# Variability of massive stars in M31 from the Palomar Transient Factory

## TL;DR

This study analyzes the variability of approximately 500 massive stars in M31 using five years of high-cadence data from the Palomar Transient Factory, revealing spectral type-dependent variability patterns and characteristic timescales.

## Contribution

It provides the first extensive characterization of massive star variability in M31, including timescales and amplitudes, using long-term, high-cadence photometric data.

## Key findings

- Red stars show larger amplitude fluctuations.
- Cool supergiants have longer variability timescales (>100 days).
- Hotter stars vary on timescales of tens of days.

## Abstract

Using data from the (intermediate) Palomar Transient Factory (iPTF), we characterize the time variability of ~500 massive stars in M31. Our sample is those stars which are spectrally typed by Massey and collaborators, including Luminous Blue Variables, Wolf-Rayets, and warm and cool supergiants. We use the high-cadence, long-baseline (~5 years) data from the iPTF survey, coupled with data-processing tools that model complex features in the light curves. We find widespread photometric (R-band) variability in the upper Hertzsprung Russell diagram (or CMD) with an increasing prevalence of variability with later spectral types. Red stars (V-I>1.5) exhibit larger amplitude fluctuations than their bluer counterparts. We extract a characteristic variability timescale, tch, via wavelet transformations that are sensitive to both continuous and localized fluctuations. Cool supergiants are characterized by longer timescales (>100 days) than the hotter stars. The latter have typical timescales of tens of days but cover a wider range, from our resolution limit of a few days to longer than 100 days timescales. Using a 60-night block of data straddling two nights with a cadence of around 2 minutes, we extracted tch in the range 0.1--10 days with amplitudes of a few percent for 13 stars. Though there is broad agreement between the observed variability characteristics in the different parts of the upper CMD with theoretical predictions, detailed comparison requires models with a more comprehensive treatment of the various physical processes operating in these stars such as pulsation, subsurface convection, and the effect of binary companions.