LSST Science Advisory Committee meeting Monday, August 14, 2017 Westin Paloma Hotel, Tucson This meeting was part of LSST2017, the LSST Project and Community annual meeting. Attending: Renu Malhotra, Michael Strauss, Lucianne Walkowicz, Risa Wechsler, David Kirkby, Jason Kalirai, Mansi Kasliwal, Niel Brandt (remote), Timo Anguita (remote), Anze Slosar (remote), Nelson Padilla (remote) and Josh Simon (remote) Project personnel attending and presenting included Beth Willman, Steve Kahn, Zeljko Ivezic, Mario Juric, Eric Bellm and Amanda Bauer. The meeting was open to all who wished to take part, and there were several dozen LSST2017 participants who joined us. Our agenda covered three broad topics: -Project-community interface and communication Presentations by Beth Willman (overview: willman_communication.pdf) Amanda Bauer (EPO construction effort: bauer_EPO.pdf) Mario Juric (Interface between data management and the science community: juric_communications.pdf) -Community alert brokers Presentations by Beth Willman and Eric Bellm (willman_bellm_brokers.pdf)_ -Status of and plans for observing strategy Presentation by Zeljko Ivezic (ivezic_cadence.pdf) The presentations are available on the SAC web page. The following is a summary of our findings and recommendations. We give more details in the minutes below. -We are supportive of the on-going effort to improve communications within the project, and between the project and the broader community. We are enthusiastic about the plans to define liaisons from the Data Management group for each science collaboration. We recommend that the broader community, including international contributors, also have access to information about DM's activities. -We would like the EPO team to explore ways for the science collaborations to get involved in outreach activities now, even during construction. We would also like clarification on the ramp-up of EPO activities during commissioning: when in practice ill the Project be actively doing outreach? -The plans for giving guidelines for and selecting event brokers from the scientific community need to reflect the fact there is likely to be a continuum of needs: from community brokers designed to serve the world, to specialized filters aiming to do specific scientific projects. We recommend that the call for brokers reflect this continuum. -We recommend that the project start a conversation with the Transients and Variable Stars science collaboration about use cases for brokers, to understand the extent to which the LSST simple broker, and a few general community brokers, will or will not satisfy the needs of the community. -The SAC remains unclear on whether the outputs of brokers can, or should, be world-public. This needs to be clarified. -The SAC looks forward to reviewing the document which will describe the requirements and review process for brokers. -The SAC looks forward to reviewing the call for white papers from the community for mini-surveys and deep drilling fields. We strongly recommend a single such call, and in the detailed notes below, make a series of recommendations about what that call should include. It will be particularly important that this call be widely distributed in the community, via, e.g., the AAS newsletter, presentations at AAS meetings, social media, and so forth. **********************Meeting minutes******************** These minutes do not attempt to capture and repeat all the material available in the presentations listed above, and we recommend that those interested in the details read the presentation slides. We are happy to welcome Josh Simon of the Carnegie Observatories as the newest member of the LSST Science Advisory Committee. [Note that Rachel Bean of Cornell University joined the SAC after this meeting.] ********Communications*********** The SAC is excited about the continued emphasis on improving communication both within the project, and between the project and community. We are eager to see the planned improvements in the LSST website, which is the first place that community members often go to find out information about the LSST. We recognize the challenge of balancing the time of Project personnel between communication tasks and their immediate responsibilities for the Project itself. Communication happens in many directions: between Project personnel, from the Project to the scientific community (and back), and from the Project to the general public. Each has its own challenges, and we urge the Project to continue to work through use cases in each of these areas to make sure that distributed information is timely, up-to-date, and accessible. We also recognize the role of the SAC itself as a communication channel between the Project and the community, and will continue to look for ways to improve that channel. There is a real challenge in the plethora of different communications mechanisms currently being used in the LSST community, and we appreciate the effort underway, through questionnaires and so on, to determine which of those mechanisms are most effective. We would like to see more emphasis on using social media to communicate with the LSST science community and the general public. In this regard, it is important that the project speak with one voice; there has to be an official source of such information. We are enthusiastic about the plan that Mario described, whereby DM personnel are assigned as liaisons to each of the science collaborations to advise them and to receive feedback from them. This has been happening to some extent already, but this would formalize these activities, and properly give credit to the DM folk for the time they put into this effort. It is important to explain to the science collaborations what the role of the DM liaisons will be, and what would be appropriate to ask them to do. There is a great deal of interest and need in the science collaborations to learn how to run the LSST stack; work has been stymied in part because of lack of funds to support, e.g., postdocs in the science collaborations to develop the necessary expertise. It is also important to make sure that the broader community outside the science collaborations, especially the international contributors, also have access to information about DM's activities. We were happy to hear about the many activities of the EPO group. We understand why it is necessary, but we are somewhat frustrated by the fact that the EPO efforts are focussed solely at this point on building the system; no outreach activities per se are allowed during construction. This is a real missed opportunity! In this context, it was unclear when outreach activities could in fact start: the beginning of commissioning? The start of full survey operations? We would like to see more opportunities for scientists in the LSST community to get involved in EPO efforts; ideal would be if there were funds to support these scientists in this work. The science collaborations have a lot of interest and ideas in outreach; we urge the EPO team to take advantage of that expertise and energy wherever possible. Indeed, while the EPO team cannot get involved in outreach activities now, the science collaborations can do so, and this is an opportunity that should be exploited. We urge the EPO team to consider translating material into more languages than just Spanish. It appears that most of the effort of the EPO team is for constructing virtual/electronic access to the LSST data. We encourage the team to consider "hard access" opportunities for engaging the public as well: museum displays of LSST hardware, public visits to LSST sites, public talks by LSST scientists, and so on. ***Community brokers The SAC is happy to see that the project is starting to develop guidelines for the community alert brokers, and is putting together tools for testing brokers, as well as a timeline for the broker selection process. We are somewhat concerned that the focus of the discussion is on brokers designed to make data available to the world-wide community. However, in practice, there is likely to be a need for a continuum of brokers, from those designed to serve the broad community (like the Antares system being developed by NOAO), to specialized brokers searching for specific types of variables and transients. The latter may be designed by and for small groups of astronomers, who may not want the results to be shared with the world. Guidelines need to be set to describe how many of these specialized brokers can be accomodated, and what the rules will be. The SAC remains unclear to what extent the LSST's simple broker will satisfy the needs of this community; it will require exploring a variety of specific use-cases. It would be useful to have a conversation with the Transients and Variable Stars Science Collaboration to develop those use cases. We understand that the community brokers will make their results available to the world (with no limit on those that do not have formal LSST data access). The SAC remains unclear on exactly what this would include. Each alert will be packaged with some information from the database about the past history of the object in question. Will the community brokers have the ability to query the database for additional information (which may be proprietary to the US, Chile, and International Contributors)? If so, do they have the right to share the results with the world? There will also be a desire to run queries on archival data (not just on the real-time alert stream), for statistical studies of various categories of transients. It remains unclear whether this will be possible. The SAC looks forward to reviewing the document (planned for March 2018) which will describe the requirements and review process for brokers. It should include the proposal process, timescale, and review process for groups to request access to the raw data stream. Exactly what will the data stream these brokers will receive include? What level of support, if any, will the brokers receive from the Project during construction, commissioning, and operations? Will the community or specialized brokers be expected to archive their results for posterity, and if so, how should/could this be integrated with the LSST database and science user interface? Will the answers to these questions depend on the scale of the broker being proposed? ****Observing Strategy The plans for developing the LSST observing strategy were a key focus of previous LSST SAC meetings; we were pleased to see that action is being taken on many of our key recommendations. We are delighted that the Observing Strategy White Paper is now out (arXiv:1708.04058); kudos to Phil Marshall and everyone else who worked so hard on this. We are excited by the news that the group working on the Operations Simulator and Scheduler has grown substantially. We are similarly excited by the plans to use OpSim to carry out massive exploration of the space of possible LSST cadences, including the very promising rolling cadence concept. The time is ripe for new community input in the deep-drilling fields and mini-surveys, and we endorse the idea of a call for white papers from the community for these. We have several recommendations: -There should be a single joint call for white papers for deep drilling fields and mini-surveys. There is potentially a continuum of observing modes between these two, and we do not want to make an artificial distinction between them. -OpSim4 outputs of the current baseline cadence need to be made available to the community well before the deadline for the white papers, so that proposers can put their requests in the context of the existing baseline. -The call for white papers will need to be carefully crafted, explaining in some detail what is and is not appropriate to request. Among the issues that will need to be made clear: -How much time will be available for other than deep-wide-fast observing; -To what extent white papers for variants on deep-wide-fast will be entertained; -What practical or instrumental constraints there will be (e.g., the number of times filters can be exchanged, whether twilight time can be used, what range of exposure times are allowed, etc); -How target of opportunity proposals (think LIGO follow-up!) will be handled; -The fact that the Project will not take formal responsibility for specialized data reduction algorithms needed to process data, including that taken in "non-standard" modes; -The decision-making process going forward; -The fact that the data from any given specialized survey is treated exactly the same way as all LSST data: the proposers have no proprietary access to it. Indeed, the final set of deep drilling fields and mini-surveys may be based on an amalgam of ideas from different white papers; there will be no sense in which a given proposal must be accepted or rejected as-is. -The fact that the detailed cadence for the four existing deep drilling fields, and the existence and parameters for the current suggested mini-surveys (North Ecliptic Spur, the Galactic Plane, and the South Celestial Cap) need justification and finalization, and therefore are also suitable topics for white papers. The SAC would like to review the Call for White Papers before it goes out, to give feedback on these and related issues. Indeed, NASA often puts draft calls for proposals out to the public for comment before they are officially released; LSST may want to do something similar. The nominal timescale has the Call going out in December 2017, which doesn't leave us much time! Once the Call is finalized, it will be important to distribute it widely. Mentioning it in the LSST Weekly News Digest will not be enough! The LSST Communications team should develop a strategy for getting the word out, using social media, the AAS newsletter, targeted e-mailing, announcements at the AAS and other meetings, and so on. Related to this, we need to make sure that the community has the resources they need (including perhaps the ability to run OpSim themselves) to assess the results of any given observing strategy they propose. There should be a continuing effort to solicit new metrics (or ideas for them) from the community. The SAC is concerned that the metrics currently implemented may not fully span the space of important science drivers for the LSST. Finally, given that the SAC will play a major advisory role in selecting among survey strategies, we may want to bring in outside expertise to the subcommittee that will be involved in the decision-making. This will also help with potential conflicts of interest within the SAC itself. ****Looking forward The SAC is likely to be faced with a number of timely topics for discussion over the next months, including a presentation on the LSST science platform (http://ls.st/lsp). We will discuss these by phonecon; the sooner these can be scheduled, the better.