Should the same thing happen 100 years ago, it would be, roughly, 29 C SST over average of 25 C (an average for a previous long term observation period for the same location and season). (realclimate.org)
Earthling - Long term observations show continuing acceleration of SLR, with shorter term (including multidecadal) variation superimposed. (skepticalscience.com)
The only way the models can be wrong (in direction) is if they have grossly underestimated natural variability, and a global cooling trend is established in the long term observations. (realclimate.org)