ULTRACAM/HiperCAM/ULTRASPEC Throughput Models

ucam_thruput provides throughput models for the instruments HiperCAM, ULTRACAM and ULTRASPEC, and some tools for using these throughput models with the synphot and stsynphot modules.

Installation

The software is written as much as possible to make use of core Python components. The only third-party requirements are synphot, stsynphot, astropy, numpy and, optionally, graphviz

for making pretty graphical illustrations of the lightpath through the instruments.

Once you have installed these, download the module and install with the usual:

pip install .

or if you don't have root access:

pip install --user .

For more information, see:

• This packages' Github code repository

Usage

First of all, after installing synphot and stsynphot, make sure you also install some of the data files, in particular those in the HST data files section for bandpasses and the Atlases for calibration spectra such as the Vega reference spectrum. Set the PYSYN_CDBS environment variable to the location of these data files.

When you use the module for the first time, some data files must be installed. You can do this from inside Python:

You will only need to do this once.

Using the ucam_thruput models is just a matter of changing the tables used by stsynphot to create a bandpass from keywords.

import stsynphot as stsyn
from ucam_thruput import setref

setref('ntt')  # set the tables to use the ucam_thruput models for the NTT (sets telescope area)

You can switch back to using the built-in stsynphot models (including HST instruments, and Johnson/SDSS filters) at any time

stsyn.conf.reset()

Once stsynphot is setup to use the ucam_thruput models, we can define a BandPass using a string of specified instrument mode keywords:

# make a bandpass object using an obsmode string
bp = stsyn.band('uspec,tnt,g')

These bandpasses can be used with stsynphot in the usual way. See the stsynphot docs for full information.

A complete observing mode string specfies the telescope (gtc, tnt, wht, ntt or vlt), the instrument (ucam, uspec, hcam) and a filter. Additional keywords can be used that allow one to ignore the atmosphere (noatmos), insert the scintillation corrector (scint_corr), use the old NTT/ULTRACAM collimator (old) or use the NTT cube (cube). Not all combinations of keywords are valid. A full list of keywords, including all filters, can be found using the list_keywords function.

from ucam_thruput import list_keywords
list_keywords()

Models

Below are graphical representations of the instrument throughput models. The light paths are shown as a series of nodes, connected by edges (lines). Each edge represents the application of a transparency curve. If a line is labelled by a keyword, that path will only be taken if the keyword is present in the string used to define the BandPass.

Dashed lines represent "clear" transparency curves, that do not affect the throughput. Red lines represent reflections from dichroic surfaces. Unlabelled lines represent the default path.

Common - entry path followed by all instruments

ULTRACAM

HiperCAM

ULTRASPEC

Examples

A few example uses are shown below. These assume you've downloaded many of the stsynphot data files and installed them in a directory referenced by the environment variable `PYSYN_CDBS`.

The following example calculates the colour terms of USPEC/TNT g'-band.

import os
from matplotlib import pyplot as plt
import numpy as np
from ucam_thruput import setref
import stsynphot as stsyn
import synphot as syn

pickles_path = os.path.join(os.environ['PYSYN_CDBS'], 'grid', 'pickles', 'dat_uvk')
pickles_ms = (
    ('pickles_uk_1',    'O5V',     39810.7),
    ('pickles_uk_2',    'O9V',     35481.4),
    ('pickles_uk_3',    'B0V',     28183.8),
    ('pickles_uk_4',    'B1V',     22387.2),
    ('pickles_uk_5',    'B3V',     19054.6),
    ('pickles_uk_6',    'B5-7V',   14125.4),
    ('pickles_uk_7',    'B8V',     11749.0),
    ('pickles_uk_9',    'A0V',     9549.93),
    ('pickles_uk_10',   'A2V',     8912.51),
    ('pickles_uk_11',   'A3V',     8790.23),
    ('pickles_uk_12',   'A5V',     8491.80),
    ('pickles_uk_14',   'F0V',     7211.08),
    ('pickles_uk_15',   'F2V',     6776.42),
    ('pickles_uk_16',   'F5V',     6531.31),
    ('pickles_uk_20',   'F8V',     6039.48),
    ('pickles_uk_23',   'G0V',     5807.64),
    ('pickles_uk_26',   'G2V',     5636.38),
    ('pickles_uk_27',   'G5V',     5584.70),
    ('pickles_uk_30',   'G8V',     5333.35),
    ('pickles_uk_31',   'K0V',     5188.00),
    ('pickles_uk_33',   'K2V',     4886.52),
    ('pickles_uk_36',   'K5V',     4187.94),
    ('pickles_uk_37',   'K7V',     3999.45),
    ('pickles_uk_38',   'M0V',     3801.89),
    ('pickles_uk_40',   'M2V',     3548.13),
    ('pickles_uk_43',   'M4V',     3111.72),
    ('pickles_uk_44',   'M5V',     2951.21)
)

uspec_g = []
sdss_g = []
sdss_r = []
setref('tnt')
for name, spt, teff in pickles_ms:
    sp = syn.SourceSpectrum.from_file(os.path.join(pickles_path, name+'.fits'))

    bp = stsyn.band('uspec,tnt,g')
    obs = syn.Observation(sp, bp, force='taper')
    uspec_g.append(obs.effstim('abmag'))

stsyn.conf.reset()
for name, spt, teff in pickles_ms:
    sp = syn.SourceSpectrum.from_file(os.path.join(pickles_path, name+'.fits'))
    bp = stsyn.band('sdss,r')
    obs = syn.Observation(sp, bp, force='taper')
    sdss_r.append(obs.effstim('abmag'))

    bp = stsyn.band('sdss,g')
    obs = syn.Observation(sp, bp, force='taper')
    sdss_g.append(obs.effstim('abmag'))

uspec_g = np.array(uspec_g)
sdss_g = np.array(sdss_g)
sdss_r = np.array(sdss_r)
plt.plot(sdss_g - sdss_r, uspec_g - sdss_g, 'r.')
plt.xlabel("g'-r'")
plt.ylabel("uspec_g - g'")
plt.show()

---

Here is an example that plots the various contributions to a bandpass.

import stsynphot as stsyn
from ucam_thruput import setref
import os

setref('tnt')

bp = stsyn.band('uspec,tnt,g')
fig, ax = plt.subplots()
ax.plot(bp.waveset, bp(bp.waveset))
for comp in bp.obsmode.components:
    name = os.path.splitext(os.path.split(comp.throughput_name)[1])[0]
    if 'alum' in comp.throughput_name:
        continue
    thru = comp.throughput
    ax.plot(thru.waveset, thru(thru.waveset), ls='--', label=name)
plt.legend()