With this message, we want to give you some details about the SCOC deliberation timeline.
We are recommitting to the overall timeline which we established in 10/2021 (which can be found here The Survey Cadence Optimization Information | Rubin Observatory) and aim at delivering the Phase 2 recommendation in December 2022.
To achieve this, after the delivery of the v2.0 and v2.1 set of simulations (April 2022) which incorporate the SCOC Phase 1 recommendations, the committee is currently exploring several questions through the Metrics contributed by the community (see for example survey_strategy/fbs_2.0 at main · lsst-pst/survey_strategy · GitHub for guiding notebooks).
The specific topics we will discuss in this phase of our deliberation include filter pairing and alternation, detailed inter and intranight cadence, detailed footprint, rolling cadence, DDFs strategies, Early Science, ToOs and Mini-/Micro-surveys. More details on the specific questions we aim to answer with the Phase 2 recommendations can be found below.
At the 2022 PCW the SCOC will share its current considerations and organize a workshop in Fall 2022 (note that the original timeline stated Summer 2022), and we continue to request feedback from the community and the Science Collaborations in particular. Please be aware that the next two months will see intense SCOC activity so any feedback should be provided as timely as possible! The SCOC is happy to continue communicating with the SCs via their liaisons and to receive memos or reports from the SCs if they want to share their considerations this way.
Simulation Suite v2.1 Review Questions
Filter Distribution
Should the survey cadence skew towards bluer filter observations (compared to the current baseline)?
Should the same filter balance be applied to all sky (e.g EG vs Galactic)
What is the exposure time for the u-band observations? (1x30 and 1x50s or exposures)
Should the survey use variable exposure times or set exposure times in each filter?
Should the survey use a different exposure than 2x15s (or possibly 1x30s) for non-u-band filters? (see the shave filters v2.1 simulations)
Nightly Visits pairs and triplets
Should there be a third visit all the time, or on only some observations?
Should we add a third visit everywhere in the sky (e.g. Gal vs EG or by Ecliptic Latitude)
If there is going to be a third visit on a night, what is the spacing between the 2nd and third visit?
If there is no third visit, what is the time separation between visit pairs (33 minutes versus 2-7 hours)?
Footprint
What should the exact Declination and dust extinction limits for the WFD region,
What should the definition of the Galactic bulge region be (e.g. should we add the Virgo cluster to WFD)
How much time is spent observing the Galactic Plane?
How much time is spent observing the NES?
How much time is spent covering the South Celestial Pole?
How much time is spent on pencil beam surveys?
Rolling Cadence
Should a rolling cadence be adopted in the WFD?
Should a rolling cadence be adopted in the special regions of the WFD (NED, GP, SCP) and also in the minisurveys?
Which scheme for rolling should be adopted? (number of bands, other spatial region splits)
How aggressive should the rolling be in the WFD or non-WFD footprint?
When should rolling start (end of year 1 or at 1.5 years)?
DDF Strategy
How much survey time should be spent on the DDFs?
Should all DDFs be observed for the entire 10 years?
Do some DDF fields get more observations in certain years and none in others (rolling DDF strategy)? If so, which ones and how many years do those fields get observed?
Should the Euclid South DDF to be finalized as the 5th field (what else do we need to know about observing and co-observing needs)? Should the Euclid South DDF observed differently to other DDFs?
Early Science
Should the LSST cadence run in the first year or should some other cadence be executed for part of year 1?
Should incremental template generation be prioritized in Year 1?
What community feedback do we need to help decide the early science strategy?
Time allocation for ToO
How many ToOs per year should be observed?
How should time be allocated for ToOs with respect to the Ligo-Virgo runs?
How should ToO observations be coordinated with other groups?
Should ToO observations be discovery-night only or follow up on days scales?
Micro-surveys
What are our priorities for Micro-surveys?
Should any microsurveys occur in Year 1? If so, which ones?
What should happen to the even smaller sub-percent micro-surveys (nano surveys) proposed?
CARINA:
If approved, how much time should be allocated for the micro survey?
Should one week in a year be spent observing the Carina cluster (carina_* run)?
Should this micro-survey be executed every year or only some years?
ROMAN:
Should there be a micro-survey of a Roman microlensing bulge field?
If yes, how much time should be allocated?
LV GALAXIES:
Should there be deeper g-band imaging of 10 local volume galaxies?
If yes, how much time should be allocated?
Should all 10 local volume galaxies receive additional g-band observations?
SMC Microlensing
Should there be high cadence visits in the SMC for microlensing?
If yes, how many pointings should be awarded (1 or 2)?
NORTHERN STRIP
Should there be a northern stripe observed with a limited number of visits in ugrizy from the upper limit of the survey footprint to Dec=30?
If so, how much time should be allocated to the survey?
SHORT EXPOSURES:
Should there be a single short (5s) exposure survey of the sky in ugrizy in year 1 for static sky calibration?
Should there be four short exposures of the sky in ugrizy at a range of times for transient detection and static sky calibration
If so, how much time should be allocated
NEO Twilight:
Should there be a low solar elongation Solar System NEO twilight survey?
If so, how often should this mico-survey be executed?
In response, I would like to share the following reports from TVS members, which summarize our analysis of the recent opsim experiments, and do address some of these questions and offer recommendations. Feedback from the SCOC and MAF team is very welcome:
Should any microsurveys occur in Year 1? If so, which ones?
The team working on young stellar objects and their variability suggests to start with Carina and to use this test case as early science. So we suggest the microsurvey focused on young stellar objects (starting with Carina) to occur in Year 1
CARINA:
If approved, how much time should be allocated for the micro survey?
We request a week long observation with a cadence of 1 visit every 30 minutes for all of the ugri bands during a 10 hours long night, staring with the region of Carina and then repeting the same strategy for other star forming regions, SFRs, (e.g. Orion Nebula Cluster, NGC 2264, NGC 6530, NGC 6611; Cadence Note Bonito & Venuti et al. 2021)
Should one week in a year be spent observing the Carina cluster (carina_* run)?
In the microsurvey focused on young stellar objects and their variability, we propose to start the
exploration at high cadence as described above in Carina Nebula (1 week, possibly in Year 1), and then repeating the same strategy on different SFRs, as e.g. Orion Nebula Cluster, NGC 2264, NGC 6530, NGC 6611
Should this micro-survey be executed every year or only some years?
In the microsurvey focused on young stellar objects and their variability, we propose to observe Carina Nebula possibly in year 1, and then the SFRs listed above in some years of the whole survey
The SCOC emphasizes that our goal is not to present decisions at the PCW but to share the current status of our assessment. The SCOC (via Lynne) will present slides and share the questions we outlined (to reach a broader audience). Where answers are achieved by August ,we will share the current answers and invite comments; where we do not yet have answers, which we expect may be many of the questions, we will share the current considerations, roadblocks, and needed information from the community to reach a recommendation.
The SCOC is working on identifying the best dates for the 3rd workshop which is expected to happen in the Fall (October or November)
A subgroup of the SCOC presented considerations on the Q3 Footprint and associated questions. This question relies on static science and the SCOC has a significant amount of information ready to address this question. Discussions on these questions are ongoing and open SCOC-wide and will continue through August.
The timeline for discussion of the questions is impacted by the recent finding that the time domain DESC metrics are currently producing unexpected results. As most questions, perhaps with the exception of Q3, rely on time-domain considerations, the inconsistencies in these metrics have to be understood and resolved. The SCOC has resolved to organize a meeting with DESC liaisons, the OpSim team, and DESC representatives to find a path to resolving these inconsistencies.
Regarding if there should be a Twilight NEO survey, this is something we have been discussing in the Solar System Working Group. The NEO Twiligth survey will allow for detection of near-Sun objects such as sun-grazing comets, asteroids interior to the orbit of the Earth and the orbit of Venus and possibly interstellar objects like 2I/Borisov which was found in a twilight sky. These kinds of objects have general interest since they push the boundaries of our Solar System and not many are known. The detection and discovery of near-Sun asteroids and comets/interstellar objects has scientific interest in possibly revising existing asteroid population models and will serve as a laboratory for the study of the formation remnants of the original protoplanetary disk as well as interlopers from other planetary systems. @mschwamb and @sgreenstreet can attest as well.
Regarding how often the survey will occur, it should occur, perhaps alternating between evening and morning twilight. There might be a preference for morning twilight since objects discovered in morning twilight will have more time to be followed up before they go into Solar conjunction. @ljones and @yoachim have performed simulations showing the effect of the Twilight survey on the main survey. I will defer to their comment, but in general it seems that the survey will have a detrimental effect if performed every night suggesting that there should be some gap between subsequent twilight survey sessions. It should be noted that if there is too long a gap between twilight survey sessions it might result in it being more likely that objects will be lost due to not enough self follow up by LSST.
It might also be noted that a cadence where there are more exposures per foot print might be more resilient to satellite trails which could be more prevalent with starlink and other mass satellite constellations. The more exposures per footprint will allow for the recovery of detections in the event that a satellite crosses an asteroid detection.
The motivations supporting Solar System science and Planetary Defense (near-Earth objects) for a near-Sun twilight micro-survey are discussed in the near-sun survey white paper. (DocuShare License Error) They need not be belabored here, but Rubin will otherwise spend very little or zero time observing valuable near-Sun targets.
Surveying near-Sun also lengthens the duration of study of all time varying targets, Solar System or not, during any particular apparition as virgin sky appears a degree per day from behind the Sun in the morning sky and is subducted again behind the Sun a degree per day in the evening sky. It is true that many targets are best observed when they are high in the sky transiting the local meridian. While other targets are only observable near-Sun. Not every decision is a trade-off. If Rubin does not observe near-Sun, it will not see objects that remain near the Sun.
Whatever the frequency of executing the near-Sun survey, Rubin should remain flexible to adjusting this allocation seasonally or from year-to-year of operations. Early during the operations phase enough twilight hours should be scheduled near-Sun during both evening and morning hours to statistically evaluate the benefits against the actual sky. These should be scheduled frequently enough to retain rapidly moving targets like NEOs.
If LSST were operated as a pure NEO survey (with no other science considerations) it would use the well-tested NEO cadence of four or more exposures per field separated by ~8-10 minutes, and the fields would tile the sky in a regular pattern such as declination stripes from western horizon in the evening to eastern horizon in the morning. It would repeat fields containing likely NEOs a few hours later. Community tools (https://neofixer.arizona.edu) could recommend priorities. The point of making these assertions is not to debate the universal cadence or its rolling variations, but to suggest that during hours spent on the near-Sun twilight survey that more regular search patterns and higher number of per-field repeat exposures should be evaluated both on-sky and through simulations. These issues were discussed in the white paper. A universal cadence is mediocre for all science cases, but is more mediocre for some.
Also early in the Survey, and perhaps throughout, observing in twilight, especially at low elevations to the western and eastern horizons, will best sample the LEO satellite constellations when they are most illuminated. Acquiring such data in twilight may aid in characterizing, modeling, and removing these streaks outside of astronomical twilight. Tracking the anticipated year-over-year growth of the constellations will aid in strategic planning for handling this threat. Conceivably there may be additional direct benefits to Space Situational Awareness and spacecraft traffic management use cases, perhaps funded by the satellite operators or other LEO stakeholders outside the astronomical community.
The SCOC outlined a plan for working with DESC to resolve time-domain metric inconsistencies in a short scale. The OpSim team and liaisons are working with the DESC to achieve this goal.
The SCOC discussed potential dates for Workshop 3: based on the availability of SCOC members dates in late August and early October were considered. The SCOC resolved that late August is too early and too close to the PCW to make the best of this meeting and proposed to consider the second week of October or potentially early November dates. Availability of members of the SCOC and the timeline needed to generate a Phase 2 report draft will be considered to finalize these dates.
Notebooks that summarize the impact of changes on the strategy consistent with the questions the SCOC is investigating have been and are being prepared by Lynne. A notebook to investigate changes in filter balance was presented as an example.
The SCOC is strategizing to move the modality of interaction to online-asynchronous to facilitate work in the summer months. A timeline for SCOC-wide meetings closer to the PCW will be finalized.
For ease of reference I would like to summarize the various feedback from TVS. Most have already been posted to this thread, but the extra-galactic report in particular has recently changed so we recommend review.
Also, a question regarding the proposal for microsurvey observation of Star Forming Regions. Sara Bonito previously reported differences between the v1.7 and v2.0 implementations of the recommended strategy, and it is unclear why this has changed. We understand that the MAF team were investigating this. Has there been any update? Thanks in advance.
Hi Rachel, We did find that the sequences were somehow being cut in half in the new v2.0 simulation, I am not sure why but @yoachim was looking and was going to rerun this simulation. We’ll update with a new version.
Early-time observations of the Carina Nebula will be important to explore the other SFRs after the analysis of the first results for variability in YSOs at different timescales. Moreover, an early scientific analysis of the Carina Nebula will allow us to develop reliable photometric tracers of YSO variable populations in a combination of Rubin LSST filters that can then be used to conduct a blind search of the entire Rubin LSST field to identify new pre-main sequence populations, thus realizing the full discovery potential of the survey beyond the ∼ 500 pc radius within which most star formation studies have traditionally been confined.
[Notes from the meeting before the PCW were not posted as the meeting focused on PCW preparation and the PCW Opsim session was essentially a direct outcome of the meeting]
The committee thanked Lynne Jones for excellently representing our views at the PCW.
Lynne Jones reported receiving new Galactic footprint and filter balance input from the TVS and SMWLV teams and performing 4 new simulations combining two options each for footprints and filter balances. The SCOC Rolling Cadence subgroup will interact with TVS and SMWLV liaisons to understand the GP filter balance and if the best rolling cadence options for the Galactic sky differ from those for the rest of the sky.
The SCOC reviewed the SAC recommendations and is planning interaction between the Early Science subcommittee and Operation teams
The full SCOC convene on 10/11. Between meetings the SCOC has been collaboratively editing documents and subgroups dedicated to specific questions have worked both synchronously and asynchronously.
On 10/11 the SCOC reviewed the plan for needed simulations of Rolling Cadence, DDFs, and Galactic Plane coverage. The simulations were described but not yet available to the SCOC due to a minor bug in the OpSim code which was on its way to be fixed. The SCOC expected to receive the simulation within roughly one week.
The SCOC reviewed the current DDF simulations and metrics that address their effectiveness. The DDF subgroup reviewed the current status of their simulation and described pending decisions. In particular as they related to the DESC SN science. The SCOC planned on closer interaction with the DESC Observing Strategy Working Group to better understand the SN Ia performance of DDF strategies.
The SCOC expressed a desired to meet more frequently as we ramp up to Nov2-3 workshop and to the delivery of our recommandations. Accordingly, weekly meetings have been scheduled for the full SCOC while subgroups continue to be charged to meet or work asyncronously to the resolution of the remaining uncertainties in the recommandation
The SCOC met on 10/17. The discussion focused on reviewing information received recently on
the performance of recent simulation strategies on the Deep Drilling Fields as measured by the DESC SN Ia metric
the performance of rolling cadence strategies as measured by all relevant MAFs.
The SCOC has prepared recommendations on most topics, as already presented at the PCW, with a few pending issues that were discussed in this meeting:
The Deep Drilling strategy implemented in v2.0 and v2.1 simulations (e.g. baseline) produces poor throughput for the DESC SN Ia metric that measure numbers of SNe observed sufficiently well based on constraints - the performance of the metric seems impacted primarily by~days observing gap between nights when the DDFs are observed (thes gaps improve efficiency by reducing filter changes but SN Ia observed with this strategy do not pass the SN Ia metric cuts). The SCOC is assessing how “compartimentalizable” the DDF implementation can be, particularly varying the days gap between DDF sessions is to enable a simulation of our recommendation with pending improvements on the DDF strategy. That is: we expect to produce a simulation for the November 2nd workshop that will reflect our WFD and other survey’s recommended strategy without finalizing the DDF strategy in detail; the survey strategy is assessing the feasibility of this plan.
The Rolling cadence strategy decisions that are still pending are on the number of bands to be used and rolling strength. The SCOC planned on reviewing 4 more simulations that vary the number of bands and rolling strength that will help finalize the decision. We expect this decision should be achieved by late this week.
The Survey Strategy team is creating simulations ahead of the November 2nd/3rd workshop. We expect to share a single simulation with the community early next week that reflects our recommendation, with a “best guess” as to what the final recommendation may look like for all decisions that are still being finalized (as described above), and a short write up that represents a draft of the document to be released in December.
The full SCOC will meet again on 10/28 and work asynchronously to review the simulations the SCOC is producing and the associated writeup. Subgroups will meet to finalize the above questions
Is it possible for the notes from the 10/28 meeting to be posted? It may be helpful for the breakout sessions today. Thanks. I understand that might not be possible.