the cold outbreak to the southeast, the red front marks the warm airmass ahead. The blue line in the Adriatic is an area with active showers.
The 17 storm periods have been simulated with the HARMONIE model. For each storm, the simulation starts typically 4 days before the main event and extends to approximately 2 days after the maximum intensity of the storm. Fields of 10m Wind and Mean Sea Level Pressure can be found here.
Example of HARMONIE model output for 25 January 1990 18 UTC. Left: wind speed in Beaufort. Right: mean sea level pressure in hPa.
Stormlists since 1979 (plus 1953)
for reanalysis in Harmonie.
Top 17, ranked by date
Hydra or more
Explanation of the column-titels:
Motivation for the stormlist:
In terms of relevance for the stormselection the projectteam is convinced that with this selection the criteria for the choice of storms for reanalysis with Harmonie are met. The choice is not fully objective (in a quantative way), but looking at the mentioned criteria we can motivate the variations of events:
- historical relevance (like # 1, 1953-flood, 15 Allerheiligenflood),
- significant wind-events (like # 1,5,7,14),
- significant hydraulic events (like # 6, 9, 11, 15, 17,
- variation in season (10 in winter, 6 in autumn, #13 in spring),
- variation in windfield direction (like # 2, 12 (NE) versus 3, 4, 7, 10, 14 (SW) versus 1, 6, 11, 17 (NW)),
- variation in stability changes (like # 1, 7, 11, 14, 15) (situatons with cold and warm air advection),
- variations in size, width variations of width/gradient (like # 6 and 8, 10 and 13) en length of the windfield (like # 7 and 10, 6 and 16),
- variations in speed/track (like # 17 and 8, 2 and 5) between slow to fast moving storms,
- presence of sharp windchanges (front, trough) ( like # 3,4, 5, 9, 10, 14, 15, 16)
Summary of all events:
Comprehensive description is made of the 17 selected historic storms
The storm on 1 February 1953 is marked as the major water flood in the Netherlands in 20th century. The water level in nearby Hoek van Holland reached 4.096 m (This water level includes the trend adjustments to the levels of 2009). The flood was caused by a severe storm (combination of two low-pressure systems) from the northwest. The major system moved on 31 January 1953 from Scotland to the German Bight. The next day the low pressure centre moved to southern Poland, causing a long and severe north-westerly storm field on its west flank over the North Sea. In 2003, 50 years later, the disaster was remembered with various presentations, available at this website.
In the winter of 1979 a cold outbreak is occurring on 14-16 February 1979. Between a complex low over Belgium and France and increasing high pressure over Skandinavia an increasing easterly wind advects very cold air. In this cold event the temperatures in the northern part of Holland dropped at 1.5 km up to 20 degrees below standard, this event is marked as the coldest in history since 1900. Increasing rain and freezing rain in the south and intensive snow in the north, in combination with strong
winds during about three days, caused many problems for the society. Drifting snow caused snow dunes up to 3 to 6 meters in the northern part.
Remark: The anemometer of Leeuwarden was frozen, the wind was estimated (too high) by the observer in Leeuwarden. Ever since (till 2010) this event was (incorrectly) marked as a Beaufort 10 easterly storm. This case is interesting for reanalysis as a (although probably not Beauford 10) significant winter case with strong easterly winds, reanalysis might be used for (better) indications in climatological databases of the wind speed in Leeuwarden.
On 01/02 February 1983 a slow filling low with a westerly wind field moves rather quickly from Scotland to the Baltic. Temperature changes at 850 hPa are in the order of 10 degrees, indication for the presence of larger stability changes as sea-water temperatures are close to minimum in mid-winter. This storm has a rapid displacement and the wind field has major dimensions, the event is ranked in the top 3 for wind- and hydraulic events.
On 27 November 1983 a trough is developing to a major low over the Irish Sea and moving to the east over the southern North Sea and the northern part of Holland. The system is rather small and fast moving. The main wind field starts as moderate southwesterly, weakening on the passage of the centre, but increasing to a strong west-northwesterly wind. The frontal systems are occluded, the 24 hour change of the temperature at 850 hPa is about 6º, indicating little stability variation over sea.
On 14 January 1984 a major low over northern Scotland moves, whilst deepening, to central Norway. The system is extensive in dimensions, the main wind field veers from southwest to a strong west-northwesterly direction and the system movement is rather fast. The 24 hour change of the temperature at 850 hPa is about 5 ºC in cool polar air, indicating few stability variations over sea. The wind field is ranked as #1 for measuring stations in the north.
On 14 February 1989 a rapidly developing low moves from northern Scotland to western Norway. The system has a sharp trough to the south, indicating a quite rapid change of wind direction from southwest to northwest on passage of the system along the Dutch coast. The system movement is rather fast. The 24 hour change of the temperature at 850 hPa is about 3 ºC in cool polar air, indicating few stability variations over sea. The storm generated high surge along the Holland and Zeeland coast.
On 25 January 1990 a very rapidly developing low, part of a complex low on the Atlantic Ocean, moves from the Irish Sea to Skagerak. The frontal system has a warm sector, passing in the afternoon with extremely mild air for February, temperature changes at 1.5 km in the order of 12 ºC, indicating the presence of stability changes over sea (reducing wind speed over sea if temperature of the air is much higher than the temperature of the sea surface) which is illustrated by the fact that the peak wind speed at Schiphol exceeds that at IJmuiden. For this reason this storm has been part of the investigation of thermal effects in the land-sea wind speed ratio (Caires et al., 2009). Three days later a second low is passing with a strong southerly wind field, a few days later followed by a third system. The total event is lasting more than a week, and therefore a serious candidate for multiple day reanalysis and analysis
of stability effects and succeeding wind fields.
On 26 February 1990 a very rapidly developing and fast displacing low, part of a complex low on the Norwegian Sea, moves from Scotland to Sweden. The frontal system has a small warm sector, temperature changes at 1.5 km in the order of 5 ºC indicating small stability changes over sea. The storm caused high surge levels in the Wadden Sea and along the Dutch Coast, in the period between the 26th of February and the 2nd of March, the Eastern Scheldt storm surge barrier was closed four times.
On 12 December 1990 a north-northwesterly flow over the North Sea around a stationary high over the Atlantic causes an outbreak of a cold storm low over the North Sea, later integrating in a complex low over Germany. In the cold event the temperature changes at 1.5 km in the order of 4 ºC, indicating little change in stability during the event. The long fetch in the stretched wind field from the Norwegian Sea to the southern coasts of the North Sea cause a significant wave- and swell-event. Furthermore, although the highest wind period coincided with low tides, the attained water levels were rather high.
On 14 November 1993 a strong westerly flow in the upper air (jetstream) causes the rapid development of a low pressure system over southern Ireland, moving in ENE direction over the central part of the Netherlands. As the system approaches the Netherlands the southerly wind is decreasing, becoming a strong NW wind after passing the country in the second part of the day. In the mild event the temperature changes at 1.5 km are in the order of 5 ºC. indicating small changes in stability during the event. The changes in the wind field cause a significant wave event in the southern North Sea. The Eastern Scheldt and the Hollandse IJssel storm surge barriers closed twice during this storm.
On 28 January 1994 a rapidly moving and filling low moves over the Norwegian Sea to the Baltic, the pressure changes are large (about 25 hPa over 24 hr along the Dutch coast). The strong and mild south-westerly flow over the North Sea increases and veers to a colder northwesterly flow in the morning of the 28th. At the end of the day the wind decreases as the gradient weakens. In this event the temperature changes in the colder air mass are at 1.5 km in the order of 5 ºC indicating small changes in stability, if the previous day (27th) is also included the stability variations are large, about 12 ºC. The storm causes high water levels along the whole coast of the Netherlands.
On 18-20 February 1996 a rapidly moving and filling low moves over from Scotland over the centre of the Netherlands to the Alps. The mild south-westerly flow in front of the system decreases to a light and variable wind during the passage. After the passage of the centre of the low the north-easterly wind increases to a storm along the coast. The temperature changes in the colder air at the rear of the low are small, at 1.5 km in the order of max. 5 ºC, indicating small changes in stability. This case is an interesting example of a rapidly changing winter flow with relatively high wind speeds from the Northeast.
On 28/29 May 2000 a small and rapidly developing low is passing the Netherlands from the southwest, moving to Denmark and often indicated as "Kanaalrat"8 due to its origin, small size and high impact. The low develops on the southern part of a steering low in the Norwegian Sea. The temperature changes at 1.5 km in the order of 7 ºC indicating the presence of some changes in stability in spring season. The storm was also marked as a twin storm (two systems in sequence at the 27th and 28th/29th
On 27 October 2002 a storm low is moving, after its maximum development over central UK, very rapidly over the North Sea to southern Baltic. The storm field veers from W-SW to WNW. The temperature changes at 1.5 km are about 10 ºC, indicating some changes in stability. The storm is at the website of the Storm Catalogue documented in a TV-news bulletin and a case-study. This storm appears in the top-3 of wind events.
On the 1st of November 2006 a sharp trough (convergence zone in cold air) circles over the North Sea to the south, steered by a rapid eastward moving storm low in the northern North Sea. Wind variations (increase and decrease, as well as direction changes) are large during the day, temperature variations small (cold air event). The storm is marked as a surge- and SWL record in Delfzijl and well known as Allerheiligenvloed9 or the “Horses storm”, because 227 horses were isolated due to flood in Marrum, 25 horses drowned, the rest was saved in a spectacular rescue operation.
On 18 January 2007 an intensifying complex low moves rapidly from the eastern Atlantic to the Baltic. The main wind field veers from a mild southwest to a cool northwest, but temperature variations during the singular event are small. The event is part of a sequence of low pressure systems, marked as a triple storm. A reanalysis for a longer period is recommended, also for analysis of various stability changes in this winter period, due to vertical temperature variations (several fronts pass between 16 and 21 January 2007). This storm occurs in the top three of wind events (cf. Table 2.2). In the period between 12 and 20 January two significant wave events occurred in the Wadden Sea region.
On 08 November 2007 an intensifying low moves over the southern Norwegian Sea to Skagerak, the next day with decreasing intensity to the central Baltic. In its intensifying period it causes a north-westerly storm field over the central North Sea with a long fetch over water on the 9th. As a cold-air-mass-event the temperature changes at 1.5 km are small, about max. 5 ºC, the air mass is unstable (in November with relative warm seawater). This storm is part of the top three of hydraulic events and is the first storm since the building of the Nieuwe Waterweg and the Maeslantkering that it was closed due to the exceedance of the threshold water levels in Hoek van Holland. The peak water level at Hoek van Holland was the highest since 1954, but still 90 cm lower than the peak water level of the February 1953 storm.
Geert Groen and Sofia Caires, nov. 2011.