The 2.5-meter Wide Field Survey Telescope Real-time Data Processing Pipeline I: From raw data to alert distribution

TL;DR

This paper presents a complete real-time data processing pipeline for the WFST by adapting the Rubin LSST science pipelines through a WFST-specific obs_wfst package. It details end-to-end steps from raw image ingestion, calibration, single-frame processing, and image coadding to image differencing and alert distribution, with per-CCD quadrant processing across 36 detectors and GPU-accelerated SFFT-based subtraction. Key contributions include the customized WFST workflow, the integration of a PostgreSQL-based transient association and Avro alert packaging, and a well-defined set of data products (raw, science, difference images and catalogs, templates, coadds) produced in real time or nightly cadence. The work enables near-real-time discovery and characterization of transient phenomena with WFST, leveraging LSST heritage while accommodating WFST’s hardware design and data characteristics.

Abstract

The Wide Field Survey Telescope (WFST) is a dedicated photometric surveying facility built jointly by the University of Science and Technology of China (USTC) and the Purple Mountain Observatory (PMO). Since many of its scientific objectives rely on near-real-time data for effective analysis, prompt processing of WFST images is of great significance. To meet this need, we adapted the Rubin Observatory Legacy Survey of Space and Time (LSST) science pipelines to handle the data collected by WFST. This paper presents the complete data processing workflow, from ingestion of raw images to the distribution of alerts, and details the primary data products generated by our pipeline. Researchers using data processed by this pipeline can refer to this document to fully understand the data processing procedures.

Figures (2)

• Figure 1: Processing flow of the single-frame-process phase
• Figure 2: Processing flow of the image-differencing phase