The DECam Asteroid Database

Francisco Valdes, November 2017

Database Tables

The fundamental organization and linkage between tables is the dataset. A dataset is a pointing on the sky from which multiple exposures were used to search for asteroid tracklets. This field is common to all tables and is a primary key for joining.

The database is organized around tracklets. A tracklet is some number of measurements from individual exposures that are identified as moving in the field and are celestial sources (i.e. not junk). A tracklet has an identification (movid) which is uniquely derived from the exposure and detection number (groupid). When known, tracklets will have IDs from the Minor Planet Center (mpcid) which may then link multiple tracklets from different nights.

The first data release was only for the DECam NEO Survey (P.I. Allen). The second release added tracklets derived from all public DECam data that are suitable for the MODS algorithm (Valdes 2017). This version is still an early release with some database records having incomplete fields.

The following is a brief overview of the logical organization of the entities that make up the database tables. A dataset consists of a number of exposures (> 2). An exposure is characterized by an exposure id (a filename) and parameters of the exposure such as time and length of the exposure. A group is a set of observations of an object in in some or all of the exposures identified as forming a moving object tracklet. A tracklet is characterized by an id and parameters such as the rate of motion and magnitude. An observation is a individual detection of a moving object in a CCD of an exposure. It is characterized by positions in three coordinate systems (equitorial on the sky, CCD image pixel, and cutout image pixel). A filename for a cutout image is also given. Related is a table giving the MPC formated record for the observation. The set of records for a particular tracklet is what was submitted as a detection to the MPC.

Records

MOVDS4098
MOVEXP19101
MOVGRP663489
MOVOBS2983428
MOVMPC2983428

Basic Statistics

Pi Pointings Tracklets
Allen 3126 608829
Sheppard 393 49325
Forster 220 3730
Fuentes 132 1088
Munoz 82 284
Carlin 27 167
Vivas 17 118
Rich 9 89
Sullivan 35 73
Saha 24 53
Yip 12 46
Rest 27 39
Geha 9 39
Frieman 33 37
Dai 23 35
Walker 15 33
Heinze 3 27
Bonaca 3 25
Crnojevic 15 22
Konstantopoulos 4 21
Hargis 3 18
Taylor 21 16
Mackey 9 15
McMonigal 6 10
Calamida 4 7
Sheen 1 6
McCleary 2 3
Rusu 1 2
Curtin 2 2
Tucker 1 1
Schlegel 1 1
Geisler 1 1
Mamajek 1 1

Table MOVDS

This table describes the characteristics of the datasets (pointings). The time and position fields are for the start of the first observation exposure in the set. Position fields are for the center of the field.

MOVDS: Schema (postgres)

   Column   |            Type             | Modifiers 
------------+-----------------------------+-----------
 dataset    | character varying(64)       | not null
 nexposures | integer                     | 
 date       | timestamp without time zone | 
 mjd        | double precision            | 
 lst        | character varying(13)       | 
 filter     | character varying(64)       | 
 ra         | double precision            | 
 dec        | double precision            | 
 gamma      | double precision            | 
 beta       | double precision            | 
 opposition | double precision            | 
 night      | timestamp without time zone | 
 ngrp       | integer                     | 
 photref    | text                        | 
 plver      | text                        | 
 propid     | text                        | 
 pi         | text                        | 
 observers  | text                        | 
Indexes:
    "movds_pkey" PRIMARY KEY, btree (dataset)

MOVDS: Column descriptions

dataset dataset identifier  
nexposures number of exposures  
date observation start of the first exposure (UT) UT timestamp
mjd modified Julian date at the start of the first exposure fractional day
lst local sideral time at the start of the first exposure sexigesimal string
filter filter used for observations  
ra RA of the center of the field (deg) degrees
dec DEC of the center of the field (deg) degrees
gamma ecliptic longitude of the center of the field (deg) degrees
beta ecliptic latitude of the center of the field (deg) degrees
opposition angle from opposition for the center of the field (deg) degrees
night calendar night of observations timestamp
ngrp number of tracklets found  
photref filter used for magnitude calibration  
plver pipeline version  
propid proposal ID of observations  
pi principle investigator given in the exposure header  
observers observers given in the exposure header  

Table MOVEXP

The exposure table provides details of each exposure. The expname is a way to link to the image data in the NOAO Science Archive. The expid is the sequencial count of the exposures in a dataset. The separation time is a relative time for the middle of each exposure with an origin at the start of the first exposure.

MOVEXP: Schema (postgres)

 Column  |            Type             | Modifiers 
---------+-----------------------------+-----------
 dataset | character varying(64)       | not null
 expname | character varying(64)       | 
 dateobs | timestamp without time zone | 
 mjdobs  | double precision            | 
 expid   | integer                     | not null
 tsep    | integer                     | 
 exptime | integer                     | 
 magzpt  | real                        | 
 fwhm    | real                        | 
Indexes:
    "movexp_pkey" PRIMARY KEY, btree (dataset, expid)

MOVEXP: Column descriptions

dataset dataset identifier  
expname filename root of the image data used  
mjdobs MJD of the start of the observation fractional day
expid exposure id within the set of exposures  
tsep mid-time of exposure relative to start of first exposure (sec) seconds
exptime exposure time of observation (sec) seconds
magzpt estimated magnitude zeropoint (for 1 sec)  
fwhm median FWHM of exposure over FWHM for all CCDs (pixels) pixels

Table MOVGRP

This is the main tracklet table. It provides information about each connected set of observations corresponding to a moving object.

MOVGRP: Schema (postgres)

  Column  |            Type             | Modifiers 
----------+-----------------------------+-----------
 dataset  | character varying(64)       | not null
 groupid  | integer                     | not null
 movid    | character varying(8)        | not null
 nobs     | integer                     | 
 rate     | real                        | 
 pa       | real                        | 
 mag      | real                        | 
 digest   | integer                     | 
 mpcsent  | timestamp without time zone | 
 mpcid    | character varying(24)       | 
 mpcsubid | character varying(8)        | 
 jpg      | character varying(24)       | 
 rate_rms | double precision            | 
 rate_pct | double precision            | 
 ratec    | double precision            | 
 checked  | text                        | 
Indexes:
    "movgrp_pkey" PRIMARY KEY, btree (dataset, groupid, movid)

MOVGRP: Column descriptions

dataset dataset identifier  
groupid tracklet ID number within dataset  
movid unique tracklet ID assigned by the survey  
nobs number of detections associated with the tracklet  
rate apparent rate of motion from measured positions (arcsec/hr) arcsec/hr
pa apparent position angle from measured positions (deg) degrees
mag estimated average magnitude over detected observations  
digest rarity/interest score from the digest2 program (0-100)  
mpcsent indication of when measurements were submitted to the MPC  
mpcid identifier returned by the MPC as initial designation or, sometimes, later more final assignment  
mpcsubid tracklet ID used in MPC submission  
jpg filename for cutout graphic of the tracklet  
rate_rms (for internal use) arcsec/hr
rate_pct (for internal use) arcsec/hr
ratec approx. projection corrected rate (rate/cos(opposition)) arcsec/hr
checked indicate of verification (e.g. visually)  

Tables MOVOBS and MOVMPC

These tables provide information for each separate detection of a moving object. They provide links to associated files (see below) and to what was reported to the MPC. (The MPC records are sometimes the only way to find information from the MPC website). Note that the MPC records are basically an encoding of the information in the exposure and observation tables in a specific format for reporting the time, position, and estimated magnitude for an observation.

MOVOBS: Schema (postgres)

 Column  |         Type          | Modifiers 
---------+-----------------------+-----------
 dataset | character varying(64) | not null
 groupid | integer               | not null
 movid   | character varying(8)  | not null
 expid   | integer               | not null
 ccd     | integer               | 
 ra_obs  | double precision      | 
 dec_obs | double precision      | 
 x       | real                  | 
 y       | real                  | 
 xco     | real                  | 
 yco     | real                  | 
 cutout  | character varying(64) | 
Indexes:
    "movobs_pkey" PRIMARY KEY, btree (dataset, groupid, movid, expid)

MOVOBS: Column descriptions

dataset dataset identifier  
groupid tracklet ID number within dataset  
movid unique tracklet ID assigned by the survey  
expid exposure id within the set of exposures  
ccd CCD number in DECam data  
ra_obs RA of detected source (deg) degrees
dec_obs DEC of detected source (deg) degrees
x column pixel coordinate in CCD image (pix) pixels
y row pixel coordinate in CCD image (pix) pixels
xco column pixel coordinate in cutout image (pix) pixels
yco row pixel coordinate in cutout image (pix) pixels
cutout filename for cutout image (file[extension] MEF syntax)  

MOVMPC: Schema (postgres)

 Column  |         Type          | Modifiers 
---------+-----------------------+-----------
 dataset | character varying(64) | not null
 groupid | integer               | not null
 movid   | character varying(8)  | not null
 expid   | integer               | not null
 mpcrec  | character(80)         | 
Indexes:
    "movmpc_pkey" PRIMARY KEY, btree (dataset, groupid, movid, expid)

MOVMPC: Column descriptions

dataset dataset identifier  
groupid tracklet ID number within dataset  
movid unique tracklet ID assigned by the survey  
expid exposure id within the set of exposures  
mpcrec MPC record for observation  

Exposure Data

The survey used the NOAO Community Pipeline "resampled" data products. These are multi-extension FITS files where each of the 60 CCDs has been warped to a common pixel sampling. These files are in the NOAO Science Archive, though a few nights are missing. The exposure rootname in the database points to the file in the NSA. A rootname is used because there are associated files for data quality and weight. The image data has extensions "_opi" (processing type "resampled" and product type "image").

The database fields ccd, x, y can be used to find the particular observation. A service based on the ra_obs and dec_obs fields can also locate the observation.

Cutout Data

Each source observation also may be found in the "cutout" data. Each dataset has a multi-extension FITS file containing 300x300 pixel cutouts for all observations and all tracklets in the dataset. In addition cutouts are included from the estimated location of the source in those cases where the automated detection algorithm failed to produce a detection and measurement. The cutouts are centered on the position of the source except when the source is near the CCD edge.

The database contains a reference to the specific cutout in the form of the FITS filename and the extension name in the file. The header information in the global header and the extension header contain much of the same information as the database.

Graphics Data

Each tracklet has a JPG file produced from the cutout file where the cutouts are tiled together and the source position indicated by an overlay. Because the cutout file includes images for the estimated position in the images for which an observation detection was not made, these are included and are useful validity checks. The filenames are included in the database and these filenames use the unique survey identifier.