VVLBA question...

On 09/12/2013 15:15, dlzc wrote:
Dear Martin Brown:

On Monday, December 9, 2013 3:13:31 AM UTC-7, Martin Brown wrote:
On 08/12/2013 14:21, dlzc wrote:
I understood that at one time, VLBA
interferometry involved carrying tapes from
place to place for processing.

The tapes also contain precise time references
linked to local H-maser clocks which are used
to find the white light fringe.

Could they be programmatically offset, to produce minimum noise on a few key stars?

You can do anything, but finding the interference fringes is or rather
was extremely difficult. If any path length compensation is out even
slightly you end up with nothing but dim faint useless noise.

Increasing aperture size increases resolution.

Could tapes be compared from 6 months before,
but sighted on the same location in space, for
an aperture of 2AU instead of just the size of
the Earth?

Not unless you could persuade the Earth to be
in two places at once. The measurements have to
be made simultaneously and in VLBI with N antennas
you get N(N-1) baselines which gives N(N-1)(N-2)
good closure phase observables and N(N-1)(N-2)(N-3)
closure amplitude observables.

It is for this reason that when big international
VLBI observations are in progress the
observatories try to stay on track for as long as
possible since dropping one dish out loses a lot
of useful data.

I accept that this is what we know, and what our model tells us. I just wonder if someone has tried this, or is there simply no long term storage of those tapes?

There would be no point. Trying to correlate signals where the path
compensation isn't quite right can occur when there is a malfunction and
the result is nothing - no fringes where fringes should be found.

Your question is a bit like asking can you do Young's slit experiment
one slit at a time and then combine the results later.

Big dishes have been damaged by wind loading
the past when they attempted to track objects
during VLBI and not drop out of lock.

I understand we aren't using the "same light
wave detected at different separated
locations", but...

And may already be being done, or have already
been tried.

There are proposals to put a VLBI antenna onto
a satellite in a very well characterised
orbit(s) with a dedicated ground station each.

http://www.gb.nrao.edu/ovlbi/OVLBI.html

Trojan to Earth would be nice, L4 and L5 would be line-of-site.

Be careful what you wish for. There is a limit to how far away from the
rest of the antennae you can place a new one and still get useful
behaviour. The U-V coverage of the interferometer determines how well
you can determine the sky brightness distribution.

There is a sample image from data using HALCA
further down the page.

But you have to be inside the white light (or
whatever bandwidth you are using fringe) to
observe interference - which roughly speaking
means contemporaneous measurements give or
take a few tens of ms.

I know why it hasn't been tried before, if if hasn't been.

I know we "interfere" two temporally-separated images to look for source movement... discovery or planets and NEOs.

That is more a differencing process - in the old days on a blink comparator.

Would it be possible to try? Or are there no long-term archives?

I don't know what policy is these days, but in the old days VHS tapes
were used in bulk quantities and shipped to the correlators later. I
doubt that anyone would bother trying to correlate two tapes taken
months apart - you have no idea how hard it was to get the white light
fringe even when starting from a pretty close guess.

Basically if anything is even slightly out of kilter you see nothing
interesting at all. The signals have to be coherent across both ends of
the baseline for the system to work at all.

Hanbury-Brown & Twiss's intensity interferometer for measuring stars
angular diameter was the most extreme optical interferometry for a while
but now COAST and various other systems are using radio astronomy
techniques in the optical and near infrared bands.


--
Regards,
Martin Brown