Would a similarly extreme 5 - sigma event in July (average temp. (realclimate.org)
Even if the distribution is fairly close to Gaussian, I would not be surprised if the probabilities for 4 + sigma events to be considerably larger («heavy tails») than Gaussian probabilities. (realclimate.org)
Trying to find a five sigma event not caused by warming is not really relevant to the argument. (realclimate.org)