Observational Analysis

Central England Temperature

The Central England Temperature (CET) record (Parker et. al. 1992) is the longest instrumental record of temperature in the world, going back to 1659. For the analysis of extremes we only use data from 1920 onward, as the variability is significantly larger before that date. This provides a useful dataset to investigate the anthropogenic influence on temperatures such as in King et. al. 2015. It also provides accurate and up to date data against which to compare temperatures simulated by climate models. December 2015 was the warmest ever recorded, at 9.7 ºC, a remarkable 1.6 ºC warmer than the previous record, set in 1974. The deviation from normal was also the largest one in the whole dataset.

1) Connection with global warming.
The Central England Temperature does not show a significant trend in December. As November, January and February do show trends and winter variability is high we ascribe this absence of a trend in December to chance and assume that the background trend is the same for all these months, rising at about the same rate as the global mean temperature. This gives a return time of the value observed in December 2015 of about 400 years, with the lower limit of uncertainty at around 150 years (assuming a normal distribution, which fits the upper tail relatively well).

Figure 3. Return time of the December 2015 Central England Temperature after subtraction of the November-February mean trend, 0.92 time the smoothed global mean temperature.

We conclude that the very significant winter trend has made these high temperatures much more likely, but they still represent an extraordinary departure from the mean with a return time of a few hundred years.

2) Connection with El Niño
In general, winter teleconnections from El Niño to Europe are very weak (eg, van Oldenborgh et al, 2000). The correlation with the Niño3.4 index of the strength of El Niño (taken from ERSSTv4) gives a just-significant positive influence in December. However, the influence disappears when we restrict ourselves to the period starting in 1920. A very small influence is supported by an analysis of the EC-Earth model, in which the teleconnection in December is r=0.08.

Figure 4. Correlation of the CET against the Niño3.4 index over 1920–2014, showing that the deviations from zero are not significant at 95%.

Precipitation (Northern England) :

The results shown here represent Northern England precipitation for which we used the average of precipitation in three smaller regions, North-west England, North-east England and Southern Scotland.

Figure 5. a) Met Office precipitation regions. We averaged the regions Northwest England, Northeast England and Southern Scotland. The December 2015 precipitation is the highest one in the series (b).

As in the observations for temperature, the most extreme precipitation events in the past occurred in months other than December. Taking only December values may therefore give a misleadingly large return time for the observations of December 2015. We therefore consider all the months with high precipitation in this area, October through January in this section. These months encompass the season of heavy large-scale precipitation in this area and exclude the season of heavy thunderstorms in early autumn.

1) Connection with global warming.
There is a very significant positive trend in October–January northern England precipitation of about 40% up to 2014. This is much more than the increase from the Clausius-Clapeyron relation that describes the maximum amount of water vapour in the atmosphere. For a fixed relative humidity this gives an increase of about 6 to 7% per degree warming. The larger increase in the observations is due to an increase in westerly circulation types (van Haren et al, 2012).

This trend also increases the probability of extreme monthly mean rainfall, see Figure 6. Here we fit a simple statistical model to the observed data: scaling the precipitation with the smoothed global mean temperature and describing the upper tail with a normal distribution, which fits quite well. This gives a return time in the current climate of about once every 200 years, with a 95% uncertainty range that starts at 60 years. In this model the probability of observing such a precipitation extreme would have been extremely small in the climate of 1920.

Figure 6. Extreme value fit of monthly precipitation in northern England. The lines indicate a fit to a normal distribution assuming the distribution scales with the smoothed observed global mean temperature. Red values indicate the climate of 2015, blue lines the climate of 1920. The stars denotes the observed block maxima, shifted up with the fitted trend to 2015 (red) or down to 1920 (blue). The purple line denotes the value observed in December 2015.

2) Connection with El Niño
Statistically, the northern England precipitation has a weak connection with El Niño in December. As with the temperature, the signal disappears in January and February, and reappears in March–April (van Oldenborgh et al, 2000). In the coupled climate model EC-Earth, these correlations are absent (r=0.01), so again our confidence that this is a physical signal and not a coincidence very low.

Assuming the signal is real, the probability of an extreme winter month like the one observed increases somewhat due to the very strong El Niño of December 2015, see Figure 7. However, the change in probability would be much less than the change associated with global warming.

Figure 7. As the previous figure but for the strength of El Niño.