Configuration telecon. 2005 Feb 16 Conway, Wootten, Beasley, Holdaway Purpose: To get configuration redesign going. Do this for fifty antennas. ACTION: AW to report on progress in biweekly reports. ACTION: ESO to renew JC's contract appropriately. ACTION: Tom Wilson ACTION: Get into a good form the DEM file with a readme, talk to Angel. Also update the topography mask with AOS coordinates, ACA coordinates, and updated information on the shallow quebrada disussed below. Responsible: Holdaway. Mark can you place it on the characterization website in Tucson noted above? Also on Tony's alma.cl site but since we've had coordinates in the Tucson website for a decade I think we should include it there, and then mirror that site to alma.cl. I think Kevin Long (does he have a JAO phone yet?) has access to both web locations and can accomplish this. ACTION: JC to send MH the inner mask. TB: Let's put it in alma.cl website. Responsible: JC ACTION: John or Mark--what are the approximate coordinates? I'll ask Lars-Ake, Roberto and Jorge to lead an expedition of discovery to obtain digital photos and a ground truth description of this area as soon as possible. Coordinates of this feature as proposed by Mark: UTM NW 628630 7453370 SE 629070 7452920 John do you agree? Responsible: AW will send query on ground truthing expedition. ACTION: JC will try to modify current compact array in such a way that it would be optimized for 50 antennas while allowing additional pads to be built should another dozen antennas become available. Responsible: JC ACTION: MH will get matlab and attempt to understand JC's software. MH will work on outer array design, utilizing a reasonable number of new pads in addition to those of JC's inner pad design. Responsible: MH Meeting Notes: TB: Buildings to start September. Pads and roads could begin as early as August, not likely but if nothing else is ready, we should start then. Configuration process was incomplete, as some were not reviewed. What range of spatial scales, divvying up of work. Access roads have restrictions. Pads can be approached by transporter from any angle by assumption. We would prefer to have two documents. Would prefer to have a small group dedicated to this. JC: Need someone from site who is dedicated to this to avoid confusion on coordinate systems. There is considerable confusion on coordinate systems on the site. Documents say north is astronomical north but this is not so on the UTM system, with a 1-2 degree offset. Lines of constant UTM East and UTM North are not orthogonal. Particularly for the compact array, should define in offsets and make sure we have the correct North. TB: Eschwey--with a University put geodetic markers around the site. Measure relative to these with theodolites. Would we be better off with a differential GPS. We need to agree on the absolute coordinate of one point on the site. We have a monument for this. The offset between true siting and desired siting won't affect imaging. MH: We have two grid markers, Master 0 and Master 1. One is from Onsala, the other defines north relative to the first. N. B. from AW--coordinates for things are given at the link on www.alma.nrao.edu which leads to: http://www.tuc.nrao.edu/mma/sites/Chajnantor/maps/coordinates.html JC Center of array defined relative to one of the markers. UTM measured for one of markers. This is how the array was defined. North is then UTM north. But the official ALMA coordinate system says astronomical north is north. We just need one coordinate system. FITS file has UTM coordinate and pixel size. The mask, derived from the DEM, was crosscorrelated with old DEM model. It was good in the center and poorer on the edges. Discussion with the companys resulted in new UTM coordinates. DEM now has good coordinates but is not generally available. ACTION: Get into a good form the DEM file with a readme, talk to Angel. Also update the topography mask with AOS coordinates, ACA coordinates, and updated information on the shallow quebrada disussed below. Responsible: Holdaway. Mark can you place it on the characterization website in Tucson noted above? Also on Tony's alma.cl site but since we've had coordinates in the Tucson website for a decade I think we should include it there, and then mirror that site to alma.cl. TB: For CARMA TB produced a model for Mel and Tam. Mel made configuration ~1m, TB converted to real coordinates. Then he went and looked at location with a GPS, ~5m. Probably within 10-15m of pad location. Then an iteration provided Mel with a list of which places weren't appropriate. This was iterated with Helfer. In the end, perhaps the configuration was within a few percent of the perfect no-topology array. We did this. The pads were ground-truthed, surveyors said no changes needed, we think, but no information actually came back. TB: Job 1 is to look at locations. There surely will be some corrections. JC: Perhaps civil engineers need to do this. I visited the site a few times and used that knowledge (i.e. putting the pads on ridges). TB: We noticed that the road system just went across everything. You were correct that we should avoid the quebradas. JC: We need someone to interact on us about what is reasonable. TB: I think you have the correct approach. All we need to change in ground truth is the discovery that perhaps there is rock which would be expensive to move at the location. This is simple ground truthing. For road access, we can look at it and try to make a sensible network high on the ridges. We can take the design we come up with and suggest it to a designer. First step--in house evaluation of pad locations. JC: In some particular regions--directly East of compact array--we blanked. The broad quebrada there results in poorer imaging. I would like some input on that particular area. This is the one which runs to the SE diagonally across the site and then runs into some others. If we could ground truth that area it would be very helpful. ACTION: John or Mark--what are the approximate coordinates? I'll ask Lars-Ake, Roberto and Jorge to lead an expedition of discovery to obtain digital photos and a ground truth description of this area as soon as possible. Coordinates of this feature as proposed by Mark: UTM NW 628630 7453370 SE 629070 7452920 John do you agree? JC to send MH the inner mask. TB: Let's put it in alma.cl website. JC will try to modify current compact array in such a way that it would be optimized for 50 antennas while allowing additional pads to be built should another dozen antennas become available. 1/r^2 is the better choice than circles or rings when you may have to descope antenna number. MH: On larger array, there should be some further optimization. We need to determine how many pads to share. I use John's inner configuration, select the most appropriate ones. The number of shared pads was a constraint. I could use more than I was allowed. I was allowed 44 additional pads. JC: The out and in methodology still awaits discussion. AW: That is now an approved document but the timing details which were in the almaedm have vanished from the approved document. JC: Always optimized for a snapshot. If one optimizes for a specific declination one can create an array which produces uv holes at other declinations. I do intend to optimize the compact configuration. After I finish I do look at long tracks. MH: Some of the inner sidelobes can remain in this process. JC: Yes but they are already very small. Inner sidelobes don't change much going from snapshots to long tracks. MH: Y+ does have problems; the inner sidelobes are the name of the game owing to the Y configuration. I suggest you send me your software. I'll need to get matlab. JC: You'll be shocked at the kludginess. MH: I am a kludginess aficionado. ACTION: MH will get matlab and attempt to understand JC's software. MH will work on outer array design, utilizing a reasonable number of new pads in addition to those of JC's inner pad design. AW to circulate minutes, action items. Notable quote of the day: JC: 1/r^2 is the better choice than circles or rings when you may have to descope antenna number.