Extended Science Verification as an alternative to a full Call for Early Science Proposals - draft for discussion RAL, 2006 July 4 1. Introduction 1.1 Definitions and abbreviations [These closely follow the CSV Plan] AIV = Assembly, Integration and Verification SV = Science Verification CSV = Commissioning and Science Verification The purpose of the commissioning and science verification activity is to test and optimize the elements of the ALMA system and to verify that the array meets its science requirements. The CSV phase starts when a mode or component is handed over by AIV and ends with acceptance on behalf of the ALMA operation. Commissioning and Science Verification are carried out by the same team, working under the direction of the Project Scientist. Commissioning covers initial testing, interaction with AIV and other construction IPTs to identify and resolve faults, optimization, training and documentation. Science Verification is done to verify and document the performance of a particular observing mode for an outside observer. It is a continuing and incremental activity, as new modes will be added. It tests the end-to-end system, from proposal submission to final science, with the involvement of outside users. 1.2 Current schedule The current schedule for commissioning and early science includes the following. 2008 Nov 29 Ready to commission Antenna 3 2009 Mar 29 Decision point for Early Science Call for Proposals 2010 Jan 27 Start of Early Science The decision point has been set at 10 months before the start of Early Science to allow time for the proposal cycle to operate for the first time. The consequence is that we have only 4 months between the first chance to operate a 3-element interferometer and the need to decide on the modes to be offered for Early Science. Any decision made at the end of this period is therefore likely to be badly informed. There are risks to the credibility of the project if we offer either a very rudimentary capability or (much worse) promise something we later fail to deliver. During the commissioning period prior to the start of Early Science, it is planned to carry out a programme of Science Verification (SV) Calls for SV would be targeted in order to test one or more recently-commissioned modes (i.e. combinations of band, array configuration and correlator setup). Proposals would be sought from the user community, assessed not only for feasibility and scientific merit, but also to provide a comprehensive test of the current performance of the array. The aim of the process would be to get rapid feedback from the proposers in order to inform the commissioning programme. There would inevitably be an element of shared risk. It is likely that the amounts of time scheduled for individual proposals would be fairly short, although the very limited snapshot uv coverage at the start of the process may dictate longer observations initially. At its meeting in Oct 2005, the ASAC endorsed the concept of Science Verification and recommended that "a comprehensive programme of Science Verification should be viewed as the start of "Early Science" and thus allow the formal call for Early Science observations to take place a little later than is currently planned." This note is intended to initiate a discussion on the proposal to delay the start of Early Science and to increase the scope of the SV programme. There are important implications for staffing profiles, which need to be settled soon. In addition, such a change will require Board approval. 2. Advantages and Problems Delaying the full Call for Proposals has several obvious advantages: (a) When the Call is made, substantially more information will be available about the performance of the array and the modes which can be expected to operate reliably can be defined with more certainty. This should avoid both the temptation to advertise a very limited selection of modes (in which case the first set of proposals will not take advantage of ALMA's unique capabilities) or to promise far too much (so that proposals cannot be executed, and users lose confidence). (b) The capabilities of ALMA when full Early Science begins will be substantially greater than is currently anticipated (especially in the number of antennas available) and the increase in performance over the present generation of mm arrays will be much larger. (c) SV can be managed to give users access to ALMA as new modes become available, thus giving them prompt access to new capabilities which might not have been anticipated in a full CfP. The process is inherently flexible, allowing changes of plan (e.g. the order in which receiver bands become available) at fairly short notice. The main disadvantages of the proposal are: (a) There is a substantially increased load on the commissioning team, which has to handle interactions with the users, including assessment of the proposals (essentially operating a fast version of the TAC), delivery of the data and dealing with feedback. To this extent, the CSV team would have to take on some of the ARC functions (note, however, that it has always been the intention for the CSV team to perform end-to-end tests of the full ALMA system from proposal preparation to data delivery). (b) There is the risk that an over-long SV period will be seen as unfair to new users, in the sense that those willing to take the risk of observing in this period are likely to be experienced (indeed, the ability to reduce the data rapidly and provide informed comments to the CSV team is one criterion for a successful SV proposal). In addition, whilst an element of independent peer review would be included (ASAC recommended a proto-TAC), the need for speed and relevance to the commissioning process probably precludes a strict ranking by scientific merit. 3. Implementation In order to assess the impact of this proposal, two possible delays might be considered - 6 and 12 months. My personal view is that 6 months is the minimum, but that a year may be too long to sustain such a flexible operationa model. This clearly needs more discussion. All planning dates are, of course, incremental from the handover of a verified 3-element interferometer. SV would be interleaved with commissioning throughout the period, taking a larger fraction of time towards the end when the system becomes more stable. The attached IPS view shows the current schedule (above the blue line) and a one-year slip (below it), together with the antenna delivery schedule from AIV to CSV. Alternatively, the timeline for a 6-month slip might then look something like: 2008 Nov Ready to commission Antenna 3 (2+1) 2009 Mar? First SV call (8+3) 2009 Sep Decision point for Early Science Call for Proposals (11+2+2TP) 2010 Apr? Start of a more comprehensive SV programme ("dry run") (19+2+2TP) 2010 Jul Start of Early Science (22+2+2TP) Bracketed figures are the numbers of 12m antennas handed over from AIV to CSV, although not necessarily fully tested (bilateral+Japanese). I have assumed that 3 or 4 of the ALMA-J 12m antennas are available for interferometric operation initially with the main array correlator and that they subsequently move to their final role as total-power/compact array antennas (TP). At the recent AICSV meeting, the role of the ARCs in commissioning was discussed. It is clear that a model in which ARC staff come to Chile for short (e.g. single-turno) visits to act as astronomers on duty will not work during the commissioning phase: more substantial trips will be necessary (see separate discussion paper). One way in which ARC astronomers could participate very effectively before the first full CfP is to take the main responsibility for the SV programme. The CSV team must still operate as a single unit, so a model in which one group commissions and the other runs SV is not feasible. Nevertheless. individuals will inevitably have to specialise, and ARC staff could take the lead in the SV area. This provides more manpower to carry the additional load and also allows the ARC staff to concentrate on what will become their area of responsibility in normal operation - interaction with the users, the proposal and observing tools and quality assurance. It also avoids too much blurring of the distinction between construction and operations. I comment further on the issue of the role of ARC staff in CSV in the accompanying note. 4. Actions I have circulated this to the participants in the AICSV meeting, for discussion. If there is general agreement that the idea is sound, then it would be appropriate to discuss it at the ASAC meeting in mid-September with the aim of asking for a reiteration of their October 2005 recommendation to the Board . I will coordinate any responses I receive at the end of July, when I get back from holiday.