"SDSS-IV DR16 Quasar Catalog: Spectroscopic Data for Quasar Analysis and Classification"

Published: 23 December 2024| Version 1 | DOI: 10.17632/cg3vwcc8j5.1
Contributor:
sen pang

Description

Research Hypothesis The primary hypothesis of this research is that deep learning techniques, specifically convolutional neural networks (CNNs), can effectively classify quasars into Broad Absorption Line (BAL) and non-BAL categories based on the spectroscopic data provided by the Sloan Digital Sky Survey (SDSS) IV DR16. Given that the SDSS dataset includes a wide range of quasar observations with varying spectral characteristics, it is hypothesized that machine learning methods, when trained on sufficiently large and diverse data, can achieve high classification accuracy in differentiating between these two types of quasars. Description of the Data The data comes from the SDSS-IV DR16 Quasar Catalog (DR16Q), which is part of the SDSS’s ongoing observational efforts to create detailed astronomical catalogs. DR16Q provides spectroscopic measurements for over 750,000 quasars, with more than 920,000 individual observations spanning a spectral range from 3600 Å to 10400 Å, with a resolution of approximately 2000. These quasars have been classified based on their spectral features, and the dataset contains valuable information such as redshift values, spectral fluxes, and other relevant astrophysical parameters. The dataset includes key features: Spectral Wavelengths: Range from 3600 Å to 10400 Å. Spectral Resolution: Approximately 2000. Quasar Classification: Information about whether the quasar is classified as a Broad Absorption Line (BAL) quasar or a non-BAL quasar. Observational Data: Quasar magnitudes, flux densities, and redshift values. Data Gathering Process The SDSS-IV DR16 dataset was obtained through the SDSS collaboration's use of telescopes at Apache Point Observatory (APO) and Las Campanas Observatory (LCO). The data was collected using the Sloan Foundation 2.5-meter telescope and the DuPont 2.5-meter telescope, with the aim of mapping a wide variety of astronomical objects, including quasars. The quasar spectral data in this dataset is from the SDSS-IV's extended Baryon Oscillation Spectroscopic Survey (eBOSS), which focuses on obtaining accurate redshifts and spectral characteristics of quasars. The DR16Q catalog was publicly released in 2020, and includes spectroscopic data of quasars observed during the SDSS-IV phase of the survey. It represents the final catalog of the SDSS-IV/eBOSS era and is the largest such collection of quasars to date. Because the original data is too large, it is about 260Gb. Therefore, only some of the data samples are uploaded here

Files

Institutions

City University of Macau

Categories

Quasar Absorption Line, Deep Learning

Licence