posted on Dec, 17 2007 @ 07:50 AM
Indy, you are showing your lack of statistical knowledge here. That line provides a regression analysis of the temperature data, it's not 'connect
any two dots'.
When you say the trend is downwards from 1998, the only way to assess this correctly is using regression statistics. When such stats are used, we
still see an upwards trend, it is not 'flat'.
The reason that the data before 1998 has been left out is because the argument is that 'since 1998, the trend is down'. So, to assess this, we use
the data from 1998+, this clearly shows the trend is still upwards. The data from before 1998 is irrelevant for your argument.
This is completely ignoring the fact this is just one big cherrypick, using an anomalous point as the starting point in any assessment is bad analysis
- and 1998 was anomalous. But even when we do, it still doesn't show what you want it to.
If you want to say that 1998 temps were higher than those after, fair enough, this applies for one set of data (Hadley CRUT), but not the other
(NASA-GISS). But you can make no reliable assessment of a trend without some sort of statistics.
With such data, the longer the data series, the more reliable the analysis. Yeah, pick 2002 to 2003; forget that, lets pick august to december; no
scrap that, lets do yesterday vs. today...
Yesterday was warmer than today, therefore teh glowb iz kewling!
Even if we find several years of downward trend, it doesn't mean that the science is wrong, we expect noise in the data, we expect natural
variability to be present.
[edit on 17-12-2007 by melatonin]