We have analysed observational data from the fourth phase of the Optical Gravitational Lensing Experiment (OGLE-IV), covering the Magellanic Clouds, the Galactic Bulge and Disk, in order to identify all long period variables (LPVs), including LSP stars. This required processing 2.4 billion light curves. To achieve this, we adapted software designed for detecting periodicities in time-resolved observations using Fourier transforms, to efficiently process large datasets. We are currently working on developing efficient methods to classify variable stars with machine learning techniques.
Additionally, we identified bright LSP variables stars in the All Sky Automated Survey (ASAS) data as candidates for interferometric follow-up. Such observations could provide valuable insights for testing the hypothesis proposed in the grant, that is, that there is a spiral dusty structure following the companion on its orbit around the red giant host star.
During our analysis of the radial velocity data of 9,614 LPV stars published as a Focused Product Release by the Gaia Collaboration, we found unexpected phase shifts between the light and radial velocity curves. These shifts differ from predictions made by existing LSP system models, and this investigation is ongoing. In the course of this analysis we also discovered a classical Cepheid with a pulsation period of 78 days, which makes it the first ultra-long period Cepheid discovered in the Milky Way. Previously, such stars had only been observed in neighbouring galaxies, raising questions about their apparent absence in our own Galaxy.
In parallel, we have been working on hydrodynamic simulations of the LSP systems using the Smoothed Particle Hydrodynamics code Phantom, focusing on cooling processes which are crucial for dust formation around the red giant. We have identified and resolved problems in the existing implementation of the cooling processes and are working on further improvement to the numerical integration methods.
