The SBU ANDICAM database

This is a guidebook for Stony Brook users of ANDICAM images. It requires access to the raid array on osiris, and some knowledge of IDL. The IDL procedures are located in /home/fwalter/LIBS. This directory must be in the IDL_PATH variable in your .cshrc file.

Contents

I. Introduction

The ANDICAM instrument is a dual channel imager used mostly for obtaining long time series on variable objects for relative photometry. Instrumental details are available at the Ohio State University ANDICAM web site. The instrument is mounted on the 1.3m telescope (formerly the 2MASS telescope) at Cerro Tololo.

Images are taken by service observers. The data are downloaded to Yale each morning, where the data are archived. The optical (CCD) images are processed, by trimming the images, subtracting the bias, and dividing by the flat fields. The infrared images are rebinned to a manageable scale, but are otherwise archived raw. The images are placed on the SMARTS ftp site after processing. We download the data from there.

We utilize both the optical (CCD) and infra-red (IR) channels.

II. Data Storage

The images themselves are stored on the raid array on osiris. The catalog of images is maintained in a pair of IDL databases, one for each channel. For manageability, a new database is created for each observing year. SMARTS began operations in 2003. Note that the observing year generally, but not always, runs from February through January. After downloading the images, we update the databases.

In order to access the databases, you must have the following line in your .cshrc file:

setenv ZDBASE /data/fwalter/DB/

In order to access the IDL software, you must have a line in your .cshrc file that looks like the following:
setenv IDL_PATH +/usr/local/rsi/idl_5.3/local_pro:+/home/fwalter/ASTROLIB:+/home/fwalter/LIBS:+/usr/local/rsi/idl_5.3/lib:

The databases are named s13_yyyy and s13ir_yyyy, for the optical and near-infrared data, respectively. yyyy is the year (e.g., 2006)

While you can in principle query the databases using any IDL databasing routines, most routine queries are handled through the software described below.

III. Observing Programs

The long term monitoring programs we are engaged in include: Finding charts for most of the targets are available here.

IV. Querying the Databases

There are two primary ways of querying the databases.

  1. When was my target observed?
    Use the IDL procedure GET_ANDICAM (note: IDL is not case-sensitive). You must know the name of the target. The basic call, to find a target called V1118 Ori (an Exor), would be
    GET_ANDICAM,'V1118 Ori'
    Note that the name of the target is not case-sensitive, and may be abbreviated. Spaces are important. GET_ANDICAM,'V1118' will yield the same result. However, GET_ANDICAM,'V1' will list all targets whose names include the characters V1.

    GET_ANDICAM lists only the dates (in yymmdd format) that the observation was made: it doesn't provide any other information. For more you'll need to run S13DB_TARGET (see below).

    GET_ANDICAM can be run with a number of optional keyword parameters. Type GET_ANDICAM,/help for a full listing. The most useful keywords are:

    • /Yxxxx: use to search the database for the year xxxx. The default is the current year.
    • /IR: set to query the infrared database rather than the optical database.

    For example, to find out when we got near-IR images of the polar QS Tel in 2005, you would type GET_ANDICAM,'QS Tel',/y2005,/ir

    Note that GET_ANDICAM can also be used to extract data from the database. The full list of option is:

    * GET_ANDICAM - get ANDICAM data
    * calling sequence: GET_ANDICAM,target,date,im,filt,med
    *    TARGET: name of target, as recorded in s13_* database
    *    DATE:   date of data to be grabbed.
    *            If omitted, a list of dates is returned.
         EXPT:   returned exposure times
    *    IM:     a 1024x1024xn array containing all the images for this date
    *    FILT:   list of filters corresponsing to the n images
    *    MED:    the median-filtered image, if only one filter exists OR
    *            keyword FILTER is specified
    *
    * KEYWORDS
    *    ALL:    set to return all data when both ccd and rccd files exist
    *    DB:     data base to be searched, def=s13_2003
    *    FILTER: set to one of U,B,V,R,I to return image(s) in only one band
    *    IR      set to read IR database, def= CCD
    *    JD:     output Julian date of start of obs
    *    MARK:   filter names for interactive shifting, if needed
    *    MERGE:  set to merge CCd images, or sky-subtract and merge IR images
    *    MSC:    upper TV scaling parameter, def=10
    *    NOBINDOWN: if not set, rebin to 512x512
    *    PROPID: set to search on propid rather than target name
    *    RAW:    set to return unprocessed data when both ccd and rccd files exist
    *    SAVE:   set to save merged image in .fits file
    *    SET:    set number in multiobservational set, def=1
    *    THISDIR: use files in this directory rather than database
    *    TIMES:   output JD hours at start of observation
    *    TV:     set to display median or reddest image
    *    Y200x:  search database for year 200x; def=this year.
    *    YSTART: starting Y pixel, def=12, height=952
    
  2. Details of an observation
    To find out the details of an observation on a particular day, use the procedure S13DB_TARGET. Type S13DB_TARGET,/help for on-line help. The main call is S13DB_TARGET,target,date. If you omit the date you'll get a listing for the entire year. The date is a string of the format yymmdd.

    For example, to get the details of the V1118 Ori observations on Jan 16 2006, you would type S13DB_TARGET,'V1118 Ori','060116'. The output is

          Catalog of SMARTS ANDICAM Observations  2006
     ----------------------------------------------------
      civil    Date        Time      filter Exposur air
      date                                   time   mass
                                            (sec)
     ----------------------------------------------------
     060116 2006-01-17 03:01:39     B        66.00 1.110
     060116 2006-01-17 03:03:35     B        66.00 1.110
     060116 2006-01-17 03:05:30     B        66.00 1.110
     060116 2006-01-17 03:07:45     V        23.00 1.110
     060116 2006-01-17 03:08:58     V        23.00 1.110
     060116 2006-01-17 03:10:10     V        23.00 1.110
     060116 2006-01-17 03:11:38     R        18.00 1.110
     060116 2006-01-17 03:12:45     R        18.00 1.110
     060116 2006-01-17 03:14:16     I        10.00 1.120
     060116 2006-01-17 03:15:15     I        10.00 1.120
    
    The Civil date is the date at the start of the night in yymmdd format. The Date and Time are in UT. In this example, there were 10 individual exposures in the 4 filters.
  3. What was observed on a particular day?
    To find out what targets were observed on a particular day, use the procedure S13DB_TODAY. A call with no arguments looks for data taken last night. Unless we're really on the ball, it probably won't have been downloaded and archived yet. To see the most recent data, type S13DB_TODAY,/last. To see a particular date, type S13DB_TODAY,date where date is a string of the format yymmdd. For example, to see what else was oberved on Jan 16 2006, you would type S13DB_TODAY,'060116'. The output should look like this:
             Catalog of SMARTS ANDICAM Observations  2006
     -----------------------------------------------------------
      civil     Proposal           Object         filter Exposur
      date         ID                                     time
                                                         (sec)
     -----------------------------------------------------------
     060116 SUNY-03B-0001   BL Hyi               B       100.00
     060116 SUNY-03B-0001   BL Hyi               V       100.00
     060116 SUNY-03B-0001   BL Hyi               I       100.00
     060116 SUNY-03B-0002   V1118 Ori            B        66.00
     060116 SUNY-03B-0002   V1118 Ori            B        66.00
     060116 SUNY-03B-0002   V1118 Ori            B        66.00
     060116 SUNY-03B-0002   V1118 Ori            V        23.00
     060116 SUNY-03B-0002   V1118 Ori            V        23.00
     060116 SUNY-03B-0002   V1118 Ori            V        23.00
     060116 SUNY-03B-0002   V1118 Ori            R        18.00
     060116 SUNY-03B-0002   V1118 Ori            R        18.00
     060116 SUNY-03B-0002   V1118 Ori            I        10.00
     060116 SUNY-03B-0002   V1118 Ori            I        10.00
     060116 SUNY-04A-0011   Nova LMC 2005        V        90.00
     060116 SUNY-04A-0011   Nova LMC 2005        R        30.00
     060116 SUNY-04A-0011   Nova LMC 2005        I        90.00
     060116 SUNY-04A-0011   Nova LMC 2005        B        90.00
     060116 SUNY-04B-0012   V1647 Ori            I       101.00
     060116 SUNY-04B-0012   V1647 Ori            I       101.00
     060116 SUNY-04B-0012   V1647 Ori            I       101.00
     060116 SUNY-04B-0012   V1647 Ori            R       101.00
     060116 SUNY-04B-0012   V1647 Ori            R       101.00
     060116 SUNY-04B-0012   V1647 Ori            R       101.00
     060116 SUNY-04B-0012   V1647 Ori            V       197.00
     060116 SUNY-04B-0012   V1647 Ori            V       197.00
     060116 SUNY-04B-0012   V1647 Ori            V       197.00
    
    Four of our targets were observed on that night.

    The formal ANDICAM observing logs are maintained at this site. These logs are not easily searchable, and some are missing, but they are a good source of information. Data that we have not downloaded will not be listed in our database.

    V. Doing the Photometry

    We do aperture photometry here. Basically, one sums the number of counts within some radius of the star (the aperture). The background is measured in an annulus surrounding the aperture. The background is the median value of the counts in the aperture, so that background stars inadvertently in the aperture have little effect. The background level is subtracted off, and uncertainties are determined. The net counts are divided by the exposure time to yield the net counts/second. This is converted to an instrumental magnitude, which is merely -2.5 times the log of the count rate,

    If we had an unchanging sky, so that the atmospheric transparency was constant,this would be enough. But the sky varies. Clouds drift over. Seeing changes. Therefore, we also do photometry on a sample of other stars in the field. These are the comparison stars. The assumption behind relative photometry is that the ensemble average of the brightness of the other stars does not change. Any variations we see must be due to the atmosphere. This will affect out targets and all the other stars in the field equally. We can therefore use the ratio of the counts in our target to the counts in the comparison stars to study the true variations in our target. The ratio of counts is the difference of instrumental magnitudes.

    V.a. CCD Photometry

    The procedure ANDICAM_PHOT is used for the CCD photometry. In principle, you merely type ANDICAM_PHOT,target,date, where target is the name of the target and date is the date in yymmdd format. However, you need to ensure reproducibility of the results. This means that you must measure the same comparison stars every time. This is accomplished by selecting the stars, and writing their positions to a .compos file. You can then use this file to select the comparison stars, either by specifying the file name with the COMPOS keyword, or by using the TARGET keyword. The TARGET keyword reads a file that contains additional initialization information keyed to the target name.

    Generally the .compos file will exist. You will need to know where the target is, because all positions are measured relative to that. The target can be identified on its finding chart.

    The usual procedure is to type

    ANDICAM_PHOT,target=name,date
    where name is a unique string identifying the target, generally the same as the name you use in GET_ANDICAM, and date is the date in yymmdd format.

    ANDICAM_PHOT will pause and ask you to mark the target. Move the cursor to the target position and click the left mouse button. After that, the program should continue autonomously.

    If you click the right mouse button additional information about the target will be written to the terminal, and the radial counts plot will be shown in another window. This additional data will look like this:

     Filter = V   ecor=      1.14673
     Centroid at       474.940      558.464
    V: Source: Max =  1204.41  Net =  13085.13
    Background: Median =   335.97   Mean =   336.03 +/-   10.95  Area=  876
     Gaussian fit:
    Source: peak =   812.17 +/-  16.41 counts;  Sigma =   1.47 +/-  0.02 pix
    Net in image =   11783.86  Net in fit =   11078.55
    Background =   336.71 +/-    0.46
    
    The most important values are:
    • Source: Max (line 3): If this exceeds about 45,000 the image may be saturated. This should be noted.
    • Sigma =: This is a measure of the seeing.
    After showing the additional data, the program will pause: hit any key to continue.

    When done, the output is sent to the terminal. It will look like this (the lines may be much longer and wrap around):

    060112 B 53748.732  5.568  -5.692  0.019   -4.372   0.061   -5.824   0.017   -5.213   0.029   -5.394   0.025   
    060112 V 53748.734  5.614  -5.475  0.020   -4.329   0.055   -6.271   0.010   -5.527   0.019   -5.911   0.014   
    060112 I 53748.736  5.662  -5.651  0.015   -4.127   0.058   -6.296   0.009   -5.564   0.016   -6.023   0.011   
    

    There will be one line for every image measured. Each line consists of the following: the civil date, the filter, the Julian date (-2400000), the starting time in hours UT (which give more accuracy than the Julian date), and then pairs of instrumental magnitudes and errors. The first pair is the target; subsequent pairs are the comparison stars.

    You should cut and past the output lines into a text file called target.mag (where "target" is the name of the target. You should also check the IMEXAM output, and write the Source: Max and Sigma values on a commented line (one starting with a semi-colon).

    ANDICAM_PHOT has lots of other options that you should not have to invoke. The online help file is:

    * ANDICAM_PHOT - ANDICAM Photometry: combines get_andicam and apphot
    * calling sequence: ANDICAM_PHOT,target,date
    *    TARGET,DATE: required
    *
    * KEYWORDS:
    *    COMPOS:    file containing comparison star positions
    *    IMAGE:     optional output image array
    *    LOUD:      set for full apphot output
    *    MARKALL:   set to mark stars in each image
    *    MARKTARGET: set to mark target in each image
    *    MERGE:     set to merge multiple images in same filter
    *    PDM:       set to compute and print differential magnitudes
    *    RADIUS:    apphot radius, def=9
    *    RADT:      apphot radius for target, def=RAD
    *    RANGE:     plot range, def=100
    *    SET:       data set to read
    *    SHIFT:     set to shift images prior to merging
    *    STARTFILT: filter to start with
    *    TARGET:    name of target, if defined for automatic extractions
    *    USEIMAGE:  name of input image - bypasses database
    *    ZOUT:      output magnitude differences
    

    V.b. IR Photometry

    ANDICAM_IR is the procedure that you use to do aperture photometry for the IR images. It works in much the same way as ANDICAM_PHOT, but has a section that first processes the IR images. Read the previous section on CCD photometry.

    By default, ANDICAM_IR tries to shift and add the images autonomously (the /MERGE keyword). It often fails: the consequence is images with doubled or trailed stars. To help the program out, use the keyword IDT=2. The procedure will then display each image, and you will click on a bright object (it need not be the target, but it must be in all images in a given filter). The software will then center up on that. Note that these images will be rotated by 90 degrees, and will not look like the finding chart.

    After the merger process, the properly rotated, shifted-and-added image is displayed. Click (L or R, as desired) on the target. Be aware that in many cases the target is very faint in the IR bands.

    Things to note: in the IR images the non-linear response sets in at about 4000 counts. Please note if the Source: Max: value exceeds this. Also, in the IR images the sky subtraction should result in a background very close to zero. Please note if this is not the case.

    The full range of options are:

    * ANDICAM_IR - ANDICAM IR Photometry: combines get_andicam and apphot
    * calling sequence: ANDICAM_IR,target,date,zout,image
    *    TARGET,DATE: required
    *    ZOUT: output magnitude differences
    *    IMAGE: optional output image array
    *
    * KEYWORDS:
    *    COMPOS:  file containing comparison star positions
    *    IDTARG:  set to mark target in individual images in merge_ir
    *    LOUD:    set for full apphot output
    *    MARKCOMP: set to mark target and comps in all images
    *    MARKT:   set to mark target in all summed images (apphot)
    *    MED:     set to return median images, def=sum
    *    MERGE:   set to merge multiple images in same filter (def)
    *    RADIUS:  apphot radius, def=9
    *    RADT:    apphot radius for target, def=RAD
    *    RANGE:   plot range, def=10
    *    TARGET:    name of target, if defined for automatic extractions
    

    V.c. The Photometry Output File

    This is a text file with the extension .mag. The first 3 lines are comments. The first line is read by the plotting program RD_MAGDAT, and should contain the name of the target. Following that are lines output by ANDICAM_PHOT or ANDICAM_IR. For human readibilty, put blank lines between successive days. Blank lines are ignored, as are comment lines. Comment lines start with a semicolon. Nothing will be read past any line beginning with .--- (three dashes).

    VI. CCD Calibrations

    The CCD data (prefix rccd) come calibrated for instrumental effects, i.e, they are trimmed, and the bias level has been subtracted. They have also been flat=fielded. Details of the process are this web page. There is a significant region on the left and bottom sides of the image that is unuseable.

    Raw CCD data (prefix ccd) are not generally available.

    The only calibration remaining is the photometric calibration. This is not normally important, as the data are generally used for relative photometry. However, on photometric nights 1 or 2 photometric standard stars are observed. These do not yield a full photometric solution, but can be used to establish the photometric zero-point to within probably 10%.

    To use these, you must download CCD standard star observations to /raid1/DATA/SMARTS/1.3m/CCDSTAND/processed

    Then use ANDICAM_CAL to determine the zero point on a particular night.

    * ANDICAM_CAL: do crude absolute calibration for ANDICAM
    * calling sequence: ANDICAM_CAL,date,meandm
    *    DATE:   6 digit civil date
    *    MEANDM: nx5 array of delta mags, n sets, UBVRI
    *
    * KEYWORDS
    *    CHECK:  check what fields exist, do not reduce
    *    NEW:    set to re-reduce calibration fields
    *    RADIUS: aperture radus, def=13 pix
    

    Use ANDICAM_ZP to apply zero-point correction to magnitude data file. mean magnitudes amd standard deviations are printed, and returned in output MMAG and SMAG arrays (nstars x nfilters).

    * ANDICAM_ZP: apply zero point correction to ANDICAM photometry
    * Calling sequence: ANDICAM_ZPfile,mmag,smag,filt
    *    FILE: name of input magnitude file
    *    MMAG: mean magnitudes of all stars
    *    SMAG: stddev of all stars'
    *    FILT: filter names
    

    VII. IR Calibrations

    IR images (prefix binir) are rebinned 2x2 to make the 512x512 images, but are otherwise not processed. See this web page for details. Unbinned images have the prefix ir.

    The user must do the flat fielding, dark and sky subtraction, and shifting-and-adding. These tasks are built into the ANDICAM_IR procedure. For this to work correctly, you must make and use the most recent flat and dark frames. To make them from the raw calibration data, use IRFLATS.

    * IRFLATS - make ANDICAM IR flats
    * calling sequence: IRFLATS,dates
    *    DATES: list of dates to process (string)
    *
    * KEYWORDS:
    *    MON: set to yymm to process entire month
    /
    
    
    

    To find nearest IR flats on disk, use GET_NEAREST_IRCAL()

    * GET_NEAREST_IRCAL - find nearest ANDICAM IR flat files
    * calling sequence: FILES=GET_NEAREST_IRCAL,date,files
    *    DATE:  date in question
    *    FILES: output list of cal files
    

    To do a crude photometric calibration, use ANDICAM_IRCAL

    VIII. Examining the Light Curve

    Use RD_MAGDAT
    * RD_MAGDAT - read ANDICAM extracted photometry file
    * calling sequence: procname,file,jd,dm
    *    FILE: def=mag.dat
    *
    * KEYWORDS
    *    ABSOLUTE: set for absolute photometry (requires *.cmpmag or *.magcor file)
    *    COMP:     comp stars to compare: 2-element vector or mean
    *    IGNORE:   comparison star numbers to ignore  (1-N)
    *    JD:       Julian dates, 2003, 2004, range(2),min, -max
    *    LABEL:    label stars
    *    STARS:    select comp stars, def=all. Supercedes IGNORE
    *    STOUT:    named structure for output variables
    *    NOCONNECT: if not set, dots are connected
    *    OPTREND:  set to overplot linear trend
    *    PLT:      set to plot delta mag
    *    PMM:      set to use mean comparison magnitude (OPMM to overplot)
    *    PWM:      set to use weighted mean comparison magnitude (OPWM "" )