Mime-Version: 1.0
Message-Id: <p06020411c9777084402a@[129.116.190.133]>
In-Reply-To: <20110208174120.B4B72B754F@santafox.mns.ru>
References: <20110208174120.B4B72B754F@santafox.mns.ru>
Date: Tue, 8 Feb 2011 17:42:06 -0600
To: malkin@gao.spb.ru, zmalkin@zmalkin.com, ids.analysis.forum@ids-doris.org
From: "John C. Ries" <ries@csr.utexas.edu>
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Subject: Re: [ids.analysis.forum] TR : Long term trends in the gravity field
  (forwarded)

At 8:41 PM +0300 2/8/11, Zinovy Malkin wrote:
>Content-type: text/plain
>Content-description: Mail message body
>
>Dear Colleagues,
>
>It worth mentioning that the new Conventions model for the mean pole seems
>to have a bias w.r.t. actual pole motion. In the plot attached one can see
>IERS C04 data (black line), mean pole table reported by IERS (green line,
>mainly masked by blue line), old and new Convention models, and PUL mean
>pole motion computed at Pulkovo Observatory (red line).  The latter can
>be found at http://www.gao.spb.ru/english/as/persac/meanxy.dat.  It would
>be interesting to try it in your test if feasible.
>
>Regards,
>Zinovy

Hi Zinovy,

I would argue that the agreement between your mean pole and what was 
adopted for IERS2010 is actually quite good, considering the size of 
the periodic variations that have to be filtered out first. You may 
recall that when we looked at fitting the mean pole for the updated 
conventions, there were a variety of different fits to the time 
series, as each analyst had different ways of filtering out the 
periodic variations to arrive at a 'mean' pole. In the end, I 
recommended Daniel's solution primarily because it had the best 
backward compatibility with the older conventions (at least over the 
1976-1999) span.

If someone is able to test the sensitivity to a different mean pole 
model, it would be an interesting result. It might help us decide how 
accurate the mean pole needs to be for the next Conventions update.

Best regards,
John


                           Cliquez sur l'url suivante 
https://www.mailcontrol.com/sr/TTJ99pZmSU3TndxI!oX7UixArsuB8jk7lw!rUxXtPscEEhsnfAQJTazezSADyPya3HoUXzjm!9UsWPB8dn7asg==  
                    si ce message est indésirable (pourriel).