Matthew Paz | Searcharxiv

arXiv2026

VarWISE: Infrared Variability via NEOWISE Single Exposure Photometry

The Near-Earth Object Wide-field Infrared Explorer (NEOWISE) mission provides a decade of all-sky time-series data at 3.4 and 4.6um and an unprecedented opportunity for the discovery and characterization of variable objects. This paper presents VarWISE, a catalog of infrared-variable objects discovered within the NEOWISE single-exposure data. We employ unique methodologies, including the spatial clustering of apparitions and the adoption of novel machine learning-based variable detection (VARnet) and classification (XGBoost) to identify and characterize significant variability. The catalog includes a prediction of variable object type and best-fit period values for each object, if its variations are cyclical, along with other calculated parameters to characterize the nature of the variability. The VarWISE Pure Catalog, containing only variables of highest confidence, has 457,080 objects, 49.81% of which are new discoveries; the VarWISE Extended Catalog, containing all sources, has 1,918,082 objects, 82.02% of which are new. We discuss caveats for each variable type and highlight a few new objects found during a quick perusal of the catalogs' contents.

astro-ph.SR↗

arXiv2024

A Sub-Millisecond Fourier and Wavelet Based Model to Extract Variable Candidates from the NEOWISE Single-Exposure Database

This paper presents VARnet, a capable signal processing model for rapid astronomical timeseries analysis. VARnet leverages wavelet decomposition, a novel method of Fourier feature extraction via the Finite-Embedding Fourier Transform (FEFT), and deep learning to detect faint signals in light curves, utilizing the strengths of modern GPUs to achieve sub-millisecond single-source runtime. We apply VARnet to the NEOWISE Single-Exposure Database, which holds nearly 200 billion apparitions over 10.5 years of infrared sources on the entire sky. This paper devises a pipeline in order to extract variable candidates from the NEOWISE data, serving as a proof of concept for both the efficacy of VARnet and methods for an upcoming variability survey over the entirety of the NEOWISE dataset. We implement models and simulations to synthesize unique light curves to train VARnet. In this case, the model achieves an F1 score of $0.91$ over a 4-class classification scheme on a validation set of real variable sources present in the infrared. With $\sim2000$ points per light curve on a GPU with 22GB of VRAM, VARnet produces a per-source processing time of $<53\mu s$. We confirm that our VARnet is sensitive and precise to both known and previously undiscovered variable sources. These methods prove promising for a complete future survey of variability with WISE, and effectively showcase the power of the VARnet model architecture.

astro-ph.IM↗