Conventionally the above chart (fig 1) is how we have always thought is the best way to display the vagaries of the CET series; by plotting a fixed yearly bar chart, of either mean CET temperature or anomalies; of course we could add an accompanying table, sorted and ranked to show the warmest years (fig 2).
Another way of getting a measure of just how warm or a cold it’s been over the years, is to plot a chart of 365 day rolling mean anomalies (fig 3), this gets round the problem of only looking at a single ‘annual’ temperature, that’s fixed to the period between the 1st of January and 31st of December. As you can see from the graph, both higher and lower temperatures occur outside these fixed ‘annual’ periods, and the time period between the 6th of June 2006 and the 5th of June 2007 is a case in point. That particular 365 day period produced a mean anomaly of +2.15°C, the highest in the CET series since 1878, and probably since 1659, even though statistically the warmest year occurred seven years later in 2014, with a calendar year mean anomaly of +1.44°C, which is significantly over 0.7°C lower.
At the moment 2017, as of the 1st of June, has a 365 day mean anomaly of +1.23°C, which is a long way short of either the warmest calendar year, or the warmest 365 day period. But if we compare the first five months of the year, as I’ve done in the table below (fig 4), you will see that 2017 is sixth warmest, a little behind both 2007 and 2014. But as the year progresses, the value and ranking for 2007 will fall, because the year gradually cooled. But we already know that 2014 is the warmest calendar year in the series, and this one will be the one to beat. With the gap at the moment between them of 0.2°C, 2017 has it’s worked cut out for it, if it’s going to catch it, in what’s left of this year.