I did it two different ways. In one I simply totaled up all the data over the years by latitude and divided by the latitude data count to arrive at a mean value. That way no individual data point gets any more weight than any other.

The other way I took the mean for each year by latitude, totaled them and divided by the number of years of data for that latitude. That way each year gets the same weight. I’ve posted links to the charts below.

Point of interest. 62% of the data points are in the last 1/2 of the record.

To not bias your perspective, I suggest you decide which way you think would be a more appropriate rendering of the data before looking at it.

]]>I infer that he considers independence of no importance. But my mind goes back to the days when Mr Yule sprang a mine under the contraptions of optimistic statisticians by his discovery of spurious correlation. In plain terms, it is evident that, if what is really the same factor is appearing in several places under various disguises, a free choice of regression coefficients can lead to strange results. It becomes like those puzzles for children where you write down your age, multiply, add this and that, subtract something else and eventually end up with the number of the Best in Revelation.

Keynes and Yule were coauthors in around 1910.

]]>*The government are very keen on amassing statistics. They collect them, add them, raise them to the nth power, take the cube root and prepare wonderful diagrams. But you must never forget that every one of these figures comes in the first instance from the village watchman, who just puts down what he damn pleases. (quoting an anonymous English judge.)*

Wikipedia

I’ll tidy up the various scripts and post on my web site when I get a chance. The result of my attempt at visualizing TSurf1200 and SSTHadR2 combined is available on Google Video.

Enjoy.

Sinan

]]>I assume the data is the same Giss data as that which Steve linked to, just saved

in binary format.

I converted the binary yearly Fortran files into individual monthly text files.

Each file contains 3 columns. Latitude, longitude, and anomaly.

All data was 2×2 degree cell size. 16200 total cells.

From the 1880-2004 Ts files(surface air temperature). 1500 months of data.

1674 cells have one month or less with data. The next lowest count is 187 months.

3217 cells have data for all months. I get an anomaly from 1880-2004 of 0.046C.

From the 1950-2004 LOTI files(land ocean temperature index). 660 months of data.

78 cells have 11 months or less with data. The next lowest count is 148 months.

10885 have data for all months. I get an anomaly of 0.135C.

I calculated the anomaly by keeping a running total of monthly anomalies for each cell

that had 120 months of data. Calculated the mean for each cell. Totaled those values

and divided by the number of cells with 120 months.

If that’s not the correct way to do it, someone please speak up and clue me in.

I don’t have great statistical skills.

Can’t say I’m surprised by the color of LOTI image. The data starts during the cold part

of the 20th century. Still nothing extraordinary.

I created a couple maps of the data.

Colored coded: yellow > 0.5C. red > 0.0C. Green 0 to -0.5C. Blue LOTI 1950-2004

TS 1880-2004

Full-size 3600 pixels wide.

LOTI 1950-2004

TS 1880-2004

I certainly don’t have your R-skills… so I’m still fiddling with the re-collated data part (GHCN Station + HadCRUT2 *.tab).

The rest however is a very straightforward thing to manage … Good work!

If someone is interested, there is a Scandinavian (up to 2002) data collection available here:

http://www.smhi.se/hfa_coord/nordklim/

A collection of excellent Denmark/Greenland data sets is available from

http://www.dmi.dk/dmi/index/viden/dmi-publikationer/tekniskerapporter.htm