Observing Plan for the November 15-16 2013 Engineering Nights

Proposal Id: 2013A-9999
Exposure Scripts: Most of the exposure script for this night are available in ~DECamObserver/ExposureScripts/engineering_20131115
(from the Observer Console GUI select Exposure Control --> Load Exposure Script --> Choose File)

(Late) Afternoon

Full set of zeros and dome flats
Script: prenight_calibration.json (note that this is a (standard) DES script. From the the Observer Console GUI select Exposure Control --> Load Exposure Script --> Select DECam Script then choose DES/prenight_calibration.json in the dialog window)

Science projects

We have two science time requests. N.B. Engineering tasks have priority for Friday and Saturday nights. But it is assumed that there is time to fit the science in. If not, then as per instructions from Nicole, the two projects are to be fitted in, along with DES observations, on the Sunday DD night.

Friday after twilight and at around 1 am.

David James will need about 2 x 45 min for propid 2013B-0531 (David will be observing that night). David - these have been written in for Friday but you can re-schedule if that seems sensible. The conflict with early on Saturday is with Rolo and the TCS tests.

Saturday at the end of the night

Chris Smith will need about 40 minutes for propid 2013B-0421. Chris will provide instructions (or be online/on the mountain himself). High priority Friday night late is the Linearity study below.

Telescope Pointing Verification - high priority

With the hexapod locked at the zenith values with only the focus updating (AOS-1 mode).

About 3 hours. The proposal is to do this Saturday night, after twilight.

TCS team would like to take data to study the pointing model produced out of a pointing map done with AOS off. We are planing to use the active system (AOS) off and the hexapod values fixed to the values from zenith. The new pointing model will be installed before the tests and then replaced back by the one in use after the test.

From Rolo:

_ Move telescope to zenith
Switch to AOS-1 mode
On the Architect Console select the AOS role (on the left) and then enter the command init_pid 1 in the command field (upper right)
On the ICS Hexapod GUI (see screen shot below) click on the Trim button and in the pop-up window enter 0 in all fields except for focus (3rd field from the left)
Take a few exposures (30 seconds each in r filter) to allow the AOS to adjust the focus.
On the ICS Hexapod GUI uncheck the Telescope (Lookup) box. This will keep the hexapod fixed at the current values (except for focus).
Slew telescope by a few degrees, take an exposure and verify that the hexapod values do not change. (except for focus)
Run the PointingMap.json script from the engineering_20131116 directory.
Verify that the trim values stay at 0 (except for focus)
Verify that the hexapod position doesn't change (except for focus) - if this does change stop and call Klaus or Ann
On observer2 go to the image directory (/data_local/images/fits/2013A_9999) and for each exposure run the check_offset scripts.
check_offsets -x <exposure number>
For example:
[DECamObserver@observer2 2013A-9999]$ check_offsets -x 228827
Expid 228827 tim: ['-6.9', '-19.9'] center: ['-5.2', '-19.8'] Hexapod: [1000.941, 2447.446, 2278.99, -136.057, -95.603, -0.0]
this runs the kentools tim command, center command and extracts the hexapod values [Note that the kentools results are presented as error, not offsets (there is sign difference to the normal center command)]
record these values in a text file_

Plus - Rolo emailed David a text file of positions.

Linearity Study - high priority

We will take exposures of a rich star field in g band with a range of exposure times, so we can test that the linearity corrections derived from dome flats are properly correcting the sky data. Instructions are:

  • Move to the desired field (see coordinates below) and put the g filter in the beam. AOS and guider turned on.
  • Start script ??? to take series of 45 exposures of different exposure time at same pointing. Should take 40 minutes.

For the linearity tests, the darkest possible skies are advisable. It looks like the best available opportunity will be the very end of the first night. We will have a 40-minute program which we may want perhaps to start 20 minutes before astronomical twilight when the Moon is about 12 degrees above the horizon, then continue as the moon sets and twilight sets in.

A good target position for this time of night:
RA = 07:30 DEC = -50

If this time is unavailable or not photometric, then the start of the final night is the preferred time.
This test requires photometric conditions. Don't bother if clouds are visible in the sky.

Starflats - high priority

We would like to repeat the star flats measurements in the grizY filter bands.

Instructions from Gary:

Low Galactic latitudes are difficult to access early on the night so we will be aiming for the 2nd half of the night.

Here are two possible starting positions for the star flat sequence:

   RA=06:40:00  DEC=-34:00:00 (available after about 04:30 UT)
   RA=07:30:00  DEC=-50:00:00 (preferred but not availabel until 05:30 UT)

There are 22 exposures in the sequence, so 5 x 22 x (30+30)s = 110 minutes to do this whole thing.

The pointing is not critical (1 arcmin accuracy is fine), but you will need to move the telescope back
to the starting position when beginning each filter.

If time is short you can omit one or more of the filters. The star flats should be done in clear weather, although seeing as poor as 1.5-2" will give data of at least some value. Better seeing is better...

The 5 scripts can be found in the engineering_20131115 folder. They are named Starflat<filter>.json_

AOS/Donuts tests - high priority - this is pointing towards Orion or thereabouts at ~ 3 am I think, for an hour.

Test AOS script to take nightly out of focus image using new SISPI features.

AOS Test (about 1 hour). Requires seeing <1.2" and galactic plane visible (or other dense star region)

  • Move the telescope to a location with high stellar density (ideally galactic latitude around 15 degrees). Telops can usually assist with this. We will be tracking the same location on the sky for around an hour. Otherwise contact Kevin Reil ().
  • Run the script aos_focus_align.json - this will take 5 * 30 second 'r' band exposures to get the telescope in focus and aligned. (10 minutes)
  • Run the script aos_on_grizy.json - this will take 2 * 30 second exposures in focus for 5 (grizy) bands with the AOS on (normal operations). (20 minutes)
  • Run the script aos_defocus_grizy.json - this will take 2 * 30 second exposures out of focus (+1500 um) for 5 (grizy) bands with the AOS off. Camera will be refocused (to prior setting) at end of script. (20 minutes)

If at anytime the above plan is not working as expected, focus can be restored by clicking the "Init. z trim" button on the ICS/Hexapod control and running the aos_focus_align script.

Guider CCD astrometry - this requires special afternoon calibrations, suggest leave this for Saturday.

Use a special readout mode to save long-exposure data from the guider CCDs so that we can determine their astrometric solutions relative to the science array to better than 1". There will be 2 components needed for this test:


During the daytime, take zeros and dome flats in each filter using the normal dome-flat exposure times, but with SISPI instructed to save the guider readouts.
The guider calibration script is called guider_zeros_flats.json and can be found in the engineering20131115 directory.
This script does not have the standard action commands to turn of vsub and such. Nor does it include the vsub flush exposures.
Before you run it you have to tell the GCS to use a different directory so that we can keep these images.
On the architect console select the GCS. Then, in the command box enter:

set image_directory /data_local/images/fits

Once this script is done you need to switch back using the command (again with the GCS selected)
set image_directory /dev/shm/guider

Also when you configure (or restart) the image directory is switched back. If the timing works out Klaus would like to be online for this to make sure it works.

Guider Exposures

During the nighttime, move to any one of the three coordinates given for the linearity test. Then take 3x20 s exposures in each filter, moving 1 arcmin N of original pointing for 2nd exposure and 1 arcmin E for 3rd exposure in each filter. The guider_astrometry.json script in the engineering20131115 directory is set up to do this.
Like in the previous test you need to switch the image directory so that the images are kept. Using the architect console select the GCS and enter:

set image_directory /data_local/images/fits

Once this script is done you need to switch back using the command (again with the GCS selected)
set image_directory /dev/shm/guider

Calibrate Stars around BD+17 4708 (before 01:30UT): Lower priority for this observing run.

  • For today's date, the ideal time is around 0:00UT, when BD+17 is crossing the meridian and is at its lowest airmass (X~1.5)
  • Conditions must be clear; otherwise, skip this task
  • Observe a set of 3 standards fields (low-, intermediate- and high-airmass), make sure they include u-band standards (these can just be the usual evening twilight standard fields, but with all 6 ugrizY DECam filters).
  • Then observe the attached json script (attachment:BD17_4708.json) to calibrate stars around BD+17 4708
  • Then observe a second set of 3 standard star fields (again include u-band)
  • The same task was performed on the night of 20 September 2013; for reference, here is that night's nightsum

Deep Y band observations, fringe tests - This is considered lower priority for this engineering run

(from Peter Nugent and Paul Martini)


We'd like 60 Y-band exposures (each 90sec) to measure the amplitude of the fringe pattern. 10 dithered images per deep SN field and the corresponding 2 adjacent shallow sub-fields would be ideal, as that may be useful for calibration and photo-z work too. This should take a total of 2-3 hours.

  • Point at an SN field C1, C2, C3 and X1, X2 and X3. (coordinates are on this page: SN-C3 and SN-X3 are top priority
  • All of these fields are up in the second half of the night, they do not have to be directly overhead, secz < 2.0 is fine
  • Run the above json scripts (linked to field names) to take 10 exposures for each one, these are in a 3-4-3 pattern of equilateral triangles separated by 30" on a side. Please check these json files, it's my first attempt at them.
  • Repeat for the next field

Regression Script - Low priority. There won't be time for this.

Regression_Observations_All.json - the standard regression script - about 5- 6 hours