When the numbers were counted several prominent observers of climate claimed that the year 2014 was the hottest ever. For example, Scientific American reported:

It’s official: 2014 has taken the title of hottest year on record. That ranking comes courtesy of data released Monday by the Japan Meteorological Agency (JMA), the first of four major global temperature recordkeepers to release their data for last year.

The climate skeptic community responded with a critique. For example, Lewis Page (an author who has often reported climate skeptical views) in The Register argued (my emphasis):

But in fact, detecting a global average temperature rise - of less than a degree since the 1880s, as all sides agree - among thousands upon thousands of thermometer readings from all over the world and spanning more than a century is no simple matter. The temperature at any given location is surging up and down by many degrees each day and even more wildly across a year. It can be done, across a timescale of decades, but trying to say that one year is hotter or colder than the next is to push the limits of statistics and the available data. This sort of thing is why the battle over global temperatures tends to be so hotly debated.

They argue that, when measurement uncertainty is taken into account, we can't clearly say that 2014 was the warmest. The source is what sound like a fairly clear statement from the (not recently noted for their climate skeptical position) Berkeley Earth Surface Temperature project (BEST) (my emphasis):

Our best estimate for the global temperature of 2014 puts it slightly above (by 0.01 C) that of the next warmest year (2010) but by much less than the margin of uncertainty (0.05 C). Therefore it is impossible to conclude from our analysis which of 2014, 2010, or 2005 was actually the warmest year.

Clearly climate skeptics will want to spin this a different way from climate campaigners. But is their interpretation fair given the way disinterested scientists should quote statistical results?

  • What sort of confidence level are you looking for? As you have found, at the 95% CI that the Berkeley study used: no, you cannot conclude that 2014 was hotter than the previous 2010/2005 studies. At 50%, maybe. I can't find the data. – jmabs Jan 25 '15 at 20:46
  • @jferr Commonly accepted statistical standards would be a good start. – matt_black Jan 25 '15 at 21:01
  • Please focus this. Are you asking whether the data is correct or its interpretation? The second question cannot be settled here. – Sklivvz Jan 26 '15 at 8:41
  • @Sklivvz I don't think there is any disagreement about the data: the question is does the data justify the claim that 2014 was the hottest year ever. That isn't about opinion but a factual question about the interpretation of statistics which deserves a clear answer so skeptics can test whether what headlines say is reasonable or not. – matt_black Jan 27 '15 at 22:43
  • 1
    @matt_black This is the wrong site for that question, as it is only about interpreting numbers (or the maths). In fact, the fact that it's about climate change is irrelevant to the answer. Please try on Cross Validated. :-) – Sklivvz Jan 27 '15 at 22:46

Browse other questions tagged .