[asac] Gemini and JCMT TAC procedures

[Christine Wilson]

Hi, everyone,

Here is some additional material related to Charge #2 on the ASAC agenda.

I got a brief writeup from the current chair of the Canadian Gemini TAC
(CTAC).  I got even less information for the JCMT, which seems to have no
set procedure. The little I got is appended after the Gemini email. Both
Gemini and JCMT are examples of multi-TAC observatories.

Chris


---------- Forwarded message ----------
Date: Thu, 09 Sep 2004 20:58:22 -0300
From: George Mitchell <gmitchel at eastlink.ca>
To: Christine Wilson <wilson at physics.mcmaster.ca>
Subject: Re: Gemini TAC procedures

Hi Christine:

Let me address your three questions separately. I can only speak for
the CTAC and ITAC, but I believe that the other TACs follow basically
similar procedures.

1. Joint proposals (ie. proposals requesting time from more than one
partner):  The PI of the application to CTAC is normally one of the
Canadian members of the team. Each joint proposal has a "Principal
Contact". Gemini communicates with the proposers via this person. The
Principal Contact need not be the PI of the Canadian application. The
PI of the application to another country would normally be an
astronomer from that country. The applications (forms) sent to each
partner TAC are (unfortunately in my view) not necessarily identical.
Each TAC sees only "its" application. As Chair, I see the other
applications (i.e the other components of the joint applications) but
only after the CTAC meeting, in preparation for the ITAC meeting. CTAC
evaluates and ranks joint applications just as it does non-joint
applications. They are not put in a separate category.
	Before the ITAC meeting, Gemini staff determines the ranking of joint
proposals using a weighted average of each partners ranking. (There is
a twist here: If one of the partners ranks a joint proposal so low that
it does not make that country's time cutoff, that partners ranking is
deemed not to exist. In other words, its is as if the application to
that country was not made. I have objected to this, but nothing came of
it. Very few applications are in this category, so perhaps there is no
need to worry.) At the ITAC meeting, joint proposals are looked at
one-by-one. In cases where the partner rankings are divergent,
Jean-Rene Roy (ITAC Chair) asks us whether we want to change our
ranking. In fact this is an important procedural point: ITAC is
empowered to change the final rankings of any of the semester's
proposals in any way it sees fit. Of course, the TAC Chairs have to
justify any major changes to their TACs.

2. The issue of duplicate observations is addressed by the ITAC. There
is no set policy to my knowledge. In recent semesters, there have been
a number of duplications. These can be single-target proposals with the
same target. In my time on the ITAC, there has not been a case in which
all the technical details (e.g. integration time, filters, etc.) of the
observations were sufficiently alike to warrant one proposal being
denied. There are also multiple-target proposals which have one or more
target in common.  In these cases, the ITAC has made changes to the
observing times and has notified the PIs of the reason.

3. There is no special category for legacy-type programs. There are
certainly large proposals which return over a number of semesters, but
they are evaluated and ranked each semester.

There has been a recent policy decision by Gemini to permit a
high-ranked proposal to remain in the queue for up to two additional
semesters. A proposal must lie in band 1 in the final merged list to be
considered for rollover.  The decision on Canada's proposals is made by
the CTAC (same for others). This is called "rollover". A joint proposal
can be rolled over only if all the partners involved in that proposal
agree.

Another point that may be of interest: CTAC ranks its proposals. Gemini
takes all the partner's tanked lists of proposals and merges them
according to an algorithm which takes into account each partner's share
of the time. The result is a ranked list. This merged list is divided
into five "bands", of which the first four use up all of the available
time (assuming no losses to weather). Band 5 is backup. In essence,
then, there are only four ranks (ignoring band 5). I am not convinced
that this gross smoothing is necessary, but Gemini staff are not
interested in reexamining the issue. They argue that it greatly
simplifies the operation of the queue.

I hope this helps. Let me know if you want more!

Cheers,
George



On Thursday, September 9, 2004, at 06:28 PM, Christine Wilson wrote:

> Hi, George,
>
> Is there a summary on the web of how the Gemini TACs and ITACs
> function? I
> need the information for an ALMA meeting in two weeks.
>
> I'm particularly interested in how Gemini deals with proposals with
> Co-Is
> from two or more Gemini partners, how Gemini deals with proposals from
> different teams that want to do the same object with the same
> instrument,
> and if Gemini has any special procedure for large programs or
> legacy-type
> programs.
>
> If the info isn't available in an organized fashion, would you be able
> to
> send me a brief summary of how Gemini deals with these issues? I could
> send you a copy of the summary we got for ESO to help you see the kind
> of
> information we're looking for.
>
> Thanks,
> Chris Wilson
>
>

JCMT information from Gerald Moriarty-Schieven:

Joint (multiple queues) proposals are allowed, but each TAG will award
time separately, only from its own queue.  Aside from noting joint
proposals at the itac meeting, usually no special considerations are
needed at the itac meeting.

The ITAC consists of one member (usually the chair) of each of the
national TAGs, plus one more from the UK.

After I receive the proposals (including UH), I generate a source list of
all requested sources from all proposals, and look for duplicated sources.
I then check to see if the duplicated sources will be done with the same
or different instrument.  Where there is substantial overlap, I will
inform the TAG chairs.  If the overlapping proposals are in the same
queue, I let the national queue deal with it.  If from different queues,
then the ITAC will deal with it during their meeting.  This is usually
done on a case-by-case basis.

E.g. with the case Henry referred to, the ctag just made the two groups
collaborate.  Another case, the GRB Dutch/Canadian projects were told that
both rival groups would be given the data equally, and could publish
separately.  (Now both groups are collaborating on the JCMT project,
asking for time jointly from the NL/CN/UK queues.)  Sometimes the itac
merely suggests that the two groups (if awarded time separately by
different queues) collaborate on the overlap sources, but doesn't enforce
it.

Each TAG has its own policies on large or legacy-type programs, and the
time awarded to these (e.g. SHADES) is deducted from the national queue.
The UK and CN TAGs allow several pages for the scientific justification
for such large-scale proposals, but I'm not sure if NL does the same.

As you know, there is still debate on how large-scale scuba2/harp surveys
will be conducted.