sci.energy

In article , cornelius@ (George Cornelius) writes:
>> , writes:
>>>
>>> "Rolf Martens" < @ > wrote
>>>> or K warming per century, that still a very silly invention,
>>>> in particular considering that since 1998, there is no more warming at
>>>> all.
>>>
>>> No warming at all? You are a liar. Here are global temperatures plotted
>>> over that period.
>>>
>>> Note the linear increase in global average temperatures over that time as
>>> shown by the best fit line marked by "o"'s
>>>
>>> 1998 *********************o*****
>>> 1999 *****************>>>>o
>>> 2000 *****************>>>>>o
>>> 2001 ************************o
>>> 2002 *************************o**
>>> 2003 **************************o*
>>> 2004 *************************>>o
>>> 2005 *****************************o**
>>> 2006 ***************************>>>o
>
> Of course, if you would get rid of all the ascii art gobbledegook
> and the best fit junk, we could look at the chart and judge for
> ourselves. A few more years of context would be helpful, as well.
>
> My guess is he chose an outlier year as a starting point, skewing
> results in the negative direction for some time to come.

Let's run your graph through a filter so we can all read it:

$ awk <<__HERE__ \
> '{printf("%s %s ",$1,$2); i=($2-14)*50; i=int(i); while (( --i > 0 )) printf("*"); printf("\n");}'
> 1998 *********************o*****
> 1999 *****************>>>>o
> 2000 *****************>>>>>o
> 2001 ************************o
> 2002 *************************o**
> 2003 **************************o*
> 2004 *************************>>o
> 2005 *****************************o**
> 2006 ***************************>>>o
> __HERE__
1998 ***************************
1999 ***************
2000 ***************
2001 ***********************
2002 ***************************
2003 **************************
2004 ***********************
2005 ******************************
2006 *************************

Good luck fitting a straight line to that!

Now, if 1998 is an outlier (could be determined from
earlier data), than your trend line looks reasonable.

Still, the curve might be peaking in 2002.

If 1999 and 2000 are the outliers, then I would say
you are the one in error.

Not that seeing trends on this short a sequence of
data is anything other than reading tea-leaves.

--
George Cornelius cornelius(at)
cornelius(at)