Survey Options Session1 and 2 Particular focus - special cadence requirements or opportunities that deserve more (or less) attention than they have received - candidate “tools” for WFD and mini- surveys
Sessions by topics • Session 1 – morning – Standard visits: 2 exposures vs 1 – Non-standard visits: greater depth in u? – Survey uniformity: depth, seeing, … . – Survey area – trade against number of visits • Session 2 - afternoon – Rolling Cadences: what are the objectives, trades, and constraints? – Dithering: translation, rotation – Length of the observing season: denser sampling vs longer time series
Probable contributors • DESC cadence needs (Dan Scolnic, Humna Awan) • 1 vs 2 image visits (Chris Stubbs) • Strong Lensing (Phil Marshall, Aprajita Verma, ....) • Galaxies (Eric Gawiser) • Simulations for cadence options (Lynn Jones, Owen Boberg, Tiago Ribiero) • Rolling cadences (Peter Yoachim) • Solar system cadences (David Trilling, Henry Hsieh) • AGN (Gordon Richards, Neil Brandt) • Intelligent exposures, dithering (Tony Tyson) • Dwarf galaxy cadence needs (Steve Ridgway) • SN cosmology (Renée Hlozek, Nicolas Regnault) • Comments (Robert Lupton) Variables (Fed Bianco, … .) •
Overview of Session 1 Topics Exposure times and Visit Counts • Standard visits: 2 exposures vs 1 • Non-standard visits: greater depth in u? • Survey uniformity: depth, seeing, … . • Survey area – trade against number of visits
1x30 sec exposure? • Typical 2x15 visit interval 39 sec – Save 1 readout (2 sec) and 1 shutter cycle (1 sec) – Efficiency gain – 7.5% • As a fraction of LSST construction cost ~$50M • Considerations – Reduce data bandwidth and archive volume – Lose science potential of very short gap images – Data loss due to cosmic rays, satellites, glitches • Comments – Chris Stubbs – Robert Lupton – Tony Tyson
Increase read time? • 3-4 sec instead of 2 sec • Improved detector performance • Comments? – Robert Lupton
Visit pattern? • Multiplicity strategy – Facilitate study of moving, rapidly varying targets – Confirmation vs characterization – Same filter, different filter – Temporal pattern – gap length • Multiple visits compete with cadence frequency • Visit pairs reduces number of “epochs” by 2X
Fans of Multiplicity • Solar System (David Trilling, Henry Hsieh, … ?) • Variables, fast transients ………
Increase u exposure - rationale • Improved photon detection efficiency and observing efficiency • Some science benefits from improved u depth – e.g. improves discrimination between faint stars and distant galaxies • Compromise – either decreased number of u-band visits, or reduced time for other filters
Increase u-band depth • With fewer visits* – Yes • Milky Way, Variable Objects – No • Astrometry, Transients, GRB, AGN – Maybe • MW Halo, Cepheid ML, Variable Objects, SN, Large Scale Structure, Cosmology • Preserve the number of visits – Yes – Determining impact on schedule efficiency requires simulation * “Votes” from Survey Strategy paper
Increase u exposure - Simulation kraken_1045 • Double exposure time, retain number of visits • Increase u depth 0.5 mag • Decrease other bands 0.05 mag
Possible champions for increased u-band sensitivity? • Variables? • Milky Way?
Survey Uniformity • Uniform data sets are convenient for science • Observing conditions are variable • Strategies for achiving uniformity – Statistical (no selection or control) – Control of cadence for conditions – Selection of filter/field for conditions • Example – uniform depth
Baseline2018, WFD Square Degrees at Visit Depth
Uniformity by control - depth • Can be actively controlled by adjusting integration time
Relative number of visits 0.57 for median = 0.72 sec 0.95 for mean = 0.85 sec
Baseline2018, Square Degrees at Stacked Depth
Baseline2018, WFD Square Degrees at Visit Depth
Baseline2018, Square Degrees at Stacked Depth – Overlay Random Dithered Healpix
Uniformity by Selection • Uniformity in other parameters: image quality, sky brightness, zenith distance, parallax • Dithering is a profound complication to selection • One approach - reserve specific dithers for particular observing values, and/or to track conditions at all dithers and constrain or repeat dithers as needed. • If rotation of dither must be considered, it is more challenging and probably less efficient to achieve
Interests in pursuing uniformity by selection? • DESC?
Survey Area • Some science benefits from increased sky coverage possibly with reduced cadence – Higher count of targets vs incremental gain from stacking more visits – Better sky coverage • Study distributions on sky - e.g. MW dwarf galaxies – Not all science needs full power of WFD survey • Extend area with limited filter set/cadence
Overview of Session 2 Topics • Rolling Cadences: what are the objectives, trades, and constraints? • Dithering: translation, rotation • Length of the observing season: denser sampling vs longer time series
Rolling Cadences • Why needed? – Low revisit rate in universal cadence • Some of the trades – Shorter inter-visit gaps vs longer seasonal gaps – Rolling cadences can be very complex • Considerations – Cadence less “universal” – Or even heterogeneous – Survey “closure” interval
Why Rolling Cadence? • Ten year survey, 800 visits in pairs means – 40 epochs/year (all filters) – 10 epochs/year (r or i filter) • For an observing season of 8 months – 6 day phase gaps (all filters) – 24 day phase gaps (r or i filter) • Concept – redistribute visits for more dense coverage some time and less dense coverage other times
Potential reduction in phase gaps • Example – assume that half of all visits to a region are available and are deployed to enhance sampling – For one pass in 10 years, 6x reduction – For two passes in 10 years, ~3.5X reduction – For 3 passes in 10 years, ~2.7X reduction • Additional flexibility – ½ season length, no multiplicity – 24x reduction
Comments on Rolling Cadence • SN cosmology requrements (Renée Holzek, Nicolas Regnault, … .) • Rolling cadence simulator developments (Peter Yoachim, … ..)
Rolling Cadences can be complex • Spatial region definitions • Duration of cadence segments • Selection of filters • Cross-talk with regular WFD • Multiplicity of visits • Survey status for annual releases • Some science needs both small phase gaps and long time series • But implementation can be simple – candidate for scripted schedule segments
Phase coverage desert • Time constants larger than 30 minutes and less than 4 days (any one filter) are not well served by uniform cadence or by general purpose rolling cadences • Option – visits deployed as “micro- surveys”. – E.g. rolling cadence season of 30 days applied to each sky region for one roll
Dithering • Relevant discussions at LSST2017 breakout on Sky Tiling
Overview of LSST2017 Sky Tiling Breakout
TESTING LSST DITHER STRATEGIES FOR SURVEY UNIFORMITY AND LARGE-SCALE STRUCTURE SYSTEMATICS Awan, Gawiser, Jones, Zhan, Padilla, Arancibia, Cora, Yoachim • Tested random, haxagonal, spiral dithers • Examined spatial structure in coadded depth • Conclusions – Favor per-visit and per-night dithers – Most dither methods improve estimated number of galaxies – Most methods reduce spurious structure in galaxy counts below statistical
SKY TILING, ROTATIONS, OVERLAPS Chris Stubbs Attaining good sky coverage (<1% gaps) with fixed centers implies roughly 80% of sky gets single-coverage roughly 20% of sky gets double-coverage Open Question • What’s the interplay between photometric calibrations, frame subtraction artifacts, and dithering/rotations?
LSST FoV in MAF Peter Yoachim • Addressed FOV placement as packing problem • Dithering in simulations • Is there any science planned for the overlap regions? Variability on 30s-10min timescales? Need metrics.
LSST FoV in MAF Peter Yoachim
SEVERAL TOPICS Steve Ridgway • Possible efficiency loss with dithering • Difficulty of achieving uniformity in various observing parameters with dithering • Benefits/costs of achieving uniformity in stacked images, and when during survey • Randomizing optics angles
Open dithering issues? • Dither memory and make-up rules • Defining quantitative requirement for rotational dither • Dithering on field edges (1-2% of WFD) • Cross-talk between dither and temporal sampling for short time scale events
Comments on Dithering • Requirements (Tony Tyson) • Simulations …… ..
Observing Season • Trading intervisit gaps against length of season is quite particular to, e.g. – target characteristics – fast, slow – science objectives – catalog, characterize – likely follow-up strategy – lsst follow-up vs external facilities • Some survey objectives tend to shorten season – Improve temporal sampling, lower airmass • Some tend to lengthen it – Characterize slow events – Maximize certain discoveries – Maximize parallax baseline
Indicator of season length
Competition with season length?
Recommend
More recommend
