The Unhurried Universe: A Continued Search for Long Term Variability in ASAS-SN

TL;DR

The paper extends the search for long-term stellar variability in ASAS-SN by targeting the faint regime $14.5<g<15$ and selecting sources with slow brightness trends exceeding $0.03$ mag/yr over roughly a decade. It employs cross-matched multiwavelength data (Gaia DR3, WISE, AAVSO) and Lomb–Scargle period analysis to classify 426 slow-variable candidates (200 new, 226 known) into five CMD groups, with a preponderance of AGB stars; about $259$ show longer-period periodicity, and dust-related variability is inferred from mid-IR diagnostics in several objects. When combined with the prior PK25 sample, the total slow-variable census reaches about $1208$ objects, highlighting a population dominated by long-period pulsators and dust-affected giants, with a subset showing short-timescale periodicity. The work lays groundwork for extending the search across magnitude ranges, improving false-positive rejection, and studying dust dynamics and deep eclipses in long-term variability.

Abstract

We search a sample of 5,685,060 isolated sources in the All Sky Automated Survey for SuperNovae (ASAS-SN) with 14.5<g<15 mag for slowly varying sources with brightness changes larger than ~0.03 mag/year over 10 years. We find 426 slowly-varying systems. Of these systems, 200 are identified as variables for the first time, 226 are previously classified as variables, and we find equal numbers of sources becoming brighter and fainter. Previously classified systems were mostly identified as semi-regular variables (SR), slow irregular variables (L), or unknown (MISC or VAR), as long time scale variability does not fit into a standard class. Much like Petz & Kochanek 2025, the sources are scattered across the color magnitude diagram and can be placed into 5 groups that exhibit distinct behaviors. There are also six AGN. There are 262 candidates (~62 percent) that also show shorter time scale periodic variability, mostly with periods longer than 10 days. The variability of 66 of these candidates may be related to dust. Combining the new slow variable candidates with the candidates from Petz & Kochanek 2025, we have found a total of 1208 slow variables.

Figures (14)

• Figure 1: Distribution of candidates after the first bright star cut. We keep stars above the red solid line. Green squares show the candidates that make up our final sample.
• Figure 2: The Gaia DR3 $M_{G}$ and $B_{P} - R_{P}$ color-magnitude diagram of the final candidates divided into groups based on the dashed lines. The curves are Solar metallicity 1 and 10 Gyr MIST isochrones. The empty points show the CMD positions of candidates from PK25 and the filled in points show the new candidates from this work. One example of light curve is shown for each group.
• Figure 3: The Gaia DR3 $M_{G}$ and $B_{P} - R_{P}$ color-magnitude diagram of all variables in SkyPatrol V2.0 with $13<g<15$ mag that are classified as SR, L, MISC, and VAR (small points) as compared to the candidates with these classifications (large points). Previously unclassified candidates are marked by squares. The empty points show the CMD positions of candidates from PK25 and the filled in points show the new candidates from this work.
• Figure 4: Total period distribution for the candidates by group with $13<g<15$ mag.
• Figure 5: Left: Optical variability slope as a function of the W1$-$W2 color variability slope. Red and blue shaded regions indicates candidates that are likely getting redder and dimmer, or bluer and brighter. Right: $G_{BP} - G_{RP}$ color as a function of mean W1$-$W2 color. Candidates in the orange shaded region have a mid-IR excess due to dust emission. Candidates from the shaded region of each panel are shown with colored symbols on the other panel. The empty points show the candidates from PK25 and the filled in points show the new candidates from this work.