last update: 150216 ------------------- The first U.S. Naval Observatory Astrometric Robotic Telescope Catalog (URAT1) Table of contents 1) Introduction 2) Details about observations and reductions a) Instrument b) Observations c) Pixel data reductions d) Astrometric reductions e) Proper motions f) Photometric data 3) Properties of the catalog and important notes for the user a) Basic numbers b) Sky coverage c) Completeness and contamination d) Reference frame e) Magnitudes f) Additional photometry g) GSC flag h) Non-stellar data i) Provided utility software j) History k) Arrangement of the data files l) Star identification numbers 4) Files distributed with this data release 5) Data formats a) Main catalog zone files format b) Notes to zone file data c) min / max data values of zone files d) Format of index files e) Description of other files 6) Description of URAT1 access software 7) People a) The URAT team b) Acknowledgments c) Testers 8) References ====================================================================== 1) Introduction --------------- This is the "readme" file of the first U.S. Naval Observatory (USNO) Astrometric Robotic Telescope Catalog, URAT1. This data release is available free of charge through astronomical data centers or DVDs. URAT is a follow-up project to the successful UCAC project using the same astrograph but with a much larger focal plane array and a bandpass shifted further to the red. Longer integration times and more sensitive, backside CCDs allowed for a substantial increase in limiting magnitude, resulting in about 4-fold increase in the average number of stars per square degree as compared to UCAC. Additional observations with an objective grating largely extend the dynamic range to include observations of stars as bright as about 3rd magnitude. Multiple sky overlaps per year result in a significant improvement in positional precision as compared to UCAC. A URAT1 release paper for the Astronomical Journal is in preparation. What URAT1 is: - an observational catalog, i.e. mean positions of stars at mean epoch of URAT observing (between 2012.3 and 2014.6), on the International Celestial Reference System (ICRS) by using UCAC4 (8 to 16 mag stars only) as reference in conventional "plate" adjustments - covering the about R = 3 to 18.5 magnitude range with positional precisions of 5 to 40 mas depending on observing history and brightness of stars - covering most of the northern hemisphere and in some areas as far south as -24.8 deg Dec (Pluto field), see urat1-cov14.pdf - observed in a single bandpass (680 to 762 nm = between R and I) - constructed from 2 years of operations of the astrograph at the Naval Observatory Flagstaff Station (NOFS), with multiple sky overlaps per year of regular exposures (240 sec and 60 sec) and objective grating exposures (30,20,10 sec) - small systematic errors due to higher quality of the CCDs as compared to UCAC (expected error floor is about 5 to 10 mas) - URAT1 is supplemented by preliminary proper motions derived exclusively from 2-epoch URAT1 and Two Micron All Sky Survey (2MASS) data (at around epoch 2000) resulting in proper motion errors of typically 5 to 8 mas/yr - contains 228 million objects (mostly stars) of which 188 million are common to 2MASS - photometry from 2MASS and American Association of Variable Star Observers Photometry All-Sky Survey (APASS) is added for stars in common What URAT1 is NOT: - a compiled catalog (as UCAC2 and UCAC4 are) - complete regarding sky coverage or multiple stars or any "difficult" or "problem" cases We plan for a second URAT2 data release in less than a year, which will include 3+ years of operations, proper motions and parallaxes (of nearby stars) derived from URAT data. After that Gaia data is scheduled to be available and will supersede most URAT data. In the meantime URAT will serve as the most accurate representation of the optical reference frame in the covered areas, significantly outperforming UCAC and going about as deep as 2MASS. The latest update on astrometry related projects at USNO can be found at: www.usno.navy.mil/usno/astrometry . 2) Details about observations and reductions -------------------------------------------- 2a) Instrumentation ------------------- URAT uses the same objective "redlens" as was used for the UCAC program. The 5-element "redlens" has a design bandpass of about 550 to 720 nm for a 9 degree diameter field of view. Optical ray-tracing indicates that acceptable image quality can be obtained for even redder wavelengths with modest increase in differential color distortion. The 3rd order optical distortion term of this lens is basically zero. Irregular, higher order distortion effects from the dewar window are on the same about 10 to 50 mas level as the residual optical distortion of the lens. The 300 mm clear diameter, flat dewar window serves as filter to provide the fixed 680-762 nm bandpass for URAT observations. Instead of a single 1 square degree detector as used for UCAC, a large LN2 cooled dewar with 4 of the world largest CCDs (STA1600) is used for URAT. Each CCD has 10,560 by 10,560 pixels at 9 microns, providing 28 sq.deg of sky coverage with a single URAT exposure at 0.9 arcsec per pixel resolution. The gaps between the 4 main CCDs is 20 arcmin and each individual CCD covers 2.65 by 2.65 deg of sky. For guiding and focus control 3 small CCDs (4k by 2.5k) with 8 micron pixels are placed around the edges of the field of view. The camera electronics was also built by STA. The main CCDs are operated with a clocked anti-blooming (CAB) feature. Electrons beyond the pixel full-well capacity are drained (lost) leaving the image profile of saturated stars symmetric without bleeding columns. Stellar images up to about 2 magnitudes brighter than traditional saturation can thus be utilized for precise astrometry, extending the dynamical range of URAT observations significantly. A completely new tube structure was built by the USNO DC instrument shop to connect the single lens and dewar, replacing the 2-tube double-telescope design of the original astrograph. The same B&C mount as for UCAC was used, however, with significant upgrades toward automation. The telescope control system was developed and implemented by Greg Bredthauer and the USNO instrument shop. The entire instrument was tested at USNO in Washington DC and then deployed at the Naval Observatory Flagstaff Station (NOFS) for survey operations. At NOFS upgrades, mainly with respect to lightning protection and dewar autofill were added in a joint effort of USNO DC and NOFS personnel. Operations code was development and implemented in house in Fortran utilizing the lower level instrument interfaces written in C by Greg Bredthauer. In front of CCD "D" 2 small neutral density spots were placed on the dewar window. These were used for test observations of bright stars. However, the grating observing scheme was adopted for this survey and the 2 ND spots just produce "blind spots", about 50 pixels in diameter. A single Linux PC operates the telescope, camera and auxiliary inputs like weather data and power control. A second Linux PC performs the pixel data reductions and generates backup copies. All runs robotic. Human interaction is required for: - connecting / disconnecting LN2 dewar fill line, exchange of bottles - change of data hard disk drives - kick off observations in the afternoon subject to sky conditions - monthly change of objective grating on/off - infrequent change of desiccate to control humidity in lens cell The dome opens and closes depending on sky brightness and weather conditions. An automatic shutdown including dome rotation and dome shutter operation can be performed with UPS battery banks even in case of loss of power. 2b) Observations ---------------- URAT survey operations begun in April 2012 at NOFS and are still ongoing. For about 3 weeks every month URAT operates in regular survey mode: each field is observed with a 240 sec and a 60 sec exposure without objective grating. During the week around full Moon a short exposure survey is run with objective grating and 30 and 10 sec exposures (30 and 20 sec during early part of survey). The grating constant is about 4.5 magnitudes, while the gain in dynamic range is close to 5.0 mag due to the wider profile of the diffraction images. Besides the central images, only 1st order diffraction images of the grating survey are used, which appear only slightly (typically 10 to 15%) elongated and are well separated from the central image. Higher order diffraction images are strongly elongated and would provide only small additional benefit regarding the dynamic range and thus were not used at all. In order to cover the gap areas between the 4 main CCDs the pointing of the telescope is diagonally offset (3 different pointings). Furthermore the same area of sky is observed at different times of the night over several months, splitting the observing pattern into 5 lists during the course of every night. This allows for observations at parallax factors as different as possible for a uniform all-sky survey. Thus any given field in the sky is observed 3 x 5 = 15 times (minus losses from gaps between CCDs) per year, each with a long and short exposure, and all of this is repeated for the grating survey observing. The number of available clear sky hours determines how far south observations progress. The north celestial pole is not accessible with URAT observations because it always falls into the gap area between the main CCDs. Furthermore hardware restrictions (undersized dome, inclination limit switch settings) during a significant period of time lead to avoiding the sky area north of about +87 deg declination, which then lead to incomplete coverage of the few square degrees around the pole. A single main CCD has 111 million pixels which are read out with 8 ports and 16 bit in about 20 sec. Images are stored as 2-byte integers in FITS files resulting in about 1 GB of data per exposure. Depending on the season, about 120 to 200 regular survey exposures are taken per night. In grating mode up to about 550 exposures were obtained in a single winter night. All exposures are taken close to the meridian (within about 5 deg) and the observing schedule prioritizes fields from north to south. The guide CCDs take 4 or 5 sec exposures and data are processed immediately, providing feedback to the telescope drive before the next guide exposure. Exposures of less than 30 sec are not guided. One of the guide CCDs is mounted above and another below the focal plane. Focus changes are detected from the mean image profile width by comparing data of the 3 guide/focus CCDs at the end of a field exposure. The telescope focus is thus adjusted on the fly if needed. Due to the small aperture (20 cm) and red bandpass the diffraction images are already about 1.5 pixel wide (FWHM). Adding some seeing, guiding effects, and focus tolerance lead to typical observed image profiles of about 1.8 to 2.3 pixels wide. This is a bit larger than the best images which could be obtained in the UCAC program, however the roundness of URAT images is noticeably better than for most of the UCAC data. Typical URAT images are elongated by not more than about 3% (ratio of major to minor axis <= 1.03), averaged over the entire field of view. 2c) Pixel data reductions ------------------------- Pixel reductions were performed with custom code utilizing dark exposures taken for every standard survey exposure time and occasional twilight sky flats. Bias correction was performed line-by-line using overscan data. No additional bias frames were taken or applied. Pixel-to-pixel sensitivity changes were found to be very small and stable. There are only a few column defects, mainly on CCD C and D, while CCD A does not have a single column defect. Differences in overall sensitivity, noise and mean image profile width between the 4 main CCDs lead to a spread of limiting magnitude of about 0.5 mag among the CCDs for any given long exposure. Object detection was performed with custom code adopting a 4-sigma threshold above the background. Each main CCD was handled as a "separate telescope" from start to finish in the reduction pipeline. A 2-dimensional Gaussian model was adopted for the image profile fits. Image centers (x,y data), instrumental model magnitudes and errors were determined for all successfully fitted objects. No double star fits of blended images were performed at this time. This would have required the construction and use of a better matching point-spread function than the simple Gaussian used here, following procedures of UCAC4 reductions. Due to time and manpower restrictions this approach could not be followed here and now. Thus URAT1 does not contain any double stars unless the pairs are relatively wide (several arcsec) and do show up as not blended, individual images. 2d) Astrometric reductions -------------------------- A subset of the UCAC4, those stars with UCAC magnitudes in the range of 8.0 to 16.0 and identified as "clean" single stars were used as reference stars for the URAT astrometric reductions. This avoids potential UCAC4 problem sources and still leaves a large number of reference stars (typically many hundreds to several thousand) per CCD. Due to the large field of view "plate" tilt terms were significant and had to be included into the model, besides the linear terms. Thus an 8 parameter "plate" model was used for the weighted least-squares astrometric reductions of each individual URAT exposure and CCD. In pilot investigations the residuals were stacked up as a function of x,y location on the CCD and mean field distortion patterns were constructed. Amplitudes of these systematic error vectors are typically in the 10 to 50 mas range, up to 100 mas in the corners. For the final reductions the x,y data were corrected for this fixed pattern, which was found not be be changing as a function of the stars brightness or time. Small systematic position corrections were derived and applied as a function of the pixel phase and image profile width of the URAT data, again using stacked up residuals of pilot investigations. As expected, a sine-curve type of position correction is seen with amplitudes of only about 5 to 15 mas depending on FWHM of the image profiles. For stars beyond the traditional saturation (CAB regime) position offsets along both coordinate are seen as function of magnitude. These vary over time and the URAT data was split into epochs to correct for the mean signature of this systematic position error, typically in the range of 10 to 100 mas. Reasonable good corrections could be derived for stars up to about 1.5 magnitudes beyond the traditional saturation limit and positional data of even brighter stars were discarded for the URAT1 catalog. Mean, observed positions were derived from sets of data covering either the times of regular survey observing between grating runs, or the times of a grating survey run (typically 3 and 1 week, respectively, sometimes adjacent data combined). A total of 19 + 19 sets of data were defined this way over the 24 months period of observing included in the URAT1 release. In addition data of the Pluto area were included from a single night in 2013 and another night in 2014. The URAT instrumental magnitudes were calibrated per individual CCD and exposure utilizing the APASS photometry given with the UCAC4 reference star catalog. A zero-point correction was derived between the instrumental and calibrated URAT magnitudes from a linear photometry fit using R-I colors. Finally, these mean positions, magnitudes and errors of the individual sets of data were combined to overall, weighted mean positions and magnitudes resulting in the URAT1 catalog data using the individual set data errors for weights and adding an RMS error floor of 5 mas per position coordinate and 0.01 mag photometry error at the end. Only objects with positions based on at least 3 observations or a match with a 2MASS source within 3 arcsec were propagated to the final URAT1 catalog. 2d) Proper motions ------------------ Preliminary proper motions were determined exclusively using the 2MASS data as early epoch for about 188 million sources in common with URAT1. The mean URAT observational epoch is between 2012.3 and 2014.6 while the typical 2MASS epoch is around 1999, providing about 14 years of epoch difference. A fixed positional error of 80 mas per coordinate was assumed for the 2MASS data and the individual stars (model) URAT position errors were used to derive errors of proper motions (typically 5 to 8 mas/yr) from these 2-epoch data. A match between URAT1 and 2MASS is assumed when the position difference per coordinate did not exceed 3 arcsec. This leads to a maximum of about 200 mas/yr per coordinate for a proper motion derived in this manner. Stars with larger proper motions will not show up as having a proper motion or 2MASS match in the URAT1 catalog. However, those stars should be present in URAT1 with a URAT1 position at their current, mean epoch. 2e) Photometric data -------------------- Similar to UCAC4, photometric data from 2MASS and APSS were tagged on to the observed URAT data for the URAT1 data release. The APASS DR8 data plus single photometric observations (priv. com. A.Henden) are given with this public URAT1 catalog data. 2MASS provides J,H, and Ks magnitudes and errors for about 85% of the URAT1 stars, while APASS data gives B,V,g,r,i magnitudes and errors for about 37 million URAT1 stars (= 16%). 3) Properties of the catalog and important notes for the user ------------------------------------------------------------- 3a) Basic numbers ----------------- total numb. URAT1 stars = 228276482 number stars with 1 obs = 10309229 number stars with 2 obs = 8875122 average numb. obs/star = 24.3 number valid 2MASS data = 188656145 82.64 % of URAT1 stars no 2MASS match stars = 39620337 stars >=3 obs., no 2MASS= 39079551 number valid APASS data = 37010348 16,21 % of URAT1 stars APASS stars valid B mag = 29313850 APASS stars valid V mag = 30057593 APASS stars valid g mag = 32340624 APASS stars valid r mag = 32474206 APASS stars valid i mag = 28052917 Histograms and mean data values by magnitude bins are given on the file chkv1d.log (see below). 3b) Sky coverage ---------------- URAT1 covers most of the northern sky and extends south of the equator to about -15 deg Dec for some RA areas, plus the Pluto field area around RA = 18h30m Dec=-20. However, the number of observations per star varies largely with best coverage for the Dec = 30 to 80 deg zones. Very near the pole, no URAT observations could be made due to the gap between the main CCDs. 3 sky plots distributed with this release show the sky coverage color coded for mean number of observation per area (overlaps) for stars near magnitude 14 and 18, and number of stars per square degree (density), respectively. 3c) Completeness and contamination ---------------------------------- Contrary to UCAC4, only objects observed by the URAT astrograph were included in the URAT1 catalog. URAT is *not* complete. For sky coverage see previous section. Furthermore, all "problem" sources are not included. If at any point in the reduction data or objects were dropped, no attempt was made to recover those targets or investigate the nature of the problem. Thus all blended images and most double stars will not have made it into the URAT catalog. Most of possible hot pixels or defects of the CCDs are hopefully excluded due to the requirements to enter the URAT1 catalog (match with 2MASS or at least 3 observations). However, those criteria might not have been enough to prevent some defects to enter the catalog. Likewise many asteroids will likely have made it into the catalog, although an attempt was made to exclude those due to the 3 observations minimum criteria. To assist in contamination control, a flag was added whenever a URAT1 object has no match with the GSC 2.4 catalog. Most of the Hipparcos stars in the URAT1 area could be observed due to the large dynamic range provided by the grating data. A separate investigation is underway to compare URAT positional results with Hipparcos and Tycho catalog data. Due to the much fainter limiting magnitude of URAT versus UCAC, many extragalactic sources will be in the URAT1 catalog. At this point no galaxies are identified, however a separate investigation about ICRF optical counterparts found in URAT1 is under way. 3d) Reference frame ------------------- The astrometry provided in URAT1 is on the Hipparcos system, i.e. the International Celestial Reference System (ICRS), as represented by the UCAC4 catalog. Positions are given on the ICRS at mean epoch of observations. Thus URAT1 is an observational catalog, not a compiled catalog. The preliminary proper motions are only tagged on, while the URAT1 catalog positions are *not* the weighted mean positions between URAT and 2MASS observations. Systematic errors are seen in the URAT reductions with respect to UCAC4 reference stars as a function of brightness of the stars. However, those about +-10 mas effects are believed to be inherent in the UCAC4 data and no corrections for these were applied in the URAT1 data. Due to the large field of view and large dynamic range, URAT should "average" over many of the UCAC4 remaining systematic errors and should locally be better on the ICRS than UCAC4 is. 3e) Magnitudes -------------- URAT1 bandpass (680-762 nm) photometry is derived from the volume of fitted image profiles (2-dim Gaussian) above the local background. Thus the individual observed magnitudes are model magnitudes, not aperture photometry magnitudes. Calibration from instrumental to true magnitudes was performed with a linear model using APASS R magnitudes and R-I colors. Note, many URAT observations were performed in non photometric conditions. URAT1 is primarily an astrometric catalog and URAT photometry need to be taken with caution. Photometric errors were derived from scatter of individual observations per star. In order to avoid unrealistically small accidental errors, an error floor of 0.01 mag was RMS added to the scatter error. The resulting values are given as individual star URAT bandpass photometry error in URAT1. Due to small number statistics these errors given for individual stars are not very reliable. A mean error over many stars of similar brightness should give a more reliable photometric error estimate. 3f) Additional photometry ------------------------- The URAT1 observational data are supplemented by 5-band photometry (B,V,g,r,i) from the APASS project as well as with near-IR photometry (J,H,K_s) from the Two Micron All Sky Survey, 2MASS (Scrutsky 2006). In addition to the magnitude values errors and some flags are provided. For APASS the DR8 data plus single observations were used (Henden, private comm.). For more details see http://www.aavso.org/apass and http://www.ipac.caltech.edu/2mass/releases/allsky/ . 3g) GSC flag ------------ As part of the testing and data verification the URAT1 data was matched to the GSC 2.4 catalog (priv. comm. Ricky Smart, Torino Obs.). A total of 1123782 URAT1 stars had no match to the GSC 2.4 catalog entries, although its limiting magnitude is even a bit deeper than that of URAT1. Of these 1123782 stars, 94384 do have a unique match to 2MASS, while 951006 do not match with 2MASS at all. The remainder of stars have multiple or uncertain matches. A spot check of the stars not matched with GSC 2.4 revealed different cases. Some stars are clearly seen on the digital sky survey images, others not at all or are associated with a very faint object. In other cases the URAT1 object is close to another star, indicative of a relatively large proper motion. All these about 1.1 million objects are flagged (negative "number of sets used", see data format in section 5) in this public release of URAT1. It is up to the user to use these objects or not. 3h) Non-stellar data -------------------- Non-stellar object, like minor planets, galaxies, or QSOs are not identified as such in the URAT1 catalog. The goal of URAT1 is to provide the users with a reference star catalog going significantly deeper and being more accurate than UCAC4 as soon as possible. Some data mining effort of URAT1 is already in progress particularly regarding Hipparcos stars and extragalactic targets. 3i) Provided utility software ----------------------------- Only some basic access code is included with this release to illustrate the data format and allow conversion of the main binary files to ASCII tables. It is anticipated that 3rd party companies specialized in providing interfaces between various star catalogs and the user will integrate URAT data into their products, as has been the case for UCAC catalogs. 3j) History ----------- The URAT project idea was developed around 2000. Originally a larger telescope of 0.85m aperture was envisioned for this project. However, the dedicated astrometric design with full-size aperture corrector plates proved to be too expensive to realize. On the other hand the detector development sponsored by the Navy through a SBIR contract progressed well leading to the successful manufacturing of the world largest monolithic CCD detector. In 2008 funding became available for the huge focal plane array. Fortunately the "redlens" astrograph was a valid "plan B" option to utilize the 474 million pixel camera and push the project forward. Some pictures and milestones are shown on our web page (www.usno.navy.mil/usno/astrometry) and a complete list of (mainly poster paper) publications about the URAT project so far are given in the References section below. 3k) Arrangement of the data files --------------------------------- The main catalog data are arranged in declination zones of 0.2 degree width. Zones are numbered from 1 starting at the South Pole and increasing toward north. The first zone with data in URAT1 is 326 for -25.0 to -24.8 deg Dec. There is a separate file for each zone up to zone 900 near the north celestial pole. Records on the individual data files are sorted by right ascension. Each record contains the data for one star, 80 bytes of binary integers as described below. An index file is provided for fast access. The step size along RA for that index file is 0.25 degree = 1 arcmin, for a total of 1440 bins along RA. The value of that index file gives the largest running star (record) number along that zone for that RA bin. If there are no stars in a bin, the index values is the largest record number of the nearest, earlier bin which does contain stars. 3l) Star identification numbers ------------------------------- Official URAT1 star ID numbers consist of 2 parts, the 3-digit zone number (zzz) and the 6-digit running record number (nnnnnn) along a zone, which can be as large as 674123 (for zone 551). Thus a URAT1 star number is given by: URAT1-zzz-nnnnnn (or zzznnnnnn for short as a single integer) Cross identifications to other catalogs are not given at this time, except for the 2MASS running ID name, which is part of the 80-byte URAT1 record for each star. The provided access code which dumps the contents of a binary zone file to an ASCII file automatically attaches that zzznnnnnn number to the output records, however, that number is not part of the 80 byte binary data. 4) Files distributed with this data release ------------------------------------------- The URAT1 data release contains the following directories: access = example access code (Fortran), and sample output (section 6) info = this "readme.urat1" file, other info files, all-sky plots v12 = binary zone files and index files of URAT version 1.2 The main catalog data are contained in zone files "z326" to "z900" inside the "v12" directory. The format of these binary data files is explained below. The "v12" directory also contains the index file in 2 different formats: v1index.asc = ASCII table as explained below v1index.unf = 2 unformatted, binary records as explained in access code 5) Data formats --------------- Binary files in the URAT1 distribution use Linux-style sequence of bytes. Other computers might need a flip of bytes for 2 and 4 byte integers. An option for this is build into the access code provided here. 5a) Main catalog zone data files -------------------------------- The data of the 147 and 14184 stars in the zones z326 and z327 respectively are given in ASCII fixed format and comma separated value format each (z326.asc, z326.csv, z327.asc, z327.csv) to illustrate the format and allow verification of binary file access by the user. These files was generated with the access/v1dump.f code. The "v1" stands for "version 1 of URAT" to distinguish these software files from "u" code used in the UCAC program. The last column of these files contain the official URAT star ID number with format zzznnnnnn, where zzz is the 3-digit zone number and nnnnnn the 6 digit running record number along the zone. The following shows sample code (Fortran 77) to illustrate the data format, followed by a table describing each data item (column, same order also on the ASCII sample output). There are 45 columns of URAT1 data. The min/max values for each column over all URAT1 data are given below. INTEGER*4 ra,spd, id2 INTEGER*2 sigs,sigm, epoc, mmag,sigp, nit,niu . ,pmr,pmd,pme, jmag,ejmag, hmag,ehmag, kmag,ekmag INTEGER*2 abm,avm,agm,arm,aim, ebm,evm,egm,erm,eim INTEGER*1 nst,nsu, nsm,ref, ngt,ngu, mf2,mfa . ,iccj,icch,icck, phqj,phqh,phqk, ann,ano OPEN (11,FILE=fnurat,ACCESS='direct',RECL=80) READ (11,REC=n,ERR=99) ! 80 byte . ra,spd, sigs,sigm, nst,nsu, epoc ! 8 + 4 + 4 . ,mmag,sigp, nsm,ref, nit,niu,ngt,ngu ! 4 + 2 + 6 . ,pmr,pmd,pme, mf2,mfa, id2 ! 6 + 2 + 4 . ,jmag,hmag,kmag, ejmag,ehmag,ekmag ! 6 + 6 . ,iccj,icch,icck, phqj,phqh,phqk ! 3 + 3 . ,abm,avm,agm,arm,aim ! 10 . ,ebm,evm,egm,erm,eim, ann,ano ! 10 + 2 column item type unit description notes ------------------------------------------------------------------------ 1 ra I*4 mas mean RA on ICRF at URAT mean obs.epoch ( 1) 2 spd I*4 mas mean South Pole Distance = Dec + 90 deg ( 1) 3 sigs I*2 mas position error per coord. from scatter ( 2) 4 sigm I*2 mas position error per coord. from model ( 2) 5 nst I*1 -- tot. number of sets the star is in ( 3) 6 nsu I*1 -- n. of sets used for mean position + flag ( 3) 7 epoc I*2 myr mean URAT obs. epoch - 2000.0 ( 1) 8 mmag I*2 mmag mean URAT model fit magnitude ( 4) 9 sigp I*2 mmag URAT photometry error ( 5) 10 nsm I*1 -- number of sets used for URAT magnitude ( 3) 11 ref I*1 -- largest reference star flag ( 6) 12 nit I*2 -- total number of images (observations) 13 niu I*2 -- number of images used for mean position 14 ngt I*1 -- total number of 1st order grating obs. 15 ngu I*1 -- number of 1st order grating positions used 16 pmr I*2 0.1mas/yr proper motion RA*cosDec (from 2MASS) ( 7) 17 pmd I*2 0.1mas/yr proper motion Dec ( 7) 18 pme I*2 0.1mas/yr proper motion error per coordinate ( 8) 19 mf2 I*1 -- match flag URAT with 2MASS ( 9) 20 mfa I*1 -- match flag URAT with APASS ( 9) 21 id2 I*4 -- unique 2MASS star identification number 22 jmag I*2 mmag 2MASS J mag 23 hmag I*2 mmag 2MASS H mag 24 kmag I*2 mmag 2MASS K mag 25 ejmag I*2 mmag error 2MASS J mag 26 ehmag I*2 mmag error 2MASS H mag 27 ekmag I*2 mmag error 2MASS K mag 28 iccj I*1 -- CC flag 2MASS J (10) 29 icch I*1 -- CC flag 2MASS H 30 icck I*1 -- CC flag 2MASS K 31 phqj I*1 -- photometry quality flag 2MASS J (10) 32 phqh I*1 -- photometry quality flag 2MASS H 33 phqk I*1 -- photometry quality flag 2MASS K 34 abm I*2 mmag APASS B mag (11) 35 avm I*2 mmag APASS V mag 36 agm I*2 mmag APASS g mag 37 arm I*2 mmag APASS r mag 38 aim I*2 mmag APASS i mag 39 ebm I*2 mmag error APASS B mag 40 evm I*2 mmag error APASS V mag 41 egm I*2 mmag error APASS g mag 42 erm I*2 mmag error APASS r mag 43 eim I*2 mmag error APASS i mag 44 ann I*1 -- APASS numb. of nights (12) 45 ano I*1 -- APASS numb. of observ. (12) ------------------------------------------------------------------------ 80 = total number of bytes per star record 5b) Notes to zone file data --------------------------- Note (1): Positions are on the International Celestial Reference System (ICRS) as represented by the UCAC4 catalog. Mean observed positions are given at mean epoch of URAT observations (column 7). Thus the epoch is slightly different from star to star, but it is always in the range between 2012.311 and 2014.679. Instead of Declination the always positive south pole distance is given. To obtain Declination subtract 90 degree (= 324000000 mas). Note (2): Column 3 gives the mean position error of the mean position as given in columns 1 and 2 as derived from the scatter of individual observations. A mean was taken over RA and Dec component errors because they are very similar for most stars. In case only a single observation was made a conservative estimate from error propagation is made. An error floor of 5 mas is added RMS to the scatter error to arrive at the value given in URAT1. Column 4 gives another estimate of the error of the mean position components as given in columns 1 and 2. Here a model is used which include image profile fit (x,y data) errors, atmospheric turbulence, and astrometric reduction error propagations. As for the scatter error, a systematic error floor of 5 mas was added RMS. The model error is likely a better estimate of the true positional errors than the scatter error, at least for small numbers of observations. Note (3): The 2 years of observing used for this catalog release was split into a total of 38 sets of data (19 each for regular survey and grating survey). Unless very bright or very faint or variable, a source on the sky should show up in several sets (adjacent sets, 1st and 2nd year at same season...). If an object is within the about 5 to 17.5 mag range and does not show up in several sets it is suspicious and might be a variable or moving source or sometimes blended image. Item 6 (nsu) is set negative for stars which were not matched with the GSC 2.4 catalog (about 1.1 million objects, priv. comm. R.Smart). Some of these are fake entries in URAT (like a minor planet at a certain epoch), others appear to be slightly offset from a real star (possible large proper motion object), and others are at the limiting magnitude thesholds. In order to pick a somewhat cleaner sample of URAT reference stars these objects could be avoided, or at least those which also do not match up with 2MASS. Note (4): This is the mean, observed magnitude in the 680-762 nm URAT bandpass, calibrated by APASS photometry. This bandpass is between R and I, thus further into the red than UCAC. Observations in non-photometric nights *are* included thus the URAT magnitudes need to be taken with caution. Unknown or unrealistic magnitudes are set to 30000 milli-mag. The faintest maybe real celestial object magnitude is about 19.0, while the URAT1 catalog should be complete to about magnitude 18.0. Note (5): The photometric error of URAT bandpass observations is derived from the scatter of individual observations. A systematic error floor of 0.01 mag has been RMS added. Unknown errors are indicated by 900 mmag. Note (6): Each object on every individual exposure and CCD gets a reference star flag assigned during the conventional, astrometric "plate" reduction. The largest value of this flag for a given star over all its observations is carried into this URAT1 catalog flag. The meaning of the values of this flag is as follows: 0 = not a reference star 1 = o.k, both coordinates used as reference star 2 = x-coord. excluded in adjustment 3 = y-coord. excluded in adjustment 4 = x and y-coord. excluded in adjustment 5 = not selected due to astrometry flag 6 = too bright on URAT exposure (amplitude limit exceeded) 7 = no valid sigma_xy from profile fit on URAT exposure 8 = excluded by distance from URAT center (accepted field of view) 9 = excluded because outside of adopted UCAC4 magnitude range The UCAC4 was used as reference star catalog, however, only "clean" (astrometry flag) UCAC4 stars within the 8.0 to 16.0 mag were used. Thus there are some UCAC4 stars with reference star flag 5 and 9. These are identified, valid UCAC4 stars but were not used as reference stars in the URAT1 astrometric reductions. Note (7): These are strictly 2-epoch proper motions using the 2MASS point source catalog as early epoch and the URAT observational data as second epoch. A match radius of 3 arcsec was used to identify common sources. With a mean epoch difference of about 14 years, the largest valid proper motion can be only about 200 mas/yr. Stars with higher proper motions were not attempted to match for this release, neither were other catalogs used to improve the proper motions. The goal here was to quickly provide the community with a useful product: accurate reference stars at current epoch. The provided preliminary proper motions allow to use the URAT1 data for a window of +- several years of its observational epoch without significant degradation of positional accuracy. Note (8): The error on a proper motion component was derived from error propagation of assumed precision of the input data. A fixed position error of 80 mas was adopted for 2MASS and the model position error of individual URAT stars was used to derive the error in proper motion. No systematic error floor was added here. Typical errors in proper motions are 5 to 8 mas/yr. Note (9): The meaning of the match flag is as follows: 1 = unique match 3 = multiple match but only 1 match with high confidence 5 = closest positional match used in case of multiple matches 7 = 2nd catalog multiple match, use closest positional match 9 = likely wrong match 11 = no match at all Thus match flags up to 3 are considered "good", 5 and 7 "maybe". Note (10): Values of columns 21 through 33 are just copied from the 2MASS point source catalog (PSC). In case no 2MASS data are available, the magnitudes are set to 30000 mmg, photometric errors to 9000 mmag, and flags to 0. For each 2MASS bandpass 2 flags are given, the cc_flg and the ph_qual flag with the following explanations as provided with the 2MASS documentation: 0 = cc_flg 2MASS 0, no artifacts or contamination 1 = cc_flg 2MASS p, source may be contaminated by a latent image 2 = cc_flg 2MASS c, photometric confusion 3 = cc_flg 2MASS d, diffraction spike confusion 4 = cc_flg 2MASS s, electronic stripe 5 = cc_flg 2MASS b, bandmerge confusion 0 = no ph_qual flag 1 = ph_qual 2MASS X, no valid brightness estimate 2 = ph_qual 2MASS U, upper limit on magnitude 3 = ph_qual 2MASS F, no reliable estimate of the photometric error 4 = ph_qual 2MASS E, goodness-of-fit quality of profile-fit poor 5 = ph_qual 2MASS A, valid measurement, [jhk]snr>10 AND [jhk]cmsig<0.10857 6 = ph_qual 2MASS B, valid measurement, [jhk]snr> 7 AND [jhk]cmsig<0.15510 7 = ph_qual 2MASS C, valid measurement, [jhk]snr> 5 AND [jhk]cmsig<0.21714 8 = ph_qual 2MASS D, valid measurement, no [jhk]snr OR [jhk]cmsig req. Note (11): A custom set of APASS (The AAVSO Photometric All-Sky Survey) data was kindly provided to us by Arne Henden to include the DR8 data plus single photometric observations. Columns 34 through 45 contain a copy of these APASS data. For a total of 71614 stars with no DR8 data the DR6 data was used. In case no APASS data are available, the magnitudes are set to 30000 mmag and the photometric errors to 9000 mmag, consistent with 2MASS "no data" entries in URAT1. Note (12): A few stars had more than 127 APASS nights or observations. These numbers were trimmed to 127 to fit into the signed I*1 variable. 5c) min / max data values of zone files --------------------------------------- The following table gives the total range of values in URAT1 data columns: col min max data description ---------------------------------------------------------------------- 1 37 1295999997 mean RA [mas] 2 234696814 647515125 mean SPD [mas] 3 5 818 position error scatter [mas] 4 5 429 position error model [mas] 5 1 26 tot. number of sets 6 -24 26 number of sets used for mean position 7 12311 14679 mean URAT obs. epoch - 2000.0 [1/1000 yr] 8 726 30000 mean URAT model fit magnitude [mmag] 9 10 900 URAT photometry error [mmag] 10 0 26 number of sets used for URAT magnitude 11 0 9 largest reference star flag 12 1 397 total number of observations 13 1 353 number of observ. used for mean position 14 0 127 total number of 1st order grating obs. 15 0 127 number of 1st order grating positions used 16 -2570 2558 proper motion RA*cosDec [0.1 mas/yr] 17 -2570 2588 proper motion Dec [0.1 mas/yr] 18 48 900 error proper motion [0.1 mas/yr] 19 1 11 match flag URAT with 2MASS 20 1 11 match flag URAT with APASS 21 0 1340037284 unique 2MASS star identification number 22 -2989 30000 2MASS J mag 23 -4007 30000 2MASS H mag 24 -4378 30000 2MASS K mag 25 13 9000 error 2MASS J mag 26 10 9000 error 2MASS H mag 27 11 9000 error 2MASS K mag 28 0 5 ICC flag 2MASS J 29 0 5 ICC flag 2MASS H 30 0 5 ICC flag 2MASS K 31 0 8 photometry quality flag 2MASS J 32 0 8 photometry quality flag 2MASS H 33 0 8 photometry quality flag 2MASS K 34 5494 30000 APASS B mag 35 5676 30000 APASS V mag 36 6290 30000 APASS g mag 37 5754 30000 APASS r mag 38 5011 30000 APASS i mag 39 -900 9000 error APASS B mag 40 -900 9000 error APASS V mag 41 -900 9000 error APASS g mag 42 -900 9000 error APASS r mag 43 -900 9000 error APASS i mag 44 0 127 APASS numb. of nights 45 0 127 APASS numb. of observ. ---------------------------------------------------------------------- 5d) Format of index files ------------------------- v1index.asc = index file for each 1 arcmin in RA, all zone, ASCII (in same directory as the binary zone files) This is a text file (ASCII). Each line contains data for an area of sky 0.2 deg wide in declination (indexed by zone number from 326 to 900), and 1 arcmin wide in RA (indexed from 1 to 1440). So there are a total of 575 * 1440 = 828,000 lines on this file. The meaning of the columns are: 1 zn = zone number (326 to 900) 2 j = index for bins along RA (1 to 1440) 3 n1 = running star number (index along given zone) of 1st star in this bin (or 0 if no stars) 4 ns = number of stars in that bin 5 nn = running star number (index along given zone) of the last star in this bin (or 0 if no stars) 6 ni = same as nn except for bins with no stars where ni is the index of the last previous star in the zone ni alone can be used to figure out how many stars are in a given bin and which record number along the zone they have by looking at ni of the current and previous bin. sample data: zn j n1 ns nn ni ------------------------------------ 326 1131 0 0 0 0 326 1132 1 7 7 7 326 1133 8 15 22 22 326 1134 23 12 34 34 326 1135 35 12 46 46 326 1136 47 17 63 63 326 1137 64 19 82 82 326 1138 83 14 96 96 326 1139 97 12 108 108 326 1140 109 1 109 109 326 1141 0 0 0 109 326 1142 110 22 131 131 326 1143 132 6 137 137 326 1144 138 9 146 146 326 1145 147 1 147 147 326 1146 0 0 0 147 326 1147 0 0 0 147 v1index.unf = unformatted, binary file containing 2 records, the array of ni and nn (see above, v1index.asc) INTEGER*4 ni(900,1440), nn(900,1440) OPEN (15,FILE=fnidxu,ACCESS='direct',RECL=5184000) ! 900 * 1440 * 4 READ (15,REC=1) ni READ (15,REC=2) nn Note, entries for zone numbers 1 through 325 are zero. The second record (nn) is added for convenience. 5e) Description of other files ------------------------------ chkv1d.log in the "info" directory contains histograms and mean data values for columns of the zone file data by magnitude. It is an ASCII text file with some explanations given in the file. v1zone.tab = table with info for each zone ("info" directory) This is a text file (ASCII) giving summary information about stars in each zone. There are 575 lines, 1 each for a 0.2 deg wide zone in declination, beginning at zone 326, ending at 900. The meaning of the columns are: zn = zone number max_dec = largest declination [degree] of zone nsz = number of stars in zone n0z = number of bins along RA with no data (1440 bins, 0.25deg) ndz = number of bins along RA with data nopmz = number of URAT1 stars in that zone without proper motion nst = total number of stars accumulated at end of zone sample data: zn max_dec nsz n0z ndz nopmz nst --------------------------------------------------------------- 326 -24.80 147 1427 13 0 148 327 -24.60 14184 1411 29 23 14332 328 -24.40 19058 1407 33 58 33390 329 -24.20 27058 1402 38 134 60448 ... 897 89.40 2159 335 1105 48 228274170 898 89.60 1478 516 924 16 228275648 899 89.80 800 824 616 3 228276448 900 90.00 35 1406 34 0 228276482 urat1-cov14.pdf (in "info" directory) is an all-sky plot showing the number of URAT1 observations used per star (average over small bins on the sky and color coded). This plot is for stars around magnitude 14 which are visible in both the regular and the grating survey data of any 20 to 240 sec exposure. urat1-cov14.pdf (in "info" directory) shows the same as the previous plot, however, for stars around magnitude 18, i.e. the faint end only seen on the longest (240 sec), regular survey exposures. urat1-skyden.pdf (in "info" directory) shows the stellar density (number of URAT1 stars per square degree) averaged over small bins on the sky and color coded. 6) Description of URAT1 access software --------------------------------------- Fortran code and sample output is provided in the folder "access". There are 2 main programs (v1dump.f and v1access.f) and an additional file with subroutines (v1sub.f). Compile string examples are given in the comment section on top of each main program file. The programs can be run from the command line interactively with instructions provided. v1dump.f = read binary zone file(s), output all data items to ASCII file with optional column separator character, and star ID name appended at the end of the output record v1access.f= example code to access UCAC4 data as function of RA,Dec,mag and output with various formats, utilizing the index file, and with option to run over a list of targets v1sub.f = file with subroutines used in v1dump and v1access urat1.tab1= sample output from a run of v1access (format option 1) urat1.tab3= sample output from a run of v1access (format option 3) z326.asc = sample output from a run of v1dump z326.csv = same output as comma separated value table z327.asc, z327.csv = same as above for a zone with more data A translation of the access code into C can be found at Bill Gray's web page: http://www.projectpluto.com/urat.htm 7) People --------- 7a) URAT team ------------- Norbert Zacharias - Principle Investigator, occasional observer, responsible for URAT astrometry (pixel reductions and astrometric "plate" solution code). Charlie Finch - lead in operations, observing, training of observers, quality control, catalog comparisons, reduction pipeline development and data processing. John Subasavage - point of contact at NOFS, observing and operations. Trudy Tilleman - main observer at NOFS, assist with operations. Other NOFS observers include Chris Crockett, Hugh Harris and Jeff Munn. Gary Wieder (lead), Chris Kilian, Eric Furguson of USNO DC instrument shop, design and construction of URAT telescope tube, set-up in DC and at NOFS. Mike Divittorio (lead), Albert Rhodes, Mike Schultheiss of NOFS instrument shop, set-up at NOFS, dome improvements, telescope maintenance, operations. Fred Harris - consultation camera, design and built of LN2 auto-fill system. Greg Bredthauer - contractor, creator of camera electronics and low-level computer interface for camera and telescope, instrumental in getting URAT operational during testing and debug phase, continued consultations throughout the project. Ted Rafferty - contractor, instrumentation specialist, set-up and testing of hardware in DC and at NOFS. 7b) Acknowledgments ------------------- The USNO management is thanked for supporting this project: K.J.Johnston, B.Luzum - former and current scientific director of USNO, R.Gaume, B.Dorland - former and current head of the astrometry department, and P.Shankland - director of NOFS. Semiconductor Technology Associates (headed by R.Bredthauer) provided continued support for the URAT camera and dewar long after delivery. Arne Henden is thanked for providing unpublished APASS data for our project. Bill Gray (Project Pluto) is thanked for making available a C code version of our URAT access software: www.projectpluto.com/urat.htm 2MASS was used for near IR photometry and as 1st epoch of URAT proper motions. Aladin and Vizier were invaluable tools provided through CDS, Strasbourg. CDS Strasbourg is thanked for hosting the URAT1 catalog. DS9 by the Smithsonian Astrophysical Observatory was used as display tool for FITS pixel data files. NOAO is thanked for IRAF, which was used for image analysis while trouble shooting and performing spot checks. Pgplot by California Institute of Technology was used to produce plots. 7c) Testers ----------- The following people contributed to data verification, testing and upgrades going from the URAT 1.0 version of November 2014, to the published version 1.2 of URAT: Dave Herald (Australia, minor planet occulation community) Ricky Smart (Italy, comparison to GSC 2.4) Christine Ducourant (France, comparison to PM2000) Rama Teixeira (Brazil, vector point diagram, galactic model comparison) Jeff Munn (USNO, comparison to SDSS) John Subasavage, Hugh Harris (USNO, 61in data comparison) Julien Frouard (USNO, v1dump.f bug fix) Charlie Finch, Norbert Zacharias (USNO, various checks and updates) 8) References ------------- This list includes items relevant to URAT even if they are not cited in the above text. Finch,C., Bredthauer,G., DiVittorio,M., Harris,F., Rafferty,T., Wieder,G., Zacharias,N. 2012, "USNO Robotic Telescope (URAT) underway", 220th AAS meeting #135.06 Finch,C., Zacharias,N., DiVittorio,M., Ferguson,E., Harris,F., Harris,H., Kilian,C., Rafferty,T., Rhodes,A., Schultheis,M., Subsavage,J., Tilleman,T., Wieder,G. 2014, "URAT - year 3", poster paper at the Philadelphia DDA meeting Laux,U., Zacharias,N. 2005, "URAT optical design options and astrometric performance", in Proc. "Astrometry in the Age of the Next Generation of Large Telescopes", Eds.: K.Seidelmann & A.K.B.Monet, APS conf.ser. 338, p.184 Skrutskie,M. F., Cutri,R. M., Stiening,R., Weinberg,M. D., Schneider,S., Carpenter,J. M., Beichman,C., Capps,R., Chester,T., Elias,J., Huchra,J., Liebert,J., Lonsdale,C., Monet,D. G., Price,S., Seitzer,P., Jarrett,T., J.D. Kirkpatrick, J. Gizis, E. Howard, T. Evans, J. Fowler, L. Fullmer, Hurt,R., Light,R., Kopan,E. L., Marsh,K. A., McCallon,H. L., Tam,R., Van Dyk,S., Wheelock,S. 2006, "The Two Micron All Sky Survey (2MASS)", AJ, 131, 1163 Zacharias, N., 2002 "Astrometric surveys in support of large telescopes", Proceed. SPIE 4836, 279, Eds. T.A. Tyson & S. Wolff Zacharias,N. 2004 "Astrometric reference stars: from UCAC to URAT", proceed. 3rd Potsdam Thinkshop on robotic telescopes, AN 325, 631 (2004) Zacharis,N. 2005, "The URAT project", in Proceedings of "Astrometry in the Age of the Next Generation of Large Telescopes", Eds.: K.Seidelmann & A.K.B.Monet, APS conf.ser. 338, p.184 Zacharias,N., Laux,U., Rakich,A. & Epps,H., 2006, "URAT: astrometric requirements and design history", Proc. SPIE 6267E..227, astro-ph/0606193v1 Zacharias,N. 2006, BAAS 38, p.674, #12.02, "Status of UCAC and URAT projects", abstract 37th DDA, Halifax Zacharias,N. 2008, "Dense optical reference frames: UCAC and URAT", in Proc. IAU Symp. 248 (Shanghai, China, Oct. 2007): "A giant step: from milli- to micro-arcsecond astrometry". Eds. W.Jin, I.Platais, M.A.C.Perryman, Cambridge Univ.Press, p.310 Zacharias,N., Wieder,G., Bredthauer,R. 2009, "USNO Robotic Astrometric Telescope (URAT) phase 1 = U-mouse", 213th AAS meeting, Long Beach, CA, January 2009, BAAS 41, p.421 #470.11 Zacharias,N., Gaume,R. 2010, UCAC and URAT: optical astrometric catalog observing programs, proceed. Journees meeting Paris 2010, Ed. N.Capitaine Zacharias,N., Bredthauer,G., DiVittorio,M., Finch,C., Gaume,R., Harris,F., Rafferty,T., Rhodes,A., Schultheis,M., Subasavage,J., Tilleman,T., Wieder,G. 2012, "The URAT project", presentation at the IAU GA 28, Com. 8 session, http://www.ast.cam.ac.uk/ioa/iau_comm8/iau28/ Zacharias,N. 2015, "Bright Star Astrometry with URAT", proceedings of on invited talk at the ADeLA 2014 meeting, Santiago, Chile, in press Zacharias,N. et. al. 2015, URAT1 release paper in prep. for AJ ---------------------------------------------------------------------- Norbert Zacharias and Charlie Finch for the UCAC team, U.S. Naval Observatory, Washington DC, November 2014 - February 2015 ----------------------------------------------------------------------