I find personal anecdotes to be unconvincing ('when I was little, it was never this hot,' etc.). These offer no real value because of how unreliable the information is. Memory is notoriously faulty, not to mention perception bias, confirmation bias, etc. which can alter how memories are remembered. It seems to me it's best to stick with data when making any case, and at least the satellite data is still data, even if some people take issue with its accuracy or the method in which it is gathered. Anecdotes are not data that can be quantified and measured, and offer no real value to the topic.
I wouldn't agree with this. It can be faulty but it doesn't have to be necessarily faulty. If I ask 100 random people about weather conditions and they all have similar answers, would you say that is unreliable be cause it comes from personal experience?
I can clearly recall summers and how I had to endure them through years. It is getting harder, end of story.