This repository contains scripts and files provided by D. Kirkby to compute the DEPTH for DESI observations
The computation is done in two steps.
First Step
Run python desi_exposures_gfa.py sv0 (or python desi_exposures_gfa.py minisv2) to read the GFA and exposure information to build a summary table with the following columns
'expid' 'night' 'tileid' 'exptime' 'mjdobs' 'tilera' 'tiledec' 'airmass_min' 'airmass_med' 'airmass_max' 'moon_sep_deg_min' 'moon_sep_deg_med' 'moon_sep_deg_max' 'transparency_min' 'transparency_med' 'transparency_max' 'fwhm_asec_min' 'fwhm_asec_med' 'fwhm_asec_max' 'sky_mag_ab_min' 'sky_mag_ab_med' 'sky_mag_ab_max' 'fiber_fracflux_min' 'fiber_fracflux_med' 'fiber_fracflux_max' 'ngfa' 'ebv'
The summary is saved to desi_sv0_exposures_gfa.fits (or desi_minisv2_exposures_gfa.fits).
Second step
Checkout the spectrograph throughputs (svn co https://desi.lbl.gov/svn/code/desimodel/tags/0.13.0/data/throughput thru13) and run python depth_calculation.py sv0 (or python depth_calculation.py minisv2).
This script takes the output generated in the first step together with the spectrograph throughputs and two different sky models (dark_desimodel.fits and dark_eso.fits) to estimate the DEPTH for each band. This is done by comparing the observed sky levels (EXPSKY is the mean sky in electrons detected with 100A smoothing ) against the expected model (IDSKY is the fiducial sky "dark zenith" with 100A smoothing ):
DEPTH = EXPTIME x (TRANSP/1.0)^2 x (FRACFLUX/0.56)^2 x (FIDSKY/EXPSKY)
where the TRANSP and FRACFLUX values come from the results of the first script. Finally, the DEPTH value is added as a new columns and the result is written to
desi_sv0_exposures_gfa_with_depth.fits (or desi_minisv2_exposures_gfa_with_depth.fits)
This is how the final table looks like:
