- Joined
- Oct 17, 2007
- Messages
- 11,862
- Reaction score
- 10,300
- Location
- New York
- Gender
- Male
- Political Leaning
- Centrist
Re: And now it's global COOLING! Record return of Arctic ice cap as it grows by 60% [
Very badly flawed.
For example, here's one section (pp.9-10):
For the first year shown in the figure, 1880, there are two dots: one red, at -0.10 °C; one blue, at -0.36 °C. The two dots indicate the 90%-confidence interval for the global temperature in that year: [-0.10 °C, -0.36 °C]. The other dots indicate similarly for the other years. A straight line that was fit only via observational uncertainty would have to lie below almost all the red dots and above almost all the blue dots. Such a line obviously cannot exist. Hence, it is not possible to fit a straight line based only on observational uncertainty.
If one doesn't understand confidence intervals, this seems convincing. However, a basic understanding of confidence intervals easily dismisses this argument. A 90% confidence interval is + or -1.645 sigma. Hence, from the differences in the upper and lower bound of the confidence interval (blue and red dots), one can determine the value for sigma and also the central or mean value. A trend line would be based on the central or mean values.
And all the major global data sets (HadCrut, GISS, NCDC) have specific values for each year. It is from those values that the standard deviation and confidence intervals were calculated.
As a matter of fact, if one didn't know the central values, one could not have known that the differences represented 90% confidence intervals.
And the author's 22 page critique?eace
Very badly flawed.
For example, here's one section (pp.9-10):
For the first year shown in the figure, 1880, there are two dots: one red, at -0.10 °C; one blue, at -0.36 °C. The two dots indicate the 90%-confidence interval for the global temperature in that year: [-0.10 °C, -0.36 °C]. The other dots indicate similarly for the other years. A straight line that was fit only via observational uncertainty would have to lie below almost all the red dots and above almost all the blue dots. Such a line obviously cannot exist. Hence, it is not possible to fit a straight line based only on observational uncertainty.
If one doesn't understand confidence intervals, this seems convincing. However, a basic understanding of confidence intervals easily dismisses this argument. A 90% confidence interval is + or -1.645 sigma. Hence, from the differences in the upper and lower bound of the confidence interval (blue and red dots), one can determine the value for sigma and also the central or mean value. A trend line would be based on the central or mean values.
And all the major global data sets (HadCrut, GISS, NCDC) have specific values for each year. It is from those values that the standard deviation and confidence intervals were calculated.
As a matter of fact, if one didn't know the central values, one could not have known that the differences represented 90% confidence intervals.
Last edited: