cc: deparker@meto.gov.uk
date: Thu Oct 19 15:27:53 2000
from: Phil Jones <p.jones@uea.ac.uk>
subject: Re: NATURE PAPER
to: c.folland@niwa.cri.nz


 Chris,
    I don't think the method of calculating monthly mean temps has
 changed dramatically over the 20th century. English speaking countries
 tended to use the mean of min and max, while others used the means
 based on fixed hours. Canada, USA, India, Pakistan, Australia, NZ,
 Britain and South Africa and a few other countries have not changed at
 all. The difficulty will be finding enough data to assess what the
 uncertainties of the (max+min)/2 method are. To do this we'll need
 anomaly time series for both methods, hence my suggestion of using
 model runs. 
    Only a few small countries have changed to (max+min)/2. In many
 others there seem to be a few climatologists who were listened to, who
 said there would be homogeneity problems if changes were made.

 Cheers
 Phil



At 09:41 AM 10/19/00 +1300, c.folland@niwa.cri.nz wrote:
>Phil
>
>I agree. There is a possible issue. If the balance of the methods of 
>calculating the average globally changed, or kept changing globally, 
>then a residual error or uncertainty would arise. Any thoughts?
>
>Weather here much better now so I may book up a trip for Saturday 
>to see the Kauri forests and the West Coast.
>
>
>Chris
>
>
>
>On 18 Oct 2000, at 15:03, Phil Jones wrote:
>