A Northern Ecliptic Survey for Solar System Science
Megan E. Schwamb, Kathryn Volk, Hsing Wen, Michael S. P. Kelley, Michele T. Bannister, Henry H. Hsieh, R. Lynne Jones, Michael Mommert, Colin Snodgrass, Darin Ragozzine, Steven R. Chesley, Scott S. Sheppard, Mario Juric, Marc W. Buie
DDraft version December 5, 2018
Typeset using L A TEX preprint style in AASTeX62
A Northern Ecliptic Survey for Solar System Science
Megan E. Schwamb, Kathryn Volk, Hsing Wen (Edward) Lin, Michael S. P. Kelley, Michele T. Bannister, Henry H. Hsieh, R. Lynne Jones, Michael Mommert, Colin Snodgrass, Darin Ragozzine, Steven R. Chesley, Scott S. Sheppard, Mario Juri´c, and Marc W. Buie Gemini Observatory, Northern Operations Center, Hilo, HI USA Lunar and Planetary Laboratory, The University of Arizona, Tucson, USA Department of Physics, University of Michigan, Ann Arbor, MI, USA University of Maryland at College Park, College Park, MD, USA Astrophysics Research Centre, Queen’s University Belfast, Belfast, United Kingdom Planetary Science Institute, Tucson, AZ, USA University of Washington, Seattle, WA, USA Lowell Observatory, Flagstaff, AZ, USA University of Edinburgh, Edinburgh, UK Brigham Young University, Provo, UT, USA Jet Propulsion Laboratory, CA, USA Department of Terrestrial Magnetism (DTM), Carnegie Institution for Science, Washington, DC, UA Southwest Research Institute, Boulder, CO, USA (Dated: November 30, 2018)
ABSTRACTMaking an inventory of the Solar System is one of the four fundamental sciencerequirements for the Large Synoptic Survey Telescope (LSST). The current baselinefootprint for LSST’s main Wide-Fast-Deep (WFD) Survey observes the sky below 0 ◦ declination, which includes only half of the ecliptic plane. Critically, key Solar Systempopulations are asymmetrically distributed on the sky: they will be entirely missed,or only partially mapped, if only the WFD occurs. We propose a Northern EclipticSpur (NES) mini survey, observing the northern sky up to +10 ◦ ecliptic latitude, tomaximize Solar System science with LSST. The mini survey comprises a total areaof ∼ /604 fields, with 255 observations/field over the decade, split between g,r, and z bands. Our proposed survey will 1) obtain a census of main-belt comets;2) probe Neptune’s past migration history, by exploring the resonant structure of theKuiper belt and the Neptune Trojan population; 3) explore the origin of Inner Oortcloud objects and place significant constraints on the existence of a hypothesized planetbeyond Neptune; and 4) enable precise predictions of KBO stellar occultations. Thesehigh-ranked science goals of the Solar System Science Collaboration are only achievablewith this proposed northern survey. Corresponding author: Megan E. [email protected] a r X i v : . [ a s t r o - ph . E P ] D ec WHITE PAPER INFORMATION1.
Science Category:
Taking an Inventory of the Solar System2.
Survey Type Category: mini survey3.
Observing Strategy Category:
An integrated program with science that hinges on thecombination of pointing and detailed observing strategy SCIENTIFIC MOTIVATIONA foundational goal of the Large Synoptic Survey Telescope (LSST) is to map the Solar System(Ivezi´c et al. 2008; LSST Science Collaboration et al. 2009). Multiple major small body populations(described below) are key windows into understanding our Solar System’s formation and evolution,but are asymmetrically distributed on the sky. They will be partially mapped or completely missedwithout coverage of the Northern ecliptic, which is absent from the Wide-Fast-Deep (WFD) footprint.Other yet-unseen asymmetric distributions are likely to exist, and will only be found by surveyingthe entire ecliptic. To achieve the goals in the LSST Solar System Science Collaboration’s roadmap(Schwamb et al. 2018), we propose a Northern Ecliptic Spur (NES) mini survey covering up to +10 ◦ ecliptic latitude. Main-Belt Comets:
Main-belt comets (MBCs) occupy dynamically asteroidal orbits betweenMars and Jupiter, yet exhibit comet-like activity near perihelion due to sublimation of volatile ices(Hsieh & Jewitt 2006). Fewer than a dozen are currently known, where they are considered valuableprobes of volatile distribution in the Solar System’s primordial disk. They comprise a subset of theactive asteroids, which are dynamically asteroidal objects that exhibit activity due to sublimation,rotational destabilization, impacts, and other effects (Jewitt et al. 2015). Kim et al. (2018) find thatalmost all of the known MBCs reach perihelion (and thus become active) in the same direction asJupiter’s perihelion, clustering in our proposed NES survey region (Figure 1). A NES mini surveyis needed to (a) determine whether this alignment of MBC perihelia is maintained as more MBCsare discovered, and if this apparent alignment is verified to be real, (b) discover smaller MBCs,which will only be bright enough to detect when near perihelion and active, and (c) monitor knownmain-belt asteroids for activity at times when they are most likely to become active.
The Kuiper Belt’s Structure and Neptune’s Migration:
The detailed structure of the Kuiperbelt, the swarm of planetesimals orbiting beyond Neptune, provides important constraints on earlySolar System dynamical history. The populations now in mean-motion resonance constrain Neptune’sorbit during its outward migration. The number of Kuiper belt objects (KBOs) in each resonanceconstrain high-eccentricity phases and/or semi-major axis jumps during Neptune’s migration (e.g.,Malhotra 1995; Levison et al. 2008; Nesvorn´y & Vokrouhlick´y 2016). Detailed distribution insidewithin resonances is also valuable. For example, the ratio of KBOs in the leading and trailing librationislands in Neptune’s 2:1 resonance, can act as a speedometer for Neptune’s migration (Murray-Clay& Chiang 2005). KBOs are exceptionally distant, have a steep size distribution, and are thus faint:their discoveries are strongly biased toward detection at perihelion. Detectable resonant KBOs areasymmetrically distributed on the sky, as they come to perihelion at specific geometries relative toNeptune (e.g., Gladman et al. 2012). KBOs traverse a minute fraction of their orbit during LSST’s10-year baseline. Surveying the entire ecliptic is critical to observing enough resonant KBOs to makethese tests. In the absence of the NES, a substantial fraction of key orbital groupings within theseimportant resonances will be completely missed.
L4 Neptune Trojans:
Neptune Trojans co-orbit with Neptune around its L4 and L5 Lagrangianpoints, emplaced during Neptune’s migration. Their orbital/physical property dependencies andL4/L5 population asymmetries are important probes both of Neptune’s dynamical history and theSolar System’s primordial disk. Lin et al. (2018) discovered the first ultra-red Neptune Trojan, simi-lar to the ultra-red surfaces seen residing within the Kuiper belt (see Figure 2). With an inclination ∼ ◦ , this discovery may show that ultra-red surfaces only occur at high inclinations, but the originof this surface type remains unknown. Lin et al. (2016) also find that the larger (H <
8) NeptuneTrojans have lower inclinations. Only 19 L4 and 3 L5 Neptune Trojans are known to date; moredetections with LSST are needed to confirm these correlations. Figure 2 plots the on-sky positionsof simulated Neptune Trojans. L5 Neptune Trojans will have good coverage within the WFD sur-vey and the wide inclination distribution of Neptune Trojans make part of the high-inclination L4Trojans detectable in the WFD. However, the majority of low-inclination L4 Trojans will be missingwithout the NES. The NES is crucial to test the size-inclination and color-inclination dependenciesand symmetry in properties between the L4 and L5 Neptune Trojans.
Planet 9 and the Origin of the Inner Oort Cloud:
Inner Oort Cloud objects (IOCs) areon highly elongated orbits with perihelia beyond 45 au and semi-major axes greater than 250 auand less than 2000 au (Brasser & Schwamb 2015). IOCs are not significantly influenced by theknown inner giant planets or outside forces, but were emplaced by some sort of dynamical interac-tions, possibly from past stronger outside forces, such as would happen in a dense stellar cluster orfrom an unseen massive planet (Brown et al. 2004). As shown in Figure 3, all of the few knownIOCs come to perihelion at similar locations on the sky, which is proposed to be from a distantplanet gravitationally shepherding the IOCs onto similar orbits (Trujillo & Sheppard 2014; Batygin& Brown 2016). The IOCs appear to cluster in longitude of perihelia near an RA of 4 ± KBO Stellar Occultations:
A NES mini survey will recover most of the ∼ TECHNICAL DESCRIPTION3.1.
High-level description
We propose a mini-survey with observations covering the ecliptic plane beyond the region coveredwithin the main Wide-Fast-Deep (WFD) Survey footprint, in order to fully sample small bodypopulations throughout the Solar System. Our proposed NES contains the missing 50% of the totalarea of the ecliptic on the sky that is not contained within the WFD survey – and has about the
Figure 1. (a)
Directions of the longitudes of perihelion of outer-main-belt (OMB) MBCs whose activity isattributed to sublimation or a combination of sublimation and rotational destabilization. Adapted from Kimet al. (2018). (b)
Sky positions of sublimation-driven OMB MBCs, and sublimation and rotation-drivenOMB MBCs when at ν = 30 ◦ (i.e., when peak activity is expected) over the course of the LSST survey. Figure 2. Left:
The color-inclination relation of Neptune Trojans. The only known extra-red NeptuneTrojans has second highest inclination of ∼ ◦ (red circle). Right:
The on-sky positions of Neptune Trojansin 2022 (grey) and 2032 (yellow).
Figure 3.
Alignment of known inner Oort cloud orbits, with perihelia beyond 45 au and semi-major axes250 < a < same fraction of Solar System small bodies at any time. In the absence of the NES, the substantialfraction of objects near the ecliptic in the Northern hemisphere would be completely missed. Weare requesting 255 visits per field over ten years in grz , over our mini survey region (the ‘NorthernEcliptic Spur’ region, NES) reaching from the northernmost limit of the WFD up to an eclipticlatitude of +10 ◦ (see Figure 4). We assume LSST’s discovery and attribution performance will be asdescribed in Jones et al. (2018). The cadence of observations is important, in order to enable linkingand tracking; we are requesting 6 visits in pairs per night, for each of 5 months in 7 ‘Discovery’years, with 15 visits per year over (split between 5 months) in 3 ‘Tracking’ years; this is similarto but dramatically less densely sampled than the WFD baseline strategy. Section 3.7 describes indetail our preference for how these observations should be divided over the 10-year LSST operationalbaseline. The details in this section, including the number of fields and fraction of time required,is based on analysis of a series of simulations created with the LSST Operations Simulator ( OpSim ,Delgado et al. 2014) for the call for cadence optimization white papers. -75°-60°-45°-30°-15°0°15° 30° 45° 60° 75°
Figure 4.
Light blue represents the pointings requested in our proposed Northern Ecliptic Survey. Thesolid blue line represents the ecliptic. The dashed blue lines represent ±
10 degrees ecliptic latitude. Thesolid green line plots the center of the galactic plane. The dashed green lines reflect ±
10 degrees galacticlatitude.
Footprint – pointings, regions and/or constraints
In order to complete coverage of the ecliptic plane, we propose a mini survey footprint rangingfrom the northernmost top of the WFD ( ≈ δ = 0 ◦ ) up to an ecliptic latitude of +10 ◦ (see Figure 4).This correlates to 604 distinct LSST fields, using the pointing tessellation provided in the currentsimulations. This sky coverage is the best compromise between the northern declination limit of thetelescope and the inclination distributions of the Solar System’s small body reservoirs. Althoughorbits of all inclinations cross the ecliptic, those bodies on orbits with higher inclinations spend mostof their time away from the ecliptic plane. Additionally some of the KBO resonant populations areperturbed by Neptune such that they come to closest approach off of the ecliptic. By covering fieldsto 0 degrees declination as the lower extent of the NES, we ensure adequate coverage of all our keyNorthern Solar System populations. 3.3. Image quality
The image quality used in the NES images should be similar to what is used in the WFD. There areno special constraints on image quality beyond what has already been set in the current operationsimulations of NES fields and what is needed to achieve our desired individual image depths. In par-ticular, the image quality (seeing) constraints for northern ecliptic observations in the baseline2018aand kraken 2026
OpSim runs are sufficient for our needs.3.4.
Individual image depth and/or sky brightness
The individual image depth is important (and thus its implied constraints on image quality andsky brightness), as moving objects must be detectable in individual images. As the goal is to discoversmall bodies at all locations in the ecliptic plane without introducing significant bias, the individualimage depths in this mini survey region should be similar to the individual image depths in theWFD footprint. There are no hard cutoffs, but there is an overall preference for visits to be as deepas possible without sacrificing sky coverage. Past wide-field Solar System surveys have reached alimiting magnitude of ∼ σ limiting magnitude per exposure in r and g must be greater than 23rd magnitude with exposuretimes of 30s or more. We propose the same total exposure time per visit as the WFD (30s per visit),which currently meets our detection goals. In particular, individual image depths in the northernecliptic observations in the baseline2018a and kraken 2026 OpSim runs are sufficient for our needs.Specific details on the Solar System Object Differential Completeness Goals that are desired for boththe WFD footprint and this proposed NES mini survey are detailed in the Community ObservingStrategy Evaluation Paper (COSEP; LSST Science Collaboration et al. 2017). Additionally, we notethat sky brightness can be a particular concern for this region, as the distance to the moon will tendto be small, but is not a constraint in and of itself.3.5.
Co-added image depth and/or total number of visits
There are no constraints on the co-added image depth. We do have constraints on the totalnumber of visits, due to cadence preferences and requirements for identifying moving objects andcharacterizing their orbits and physical properties. These are discussed below in more detail, butresult in a total number of visits on the order of 255 per field. 3.6.
Number of visits within a night
At least two visits per night to a field are required to detect and identify moving Solar Systemobjects. At least three separate nights are required to identify and link newly discovered movingobjects (see the high-level requirements for LSST project’s Moving Object Pipeline System (MOPS)defined in LSE-30 and LDM-156 ). To guarantee discovery of Solar System bodies at 95% confidenceby MOPS, three tracklets (a pair of images in the same night, acquired no more than 90 minutesapart) acquired within 15 days are needed. The most distant objects in the Inner Oort cloud regionbeyond 200 au, do not move appreciably within a single night, but with two visits per night for eachfield we can use those the two epochs to weed out the faster/closer moving Solar System bodies inorder to optimize and speed up the search algorithms. Thus, we propose at least two visits per nightto each field when possible. 3.7. Distribution of visits over time
To balance our key science cases for the NES we spread the ‘Discovery’ and ‘Tracking’ years asdescribed in Table 1, below. This cadence allows orbital recovery for the different populations andmaximizes temporal coverage for main-belt comet and asteroid collision discovery. An additionalrequirement during the ‘Discovery’ years is an observation schedule that supports MOPS discoveryof main-belt objects. MOPS requires a pair of visits per night, in at least three separate nights,within 15 days in order to identify and link minor planets (resulting in a 95% confidence of discovery;LSE-30; LDM-156). Each pair consists of two visits within the same night, no more than 90 minutesapart. Once the discovery criteria are satisfied for a given field in the ‘Discovery’ year, the remainingobservations can be scheduled more flexibly. During ‘Tracking’ years, the measured orbits of previousdetected asteroids found in ‘Discovery’ years will be used to predict the locations of these bodies inthe tracking observations to check on cometary activity.
Year Observation Type Summary1 Discovery 6 observations per field per month for 5 months2 Discovery 6 observations per field per month for 5 months3 Discovery 6 observations per field per month for 5 months4 Tracking 15 observations per field divided over 2-3 months5 Discovery 6 observations per field per month for 5 months6 Discovery 6 observations per field per month for 5 months7 Tracking 15 observations per field divided over 2-3 months8 Tracking 15 observations per field divided over 2-3 months9 Discovery 6 observations per field per month for 5 months10 Discovery 6 observations per field per month for 5 months
Table 1. Proposed NES Year-by-Year Summary
The population of MBCs we have discovered in the past decade are likely the brightest and youngestmembers. With LSST’s superior sensitivity, we will search for activity around all asteroids, with http://ls.st/LSE-30 http://ls.st/LDM-156 additional scrutiny for those in the outer main-belt. For known asteroids, we only require a fewobservations of MBC-like asteroids near perihelion, rather than the more strict circumstances neededfor discovery. This is readily satisfied in both our ‘Discovery’ and ‘Tracking’ years as described above.The ‘Discovery’ years are also crucial for the MBCs. Cometary activity enhances the brightness, andtherefore discoverability of MBCs. However, detectable activity for the currently known MBCs hasbeen limited to brief periods near perihelion, typically only a few months long but sometimes evenshorter (Hsieh et al. 2015). Being outer main belt objects, MBC have orbital periods of ∼ ∼ σ semi-major axis uncertainties below ∼ ∼ Filter choice
Solar System minor planets, visible from reflected sunlight, are brightest in the mid-optical wave-lengths. Inner Solar System objects will move sufficiently over LSST’s 10-year operational baselineto be imaged at some point in all 5 filters in the WFD footprint. Over a decade Outer Solar Systembodies will not move much in their orbits, such that nearly all of the KBOs imaged in year 1 inthe NES survey will remain within the NES footprint in year 10. Thus, our filter choice maximizesthe return on Outer Solar System science. Since the NES fields would receive fewer visits thanWFD fields, we prioritize the bulk of observations in r , with some g and z -band imagery for surfacecolor/composition studies.As LSST does not have a wide-band gri -type filter as used, e.g., by Pan-STARRS (Chambers et al.2016), optimum detection efficiency will be in the r band; we thus prioritize observations in this filter,requesting 60% of the total observation in each field be taken in r . Spectral slope from solar-neutral tosolar-reddened can be minimally parameterized with the use of g . However, red surfaces will be veryfaint in g , thus a comparatively substantial 25% of time must be allocated to obtain sufficient SNRto ensure moderate-quality and (potentially) non-simultaneous g − r colours on LSST’s detections.Pike et al. (2017) showed that the cold classical population of TNOs display a distinct colour in thecolour-space of the filters g, r, z . We thus add z as our third filter for the remaining 15% of time, inpreference to i , where no such population distinctiveness is seen. Most minor planets are very faintreflectors in u , so we (reluctantly) omit it from our mini survey.Observations within a single night do not necessarily need to be in the same filter, however we willbe constrained in detection efficiency by the shallower limiting magnitude of the pair. Outer SolarSystem objects have much redder surfaces compared to the inner Solar System bodies. ( g − r ) colors1range to 1 for the very red surfaces in the dynamically excited and cold classical Kuiper belt. Tomaximize the detection of the reddest KBOs and IOCs, we ideally request that when possible thenightly pairs be taken in r . When this is not possible, we request the second nightly visit should be in g since Solar System objects will be faintest in z-band. We request that the various g, r, z nightly filterpair combinations in a NES mini survey observations with a cadence as proposed here be simulatedto better quantify the impact on discovery metrics with an improved KBO SED (spectral energydistribution) in the OpSim moving object package.3.9.
Exposure constraints
The goal of the NES is to detect sufficient numbers of Northern Solar System objects to characterizethe asymmetric distributions of MBCS, resonant KBOs, Neptune Trojans, and IOCs. Ideally, ourrequirement would be that the proposed NES mini survey is to have the same or better sensitivityto Solar System objects as in the Southern Ecliptic covered in the WFD footprint. This suggeststhat we aim for a similar NES detection threshold as in the Southern Ecliptic that will be coveredby the WFD footprint. There are trade-offs between exposure time and coverage (discussed furtherbelow), but our nominal plan uses the same 30s exposure time in the NES as in the WFD survey.Though longer exposures go deeper, the loss in coverage is probably detrimental to our coverage goalswhich are crucial since these populations are relatively sparse on the sky. Future work could employmetric-driven optimization to investigate these details.3.10.
Other constraints
None noted. 3.11.
Estimated time requirement • In total 604 fields that have field centers greater than or equal to declination of 0 up to anecliptic latitude of +10 degrees. • • • In total 255 visits for each fields in ten years. • Time required per visit is (30 second exposure time + 3 seconds slew/settle + 2 seconds shutteropen/close) = 35 second for one snap (see 3.12.6). It will be (2 ×
15 second exposure time +3 seconds slew/settle + 2 × ×
255 visits ×
35 second) = 5,390,700 second = 1,497.4hours for one snap or (604 fields ×
255 visits ×
37 second) = 5,698,740 second = 1,583.0 hours fortwo snaps. This is approximately equivalent to 187 nights total, over the lifetime of LSST, or about5-6% of the total available time. Noted that the NES will request 176.2/186.2 hours (one/two snaps)for a ‘Discovery’ observed year and 88.1/93.1 for a ‘Tracking’ year.3.12.
Technical trades
What is the effect of a trade-off between your requested survey footprint (area) and requestedco-added depth or number of visits Properties ImportanceImage quality 1Sky brightness 3Individual image depth 1Co-added image depth 3Number of exposures in a visit 3Number of visits (in a night) 1Total number of visits 2Time between visits (in a night) 1Time between visits (between nights) 1Long-term gaps between visits 2Separation between First and Final observation 2Filter Selection 1Number of Snaps in a Visit 3
Table 2. Constraint Rankings:
Summary of the relative importance of various survey strategy con-straints, ranked from 1=very important, 2=somewhat important, 3=not important.
Trading survey area for co-added depth/number of visits will lead to an increasingly biased sampleof Solar System object discoveries or a decreased number of discoveries. Decreasing the NES surveyarea will likely decrease our longitudinal or inclination coverage, adversely affecting the observeddistributions of objects in the Kuiper belt resonances, Neptune Trojan clouds, MBC reservoirs, andthe Inner Oort Cloud. Increased co-added depth or number of visits per field does not make up for themissing orbital phase space. In addition, discovery of moving targets requires multiple observationsof the same field. Reducing the number of visits per field in order to increase areal coverage willaffect the moving object pipeline’s ability to discover unknown moving objects.3.12.2.
If not requesting a specific timing of visits, what is the effect of a trade-off between the uniformityof observations and the frequency of observations in time? e.g. a ‘rolling cadence’ increases thefrequency of visits during a short time period at the cost of fewer visits the rest of the time, makingthe overall sampling less uniform.
Our science goals can be carried out with the NES being observed with a higher cadence of ob-servations in 7 years for discovery and orbit characterization with 3 years of sparser monitoringobservations in between these discovery/orbit characterization years. There is some flexibility inwhen the ‘Tracking’ lower number of observation years are scheduled as described in the Sectionsabove. We have some flexibility between the frequency of observations and uniformity of observa-tions, as long as the cadence of the visits are such that during ‘Discovery’ years MOPS is able tosuccessfully run and detect moving objects. MOPS needs three tracklets (a pair in the same nightmade no more than 90 minutes apart) over 15 days, to guarantee minor planet discovery with 95%confidence (see LSE-30 and LDM-156). Once the discovery criteria are satisfied, additional observa-tions should be scheduled in nightly pairs when possible, but the frequency of the observations fromyear to year can vary.33.12.3.
What is the effect of a trade-off on the exposure time and number of visits (e.g., increasing theindividual image depth but decreasing the overall number of visits)?
Increasing the image depth would increase the 5- σ limiting magnitude which would increase thenumber of objects detected, but we require at least 2 observations per night in 3 pairs for MOPSidentification in our ‘Discovery’ years. The additional visits proposed are to characterize and confirmthe orbit of the new NES discoveries. At least two years of observations are needed to fully secure theouter Solar System orbits. Decreasing the number of observations will also lower the opportunitiesto detect main-belt comets. MBCs are mainly expected to be visible when active (i.e., brighter), sodecreasing the number of visits will decrease the chances of finding main-belt comets.3.12.4. What is the effect of a trade-off between uniformity in number of visits and co-added depth?
All of our science goals are constrained not by co-added image depth, but by the 5-sigma detectiondepth in individual LSST frames. There is no significant gain in trading off the uniformity in numberof visits for increasing co-added depth.3.12.5.
Is there any benefit to real-time exposure time optimization to obtain nearly constant single-visitlimiting depth?
There would be a small benefit for real-time exposure time optimization to obtain a nearly constantsingle-visit limiting magnitude, as the 5-sigma limiting depth of exposures drives what Solar Systemobjects will be detectable in the proposed NES observations. Given the high airmass ( >
2) that theseobservations are normally scheduled, the expected benefit from exposure time optimization will besmall (see http://astro-lsst-01.astro.washington.edu:8080/allMetricResults?runId=2)3.12.6.
Are there any other potential trade-offs to consider when attempting to balance this proposal withothers which may have similar but slightly different requests?
Snaps:
Aside from questions of image quality and cosmic ray rejection, which we do not consider here, theSSSC finds very little benefit in having two 15-sec snaps co-added to form one 30 sec visit. Rather,the gain in survey efficiency from eliminating the time lost for the shutter throw and CCD (Charge-Coupled Device) read-out between snaps would be better used for additional observing time duringthe survey. Furthermore, combining the CCD read with the slew between visits allows for slowerread times and thus reduced read noise.There are two possible benefits to SS science from separate snaps:1. Snaps would allow us to ascertain the direction of motion of trailed Solar System detections,which could potentially ease linking to companion trails in the transient stream. However,if there is a companion, it must be near one of two obvious positions, with a known lengthand orientation. The companion can be found by searching both directions, leading to only atwofold increase in computational effort (for a relatively small number of objects with significanttrailing).2. A few small near-Earth asteroids rotate rapidly enough that they have photometric variationon the time scale of 15 s, so rotation information could be extracted from two snaps. However,this is only for a small fraction of small objects, and thus represents a tiny fraction of the smallbody object catalog. Moreover, it is not clear whether the photometric variation from snapswould be sufficient to constrain the rotation period of so-called super-fast rotators.4Based on the priorities in the SSSC’s Science Roadmap (Schwamb et al. 2018), we consider thesebenefits as minor in comparison to gaining an addition 1-2 × additional survey observations whichwould increase the number of Outer Solar System detections in key populations. Moving the Wide-Fast-Deep survey to one 30s snap per observation would add 7 percent of the operations time toon-sky observations. The Solar System metrics described in the COSEP (LSST Science Collaborationet al. 2017) and in this white paper will show no negative impact from moving to one 30s snap. Thus,we advocate for the elimination of snaps in order to accommodate observing the Northern EclipticSpur and other proposed mini surveys and deep drilling fields. Filter Selection:
As described in Section 3.8, g , r ,and z -band observations best suite on science cases, with the majorityof the proposed observations taken in r -band. If the NES is restricted to single-band observations,our minimum discovery needs require r -band. We note that observing the NES in filters without r or g observations would result in significant losses for Solar System detections based on the discoverymetrics in the COSEP (LSST Science Collaboration et al. 2017) and the metrics described in Section4. Extended WFD Footprint:
We have proposed the minimum number of observations and filters that we believe will achieve ourkey science goals. Increasing the number of filters and increasing number of visits for all or part ofthe NES in order to accommodate other science cases such as an extended WFD footprint will notnegatively impact our science goals, as long as the majority of the 604 fields in the NES region aresurveyed. Additional visits would enable better characterization of rotational variability and provideincreased sampling in the search for MBCs.
Distribution of Observations Within Tracking Years:
We have some flexibility in scheduling observations within Tracking years if there are strong tensionswith other proposed observing needs. Instead of evenly distributing nightly over the months thefield is observable, for example, Tracking Years could consist of three pairs of visits to a field usingthe same cadence used during ‘Discovery’ years in the month when a field is at opposition, and onepair of visits per month in the 2 months before and 2 months after opposition. This approach wouldpreserve some minimal ability to discover newly active objects (which may have been too faint dueto the absence of activity to be detectable in previous years) during ‘Tracking’ years, while alsocontinuing to monitor known objects throughout their available observing windows.
A Big Sky Approach to Cadence Diplomacy:
We note that our requested NES footprint is also part of the extended footprint proposed in the ‘ABig Sky Approach to Cadence Diplomacy” White Paper (Olsen et al) . Our proposed NES minisurvey is compatible with the observing scheme outlined in their proposal. We propose that eachNES field receive approximately 255 visits over the baseline survey. This value is very similar to thenumber of visits propose for the NES and other extended footprint regions proposed by Olsen et al. PERFORMANCE EVALUATION5We quantify the impact of including and excluding the NES in LSST survey operations. Wesimulate LSST observations for representative orbital distributions based on current observationalconstraints using existing OpSim runs and Metric Analysis Framework (MAF; Jones et al. 2014)moving object tools. We find for our metrics, astro-lsst-01 2039,which does not include the NESregion, underperforms.
Main-Belt Comets:
For outer MBCs on eccentric orbits, activity is confined to time periodsnear perihelion ( r h (cid:46) . DiscoveryMetric ); (2) Werequire discovery circumstances to occur in a 1-month period after perihelion (Hsieh et al. 2015)(custom metric). The ∼
10 known MBCs are active near perihelion, with a bias to post-perihelionepochs ( −
30 to +60 days about perihelion seems typical), which is currently based on a sample of 10objects. We base our metric on the period 0 to +30 days to allow for some additional diversity in thepopulation. Our metric tested with select available
OpSim runs is shown in 5 (part c) In summary,excluding the NES reduces the number of MBC discoveries due to the perihelion alignment noted inthe Section 2. Hsieh et al. (2015) estimated an occurrence rate of about 60 MBCs per 10 outer mainbelt asteroids. Taking this rate, the size-frequency distribution of asteroids (Jedicke et al. 2002), andthe current set of opsim runs, we estimate the number of LSST MBC discoveries to be 10–15 withoutthe NES, and 20–25 with the NES. The increased number allows us a better estimate of how MBCproperties vary, and assess the lifetime of water ice in ∼
200 m objects in the outer belt.
Resonant KBOs and L4 Neptune Trojans:
The effect of including the NES on the science return for the outer Solar System can be quantified interms 1) We need a sufficient number of detections across many different dynamical populations withaccurately determined orbits to constrain models of the early Solar System. 2) We need sufficientlyaccurately determined orbits to classify the detections into these different populations. For each
OpSim run, the Neptune Trojans and objects in Neptune’s 2:1 and 5:1 mean motion resonances areused to determine the expected number of detections for each population.
OpSim runs with no northecliptic coverage have extremely few detections for Neptune Trojans in the L4 cloud (none at lowinclinations crucial for testing the color-inclination relationship.) and few detections in the leadinglibration islands of Neptune’s N:1 resonances; excluding the north ecliptic cuts the total number ofexpected discoveries in approximately half (see Fig. 5 a and b). For the case of the 5:1 resonance(Figure 5 d), there would be too few detections to usefully constrain the ratio of leading to trailingpopulations to test Neptune’s migration speed. For distant N:1 resonances, losing half the detectionswould limit the accuracy of population estimates; we would ideally like the Poisson sample sizeuncertainties to be less than ∼ Outer Solar System Orbit Metric:
A metric to measure the orbit fit quality for detectedobjects will need to be constructed, accounting for the total number of observations assuming anappropriate color distribution. To securely classify objects, we typically need the orbit fit to havea 3 − σ semi-major axis uncertainty ∆ a/a < .
01. This is necessary to separate out resonant and Orbital distributions used in our assessments are available via a public GitHub repo https://github.com/lsst-sssc/SSSC LSST Cadence Optimization Orbit Test Populations. ∼ IOCs:
The predicted IOC orbital distribution is dependent on the formation model. Given that theIOCs can only be efficiently discovered near their perihelion and the Planet 9 model predicts perihleionclustering, we suggest a simple metric for success of IOC discovery using fractional coverage of theNES region within 20 ◦ of the ecliptic may be most suitable. a) b)c) d) Inclinations <20 deg missed
Excludes NES Includes NES
Figure 5.
Metric results for a) Neptune Trojans excluding the NES b)Neptune Trojans including the NES,c) MBCs, d) 5:1 KBO Mean Motion Resonance Libration Islands5.
SPECIAL DATA PROCESSINGFor understanding detection efficiency and characterizing the survey losses, having the same de-tection algorithm used in the Northern Ecliptic as Southern Ecliptic will be extremely beneficial. Inthe WFD, MOPS will be the primary moving object search algorithm for Solar System bodies atdistances less than approximately 200 au. The proposed observations are designed such that LSST7Project-developed MOPS will be able to generate tracklets and link them to identify moving objectsin the NES.One science case that requires an additional pipeline is the search for very distant small bodies,including additional Sedna-like objects or Planet 9. MOPS is designed to detect motion between thetwo visits of the same field within a night, separated by ∼ ∼
700 au), a separate detection pipeline will needto be developed by the planetary community. This pipeline task has been identified as one of thekey tasks in the SSSC Software Roadmap. We note that several members of the SSSC have writtenversions of a slow moving object pipeline (e.g. Brown et al. 2004; Schwamb et al. 2010; Sheppard& Trujillo 2016; Bannister et al. 2017; Gerdes et al. 2017; Holman et al. 2018) for other outer SolarSystem surveys and have the expertise to develop such a community pipeline. We also note thatthis pipeline could reasonably work on the sources generated from individual images, rather thanrequiring the image pixels directly, and can further reject a large majority of the sources in eachindividual image immediately as correlated with (long-term) stationary objects; the relevant inputsare relatively small compared to LSST data processing.Main-belt comet science will also require specialized data processing in the form of advanced activitydetection and characterization software that go beyond the basic activity detection and characteriza-tion performed by the standard LSST pipelines. These pipelines will build upon the alert stream andLSST produced Solar System data products. This specialized software is equally essential for cometscience in general for LSST (i.e., including observations as part of the main Wide-Fast-Deep surveyin the South), and development of this software is already a high priority for the SSSC softwaredevelopment and active objects working groups. No additional special data processing requirementsbeyond what is already planned to be developed to handle comet data from LSST in general will beimposed by this mini survey. ACKNOWLEDGEMENTSThe authors thank the Large Synoptic Survey Telescope (LSST) Project Science Team and theLSST Corporation for their support of LSST Solar System Science Collaboration’s (SSSC) efforts.This work was supported in part by a LSST Corporation Enabling Science grant. The authorsalso thank the B612 Foundation, AURA, and the Simons Foundation for their support of workshops,hackathons, and sprints that lead to the development of this white paper. Elements of this work wereenabled by the Solar System JupyterHub service at the University of Washington’s DIRAC Institute(http://dirac.astro.washington.edu). This white paper has made use of NASA’s Astrophysics DataSystem Bibliographic Services. This version of our NES whitepaper was formatted using the AASTexlatex classfile and template package from America Astronomical Society (AAS) Journals http://journals.aas.org/authors/aastex/aasguide.html.REFERENCES
Bannister, M. T., Shankman, C., Volk, K., et al.2017, AJ, 153, 262,doi: 10.3847/1538-3881/aa6db5 Batygin, K., & Brown, M. E. 2016, AJ, 151, 22,doi: 10.3847/0004-6256/151/2/22 Braga-Ribas, F., Sicardy, B., Ortiz, J. L., et al.2013, ApJ, 773, 26,doi: 10.1088/0004-637X/773/1/26—. 2014, Nature, 508, 72,doi: 10.1038/nature13155Brasser, R., & Schwamb, M. E. 2015, MNRAS,446, 3788, doi: 10.1093/mnras/stu2374Brown, M. E., & Batygin, K. 2016, ApJ, 824, L23,doi: 10.3847/2041-8205/824/2/L23Brown, M. E., Trujillo, C., & Rabinowitz, D.2004, ApJ, 617, 645, doi: 10.1086/422095Chambers, K. C., Magnier, E. A., Metcalfe, N.,et al. 2016, ArXiv e-prints.https://arxiv.org/abs/1612.05560Delgado, F., Saha, A., Chandrasekharan, S., et al.2014, in Society of Photo-OpticalInstrumentation Engineers (SPIE) ConferenceSeries, Vol. 9150, Modeling, SystemsEngineering, and Project Management forAstronomy VI, ed. G. Z. Angeli & P. Dierickx,15Dias-Oliveira, A., Sicardy, B., Ortiz, J. L., et al.2017, AJ, 154, 22,doi: 10.3847/1538-3881/aa74e9Gerdes, D. W., Sako, M., Hamilton, S., et al. 2017,ApJ, 839, L15, doi: 10.3847/2041-8213/aa64d8Gladman, B., Lawler, S. M., Petit, J.-M., et al.2012, AJ, 144, 23,doi: 10.1088/0004-6256/144/1/23Holman, M. J., Payne, M. J., Fraser, W., et al.2018, ApJ, 855, L6,doi: 10.3847/2041-8213/aaadb3Hsieh, H. H., & Jewitt, D. 2006, Science, 312, 561,doi: 10.1126/science.1125150Hsieh, H. H., Denneau, L., Wainscoat, R. J., et al.2015, Icarus, 248, 289,doi: 10.1016/j.icarus.2014.10.031Ivezi´c, ˇZ., Kahn, S. M., Tyson, J. A., et al. 2008,ArXiv e-prints.https://arxiv.org/abs/0805.2366Jedicke, R., Larsen, J., & Spahr, T. 2002,Observational Selection Effects in AsteroidSurveys, 71–87Jewitt, D., Hsieh, H., & Agarwal, J. 2015, TheActive Asteroids, 221–241 Jones, R. L., Yoachim, P., Chandrasekharan, S.,et al. 2014, in Society of Photo-OpticalInstrumentation Engineers (SPIE) ConferenceSeries, Vol. 9149, Observatory Operations:Strategies, Processes, and Systems V, ed. A. B.Peck, C. R. Benn, & R. L. Seaman, 0Jones, R. L., Slater, C. T., Moeyens, J., et al.2018, Icarus, 303, 181,doi: 10.1016/j.icarus.2017.11.033Kim, Y., JeongAhn, Y., & Hsieh, H. H. 2018,ApJ, 155, 142, doi: 10.3847/1538-3881/aaad01Levison, H. F., Morbidelli, A., Vanlaerhoven, C.,Gomes, R., & Tsiganis, K. 2008, Icarus, 196,258, doi: 10.1016/j.icarus.2007.11.035Lin, H. W., Chen, Y.-T., Holman, M. J., et al.2016, AJ, 152, 147,doi: 10.3847/0004-6256/152/5/147Lin, H. W., Gerdes, D. W., Hamilton, S. J., et al.2018, ArXiv e-prints.https://arxiv.org/abs/1806.09696Lindegren, L., Hern´andez, J., Bombrun, A., et al.2018, Astronomy & Astrophysics, 616, A2,doi: 10.1051/0004-6361/201832727LSST Science Collaboration, Abell, P. A., Allison,J., et al. 2009, ArXiv e-prints.https://arxiv.org/abs/0912.0201LSST Science Collaboration, Marshall, P.,Anguita, T., et al. 2017, ArXiv e-prints.https://arxiv.org/abs/1708.04058Malhotra, R. 1995, AJ, 110, 420,doi: 10.1086/117532Murray-Clay, R. A., & Chiang, E. I. 2005, ApJ,619, 623, doi: 10.1086/426425Nesvorn´y, D., & Vokrouhlick´y, D. 2016, ApJ, 825,94, doi: 10.3847/0004-637X/825/2/94Ortiz, J. L., Santos-Sanz, P., Sicardy, B., et al.2017, Nature, 550, 219,doi: 10.1038/nature24051Pike, R. E., Fraser, W. C., Schwamb, M. E., et al.2017, AJ, 154, 101,doi: 10.3847/1538-3881/aa83b1Schwamb, M. E., Brown, M. E., Rabinowitz,D. L., & Ragozzine, D. 2010, Apj, 720, 1691,doi: 10.1088/0004-637X/720/2/1691Schwamb, M. E., Jones, R. L., Chesley, S. R.,et al. 2018, ArXiv e-prints.https://arxiv.org/abs/1802.01783Sheppard, S. S., & Trujillo, C. 2016, AJ, 152, 221,doi: 10.3847/1538-3881/152/6/2219