[daip] forwarded message from Chris Carilli

[Eric Greisen]

------- start of forwarded message (RFC 934 encapsulation) -------
Return-Path: <ccarilli at huron.aoc.nrao.edu>
Received: from zia.aoc.NRAO.EDU (zia.aoc.nrao.edu [146.88.1.4])
	by primate.cv.nrao.edu (8.9.3/NRAO/CV-2.0) with ESMTP id MAA16471
	for <egreisen at primate.cv.nrao.edu>; Fri, 7 Jul 2000 12:17:46 -0400
Received: from huron.aoc.nrao.edu (IDENT:ccarilli at huron [146.88.1.168])
	by zia.aoc.NRAO.EDU (8.9.3/8.9.3) with ESMTP id KAA18149;
	Fri, 7 Jul 2000 10:17:41 -0600 (MDT)
Received: (from ccarilli at localhost)
	by huron.aoc.nrao.edu (8.9.3/8.9.3) id KAA14306;
	Fri, 7 Jul 2000 10:17:50 -0600
Message-Id: <200007071617.KAA14306 at huron.aoc.nrao.edu>
From: Chris Carilli <ccarilli at aoc.nrao.edu>
To: egreisen at huron.aoc.nrao.edu
Cc: ccarilli at huron.aoc.nrao.edu, cchandle at huron.aoc.nrao.edu,
        dshepher at huron.aoc.nrao.edu, fowen at huron.aoc.nrao.edu,
        julvesta at huron.aoc.nrao.edu, mgoss at huron.aoc.nrao.edu,
        mrupen at huron.aoc.nrao.edu
Subject: VLA Weight Cal Transfer
Date: Fri, 7 Jul 2000 10:17:50 -0600

July 7, 2000

from: C. Carilli
to: E. Greisen
re: VLA Weight Cal Transfer
cc: F. Owen, M. Rupen, D. Shepherd, C. Chandler,
    J. Ulvestad, M. Goss

Hi Eric,

As promised, following is a summary of the use of the 'VLA weight cal
transfer' for some recent 43 GHz data from the VLA. Overall, the
process works properly, and will make the calibration, flagging, and
imaging procedures significantly easier, and improve the sensitivity
of the final image.

In review, the process is needed in order to have proper data weights
based on measured T_sys values and antenna gains. The current
implementation is very simple, requiring a value of cparm(2) = 8 in
FILLM, then follow the normal calibration procedure, then use docal =
2 in SPLIT.  This process is extremely useful at 43 GHz and 22 GHz,
where the system response varies by more than a factor two from the
best to the worst antennas, but is also useful at 1.4 GHz where the
T_sys increases due to spill-over at low elevation.

I loaded data from project AI79 (high redshift CO at 43 GHz) in both
the usual way, and in the new proper cal-transfer way.  I ran CALIB
and found that antennas 17, 20, and 24 all had much higher gains than
most antennas, by a factor between 3 and 5. After splitting with docal
= 2, the weights for baselines involving these antennas were lower by
similar factors than for other baselines. A good example of this was
antenna 26, which had a gain factor about two times higher than good
antennas. After splitting, baselines involving antenna 26 had weights
about two times lower than those for good antennas.

In terms of image noise, the results for the rms on NA weighted images
are:

All data with Cal transfer: 0.178 mJy/beam

All data without Cal trans: 0.31 mJy/beam

Without ant 17,20,24 with Cal. trans: 0.185 mJy/beam

Without ant 17,20,24 without Cal. trans: 0.22 mJy/beam


I recommend making this option generally available in the next
AIPS release. 

cc
------- end -------