#Optimal Aperture Calculation ####Parallelized Analysis of the Spitzer Telescope Dataset Team: Hannah Diamond-Lowe, Zakir Gowani

The Spitzer Space Telescope has recently been used to observe exoplanets. Datasets from Spitzer provide 32x32 images of stars over a time period comparable to the orbit of an exoplanet around the star. The time-dependence of brightness describes various qualities of the exoplanet. However, light noise is a problem. Current methods rely on manual aperture adjustment to throw out unwanted background noise in the data. Our project is to automate this process and extend the accuracy by accounting for a finer degree of time variation.

---

Our immediate purpose is to find intelligent, time-dependent aperture sizes for stellar observations using Spitzer, which reduce noise but conserve valuable star-and-planet system information. The ultimate goal is to to produce the optimal aperture size for a considerable number of stars (each star has 4x ~5 GB worth of images, when accounting for the four detectors on Spitzer). As a check, we can compare models existing in the literature to a model output based on our aperture calculations.

#####Instructions for Prototype Use The primary script for aperture calculation can be run using

python aperture_calculator.py

The goal of this prototype script is to demonstrate our aperture calculation algorithm on a single image. This algorithm calculates signal-to-noise ratios for a set of aperture radii, incurring an exponential penalty with each increase in aperture size. The weighting function is an which imitates a Gaussian point spread distribution. The current data set is contained in a single file located in the folder titled prototype_data, titled "SPITZER_I1_41629440_0000_0000_1_bcd.fits". The Astropy library is used to convert the fits image into a numpy array of flux values. We then process this table using an assortment of disk and annulus masking functions, all of which rely on the Bresenham circle algorithm to draw circles on grids.

Technologies used include:

• Astropy, a popular astronomy library for Python which deals with .fits files, the standard data format for our Spitzer star observation data.
• SAOImage DS9, an Astronomical Data Visualization Application. Allows for direct viewing of the individual frames in .fits files
• Python Imaging Library for image file I/O and resizing
• MPI for Python, a Python library for message passing and parallelization.

#####Timeline (2014):

• By May 6: Prototype will demonstrate aperture calculation algorithm on a single frame for a single star
• By May 18: Expand algorithm's operation to consider time variation over frames
• By May 25: Revise algorithm to be parallelizable in an MPI framework
• By June 3: Final implementation of project which compares models, accompanied by visualizations and a writeup