price added a comment - 26/Apr/19 5:01 PM hassan : We currently don't have a sky spectrum that distinguishes between different species — we just have a sky spectrum. Please advise whether you want it applied before or after the atmospheric extinction.

hassan added a comment - 26/Apr/19 6:06 PM - edited Let's make the application of atmospheric extinction to sky spectra configurable. price and hassan agreed that by default no extinction to sky spectra would be applied. This is based on the assumption that that the observed sky spectrum already has extinction applied.

price added a comment - 26/Apr/19 6:23 PM Using the new atmospheric model grids to generate an extinction curve as a function of zenith distance, aerosol power-law index and PWV. cloomis will be happy that I got rid of the flood of warnings from astropy. price@MacBook:~/pfs/drp_instmodel (tickets/SIM2D-114 %=) $ git sub commit a548253de2ba62e5389750004dc39a8d7ff4ee5b Author: Paul Price <price@astro.princeton.edu> Date: Thu Apr 25 17:09:52 2019 -0400 Spectrum: add __mul__ Allows multiplying spectra together. This will be used when we multiply a spectrum by the transmission. python/pfs/instmodel/spectrum.py | 86 +++++++++++++++++++++++++++++++++++++--- 1 file changed, 81 insertions(+), 5 deletions(-) commit 16e848474d1f95fd6d1e783c75dbd78b8a238ba2 Author: Paul Price <price@astro.princeton.edu> Date: Thu Apr 25 17:19:32 2019 -0400 SkyModel: refactor and add extinction Using Spectrum rather than splines (a TableSpectrum is a kind of a spline, it's just a different representation). I want to be careful with the resolution: can't just multiply tables/splines at different resolution. This includes a new Arm enum class for standardising how the arm is represented (band vs arm). Extinction data come from the data recently added to drp_instdata, provided by Nicholas Mondrik. Note that we now use the second-to-last column of the instrument throughput data, which contains the instrument only, without the atmosphere. python/pfs/instmodel/arm.py | 34 ++++++ python/pfs/instmodel/detector.py | 18 +-- python/pfs/instmodel/exposure.py | 6 +- python/pfs/instmodel/makeSim.py | 14 ++- python/pfs/instmodel/simImage.py | 15 +-- python/pfs/instmodel/sky.py | 219 ++++++++++++++++++++----------------- python/pfs/instmodel/splinedPsf.py | 5 +- 7 files changed, 187 insertions(+), 124 deletions(-) commit 08161df473adbf9526b7f30d3cd7f5fe9bd38444 Author: Paul Price <price@astro.princeton.edu> Date: Thu Apr 25 17:21:24 2019 -0400 simplify the command-line interface Introduces "--type" which replaces --imagetyp " and " --domeClosed". Does a bit more checking on the lamps for the various image types. bin/makeSimRun | 68 ++++++++++++++++++------------------- python/pfs/instmodel/lightSource.py | 5 +-- python/pfs/instmodel/makeSim.py | 28 ++++++++++----- 3 files changed, 57 insertions(+), 44 deletions(-) commit a140af04a91a868b6f0ce91bae32b4f2bef62185 Author: Paul Price <price@astro.princeton.edu> Date: Fri Apr 26 13:30:39 2019 -0400 Remove unnecessary 'examples' symlink examples | 1 - 1 file changed, 1 deletion(-) commit e5ef8f89e0d623a3d282f750c28039657f70332f Author: Paul Price <price@astro.princeton.edu> Date: Fri Apr 26 13:32:25 2019 -0400 suppress warnings from pyfits Floods of warnings like: WARNING: VerifyWarning: Keyword name 'W_AIT_SRC_QTH' is greater than 8 characters or contains characters not allowed by the FITS standard; a HIERARCH card will be created. [astropy.io.fits.card] come from applying the bias header to the output; ignore them. Also fix the warning about 'clobber' being deprecated by using the 'overwrite' kwarg. python/pfs/instmodel/exposure.py | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) commit 713a5d5015c25785636d7fc27c33c860f85dcbec (HEAD -> tickets/SIM2D-114, origin/tickets/SIM2D-114) Author: Paul Price <price@astro.princeton.edu> Date: Fri Apr 26 13:45:37 2019 -0400 add option to apply extinction to the sky spectrum I'm assuming that the sky spectrum is intended to be used as provided, and therefore the default is to not apply extinction. But if the spectrum is a model for above the atmosphere, it should be extincted. python/pfs/instmodel/makeSim.py | 6 ++++-- python/pfs/instmodel/sky.py | 8 ++++++-- 2 files changed, 10 insertions(+), 4 deletions(-)

rhl added a comment - 26/Apr/19 11:15 PM Was modtrans run with or without aerosol absorption?  It is quite likely that it was not as the model is naïve; in that case we need to include an explicit two-parameter model (optical depth and power law)

price added a comment - 26/Apr/19 11:53 PM The grid includes aerosols with multiple power-law indices (see SIM2D-106 ): Aerosol: tau=0.0075; a=0, 1, 2.5 (arXiv:1210.2619 Fig 13, Fig 19)

rhl added a comment - 27/Apr/19 2:43 AM But only one tau?  My intention had been to run modtrans without the aerosols as they are easy to add later, but if we include them in the telluric models I think we need a larger range of opacities.  We could probably back them out, but Nick might be willing to regenerate.

rhl added a comment - 27/Apr/19 2:45 AM Can hassan edit the comment on sky spectrum absorption to explain why you decided not to absorb the OH emission spectrum?  It's not crazy (assuming we're using observed sky emission spectra), but I'd like to understand the logic

price added a comment - 27/Apr/19 2:49 AM Apparently the distribution of the aerosol tau is not very wide (median=0.0075, nMAD=0.0014), so I didn't think it was worth doing multiple values. See arXiv:1210.2619 , figures 13 and 19.