Overview

Targeting in SDSS-V is a complex process that can be broken into the following steps:

1. Generate a list of unique astrophysical sources from which targets to be observed can be selected.
2. Select targets in “cartons” (groupings with matching scientific goals) and and assign them “cadences” (observing requirements, including number of observing epochs, exposures per epoch, sky brightness, etc.)
3. Select calibration targets (sky positions and spectrophotometric standards).
4. Generate assignments between available optical fibers and targets for a given sky field and epoch.

Targeting database

To handle the large amount of data required for target selection we use a PostgreSQL database hosted in the SDSS servers at University of Utah. This database contains three schemas with tables related to target selection:

• catalogdb contains the parent catalogs used for target selection and the results of the cross-matching process, which provide the list of targets from which we select those to be observed.
• targetdb contains the results of the target selection process, including the groupings of targets to be observed and the fiber assignments for those targets.
• opsdb tracks the observation, status, and completion of the science targets.

targetdb and a subset of catalogdb (including only the information for targets selected for observation) are released as part of DR19 in a SQLServer database accessible through CASJobs, with some naming convention changes.

The target selection process

SDSS-V targets are selected from a large set of publicly available parent catalogs. The catalogs are first ingested into tables in catalogdb, usually from flat files. Given this set of catalogs, we run a cross-matching process to generate a list of unique sources to which we assign a custom identifier, catalogid. Associations between each catalogid and the original entry in each parent table (if it exists) are encoded in the corresponding catalog_to_tablename table.

From this list of unique sources we define cartons, which are groupings of targets sharing a scientific goal and observing requirements. With the exception of some open-fiber cartons, cartons are defined algorithmically, based on selection constraints that are encoded as a SQL query that runs against the sources in catalogdb. The selected targets are loaded into the target table targetdb, where we include the catalogid, coordinates, proper motions, and other information needed at the time of observation. The table mos_carton_to_target contains the targets linked to a given carton, as well as the observing constraints (S/N, cadence, epochs, lunation). A single target may be included in more than one carton, potentially with different observing requirements. Both the cross-match and carton creation processes are handled by the target_selection software.

The optimal assignment of targets to fibers in the Focal Plane System is performed by robostrategy, which generates a list of designs to be observed. A design includes a sky field with the field of view of the FPS (defined by a field center in RA and Dec coordinates, and a position angle) and the assignment of robotic positioners to targets for that observation. A design usually includes targets from multiple cartons that have compatible observing requirements. Designs and other outputs from robostrategy are stored in various tables in targetdb.

Target selection and robostrategy are run several times a year in the SDSS servers at University of Utah. For night observing, a copy of targetdb is synced to the SDSS operations servers at APO and LCO, which includes all the information required for on-sky observations. During the night, roboscheduler optimally selects the next design to be observed. The design is converted to a configuration, which takes into account epoch-specific details such as atmospheric refraction, proper motion, unavailable fibers, etc. The FPS is then reconfigured to match the selected configuration (see jaeger and Kaiju). The configuration, as well as various metrics from the observation, are stored in opsdb, which is used to track survey progress. 

Targeting Content

Selection algorithms for carton creation are defined according to the scientific goals defined for the science programs to which the carton belongs. To see the description of the Milky Way Mapper and Black Hole Mapper programs please visit (BHM Programs and MWM Programs).

The full list of cartons included in this data release is attached here as text file and as csv.

In the following table we show a few example rows of the tables linked above.

carton	program	carton_ pk	target_ selection_ plan	mapper	category
bhm_aqmes_med	bhm_aqmes	731	0.5.0	BHM	science
bhm_csc_apogee	bhm_csc	1115	0.5.15	BHM	science
bhm_gua_dark	bhm_filler	711	0.5.0	BHM	science
…	…	…	…	…	…
…	…	…	…	…	…
manual_mwm_tess_ob	mwm_tessob	1134	0.5.13	MWM	science
ops_sky_apogee_best	ops_sky	1097	0.5.9		sky_apogee
ops_std_eboss	ops_std	529	0.5.0		standard_boss

In the table presented above the program column indicates the mapper along with the scientific program name for regular cartons; “open-fiber” for programs the come from a call to fill fiber exposures that are not assigned to any of the stars in the core programs of MWM and BHM; and programs with prefix “ops” are operation sky and telluric standard for the APOGEE and BOSS instrument. The carton_pk column is a unique identifier for the carton, and each combination of carton/target_selection_plan has a different carton_pk value. The target_selection_plan column tracks the different versions of the cartons, each time a set of carton target selection algorithms change we flag it with a plan value that has a one-to-one relation with the tag of the code. The mapper indicates if it is a Milky Way Mapper or Black Hole Mapper Carton, and it is empty when is not associated with any of them. Finally the category column can be “science” for science cartons, and “sky” or “standard” followed by the instrument for calibrations cartons.

Details and motivation of the different cartons can be found in MWM cartons and BHM cartons.

Calibration targets

Calibration targets, both sky positions and spectrophotometric standards, are defined and selected as cartons. robostrategy uses these calibration cartons to generate designs that, in addition to the science targets, satisfy the required number and distribution of calibrators for a program.

Details on the selection of standards and skies are given in their respective pages.