Quote:
Originally Posted by Argonavis
"excessively large" means it doesn't show the few tenths of a degree smoothed "average" temperature variations to a sufficiently scary scale?
If you would like to draw trend lines, they are down. |
Now that I understand how the baselines have been derived let me respond to this.
It reinforces the argument about the pitfalls of looking for trends in data that is dominated by noise. The problem is also magnified by using too short a time frame for analysis.
Here is an example using the UAH data.
http://www.woodfortrees.org/graph/ua.../to:1999/trend
As illustrated I can select a time frame and come up with any trend I want.
So why select the period 2001-2009 in the first place?
The fact that climate change is a long term effect one should select a longer time frame for analysis.
http://www.woodfortrees.org/graph/ha...plot/rss/trend
The trendlines now show a completely different picture. Even though the graph is dominated by noise the increasing temperature trend is apparent.
Steven