A Pipeline for the ROTSE-IIId Archival Data

TL;DR

This paper presents a new, efficient pipeline for detecting variable stars in the ROTSE-IIId archival data, capable of processing large datasets quickly and accurately, including long-period variables.

Contribution

The authors developed a fast, robust pipeline with HPC integration for variable star detection in ROTSE-III data, improving analysis speed from weeks to hours.

Findings

Successfully detected long-period variable stars like Mira and SR types.

Pipeline significantly reduces analysis time using HPC.

Effective in processing large sky areas with high performance.

Abstract

We have constructed a new, fast, robust and reliable pipeline to detect variable stars from the ROTSE-IIId archival data. Turkish share of ROTSE-III archive contains approximately one million objects from a large field of view (1.85\dgr) and it considerably covers a large portion of northern sky ($\delta>-25\dgr$). The unfiltered ROTSE-III magnitude of the objects ranges from 7.7 to 16.9. The main stages of the new pipeline are as follows: Source extraction, astrometry of the objects, light curve generation and inhomogeneous ensemble photometry. A high performance computing (HPC) algorithm has also been implemented into the pipeline where we had a good performance even on a personal computer. Running the algorithms of the pipeline on a cluster decreases analysis time significantly from weeks to hours. The pipeline is especially tested against long period variable stars with periods of a few hundred days (e.g Mira and SR) and variables having periods starting from a few days to a few hundred days were detected.