From cnl at haystack.mit.edu  Tue Oct  7 10:18:43 2008
From: cnl at haystack.mit.edu (Ching Lue)
Date: Tue, 07 Oct 2008 10:18:43 -0400
Subject: [gps-developers] Test
Message-ID: <48EB6FC3.5010807@haystack.mit.edu>


Test.

Ching

From msagini at gmail.com  Tue Oct  7 13:01:14 2008
From: msagini at gmail.com (Mark Sagini)
Date: Tue, 7 Oct 2008 20:01:14 +0300
Subject: [gps-developers] Test
In-Reply-To: <48EB6FC3.5010807@haystack.mit.edu>
References: <48EB6FC3.5010807@haystack.mit.edu>
Message-ID: <86504c090810071001j5a7c42c1y8b98d5b182b829b8@mail.gmail.com>

Test Successful.
Mark

On Tue, Oct 7, 2008 at 5:18 PM, Ching Lue <cnl at haystack.mit.edu> wrote:
>
> Test.
>
> Ching
> _______________________________________________
> gps-developers mailing list
> gps-developers at haystack.mit.edu
> http://www.haystack.mit.edu/mailman/listinfo/gps-developers
>

From brideout at haystack.mit.edu  Fri Oct 24 10:47:56 2008
From: brideout at haystack.mit.edu (William Rideout)
Date: Fri, 24 Oct 2008 10:47:56 -0400
Subject: [gps-developers] Champ receiver bias study
Message-ID: <4901E01C.8060301@haystack.mit.edu>

Anthea,

I've analyzed the receiver bias of the Champ GPS receiver using the 
minimum scallop method.  The receiver bias was estimated by examining 
the full 24 hours of data.  The only data not used was spurious TEC 
values greater than 200 TEC.  First a ten day period from 2003-06-01 to 
2003-06-10 was analyzed, and ten independent estimates were made of the 
receiver bias.  In this case the mean receiver bias was 11.27 TECu, with 
a standard deviation of 0.59 TECu.  Given the sample size of ten, this 
implies our estimate of the mean has error bars of 0.2 TECu.

A second ten day period 2 years later was then analyzed.  The standard 
deviation was again small at 0.41 TECu, but the mean was now 8.88 TECu.  
Since this change in the mean receiver bias is much larger than the 
error bar on our estimate of mean, the receiver bias is most likely 
slowly changing over time.

This implies that in processing Champ data, it would be appropriate to 
use a 10 day running average of the individual day's receiver biases to 
calculate an estimate of the true receiver bias.

Bill Rideout