From a nanoengineering reader:

Subject: WRONG Temperature in USA Report
Dear Ms. Lopez:
As can be seen in this chart and many others like it found on the web.
global mean temperature has been between -0.4 C and +0.4 C or approximately 32 F over the years 1860 to the present.
I have no idea what the USA Today chart is depicting that reports temperatures in the range 56.4 F and 57.7 F, and it is certainly not clarified in the article.

UPDATE:

Another e-mail:

In reference to your email from a “nanoengineering reader,” please note that the referenced graph refers to deviation from the mean, not the mean. The globe’s mean temperature does not range -0.4 to +0.4 C. I hope this engineer never designs anything that I’ll come into contact with, because he (and you) seem to be able to read the simplest displays of data.

And another:

Aside from your skepticism (which I share) over the accuracy of global temperature measurements in 1860, there is another huge problem with the graph which you reproduced from USA Today. It is enormously distorted to show a steep incline in temperature when, in fact, the incline is, in absolute terms, really quite modest.
You can make any increase or decrease look either large or small by the parameters that you use for the measurements on the vertical axis of the graph. In the graph used in USA Today, someone has arbitrarily chosen to only graph temperatures between 56 Degrees and 58 Degrees. They did that, I’m sure, for the sole reason of making it look like a dramatic spike. If they had chosen 0 to 100 (or even 0 to 58), the graph would have looked virtually flat.
Anyone who works in a large bureaucracy knows this game. It’s the game people play at budget time, to show spikes in their Department’s activity or similar phenomena to try to justify the budgetary increase for which they are asking. Where I work, we’ve started requiring all graphs to start with a baseline of zero in order to stop this kind of thing, because it is just plain dishonest.