Why was the start to spring 2013 so cold?

April 2013 - The Spring of 2013 started with the second coldest March in the UK record since 1910, and was associated with a negative phase of the North Atlantic Oscillation. A number of potential drivers may predispose the climate system to a state which accounts for these conditions.

Why was the start to spring 2013 so cold? A fuller briefing on the cold weather in early spring 2013 has been written by Professor Julia Slingo, Met Office Chief Scientist. (PDF, 896 kB)

UK observations

This ranks March 2013 as joint second (tied with 1947) coldest in the records. Unusually, this March was also colder than the preceding winter months of December (3.8 °C), January (3.3 °C) and February (2.8 °C). This last happened in 1975.

As well as being very cold, March has also been very snowy and joins 2006, 2001, 1995, 1987, 1979, 1970 and 1962 as years when March saw some significant snowfall.

Time series of UK temperatures for March since 1910 Figure 1: Time series of UK temperatures for March since 1910

Global context

The cold temperatures during March were part of a hemisphere-scale pattern of temperature anomalies (Figure 3), which also led to severe cold and snowy conditions across much of Europe, Russia and Ukraine and eastern and northern USA. The large scale pressure pattern was characteristic of the negative phase of the North Atlantic Oscillation (NAO), which is associated with persistent easterly and north-easterly winds bringing very cold air over the UK from northern Europe and Russia.

Distribution of temperature anomalies over the UK from the climatology for 1981-2010. Figure 2: Distribution of temperature anomalies over the UK from the climatology for 1981-2010. Surface air temperature anomalies over the northern hemisphere for March 2013. Figure 3: Surface air temperature anomalies over the northern hemisphere for March 2013.

An integral part of the negative phase of the NAO is a southwards shift of the North Atlantic jet stream, which acts to steer the storm track across southern Europe and the Mediterranean, away from its usual trajectory over the UK. So whilst the UK has been very cold, southern Europe and the Mediterranean has had unsettled and wet weather.

Further afield, although the recent La Nina event has dissipated and equatorial East Pacific ocean temperatures have returned to near neutral conditions, the West Pacific continues to be warmer than normal. In addition, the Madden Julian Oscillation (MJO) has again been active through the Indian Ocean and into the West Pacific during late February into mid-March, contributing substantially to the increased rainfall over the West Pacific in the monthly mean pattern.

A notable feature of the global ocean temperature anomalies is the extremely warm water in the Labrador Sea and down the coast of Newfoundland (Figure 4). Likewise the Arctic is mostly warmer than normal.  The pattern of ocean temperatures in the Atlantic predisposes the winter circulation to negative phases of the NAO and is likely a contributory factor to this year's cold start to spring.

Sea surface temperature anomalies for March 2013 against the climatology for 1981-2010.  Figure 4: Sea surface temperature anomalies for March 2013 against the climatology for 1981-2010.

As the plot of March monthly mean temperature anomalies (Figure 1) shows, very cold conditions in March have not been a feature of recent decades; indeed the general trend has been one of warmer and earlier springs. The last time the UK experienced a very cold March was in 1962 and again it was a countrywide event. Comparison of the northern hemisphere surface air temperature, mean sea level pressure and upper tropospheric winds for March 1962 with those for March 2013 show a remarkable resemblance.

Potential drivers

There are a number of factors that may predispose the climate system to negative NAO states in early spring and the prevalence of easterly winds over the UK, leading to anomalously cold conditions. Several of these were in force during the 2013 spring (and also in 1962).

A number of potential drivers of the cold weather are discussed in more detail in the Why was the start to spring 2013 so cold? longer briefing note (PDF, 896 kB) , including

  1. weather patterns in the Tropics - The influence of weather patterns in the tropical Pacific on the phase of the NAO in winter is well known. Particularly relevant for this year is the behaviour of the Madden Julian Oscillation (MJO). When the MJO is in its active phase over Indonesia and the West Pacific, it tends to drive negative NAO conditions 2-3 weeks later. From late February into March, during its time over the West Pacific the MJO was particularly strong, probably having a significant impact on the northern hemisphere circulation. It is very likely therefore that the MJO played a role in this year's cold spring, just as it probably did, but for different reasons, in the remarkable transition from dry to wet conditions over the UK in spring 2012.
  2. the Stratosphere - There is now a substantial body of evidence to show that during winter and early spring, sudden stratospheric warming events in the upper stratosphere over the North Pole can influence surface weather conditions over the UK some 2-3 weeks later. The easterly winds in the upper stratosphere 'burrow' down through the atmosphere to affect the jet stream and surface climate. The result is a switch from a mild westerly Atlantic flow over Europe to easterly winds with an increased risk of cold extremes. The appearance of a sudden stratospheric warming provides a 'window of opportunity' for monthly forecasts to warn of increased risk of blocking weather patterns and the development of cold, easterly conditions over the UK.  A significant stratospheric warming event developed at the beginning of January 2013 (Figure 5), with the reversal of the normal westerly winds to easterly winds in the upper stratosphere.
  3. conditions in the North Atlantic - The climate of the North Atlantic and Europe continues to be influenced by the positive phase of the Atlantic Multi-decadal Oscillation (AMO), which describes long period fluctuations in the sea surface temperatures of the North Atlantic. The warm phase of the AMO is characterised by above normal temperatures across the North Atlantic, but with a pattern of particularly warm water down the Labrador Sea and south of Greenland. This pattern was observed in 1962 and 2013, and both years lie within the positive phase of the AMO. Whilst the AMO has been linked to large-scale precipitation changes, most notably over the Sahel, the USA and Brazil, and to the activity of the Atlantic hurricane season, its influence on our winter climate is unclear. This is because variability in the North Atlantic Oscillation dominates and tends to imprint itself on the pattern of North Atlantic sea surface temperatures. However, there is some evidence that the changes in Atlantic sea surface temperatures associated with the AMO, dispose the circulation to give drier than normal spring conditions over the UK and northern Europe.
  4. the state of the Arctic - There is no doubt that the climate of the Arctic is changing and this can be seen in the seasonal cycle of sea-ice extent (Figure 6). Since 2007 and the unprecedented loss of sea ice extent that summer, the Arctic appears to have entered a new regime where although the ice fills in each winter, it is lost again very rapidly in summer. Preliminary and ongoing research at the Met Office Hadley Centre is providing increasing evidence that the loss and thinning of Arctic sea ice predisposes the winter and spring atmospheric circulation over the North Atlantic and Europe to negative NAO regimes, as was experienced at the start of this spring.

Time series of the vertical profile of the daily mean winds at 60N from the surface to the upper stratosphere, through the winter of 2012/2013. The upper panel shows the actual mean wind and the lower panel shows the anomalous wind. Yellow/red colours de Figure 5: Time series of the vertical profile of the daily mean winds at 60N from the surface to the upper stratosphere, through the winter of 2012/2013. The upper panel shows the actual mean wind and the lower panel shows the anomalous wind. Yellow/red colours depict westerly winds and blue colours depict easterly winds. Anomalies in monthly Arctic sea ice extent (blue line) and actual September sea ice extents (brown stars) from 1979 to present. Figure 6: Anomalies in monthly Arctic sea ice extent (blue line) and actual September sea ice extents (brown stars) from 1979 to present.

Concluding remarks

March 2013 was exceptionally cold in the UK, as well as the North Atlantic and European region more generally, in the context of the last 50 years. Such climate 'events' lead to increased interest from the public, media, government and businesses in both the impacts of the weather on our livelihoods and infrastructure, and in the drivers of significant weather.

As is ever the case, the conditions that led to a cold March are linked to a number of different and often inter-related factors. This can also be said of the cold winter of 2010/11, the UK drought in 2010/12 and extreme summer rainfall in 2012. This makes it difficult to definitively attribute a particular 'event' to one simple explanation, which can make communicating the science drivers more complicated and nuanced than some audiences may wish. On the other hand, this simply reflects the richness and complexity of our climate system, which drives the weather that we experience on a daily basis.

Whilst the cold March 2013 weather is certainly unusual, it is not unprecedented or outside the expected natural variability of our climate. There is particularly heightened interest in the role of the Arctic on the UK's weather, given rapid changes in Arctic sea ice, and on the likely changes we may observe given future decline. It is worth re-emphasising, however, that while changes in the Arctic are consistent with predisposing the climate system to cold weather in northern Europe, this is only one possible driver among several potential factors which could account for the cold March weather. What we have still to understand is the degree to which our changing climate may alter the likelihood and intensity of extreme events. With the rapidly changing Arctic, this is now high on the research agenda.    

Last updated: 14 April 2014