Science Operations

Calibration Overview

Baseline Calibrations

 Measurement  LBCB  LBCR
Biases 25 recommended/run 25 recommended/run
/lbt/lbto/lbc/Calib_OBs/25Bias_Bino.ob  (run time ~17 min). There is also a 50Bias_Bino.ob.
Twilight Flats 5-8 per filter x 2 PAs @ 10,000-20,000 counts each per run  (PAs should be 180 deg apart). 5-8 per filter x 2 PAs @ 10,000-20,000 counts each per run  (PAs should be 180 deg apart).
blank sky fields for sky flats
run mkskyflat
Standards *Dependent on photometric requirements, described in more detail below

Calibration Details

LBCB LBCR
Bad Pixel Map TBD TBD
 Bias

Biases are taken with the telescope stationary, lights off, and the UBessel filter in place to minimize light leaks. The Bias has structure in both the X and Y direction. The overscan is removed from all images to remove the structure in the Y direction, this also removes any variation in the zero level(1). The Bias is subtracted to remove the remaining structure.

Biases are taken with the telescope stationary, lights off, and the F972N20 filter in place to minimize light leaks.  The bias images are flat along the x and y direction, however, the zero level varies(1).

 Dark

Darks are not normally taken. For both LBC’s the average dark current is about 0.01 ADU/s which is comparable to readnoise in a 10 minute exposure. This value is considered an upper limit considering other effects like fiber glow, and other light leaks play a larger factor.

If desired, OB’s that take two or three 900second darks are available.  These should be run with the telescope stationary and the chamber lights off.

2Darks_Bino.ob (2) – 2x900sec Darks, total run time 31 min
3Darks_Bino.ob -3x900sec Darks, total run time 47 min

 Twilight Flat

The relative detector pixel response is calibrated by dividing by a scaled “flat field” calibration image. For LBC, flat field images are obtained from exposures of the twilight sky. LBC twilight flats are taken the telescope setup in TRACK mode, on a blank field.   If no blank field was predetermined, a “standard” position of HA ~ -00:30, and declination = 27 degrees can be used unless this coincides with the Galactic Plane.  Twilight flats are started/ended with the sun about 6 deg below horizon (15-20 min after sunset or before sunrise). To reduce the effect of the sky illumination gradient, it is recommended to take twilight flats at 2 or 4 PA’s and average them (PA’s 0, 90, 180, 270 deg). Twilight flats can be combined with night flats to remove this gradient as well.
The perl script mkskyflat generates a binocular OB to take a series of 5 dithered exposures about standard position and a binocular “FlatTest” OB. After creating the OBs, load and run the FlatTest OB. This will read-out a small portion of the detector to check count levels. Once the observer is satisfied with the counts, a 5-dither flat OB can be loaded. We suggest 5-8 frames of 10,000-20,000 counts each.
If skyflats for multiple filters need to be obtained, the filters with the lowest throughput (U-Bessel, USpec, B-Bessel, z-SLOAN, Y, F970N20) should be run closer to sunset/sunrise than those with the highest throughput (g-SLOAN, I-Bessel, i-SLOAN).

Standards

Standards, used for flux calibration, are only taken under clear skies.  For photometric calibration of ~10%,  flux standards can be used from other nights.  The 4-point default Dithering Pattern is designed to position the target on the center of each of the four chips. This is the pattern used in the Calibration OBs for observations of standard stars.  The Standards Table lists the Landolt Photometric Standards and other Spectrophotometric standards along with links to associated downloadable OBs (also available on site) are available in the Table of Standards.
When selecting a standard, it is important to consider the magnitude of saturation.  A table summarizing estimates of 1 second saturation magnitudes is given to help guide Standards selection.

Astrometric Distortion Corrections

Due to the optical design of LBC, astrometric corrections are necessary.  The variance in sky concentration due to the optical distortion is particularly evident, on the order of 5-6% across the FoV of the instrument. The light distribution follows a pin-cushion pattern, enhancing photon counts at the center of the focal plane and decreasing the counts near the edges. This effect depends on the filter used for data acquisition. This multiplicative effect may be removed using a correction image of the optical distortion which can be obtained or by theoretical distortion maps or by estimating the effective pixel area using the astrometric solution.
SCAMP can be used for the distortion corrections. SExtractor is run on overscanned, trimmed, bias-subtracted, and skyflat-divided data which is then fed to SCAMP for distortion correction.

(1) Varying Zero Level: The zero level decreases, stabilizing only mid-way through the sequence of 25 biases. Zero varies as the result of a safety feature installed in the cameras that causes the electronics to power off after 5 minutes of inactivity. Each time the camera powers back up the electronics take some time to stabilize again. Subtracting the overscan from each image removes the zero level.

(2) The OB’s for the Darks and Standards are also available for download as part of the Calib_OBs.tar package which contains OB’s and Readme files for Darks, Biases, Focus sequences, Standards, etc.

Sky Flats: About mkskyflat

The perl script mkskyflat creates a binocular OB that takes a series of 5 dithered exposures about a specific or “standard” position and a binocular “FlatTest” OB. The script can be run from any directory. Corrections were made on 1 Jan 2014 to wrap values of RA that, initially, are equivalent to the LST at twilight, but, after some adjustment within the script, may fall outside the range 0-360 deg. These corrections affect the program output when the “e” or “m” option is used and for times of the year when LST at twilight is close to 0 hr or 24 hr, i.e. December/January and July/August.

The “standard” position for twilight flats is at an hour angle, HA ~ -00:30, and declination = 27 degrees, close to the zenith. However, if this coincides with the Galactic plane, or if there is a preferred position, the scripts has the option to generate the OB with these coordinates. When possible, twilight flats should be performed at blank fields. A table of coordinates of blank sky fields for sky flats is available from the NOT website.

The script attempts to scale the Blue and Red exposure times to yield similar counts in both channels, but this may need to be adjusted. The command-line syntax allows flexibility in coordinates and exposure times.

The script will create a pair of OBs to take flat fields:

  • * the one starting with “SkyFlatTest…” will slew and obtain a strip to test counts.
  • * the one starting with “SkyFlat…” will slew and execute a 5-pt dither.

There are four options for input: The full syntax is:
>> mkskyflat RA_hh RA_mm RA_ss DEC_dd DEC_mm DEC_ss PA Filter_Blue Filter_Red ExpTime_Blue ExpTime_Red

However, if the “standard” flat field position and exposure time scaling are requested, the coordinates can be replaced with an “e” (evening “standard”) or “m” (morning “standard”) and no exposure times need be given. For example,

mkskyflat e 0 U Y

will create a pair of OBs at the evening “standard” position, 0 position angle, and using the U-BESSEL and Y-FAN filters with appropriate exposure time scaling to yield similar counts in both channels. The minimum exposure time is 1sec (can be decreased by changing the scale factor on the User Interface) and when the needed exposure times differ by greater than a factor 3, a warning is issued, asking whether the user would like to choose a different pair of filters (sometimes this may not be possible, e.g. for any combination which includes the narrow band F972N20 filter). The first letter of the desired filter is sufficient to specify the filter. for LBC-Blue:

S = SDT_Uspec
U = U-BESSEL
B = B-BESSEL
V = V-BESSEL
g = g-SLOAN
r = r-SLOAN

For LBC-Red:

V = V-BESSEL
R = R-BESSEL
I = I-BESSEL
r = r-SLOAN
i = i-SLOAN
z = z-SLOAN
Y = Y-FAN
F = F972N20
T = TiO_784
C = CN_817

The “standard” flat field position is: RA = LST at sunset/sunrise minus 15 minutes, and DEC = 27 degrees. mkskyflat1 uses the program ‘skycalc’ by John Thorstensen to determine the LST at sunset or sunrise on the current date. This syntax is to be used on the date (or within a few days) that the created flat field OBs will be run. The syntax is:

>> mkskyflat [e|m] PA Filter_Blue Filter_Red ExpTime_Blue ExpTime_Red

where a single letter indicates whether these OBs are to be used to take evening(“e”) or morning(“m”) flats.

Sky Flats: Exposure Time Scalings

Because the LBC user interface only permits one scale factor for both blue and red channels, it is important that the relative exposure times in the script are correct. When ExpTime_Blue and ExpTime_Red are not explicitly given to mkskyflat, the following exposure time scales will be adopted. These were determined from the set of relative count rates, s: (tB/tR = 10(sB – sR)), for which current best measurements are listed in the table below. (Note s = -1 * log10(counts/sec)). These scaling factors still need to be fine-tuned, but are provided here to show the defaults which are assumed in the script, mkskyflat, and to serve as a rough guide for calculating exposure times to be entered on the command line, if needed.

Relative twilight count rates (8-April-2011, used in mkskyflat1.pl, but to be checked)
LBC-Blue LBC-Red
filter S = -log10(cnts/sec) filter S = -log10(cnts/sec)
(S)DT_Uspec +0.55 (V)-BESSEL 0.0
(U)-BESSEL +0.60 (R)-BESSEL 0.0
(B)-BESSEL 0.0 (r)-SLOAN 0.0
(V)-BESSEL 0.0 (I)-BESSEL -0.1
(r)-SLOAN 0.0 (i)-SLOAN -0.1
(g)-SLOAN -0.35 (z)-SLOAN 0.0
(Y)-FAN +0.6
(F)970N20 +0.9

As the sky becomes darker, it will be necessary to increase exposure times. Both the Red and Blue channel exposure times can be scaled by a factor which can be entered on the OB Execution page of the LBC User Interface. If you wish to change the scaling factor while the OB is playing, click the pause button and wait until the it is paused (the play button will change from grey to black), type the new scaling factor, and then click play again to resume.

Sky Flats: When were the last flats taken?

Sometimes it is not possible to get all of the necessary twilight sky flats during a partner block. Sky flats taken a few months from the run should be fine to use, so long as no changes were made to the instrument that might affect the flat field. Specifically, cleaning (or new debris) on the filters, which are close to focus, would change the flat field, but cleaning of L1 or L2, which are far from focus, should not affect the flat field.

Since 2017B, we have maintained a table of good sky flats (i.e. flats with counts within a suitable range) here. Times when the filters were cleaned are recorded here. And on any of the obsN machines, the observer can list flats by camera, filter and month using the command, lbcsky.

Some examples of acceptable syntax (this is seen by typing lbcsky )

lbcsky camera filter -l list,
lbcsky camera filter -y YYYY for a year
lbcsky camera filter -y YYYYMM for a month
lbcsky camera filter -y YYYYMMD for any YYYYMMD*
lbcsky B -y YYYYMM any LBCB filter in YYYYMMD*
lbcsky V -y YYYYMM any LBCB or LBCR V-BESSEL flat in YYYYMMD*

camera = [B] or [R], and
filter = [S]DT_Uspec, [U]-BESSEL, [B]-BESSEL, [V]-BESSEL, [g]-SLOAN or [r]-SLOAN for LBC-Blue, and
filter = [V]-BESSEL, [R]-BESSEL, [I]-BESSEL, [r]-SLOAN, [i]-SLOAN, [z]-SLOAN, [Y]-FAN or [F]972N20 [T]iO_784 [C]N_817 for LBC-Red

If both camera and filter are indicated, the camera must precede the filter designation. Specifying only the filter or the camera, or neither, is allowed. When not indicated, the search will be over both cameras and/or over all filters.