Date: Mon, 05 Feb 2007 15:20:17 +0000
To: hashemi@brandeis.edu
From: Phil Jones <xxxxxxxxxxxxx>
Subject: FW: Raw Data
  
Dear Kevan,

	The gridded data are available on our web site.
If you click on data then on temperature, you can download
the grids as well as the hemispheric/global averages. I'm
not sure why you would want the raw station data, warts
and all. You would need to develop methods to weed out
the outliers that come in every month, weed out the
stations that have jumps/trends relative to their neighbours.

  I have talked to a number of statisticians at meetings,
telling me there are better ways of analysing the data. I even
went once to a seminar series at the Isaac Newton Institute
in Cambridge. What most don't like is the initial subtraction
of the 1961-90 means to get anomalies/departures. If you've
read the papers, you will be able to understand why we do
this.

   We've made things easier for climatologists around the
world. None of these ask for the raw data (as they know the
problems). It is always those in other disciplines.

  Apologies if this reply has come across the wrong way, but
you can get a dataset from either

NCDC in Asheville - there is it called GHCN. There should be 5000+
stations. You will need to figure out how to put them together in
single time series. It is possible, just difficult. They have a raw
and an adjusted version.

NCAR in Boulder have a dataset - within their many sets. I don't
have the number, but look for global, monthly, climate, stations.
With this one you will get many other variables as well. The format
of this isn't very user friendly - well it didn't used to be.

Cheers
Phil