-------- Original Message --------
Subject: RE: [IVS-analysis] r1139 ($04SEP07XA) Correlation report (fwd)
Date: Fri, 17 Sep 2004 15:08:01 -0400 (EDT)
From: Arthur Niell <aniell@haystack.mit.edu>
Reply-To: Arthur Niell <aniell@haystack.mit.edu>
To: John Gipson <jmg@leo.gsfc.nasa.gov>
CC: Chopo Ma <cma@ivscc.gsfc.nasa.gov>,	Ed Himwich <weh@ivscc.gsfc.nasa.gov>
References: <000101c49cbc$b0227750$1d6bb780@nvi0009>

John,

Thanks for looking into the details. I agree that the losses are small, 
but I
guess what I remember is that, when in days past I was looking at all of the
processing that came through Haystack (or at least Brian and I shared 
it) there
were occasions when a baseline had 100% 7-9s (I ususally sent a message of
congratulations to the two stations), and that it was rare to have more 
than a
few 0 codes spread among all of the baselines.

So much for the good old days.

The other side of your statistics implies that there is more wrong with the
"standard" sources than the "monitored" sources if only 22 of the 0 
codes are
for monitor sources. That leaves 115 0-codes for the standard sources. 
However,
I just noticed that most of these may be due to problems at Onsala which
apparently had an "amplifier failure". If those are taken out, only the AE
baseline is left as anomalous with 8 0-codes. Although that bothers me 
because
I remember AE, AV, and VE as being the really good baselines genereally, 
there
is probably no need for any more investment of resources looking at the
causes.

Thanks again for responding.

-arthur

------------------------------------------------------------------------

On Fri, 17 Sep 2004, John Gipson wrote:

> Date: Fri, 17 Sep 2004 09:46:08 -0400
> From: John Gipson <jmg@leo.gsfc.nasa.gov>
> To: 'Arthur Niell' <aniell@haystack.mit.edu>
> Cc: cma@leo.gsfc.nasa.gov
> Subject: RE: [IVS-analysis] r1139 ($04SEP07XA) Correlation report (fwd)
> 
>
> Hi Arthur:
>
> First off, I don't think this is a bad experiment.
> 2% 0-codes is low, and about par for our better sessions.
>
> Looking at some R1s from May of last year, the % of non-detections is:
> R1069  3%
> R1070  3%
> R1071  3%
> R1072  1%
> R1073  1%
>
> So 2% is sort of in line with the better sessions. There are some really
> bad sessions. For example, R1064 had 13% non-detections.
>
>
> To answer your specific question, here are some statistics.
>
> 3362 obs scheduled.
> 422  obs were scheduled with sources on the monitoring list.
> 400  obs were successful on this list. (Quality code >4)
> 22   obs were not successful.  (Quality code 0)
>
> So data loss due to source monitoring was about 0.66% in the worst case.
> Loss in precision is 0.33%, which is not meaningful at all.
>
>
>
> John
>
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: Arthur Niell [mailto:aniell@haystack.mit.edu]
> Sent: Thursday, September 16, 2004 8:38 AM
> To: John Gipson
> Cc: Chopo Ma; Ed Himwich; Nancy Vandenberg
> Subject: [IVS-analysis] r1139 ($04SEP07XA) Correlation report (fwd)
>
> Hi John,
>
> Are the 0-codes for sources of the monitoring list?
>
> -arthur
>
>

--