It is a very simple calculation on an Excel spreadsheet to find out the earliest date from which the slope of temperature change to the present is zero or less. Using the latest
HADCRUT4 data set it turns out to be November 2000. Prior to that date, they are all positive. Since 1850 the slope is a positive 0.0046 per year which adds up to a 0.75°C rise in temperature since that date. Since November of 2000 (2000.83) the slope is a negative -0.00020 which adds up to a drop in temperature of minus -0.0025 since then. The 2013.08 anomaly would have to have been at least 0.86 before that earliest date becomes more recent. Here are the pertinent rows from my Excel spreadsheet:
row#; yyyy.xx; anomaly; slope
...
0001; 1850.00; -0.695; 0.0046
...
1811; 2000.75; 0.209; 0.00063
1812; 2000.83; 0.151; -0.00020 [=SLOPE(B1811:B$1958,A1811:A$1958)]
...
1957; 2013.00; 0.432; 0.60
1958; 2013.08; 0.482
What does it mean?
Temperatures have been flat since November 2000.
[Posted over at Open Mind' and they tell me, "Your comment is awaiting moderation." we'll see how open they are. ]