Science Verification Jobs¶
Here is a list of science verification projects - some small, some large - that are looking for a good home. If you are interested in any of these please contact Gary or Klaus. Getting involved with one of these projects is a great way to learn about DECam - for students, post docs and senior members of the collaboration. None of these jobs require specialized astronomical knowledge.
The numbering of these jobs is arbitrary. Some already have people working on them. All of them would profit from more applied brainpower!
21. Examine variation of z and Y sky levels¶
z and Y sky levels will vary substantially with auroral activity, not just with lunar circumstances. Substantial gains in efficiency are to be had if z/Y can be taken in darker circumstances. Study the temporal behavior of z/Y sky brightness and develop strategy to predict and exploit the times of fainter airglow.
20. Collect Cerro Pachon DIMM data for comparison to DECam seeing.¶
19. Use donuts to detect occultation of the beam by the dome.¶
18. Look for correlation of stellar photometric residuals with the tree rings.¶
17. Evaluate the effects of environmental conditions on image quality.¶
We record wind speed, wind direction, humidity and a number of temperatures. We have already seen that wind affects ellipticities and the image quality. Quantifying this can lead to improved operational procedures, e.g. when to raise the wind blinds.
16. Determine science-driven thresholds for bad pixels and automate the bad-pixel-list maintenance.¶
Lists of hot & cold pixels and Funky columns are seen to fluctuate over time as defects wander above/below thresholds. Determine what thresholds should be used to make sure bad pixels are not significantly degrading the science data (maybe background-level-dependent bad pixel lists!) and get a regular regimen of BP tests into operation and implemented in DESDM.
15. Measure and model the CCD nonlinearities¶
DECam exhibits nonlinearities between ADU output vs number of photons input. Use dome flat exposures and sky data to characterize the nonlinearities of all amplifiers, and install appropriate corrections in DESDM pipeline to re-linearize. The ultimate accuracy of photometry and WL shapes will depend critically on this.
14. Measure and model the dependence of PSF size on stellar flux.¶
DECam and other CCD cameras are seen to have the PSF broaden slightly (~1/2%) for brighter stars. Presumably due to the alteration of electric fields within the CCD by the accumulated charge in the pixel well. We need to have a model of this effect to get the best accuracy in our photometric and WL measurements.
13. Calibrate the FWHM vs half-light radius relation¶
Image Health reports the robustly measured "r50" = radius containing 50% of PSF light. Most observers are accustomed to evaluating seeing by the more slippery quantity of FWHM.
Use First Cut outputs of FWHM measured by SExtractor to improve the r50-to-FWHM relation, so that the real-time estimates of FWHM are better.
12. Determine the zeropoint variation on photometric nights and the best photometric-standards practices¶
Compare the results for zeropoints/extinction derived from standards to those from ubercal (see #10 below) and decide how quickly the photometric behavior is varying, and whether there is added value from standards taken in the middle of the night.
11. Calibrate extinction vs RASICAM data¶
Compare the results for zeropoints/extinction derived from standards and ubercal (see #10 below) to the cloud structure reported by RASICAM, so RASICAM can be more reliably interpreted as an on-the-spot measure of extinction.
10. Ubercalibration and Star Flat¶
Determine the difference between dome flats and true stellar response ("star flat") use relative response in overlapping exposures to determine precise zeropoints for SV exposures.
9. Complete the crosstalk corrections¶
Removal of crosstalk between amplifiers is the first step of DECam reduction. Complete Kerstin Paech's characterization of crosstalk and make sure the corrections are implemented in DESDM pipeline.
8. Use galaxy counts to detect extinction from clouds¶
The number of galaxies detected by SExtractor in QR and in First Cut is determined by the filter, seeing, sky background, and transparency of the atmosphere. The first three are known / measured independently. Use the galaxy counts to infer the transparency: this information is then available to decide which exposures need to be re-taken (i.e. close the Survey Table loop).
7. Measure the pointing performance of the telescope¶
From the locations of stars on the Focus/Alignment images, which are being stored for every off-season exposure, determine the difference between the true pointing of the telescope and the requested one. Statistics of these pointing errors are needed to optimize the supernova observing program. The study will likely lead to improved Blanco pointing performance for all users.
6. Use the off-season image-quality monitoring system to improve DECam image quality.¶We are collecting IQ data from Image Health, the Guider, the telemetry database and the telescope database for all exposures, even during DES's off-season. Looking for correlations and patterns in these collected data will help us determine the origin of
- Persistent non-zero whiskers indicating some asymmetry in the optics
- Substantial gap between g/r band FWHM and izY FWHM
- Discrepancy between optical (donut) and mechanical (BCAM) measurements of telescope collimation.
Correlation of the IQ measurements with telescope positions, environmental parameters such as wind speed and information from the donut system will help localize the causes of these less-than-ideal behaviors, and help establish best practices for IQ.
5. Guider "DIMM"¶
Use correlations between the 4 different DECam guider CCDs to estimate the site seeing and determine whether high-altitude turbulence or local effects (dome seeing, wind shake) are contributing the most to image size. Until the CTIO DIMM instrument is back in service we need to find alternate ways to estimate site seeing.
4. Temperature Predictor¶
To achieve best possible image quality careful thermal management in the dome, near the primary mirror etc is very important. Using the data recorded in the telemetry database we need to develop a nighttime temperature predictor to optimize thermal management.
3. DECam Regression Test Suite¶
Complete the design of a regression test suite that will be run periodically to measure the instrumental signatures of the DECam and monitor for changes or performance degradation. Then make sure that software & protocol are in place to get these taken and examined. As we learn more about instrumental signature of DECam and ways to address them we need to devise a DECam regression test suite of standard images and reduction software tools to look for changes in all instrumental signatures.
2. Extended characterization of fringing¶
Continue the fringing study started during SV (P. Martini) and check more images for changes of fringe pattern. Eyeball examination of DES-DM imcorrect outputs to see that fringing is being eliminated properly.
1. Analysis of Residual Telescope Vibrations at the Time the Shutter Opens¶
We need somebody to take a close look at engineering-test images and analyzes the First Cut output to confirm that there are no ill effects in the images from residual telescope motion (settling) after the shutter opens. Every second sooner that we can open the shutter adds ~5 nights to the DES survey!!!
Details: After the telescope slewed to a new position the TCS kernel measures the difference between target position and the encoder readings from the mount and waits until the RMS variation of the difference is below a threshold for a certain amount of time. Only then SISPI is told that the telescope is in position and the shutter can be opened for the next exposure.
In order to maximize survey efficiency we would like to keep this wait time at a minimum. During the March engineering run we took a series of exposures where we varied the delay.
From the electronic Logbook (#67556):
Efficiency tests all at 60 deg S 10s z band seeing 0,6-0.8 arcsec 1. 306-311 delay=5s 2 deg in dec 2. 312-317 ditto +1 deg in RA oops 3. 318-323 ditto +2 deg in RA 4. 324-329 ditto -2 deg in RA now change delay to 3 seconds 5. 331-336 2 deg in dec 6 337-342 +2 deg RA 7 343-348 -2 deg RA now change delay to 1 second 9 349-354 2 deg dec 10 355-360 +2 deg RA 11 361-366 -2 deg RA now change delay to 0 seconds 12 367-372 2 deg dec 13 373-378 +2 deg RA 14 379-384 -2 deg RA now, delay =0 and BCams turned off I hope. I ticked the box in the OCD Exposure control hMMM i EXPECTED 26S AND GOT 28S??? 15 385-390 2 deg in dec 16 391-396 +2 DEG ra 17 397-402 -2 DEG ra
Kevin Reil analyzed information from the telescope database and found that one can sometimes see that the telescope is still moving just prior to shutter open but this motion is very close to 0, always less than our requirement of <0.2". Details can be found in the logbook (#67645).
A nightly summary of TCS information is produced (via cron) and the SISPI telemetry database is filled with values. For this study the variables ha_mean_f0 and dec_mean_f0 are of interest. Hour angle (ha) and declination (dec) errors are averaged (mean) from -0.5 to 0.5 seconds for the integer second (note rounding error) of when the shutter opened. This plot[[http://goo.gl/V82Jj]], for example, shows the |mean declination error| (dec_mean_f0) near shutter open for the past week. Here, [[http://goo.gl/ONRKl]] is |dec_mean_f1| which is rounded (int) shutter open second averaged from 0.5 to 1.5 seconds. Details at [[https://cdcvs.fnal.gov/redmine/projects/des-sci-verification/wiki/_TCS_Studies_]]