[daip] (no subject)

Tue Sep 18 15:36:31 EDT 2001

Date: Tue, 18 Sep 2001 13:34:56 -0600 (MDT)
From: Leonia Kogan <lkogan at aoc.nrao.edu>
Message-Id: <200109181934.NAA17702 at bonito.aoc.nrao.edu>
To: lincoln at play.harvard.edu, daips at cv3.cv.nrao.edu
Subject: Re: [daip] VLBA autocorrelations / ACCOR corrections

Hi Lincoln,

I agree with Jim. The variation of the gain as a solution of ACCOR is real 
independent on the reason.
Jim is right that SOLINT should be chosen less than default 60min.
The order of minute would be better.

The question is can we hope that the autocorrelation variation can be 
applied to the crosscorrelation on the level of accuracy <1%.

The accuracy of the crosscorelation correction depends on the value of 
the correlation. The theory (VLBA scientific memo #9, 1995) sais that 
this correction is better 0.1% if the correlation is less 0.1.

The theory may miss some additional effects so the accuracy can be worse.
As I remember the ACCOR type correction decrease the difference at the different
BBC from >20% to <2%.

>I am also curious about the frequency averaged amplitude of VLBA
>autocorrelation spectra that I observe in many experiments. It is my
>understanding that these spectra are normalized.  I routinely
>notice variations in amplitude with time on the order of a few percent. I
>am especially interested by (1) isolated discrepant points that are 5+%
>away from neighboring points and (2) structure in these time variations
>that includes slopes (a la wrapping between two boundary values)
>and modulations in dispersion that look envelopes degined by sync
>functions.

The answer on this question is at the same area of my previous message.
The 5% variation is too big if you applied the ACCOR correction.

I'll look at the possible problem with negative SOLINT

Leonia

----- Begin Included Message -----

>From daip-admin at donar.cv.nrao.edu Tue Sep 18 13:10 MDT 2001
From: Jim Ulvestad <julvesta at aoc.nrao.edu>
Reply-To: julvesta at aoc.nrao.edu
Organization: National Radio Astronomy Observatory
X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.4.2-2 i686)
X-Accept-Language: en
MIME-Version: 1.0
To: Lincoln Greenhill <lincoln at play.harvard.edu>, daip at zia.aoc.NRAO.EDU
Subject: Re: [daip] VLBA autocorrelations / ACCOR corrections
References: <Pine.OSF.4.05.10109181445250.5099-100000 at play.harvard.edu>
Content-Transfer-Encoding: 7bit
Sender: daip-admin at donar.cv.nrao.edu
X-BeenThere: daip at listmgr.cv.nrao.edu
X-Mailman-Version: 2.0.5
Precedence: bulk
List-Help: <mailto:daip-request at listmgr.cv.nrao.edu?subject=help>
List-Post: <mailto:daip at listmgr.cv.nrao.edu>
List-Subscribe: <http://listmgr.cv.nrao.edu/mailman/listinfo/daip>,
	<mailto:daip-request at listmgr.cv.nrao.edu?subject=subscribe>
List-Id: This is the designated AIP list for AIPS support. <daip.listmgr.cv.nrao.edu>
List-Unsubscribe: <http://listmgr.cv.nrao.edu/mailman/listinfo/daip>,
	<mailto:daip-request at listmgr.cv.nrao.edu?subject=unsubscribe>
List-Archive: <http://listmgr.cv.nrao.edu/pipermail/daip/>
Date: Tue, 18 Sep 2001 13:09:18 -0600
X-Lines: 57
Status: RO

Lincoln Greenhill wrote:
> 
> Hi Leonia,
> 
> You mention that ACCOR is supposed to fix problems that are as large
> as tens of percent.
> 
> Do you think  that the structure in the SN tables  output by ACCOR (i.e.
> systematic variations in gain on the order of < 1% on times scales of a
> few minutes) are not REAL?  Should one only use long solution intervals
> with ACCOR?

This issue has been argued about at various times in the past.  The
default
for ACCOR at some point was a 60-minute average, and I argued pretty
strongly for a 2-minute average; among other things, that provides an
opportunity to clip individual bad points using SNSMO, which would
distort
the whole solution in a 60-minute average.

The structure on time scales of a few minutes IS real.  This is caused
by
the automatic gain control--as a source rises or sets and changes
elevation,
the system temperature changes gradually, causing a trend in the ACCOR
solutions.  Then some threshold level is reached where the AGC kicks in
and changes the calibration slightly, at which time the solution appears
to jump back to near the previous level.  This sort of cycling repeats.
I assume that you see a sawtooth of one shape when the source is rising,
and essentially the opposite shape when the source is setting, though
I've
never actually verified this.

(Probably I didn't describe this in quite the correct language, but I
hope
you get the idea.)

jim

> 
> Regards,
> 
> Lincoln
> 
> Lincoln J. Greenhill      Harvard-Smithsonian Center for Astrophysics
> Radio & Geoastronomy Division, 60 Garden St, Cambridge, MA 02138, USA
> Telephone:  1 617-495-7194             FAX:  1 617-495-7345
> Internet:  greenhill at cfa.harvard.edu   http://cfa-www.harvard.edu/~lincoln
> 
> _______________________________________________
> Daip mailing list
> Daip at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/daip
_______________________________________________
Daip mailing list
Daip at listmgr.cv.nrao.edu
http://listmgr.cv.nrao.edu/mailman/listinfo/daip

----- End Included Message -----