by Alan Longhurst
Because the climate change science community habitually concentrates attention on surface data from a very short recent period – nominally a little more than 100 years – it would be very interesting to know how the pattern habitually derived from these data compares with longer data archives that have been processed independently by the observing nations.
In just a few parts of the world there exist meteorological agencies of high competence that have inherited the observations of their predecessors, processed them in acordance with modern requirements, and made them available to the community through the GHCN system operated by NOAA.
So a small investigation was made of the reliability of some early instrumental data by comparing GHCN2 data for three central European cities at which air temperature measurement extends back over a 225-year period: the observations were made and processed by different national agencies: Czech, Slovak and Austrian. They exhibit a uniform pattern of temperature change over a large part of central Europe that which gives confidence in the precision of the measurements themselves and, by inference, in similar data from other reliable national meteorological agencies.
Throughout this long period, each station responded similarly to major changes in rates of warming or cooling although, as might be expected from their geographical position, temperature changes at Prague and Vienna were closer to each other than either was to Budapest, farther to the east on the Hungarian plain. This result suggested that it would be useful to assemble a larger file of major European cities from the GHCN2 archives, so 21 such were selected (together with the Central England data, actually centered on Oxford) being grouped for use into four regions: Western, Scandinavian, Central and Eastern. Useful data are sustained back to 1700, and demonstrate the evolution of SAT in the region from Lisbon to St. Petersburg and from Trondheim to Milan over this long period; this is similar to – but more detailed than – the pattern of the 731 GHCN stations that represent the European region.
The data also conform to (and confirm) the trend of air temperature for this region that is indicated by proxy data, and many of the features in this record had impacts on human affairs that were described in historical records. Some of these were associated with a recurrent cooling and incursion of arctic air that is a characteristic feature of the European climate that may be sustained over several years: at least eight such incursions are recorded in these data.
Some of the worst European famines have been associated with these cold events, of which at least one, that of 1816 “the year without a summer”, has been historically associated with a major volcanic dust veil from the explosion the previous year of the magma chamber of Mt. Tamboura in Indonesia.
But the instrumental data show that 1816 was no colder in Europe than the previous several years and that it was the final year (see the above plot) of an incursion of cold air over Europe (indicated as #4 in the previous plot). So, although a sulphurous dust veil undoubtedly occurred, especially over North America, the cold European summer of 1816 was probably not the consequence of Tamboura, but of the state of the atmospheric circulation over northern Europe. So it is plausible that the series of the eight cold incursions that have marked European climate during recent centuries were all similarly forced, and have little or nothing to do with the more popular suggestion that at least some were the consequence of a volcanic dust-veil, but may be associated rather with trade wind failure over the Pacific Ocean.
The very similar cold incursion of 1940-43 has been analysed in detail, and no volcanic influence is required for its explanation; this event attracted special attention because of its influence on recent European history. The planners of the German invasion of Russia in 1941 were apparently not advised by their meteorologists that eastern Europe was then in the third year of an incursion of Arctic air, and the consequent problems faced by the German army in northern Russia that winter were perhaps the first steps towards their eventual defeat.
The bitter cold in eastern Europe was caused by a persistent 24-month global climate anomaly that was coincident with the strong and prolonged Niño episode of 1939-1942; this was an intense example of a now well-understood phenomenon that links atmospheric processes over the two northern oceans through changes in the Hadley circulation and in Rossby wave generation.
In 1942 an intensified Aleutian low induced warm anomalies over Alaska, while a pattern of weak Icelandic low and strong Scandinavian high led to an unusually persistent flow of cold arctic air over eastern Scandinavia. The local mechanism was apparently typical: a high pressure ridge over the eastern Atlantic and a trough over western Russia were present throughout the winter 1941-42 and steered the migratory cyclones approaching Europe to the north of Scandinavia or into the Mediterranean. Such conditions, even temporary, are well known to bring frigid arctic air down across eastern Scandinavia and over western Russia and are associated with the strength of vorticity in the mid-latitude jet stream (see below).
To what extent we may assume that each of the cold incursions noted above in the historical European temperature records were similar in origin is not clear from the historical occurrence of Nino-type events which suggests a greater frequency than the 8 cold incursions suggested from the figure on p. NN. But when Napoleon faced Moscow in 1812, he did so in the first year of another European cold wave, while the 1789 revolution was at least partially the consequence of crop failures in France during the 1780s cold spell that led to unrest in the countryside and hunger in the towns.
These events have been associated with agricultural crises, famines, civil disturbance and migrations: the great Irish potato famine (when a third of the population starved because the tubers froze in storage) occurred during the 1740s cold incursion.
Much of European history has been marked by the effects of such irruptions of arctic air into a region that is habitually under the influence of warm Atlantic air – to which, consequently, European agriculture and economy has been adapted. But at the end of the 20th century, an anomalous and very rapid warm shift in surface temperatures occurred that has been described as a “jump” in the temperature record. Over just a 3-year period from 1987-1990, SAT anomalies inceased rapidly over about a full degree.
Regional SAT was maintained through to the end of the record in 2014 at a higher mean temperature than had been recorded during the previous century. One may choose to ignore it and simply draw a trend line from 1890 to 2015 – or one may choose to interpret the record differently, as here: both positions are valid and in the present state of climate science your choice will largely depend on your confidence in the reliability of simulation modelling of complex Earth systems.
The flowering dates of plant communities in Britan, analysed for their response to long-term change in the Central England surface air temperature record, responded closely to this regime shift.  The fit between ambient temperature and flowering dates (both at community level and for individual species) is excellent over each 25-year segment of the entire record back to the 1750s. The series terminates in a very clear 15-day advance in the dates of community flowering after 1985 that was maintained to the end of the record in 2008.
Such a rapid change in surface air temperature over this large region is compatible neither with anthropogenic nor with volcanic forcing, but is consistent with the expected result of an equally major and rapid change in the distribution of atmospheric pressure over the entire North Atlantic-Arctic region.
This is indicated by change in the values of both the wintertime NAO and the Arctic Oscillation (northern annular mode of Hurrell) which together describe the state of the polar vortex north of the mid-latitude jet stream; when polar surface pressure is low (positive AO index) this flows strongly and consistently, with relatively weak meanders, so that cold polar air tends not to intrude down into mid-latitude Europe. But when polar surface pressure is high (negative AO index) the jet stream weakens and meanders more strongly, so that cold polar air is routinely carried down into mid-latitudes. Because of the existence of the western mountain ranges in North America that perturb its flow, the jet stream has a preferred number and location of southerly waves appropriate to each state of the AO. Periods of strongly negative AO are, in western Europe, associated with irruptions of cold polar air, as occurred rather commonly in the period 1935-45, discussed above.
Major warm excursions in surface air temperatures on the Japanese islands have already been noted for these same years (see plots for two rural stations on p. NN) and although these excursions were brief and cooling set in after only a single peak warm year, they were peraps related to the same rapid change in the value of the Arctic Oscillation after 1985. Changes in the AO has consequences for the strength of the winter westerlies that bring cold air from central Asia down across the Japanese islands, affecting winter temperatures generally in East Asia; this effect is modified by the strength of western Pacific cyclonic activity, and the 1985-1990 warm event over Japan (see p. NN) appears to have been the result of complex interaction between these two processes.
 Papineau, J.M. (2001) J. Climatology 21, 1577-1592
Le Mouel, J-L et al. (2008) C.R. Geoscience doi.10.1016/j.crte.2008.06.001 and see also “Nouveau voyage au centre de la terre” (Odile Jacob, 2009),
 Amano, T. et al. (2010) Proc. Roy. Soc. B, 277, 2451-2457
 Park, H.-J. and J.-B. Ahn (2016) Clim. Dyn. 46, 3205-3221.
Moderation note: As with all guest posts, please keep your comments civil and relevant.