JAO Support Group Meeting, Thursday May 05, 2011, 10:00-11:00 ET (1400-1500UT)
Meeting Logisics
Time: 10:00 - 11:00 ET = 1400 - 1500 UT
Todays location: CV-ER???
2nd Thursdays Location:
Room: CV-ER331
Video Hub: 192.33.117.19
Domestic Call In: 434-293-7109
Int'l Call In: 434-293-6691
4th Thursdays Location:
Room: CV-ER209
Video Hub: 192.33.117.17
Toll-free: 866 560 2652
Toll: 517 308 7468
Passcode: 2652628
Agenda SPECIAL MEETING! ONLY ONE TOPIC
Attendees: TH, HL, CB, AW, JH, SC, NM
OT time calculation for Cycle1: At previous meeting we discussed the formation of a "function based team" to address how the OT should calculate observing time in Cycle 1. Last week we said Harvey and Todd will represent NA on this FBT, and that we will have a special meeting to discuss the groups input to this topic. This is that meeting!
Main charge: suggested modifications for how OT calculates estimated time, and whether additional/different inputs from users are needed, e.g. if checkbox for "time not from sensitivity calculator" should be changed (entry field for user-specified time?); how to specify which spectral setup should be used for the calculation (recalling that Cycle1 allows mixed correlator modes); polarization; etc..
Mixed-modes (narrow & wide bands, both either line/continuum)
Mixed-modes (narrow for line, broad for continuum)
For continuum: adopt Tsys(freq) or <Tsys(freq)>?
Extra parameters to be collected from users? (sensitivity dynamic range, spatial dynamic range, spectral dynamic range,
If sensitivity and dynamic range solicited, what if they are not consistent?
for extended observations: adopt Tsys(transit) or <Tsys(elev)> (if algorithm allows distribution across HA)
....
We spent first 30min with CB & TH talking about recent problems discovered from TW Hyd data.
Regarding OT, the results of our discussion:
First we noted the following "bugs" in what the OT does now:
integration time does not consider correlator efficiency/smoothing factor or effective BW for wide modes
putting in a "recommended frequency" that is outside the range of the spectral setup window (even outside the band!) and trying to calculate sensitivity or time produces an error
integration time calculation does not check if sp. windows overlap (and thus have less effective bandwidth)
Next, considering mixed modes, we decided that what it does now (drop-down with four choices) probably covers Cycle 1 too, as long as "aggregate bandwidth" is fixed to mean "non-overlapping aggregate bandwidth".
We think it would be good (although maybe not high-priority) if the system provided the option to automatically place continuum spectral windows in optimal places in the bandpass (in terms of Trx and Tsky). For this to be offered, OT would need to solicit a user-supplied spectral index, or choose from a list (e.g. -1,0,1,2,3,4).
We think "reference frequency" should be moved to be in a selection area near bandwidth (so it says "Frequency and Bandwidth used for Sensitivity"), and provide several options, including:
best freq for continuum (see previous item).
freq based on table in spectral setup panel
user specified freq (constrained to be in allowable IF range)
As for whether system should adopt Tsys(freq) or <Tsys(freq)> for the time calculation, this depends on the use case.
<Tsys(freq)> for continuum or user-supplied BW.
Tsys(freq) at freq in spectral setup panel
(maybe this should be a user-selected option)
For the last two items in the agenda bulleted list, we decided that it is best to wait to see what BVV proposes before discussing the following
JH will email BVV urging him to call a meeting already!!