Project Summary
The Chandra X-ray observatory has provided an invaluable amount of data for astronomers working in the X-ray regime. In particular, Chandra has helped shed light on the properties and behavior of the elliptical galaxy Messier 87. However, the amount of time required to thoroughly review all the data and reach conclusions is far beyond what can be done via manual methods. Our intended solution to this problem is to create an automated data analysis pipeline using tools provided by the Chandra Interactive Analysis of Observations (CIAO).
Project Objective
The main goal of this project was to create a data analysis pipeline that identifies spectra to search for supernovae. In addition, we would like to publish the code on GitHub, which may aid future research on transient X-ray events.
Manufacturing Design Methods
Five data analysis modules:
1) User Input: Handles all user interaction
2) Source Detection: Matches a list of M87 sources to the input data
3) Variability: Uses an algorithm to search for variable sources, gives sources a variability “score” (0 to 10)
4) Spectral Analysis: Extracts the photon count at different wavelengths
5) Visualization: Plots and visualizes all data from each module
Analysis
After running our data through the pipeline, we were left with 178 candidate sources from 27 observations. This list was further refined by removing false positives, such as any sources with large errors or low photon counts.
From here, we arrive at a final list of 42 variable sources across 17 observations. Of these sources, 18 were transient, 6 were periodic, and the last 18 appeared to be longer-term sources that could not be identified.
We found many false positives because sources in M87 are fairly dim due to their distance from our galaxy. This low luminosity can cause photon counts to fluctuate significantly, often mimicking the behavior of variable sources.
Although none of the variable sources we found were supernovae, some produced light curves similar to those of less intense, recurring novae.
Most variable sources had a score near 6, with the highest at 8. All sources with a score above 8 were false positives due to their large errors.
Future Works
Despite not detecting any supernovae in our M87 data, we found many variable sources and some periodic sources. In future research, a larger dataset would increase the likelihood of detecting supernova candidates.
Future work on this project includes implementing a user interface, optimizing the code, and allowing it to work with objects other than M87.
This new code will be used along with X-ray data from 10 other galaxies to search for and categorize supernovae based on their spectra.
Acknowledgement
We want to thank our graduate student advisor, Meagan Porter, for providing guidance on code implementation and data analysis, as well as the Chandra X-ray Center for the data and the CIAO software package.
![]()
