AIfA telescope data reduction tutorial#

Foreword#

The aim of this tutorial is to show you a way to “reduce” and analyse imaging data acquired with the 35-cm optical AIfA telescope, that is to turn the raw data into a scientific result. We’ll do so using professional Python packages that are easy to install. Note that while the present tutorial is fine-tuned for the AIfA telescope, it should be well possible to follow it when dealing with data from other instruments, with only minor adaptations.

Example results obtained with the reduction described in this tutorial

Image reduction can often be broadly split into two separate parts, which may comprise some of these steps:

  1. Pre-reduction of the images

  • Remove instrumental signatures using calibration frames

  • Mask (or flag) defects and missing data

  • Characterize the noise

  • Calibrate the images astrometrically and photometrically (and in wavelength, in case of spectra)

  • Model the sky background

  • Produce a stack, aka coaddition (sometimes), or even a mosaic

  1. Extracting the information of interest

  • Object detection and photometry

  • Photometric calibration against a reference catalogue

  • Creation of a light curve

  • Production of a color image

  • etc …

The present tutorial goes even a bit further than the reduction itself, as it also covers the analysis of the data (e.g., the comparison between an observed color-magnitude diagram and isochrones).

Note

The name “reduction” is somewhat misleading. It won’t give you back disk space, on the contrary!

Many approaches and software solutions for data reduction exist. The focus of this particular tutorial is on:

  • keeping it short and simple, suited for a lab course,

  • preferring well documented and easily installable tools, so that the tutorial can be used on any platform,

  • getting you to code a little bit, to interact with every step (versus a black-box automated pipeline),

  • providing background information on the key concepts. Computational efficiency is less important.

Overview of what we will use#

The whole data reduction and analysis is done in Python (except for the astrometric calibration, which is based on a local installation of astrometry.net that we’ll call from within python). We make use of the following professional Python packages, all under active development:

  • ccdproc (framework for the image pre-reduction)

  • photutils (for source detection and photometry)

  • astropy (basis of the above, but also for manipulating catalogs)

  • matplotlib (for visualization)

We no longer use SourceExtractor / SCAMP / SWarp or Theli, to make the installation as easy as possible (among other reasons). See References and links if you are interested in these alternative solutions.