I have no idea how reliable this information is. But if it is correct, and if the observed data all lie outside and below the predictions of the models, then it is time for the people who model climate change to change their models.
Note that these are temperatures from the "troposphere". Also see this for comments.