New Tools for AGN Photometry with Kepler Guest Observer Python Package Lightkurve

Transcript

1. www.nasa.gov National Aeronautics and Space Administration New Tools for AGN Photometry with Kepler Guest Observer Python Package lightkurve Anastasios Tzanidakis1, Christina Hedges1, Geert Barentsen1, Michael Gully-Santiago1, Jessie Dotson1, Sheila Sagear1 Contact KeplerGO, AndyTza at3156@columbia.edu @Andy_Tzanidakis Figure 1 — Target Pixel File of AGN With its Corresponding Light Curve KIC 7680833. A raw Target Pixel File (TPF) of a AGN found in the Kepler Catalog. The left diagram shows the light curve of the AGN, while on the right side the TPF is shown. Figure 2 — Example of PSF Contamination in Target Pixel File of Active Galactic Nucleus KIC 10402746 Target Pixel File in a Kepler field of an Active Galactic Nucleus (AGN). The AGN target is marked by the blue cross. On the left of the diagram, in Pixel Column Number 219, another detected source on the edge of the TPF (shown with the red star marker) is “contaminating” the AGN PSF. Zero-Point Calibration for Stellar Priors New lightkurve Tools Cataloging Sources in TPFs Future Work When analyzing light curves, there are generally two methods for estimating the flux as a function of time: Simple Aperture Photometry (SAP) and Point Spread Function Photometry (PSF). SAP applies an aperture mask on the center of the coordinate and sums the total flux in each cadence, with the chance of overestimating the flux in crowded fields. PSF photometry on the other hand, estimates the PSF of a target with 2 positional vectors and amplitude. Such technique is useful in crowded fields, where we can separate each stars PSF given its position and flux. To generate more accurate AGN light curves, the user must be aware of the sources surrounding the Target Pixel File (TPF) in order to understand what sources are present near each TPF. As seen in figure 2, surrounding objects are able to ‘contaminate’ the TPF causing an unnecessary increase of flux when performing photometry. We have devised a new functionality in lightkurve where the user is now able to generate all surrounding sources in the TPF, with either KIC, EPIC or Gaia DR2 catalogs. Altogether, this will allow users to identify most contributing sources that are present within each TPF. Active Galactic Nuclei in Kepler Active Galactic Nuclei (AGN) are some of the most energetic galaxies in the universe. Most of the energy emitted by an AGN is caused by the accretion of gas falling toward its central black hole. Due to dynamic friction, the accreted gas is heated and is emitted in the form of X-ray radiation where most AGN are discovered. More recently, researchers have been interested in understanding AGN light curves in the optical part of the spectrum. In particular, the variability in flux is a compelling mechanism that researchers are still trying to explain. 
 Kepler provides an ideal frame for performing precise photometry on extra-galactic objects such as AGN where we can understand more about its optical variability. Figure 3 — Another TPF With a Crowded Field This following example demonstrates a crowded TPF field. The red scatter are the identified sources within the TPF using our new tool. In such cases, it is crucial to understand the contribution of flux from each source. KIC In order to improve our PSF scene model, we require priors on all surrounding sources. One common difficulty is to derive the flux of a star given only its magnitude. We thus apply a Bayesian Analysis on parametrizing the Zero-Point (i.e. the conversion factor between flux and magnitude). Gaia DR2 Figure 3 — Corner Plot of Zero-Point Calibration for 631 Light Curves from the KIC Catalog and Gaia DR2. The bottom left figure shows the probability density function of the inferred parameters in each corner plot. For each catalog, Gaia DR2 and KIC, we estimate the distribution of each parameter using a Monte Carlo Markov Chain. Figure 4 — KIC 10402746 Example of AGN Scene Modeling TPF example of AGN with a generative scene model using PSF. Using prior information from the figure on the right, we can successfully build a PSF scene model of that AGN. We expect with our new tool to accurately model the other surrounding sources in the TPF. 1Kepler Guest Observer Office, NASA Ames Research Center, Moffett Blvd, Mountain View, CA 94035 Finally, once we have parametrized the Zero-Point factor, we are able to convert the magnitude of the given stars into units of flux: To successfully build PSF scene models for a given source, we estimate its prior information (i.e. position and flux) where the user can generate an accurate PSF model based on the TPF. Based on the prior of each object near the TPF field, we can accurately build PSF for all contributing sources.