Nightly Observing Plan for MJD 57791 (starting evening of Feb 6, 2017)

Half night


  • Alex Drlica-Wagner (Obs 1)
  • Marisa March (Obs 2, RM)

CTIO Staff

  • Rodrigo Hernández - Operator
  • Manuel Hernández // Obs, Support
  • Staff calendar


mjd 57791, date 20170206, Site CTIO, UT-offset 3 hr
Current local time 14:40 CLST     UT 17:40     SidTim 22:06
Sunset     20:45 CLST  23:45 UT   SidTim   63 deg  
Eve. -10   21:22 CLST  00:22 UT   SidTim   72 deg  Zeropoint; standards
Eve. -14   21:43 CLST  00:43 UT   SidTim   77 deg  Begin obstac
Eve. -18   22:04 CLST  01:04 UT   SidTim   82 deg  Last chance standards
Midpoint   01:57 CLST  04:57 UT   SidTim  141 deg  
Moonset    03:45 CLST  06:45 UT   SidTim  168 deg  
Morn -18   05:50 CLST  08:50 UT   SidTim  199 deg  Begin astronomical twilight
Morn Std   06:11 CLST  09:11 UT   SidTim  204 deg  End obstac; begin standards
Morn -10   06:32 CLST  09:32 UT   SidTim  210 deg  
Sunrise    07:09 CLST  10:09 UT   SidTim  219 deg  
Moon position (deg): RA 88.0, Dec 19.3 Phase: 0.695

Weather Conditions

Observing Plan and Goals for the Night

* Evening Standards: Yzirg.

UT Date = '2017/02/07 00:25'     Local time = 21:25
E4-A                  ra 09:23:44  dec -45:21:02  secz 1.7
SDSSJ0200-0000        ra 02:00:00  dec +00:00:00  secz 1.6
C26202/HST            ra 03:32:30  dec -27:46:05  secz 1.0

* Obstac: Mostly wide field

%{color:blue}* Cosmos field around 1:20 CT (4:20 UT) %
  • if obstac i-band observations within 1 hour of cosmos time AND
  • if i-band deep if i<1.0" (cosmos_ideep_1.json)
  • else, more wide field
  • Have selected standards a few mins after -10. Takes a minute to do pointing.
  • This only changes the low airmass standard, but this is the last to be taken so will be taken close to '25 than '22. Better choice.

* Obstac: Mostly wide field

  • if obstac i-band observations within 1 hour of cosmos time AND
  • if i-band deep if i<1.0" (cosmos_ideep_1.json)
  • else, more wide field

* Swap with second-half observer at 01:50 CT (04:50 UT)
* Midpoint standard (using calspec) use only 1 (highest airmass): grizY (expect 05:00)

UT Date = '2017/02/07 05:03'     Local time = 02:03
SDSSJ1227-0000        ra 12:27:00  dec +00:00:00  secz 1.6

When were the last SN fields observed?

(See for database query.)

hex_name    seqtype    last_complete
SN-X3    long_z    2017-01-26 00:55:51.528351+00:00
SN-E1    short    2017-01-27 00:53:18.591071+00:00
SN-X3    long_i    2017-01-27 01:33:32.157988+00:00
SN-E2    short    2017-01-28 00:53:06.852121+00:00
SN-X1    short    2017-01-28 01:14:15.665466+00:00
SN-X3    long_r    2017-01-28 01:33:29.425048+00:00
SN-X3    long_g    2017-01-29 01:04:14.492806+00:00
SN-S2    short    2017-01-29 01:17:20.589302+00:00
SN-S1    short    2017-01-29 01:36:41.706064+00:00
SN-C2    short    2017-01-31 00:48:51.012551+00:00
SN-C3    long_g    2017-01-31 01:12:08.842422+00:00
SN-C3    long_i    2017-01-31 01:33:23.812696+00:00
SN-C1    short    2017-02-05 00:47:39.788401+00:00
SN-X2    short    2017-02-05 01:07:52.108646+00:00
SN-C3    long_r    2017-02-05 01:28:05.676930+00:00
SN-C3    long_z    2017-02-05 01:56:37.995348+00:00

Notes from 4pm meeting

Run Manager's Meeting Feb. 06, 2017
B.B., D.F., E.N., HTD, M.M., KADW, Manuel, SoS, Robert

Camera & Telescope
All OK. Flats in 20 minutes or so. People working on the dome.

Last Nights
Clouded out. Threats of snow.
Modest seeing, so a buncha SN C1, C3r, C3z, X2. NS at one altitude, SN at another.
WF and Cosmos short zY.

Issues with the gui. The gui seems to indicate the observers lose master
control over observing. The sol'n was to logout and then log back in.
This is different from just shutting down the gui and restarting.
The observers will record the sequence of events carefully for the next
few nights.

Got all the data. All processed. All WF is good.
C1 not good, X2g maybe not good.

Suggest to not accept the SN data quality so as to maintain obsevring ones
not recently done.

Cosmos short zY looks good.

Weather has blue skies overhead.

Sultan of Swami of Seeing says zY will gooing to have seeing about like last night,
maybe a little worse. More action in the lower atmosphere.

Moony tonight. Lots of zY. Maybe a little i-band. Moon is setting over the course of
the night.

If we seee obstac doing i-band within an hour, then we could select the i-band Cosmos.

Team effort to test the seeing prediction scaling with filter and airmass
equation that we are using. See
EN suggests removing the instrumental contribution before applying the correction.
Obstac assumes an ~ 0.6" instrumental contribution. SK notes instrumental correction
is a little different in g-band.
There are up to about 6% biases indicated by what they did do. More typical is 2%.
Info to be posted on "Contributed Scripts" page.
Wildlife report: uno zorro e uno condor.

WoD: G.O.A.T.

I've selected pairs of exposures taken less that 10 minutes apart and looked at the observed difference in PSF vs. the predicted difference in PSF. Here's a complicated plot showing the difference between the predicted psf using the equation:

PSF_2 = PSF_1 * (lambda2/lambda1)**(-0.2) * (secz2/secz1)**0.6


The interesting conclusion from this figure is that there appear to be different biases depending on whether you are going from (blue -> red) or (red -> blue). There are a number of causes, probably the most obvious being that human and machine observation tacticians are more likely to go from (red -> blue) when the PSF is improving.

If we try to remove this bias by combining (blue -> red) and (red -> blue) for the same filters, i.e., combining (g -> r) and (r -> g) into one data set, we find that the biases are reduced.


We can also play around with the prescribed wavelengths for each filter to further reduce the residual biases. After about 10 minutes of playing around, I find the following reduced residuals:


These use the the following wavelength solutions for each filter (I haven't solved for u-band):
LAMBDA = ('g',425.), ('r',630.), ('i',780.), ('z',930.), ('Y',1050.)
These are in comparison to the nominal wavelengths:
LAMBDA_ORIG = ('u',380.), ('g',480.), ('r',640.), ('i',780.), ('z',920.), ('Y',990.)

Of course, some MCMC guru could simultaneously fit all the wavelengths and probably come up with a better fit.