With what may be the warmest year in centuries about to close, I thought it would be fun to have a graphic comparing the march of global average temperature over several years about a century ago with the present state of affairs. This graphic is based on NASA’s data, using John Abraham’s estimate for the 2014 temperature (it might end up being a tiny bit different). There is more information about those sources here.

[click on the graphic to get to a larger version]
Just to be clear on how to read the graph … the red dot is not anywhere in particular on the horizontal scale. The X and Y axis simply plot global average temperatures estimated for 1895 to 1933, a series of years that has 1914, a century ago, in the middle of it. This early sequence of data is meant to represent “pre-industrial” temperatures, and here that is compared using the single red to today, positioned correctly on the vertical scale (of temperature). Note, however, that 1895 to 1933 is not really pre-industrial. Human produced greenhouse gases were already being added to the atmosphere by then, though not to the same degree as more recent decades.
You will hear people say that even if 2014 is the warmest year on record, that it is not statistically significantly warmer than the next warmest year. That is absurd. One would have to have a very poor understanding of how statistics works to make such a statement non-ironically. But to make the point even more clear than I might if I explained why that is a dumb thing to say, statistically, I produced this graph which shows that today it is much warmer than it was not so long ago.
ADDED: A question has been raised as to whether or not I chose the proper scale on the Y-axis. I did. My intention was to show variation and average temperatures for several decades near the beginning of the industrial period, centering on 100 years ago, and to put the current year in context of that. This graph does that nicely, with no strangeness about axes other than the carefully explained fact that the clearly labeled 2014 datum is not scaled to the time scale on the bottom. The nature and variation of the entire instrumental curve is readily available and there are dozens of graphs here on this blog and elsewhere that show this (I placed one at the top of the post for your convenience). The point of this graph was to remove the ascending values and obviate the rather absurd question of statistical difference between the highest and second highest ranked years. As explained.
But the Y-axis problem emerges as a more general climate science denial meme (other than, and beyond, the relatively valid and honest question of how to best scale the Y-axis on a graph like this). And in relation to that, I’ve made a NEW ENTRY IN MY FAQ. Please have a look. There are some fun graphs.
Added:
To demonstrate two ways in which people get this wrong. First, an actual scientist type person simply believing (incorrectly) that all scales must start at zero (maybe they do in his field), and second, a climate science denialist actually arguing that the joke graph shown in the tweet is the best way to show global temperature change.
You might have to click on the pic to be able to read it:
