Monday, November 06, 2006

Perfect Beach Weather on a Sunday Afternoon

So how was your weekend? Was a bit chili down in Melbourne, but that’s what it’s famous for – the cold and a large variability of weather, not just long term, but also in the one day.

But what about at 3pm?
The graph on the left shows Australia’s deviations from the average for temperatures taken at 3pm. It shows some interesting things. We had a big hot spell from 1875 to 1890 and have also had a recent increase in temperature in the last 5 yeas. All the time in between suggests no increase or decrease in temperature. This is confirmed with statistical analysis that there has been no significant increase or decrease (F = 3.41, p = 0.067). Interesting, that is wasn’t far off proving a long term decrease in temperature. Even if we have just the data from 1910 we still see no significant increase. (F = 0.81, p = 0.42). So at 3pm temperatures are not increasing. In fact the early and recent increases matches our analysis of Australian wide maximum temperatures which would occur at about the 3pm time in some parts of Australia.

Hey but what about the recent increases, doesn’t that prove that we are warming up? Indeed, it looks as though we have warmed at 3pm Australian wide by on average about 0.5 of a degree. But this can be attributed simple to random variation. Co2 levels were not what they are now in the 1880s, and even should we choose to ignore such old data, we still proved with a statistical test, that there is no evidence that temperature has increased significantly.

So lets recap. Have Australia’s temperatures significantly increased at 9am? No. Have Australia’s temperatures significantly increased at 3pm? No. So so far we conclude that Australia just isn’t heating up at all. 9pm data tomorrow

No comments: