What you notice right away, is that there are a lot more of the big red dots than blue (greater anomalies for heat than for cold), and of the big red and blue dots, the red dots are much larger.
Which leads me to a discussion I had once, with a co-worker. I explained that, if there was no global warming, you wouldn't expect to see heat records breaking with this large frequency in the current time period, as it would be rare that you exceeded the extremes that had been set during the previous 100 years. And in fact, one would expect that over more time, if things were relatively constant - even with cycles - you would have even fewer, not more, records broken. After all, if you have enough outliers, they no longer are outliers, are they?
But she was unconvinced, saying that, over time, you wouldn't expect less variation than in the past.
I think I understand what she's saying, that it's still a standard distribution curve. But of course, as those outliers become numerous, the distribution curve is redrawn to incorporate the outliers, such that even though the shape of the curve looks the same, the location of the mean and median has changed, or in fact, the shape of the curve altogether has changed from that of a standard distribution.
So hasn't someone come up with a diagram showing the distribution of temperatures of the first 50 years compared to the last 50 years?
Which leads me to a discussion I had once, with a co-worker. I explained that, if there was no global warming, you wouldn't expect to see heat records breaking with this large frequency in the current time period, as it would be rare that you exceeded the extremes that had been set during the previous 100 years. And in fact, one would expect that over more time, if things were relatively constant - even with cycles - you would have even fewer, not more, records broken. After all, if you have enough outliers, they no longer are outliers, are they?
But she was unconvinced, saying that, over time, you wouldn't expect less variation than in the past.
I think I understand what she's saying, that it's still a standard distribution curve. But of course, as those outliers become numerous, the distribution curve is redrawn to incorporate the outliers, such that even though the shape of the curve looks the same, the location of the mean and median has changed, or in fact, the shape of the curve altogether has changed from that of a standard distribution.
So hasn't someone come up with a diagram showing the distribution of temperatures of the first 50 years compared to the last 50 years?
No comments:
Post a Comment