Search All of the Math Forum:

Views expressed in these public forums are not endorsed by NCTM or The Math Forum.

Notice: We are no longer accepting new posts, but the forums will continue to be readable.

Topic: a problem with temperature records
Replies: 7   Last Post: Apr 15, 2013 8:05 PM

 Messages: [ Previous | Next ]
 Axel Vogt Posts: 1,068 Registered: 5/5/07
Re: a problem with temperature records
Posted: Apr 15, 2013 8:03 PM

On 15.04.2013 14:50, JohnF wrote:

> > josephus wrote:
> > > I have some data sets ENGLAND 1729-1970, NOAA 1880-2012 ?GISS ==
> > > NOAA data. I even have a data set for ICE but that is not my
> > > problem.
> > > the problem is when I run a least squares algorithm over the data
> > > set. should I change the extent, the slope increases (??).
> > > the big question is WHAT KIND of PROCESSING caused the RAW DATA to
> > > show irregular slopes like these
> > > josephus

> >
> > I tried to understand you question - but I am a bit lost
> >

> Me, too, a bit lost. But there are lots of geostatistical
> techniques to accommodate lots of less-than-obvious behaviors,
> correlations, etc, e.g., google the term kriging and/or see
> http://wikipedia.org/wiki/Kriging
> You shouldn't be surprised seeing unexpected statistical
> artifacts just using straightforward least squares.
> You've probably got lots more literature work ahead of you
> before deciding how best to reduce, analyze, etc your data.

No, I do not want to do that, but want to know what you do
(and why only for each 2nd year?).

Roughly one can say: data have 3 periods, are always a bit
oscillating. And can never be approximated by a 'line'.

http://www.axelvogt.de/temp/temp_data.xls

Date Subject Author
4/13/13 Josephus
4/13/13 Axel Vogt
4/13/13 Herman Rubin
4/15/13 JohnF
4/15/13 Herman Rubin
4/15/13 Josephus
4/15/13 Axel Vogt
4/15/13 Herman Rubin