LSST Science Advisory Committee Face-to-face meeting, Monday, August 15, 2016 Held at “LSST2016”, the LSST collaboration-wide meeting in Tucson, Arizona SAC members attending: Michael Wood-Vasey, Niel Brandt, David Kirkby, Mansi Kasliwal, Nelson Padilla, Timo Anguita, Renu Malhotra, Michael Strauss, Lucianne Walkowicz, Marla Geha, Risa Wechsler Also attending were a number of LSST Project personnel. The meeting included a number of presentations, and resulted in a series of recommendations to the LSST Director and Deputy Director. We give these recommendations here (in the form of a letter to Steve Kahn and Beth Willman), and follow them with a list of the presentations, as available on the LSST SAC webpage. Dear Steve and Beth, This letter is a summary of recommendations from the LSST Science Advisory Committee, following our face-to-face meeting as part of LSST2016 on Monday, August 15, 2016. We discuss three items which came up in our meeting: 1. The process by which we collectively converge on the survey strategy (cadence) to be used by LSST; 2. The status of the Deep Drilling Fields, and future community calls for additional DDF/mini-survey proposals; 3. Plans for making commissioning data public. As we will argue below, items (1) and (2) are closely coupled to one another. *****Survey Strategy We are excited by the development of multiple community-developed metrics and diagnostics on the OpSim outputs in the Metrics Analysis Framework. This has been an impressive effort by a large number of people, and we praise Phil Marshall in particular for his leadership of this work. The Cadence White Paper, nearing completion, will be a tremendously important document for LSST. However, we have a number of comments and concerns. Both the Project and the community have done quite a lot of work on various aspects of the survey strategy, and it has been a significant focus of the last three LSST-wide meetings, but there is not yet a clear statement of what the process going forward is, and this lack of understanding is causing confusion among the LSST Project rank-and-file as well as the scientific community. We know that the Operations Simulator is undergoing significant revisions, but the timeline for finishing those is unclear to us (and much of the community); we also don't have a clear understanding of what the capabilities of this new version will be. While there is an impressive number of metrics currently in place, we are concerned that they don't fully cover all the different science areas that depend on the details of survey strategy, and rather reflect the specific scientific interests of the people who have developed them. It also remains unclear how the results of the OpSim and the MAF outputs will be used to make decisions about the survey strategy (although we understand that we, the SAC, will play a role in this process). With all this in mind, our recommendations are as follows: -We recommend that the Project develop a timeline for the entire process, including completing the necessary remaining work on the OpSim, giving the community a deadline for additional metrics, assessing the extent to which the existing metrics span the space of relevant science cases, weighing the relative merits of different cadences, and finalizing a decision on survey strategy. We suggest a strawman schedule below, and would be happy to iterate on this with you if that would be useful. -We recommend that the Project include a description of this timeline and the decision-making process in the Cadence White Paper. This White Paper will be the first place people in the community will go to understand the current status of Survey Strategy planning, and it would be appropriate to include a detailed statement from the Project. The SAC is happy to work with the Project both in developing the timeline, and in drafting the text for the White Paper. An important tension is the need to have a definite and well-understood survey strategy at the start of the survey, but to leave flexibility for the Operations Team to respond to new scientific opportunities to refine the cadence as the survey progresses. ****Deep Drilling Fields We are excited to be revisiting the question of the Deep Drilling Fields. We agree that the time is right to prepare a call to the LSST science community for additional ideas for Deep Drilling fields. We also need to refine the cadence for the four already decided fields currently implemented in OpSim. In preparation for a call for further deep drilling fields and cadences, we make the following comments, issues, and recommendations: -We recommend that the boundary conditions for such fields be decided. The four DDF's in the current baseline OpSim runs use 4.5% of the observing time. But including the other mini-surveys (low Galactic latitudes, the North Ecliptic Spur, and the Magellanic Clouds) raises this to 15%, well above the nominal 10% number stated in the LSST overview paper. How should we (and the community) think about what fraction of time will be devoted to observing modes other than deep-wide-fast? Similarly, the implementation of the existing mini-surveys is based on decisions made over a decade ago, and it is time that these be revisited: we know that the cadence in the North Ecliptic Spur is far from optimal for finding asteroids, for example. Are there plans for addressing this in OpSim? Moreover, there is no detailed science case in hand yet for the Galactic Plane and South Celestial Pole surveys, which could allow us to determine how much time should be spent on them (and thus how much time can be used for Deep Drilling Fields). -On a related note, any call for additional DDF proposals will inevitably attract proposals that are not single fields, but are more like mini-surveys. Examples include various regions to put at high priority early in the survey (for Euclid or WFIRST, for example). We have to decide how to handle such ideas, how they should be balanced against the currently existing mini-surveys, and whether they should be included in this call, or will be addressed separately. Target of Opportunity observations are also now a hot topic, given LIGO's success. We recommend that the call for DDF proposals explicitly state that ideas for observing modes other than DDFs will be solicited in a separate call, and are not included in this call. -The current OpSim runs are far from an optimal cadence for the DDFs. Moreover, the existing metrics do not do a proper job of quantifying how well the cadence in the DDF perform scientifically. It would be very good to rectify this, and make the results public, before putting out a call for additional DDFs. We recommend that the OpSim implement improved cadences in the existing DDFs, and metrics be devised to see how well they do in addressing the science goals they are designed for. It would also be good to run OpSim runs implementing other of the DDF ideas from the 2011 White Papers, to quantify how well they work. On a related note, we recommend that the call for DDF proposals request specific metrics to quantify how well they meet their science goals. -Beth suggested that we specifically encourage proposals for DDFs dedicated to Milky Way/Local Volume science or Solar System science. We like this idea (although we would not want to restrict proposals to these two areas). Indeed, among the 2011 White Papers are several Milky Way and Solar System science proposals; again, implementing these ideas in OpSim to see how they perform would be useful, and give context to people who wish to propose additional ideas. We will point out that Milky-Way focused DDF proposals will depend in part on the efficacy of LSST software in crowded fields, for which the requirements remain undefined (currently "best-effort; no worse than DAOPHOT, but no work has happened yet"). It will be important to give information to the community on what they can reasonably expect in this regard. -Finally, there is some urgency in finalizing the DDFs, to take advantage of potential synergies with WFIRST, Euclid, JWST, and other missions. In this context, we recommend that the call for DDF proposals include a discussion of the deep drilling fields currently planned by these and other missions, and how they relate to the 4 DDFs already selected. *****Commissioning We had time to discuss commissioning only briefly. We understand that there is real tension between what the Project has decided to make public, what the public perception of this might be, and what is realistically possible given current plans for standing up the Data Management and Science User Interface during commissioning. We recommend that the Project Office make a proposal for how to proceed, and we discuss thus in a future SAC meeting. ****Strawman schedule/milestones for Survey Strategy decisions. The following is only a strawman, and no doubt reflects our ignorance of various issues. The SAC is happy to work with the Project Office to refine or rewrite this. [P] = deadline for project, [S] for SAC and [C] for community. [P] 2016 Q4 - Define specific boundary conditions for DDF (a firm commitment of time, whether or not to include mini-surveys, whether or not to include ToOs) and issue a call for proposals. [P,C] 2016 Q4 - Cadence White Paper published, including a description of the timeline and decision-making process. [C] 2017 Q2 - Community white papers on DDF due [S] 2017 Q3 - SAC meets and recommends DDF strategy to Project [P] 2017 Q4 - OpSim simulations ready for various cadence plans that have been discussed so far including rolling cadence [C] 2018 Q2 - Community Cadence (+Follow-up?) Workshop Meeting II [C] 2018 Q3 - Community White papers due on figures of merit that could weigh into a Deep-Wide-Fast (DWF) plan. It would be best if these figures of merit could be applied to all available OpSim simulations. [S] 2019 Q1 - The Project delivers a report to the SAC, laying out (and quantifying the consequences of) a number of possible cadence scenarios for Deep/Wide/Fast. The SAC will make recommendations to the Project office. In preparation for this, the Project and the SAC may iterate with some community white papers as well, and identify important science areas which are not covered in the existing metrics. [S+P] 2019 - Based on knowledge from commissioning + draft exercise, iterate and converge on a survey strategy. Project announces to the community. [C+S+P] 2021, 2023, 2025, 2027, 2029 - Every two years (or three years?), call for community white papers for proposed changes to survey strategy. SAC-appointed panel reviews these white papers and makes recommendations to the project. ***Presentations at the LSST SAC meeting: Phil Marshall: “Science Driven Optimization of the LSST Observing Strategy”. Marshall_cadence.pdf. This discussed the progress to a community-driven white paper describing the LSST cadence. Niel Brandt: Deep Drilling Fields and LSST mini-surveys. Brandt_DDF.pdf . Niel discussed the history of the LSST Deep Drilling Fields. Beth Willman led a discussion of how we may move forward on the deep drilling fields, Willman_DDF.pdf Todd Boroson described the status and progress of the Las Cumbres Observatory; they have received a major MSIP grant giving the community access to LCO for LSST-precursor work for follow-up of transients/variables. Boroson_LCOGT.pdf David Ciardi gave a presentation on work, centered at IPAC/Caltech, on the LSST Science User Interface. Ciardi_SUI.pdf