The number of US scientists working in the exoplanet microlensing field is small, and the microlensing method is considered to be obscure by many scientists in the exoplanet field. In order to overcome this shortage of exoplanet microlensing experts, the Roman Project and the Science Support Center (SSC) have begun to produce a very large set of high level data products to enable newcomers to the field to work productively with the data. The RGES PIT works very closely with the Roman Science Centers, the Roman Project team and the Community Science Collaborations relevant to the GBTDS to develop and test these data products.

Pre-Launch:

2018 Data Challenge

The microlensing community held a data challenge in 2018 with the aim of stimulating research on microlensing modeling in preparation for the Roman microlensing survey. The specific goals of the 2018 data challenge were:

  • To stimulate research effort into outstanding modeling issues
  • To stimulate development of algorithms to detect and classify microlensing events in Roman data
  • To stimulate development of software for modeling microlensing events, capable of conducting analyses of Roman-scale datasets

Results of the data challenge as well as lessons learned, evaluation metrics, and other takeaways can be found on the 2018 Data Challenge webpage.

All of the photometric data used in the 2018 Data Challenge is public, and can be downloaded from the Github repository.

2026 Data Challenge

The RGES PIT is currently holding a data challenge in 2026. Some of the objectives of this challenge are similar to the 2018 challenge, however there is added emphasis on the realism and total data volume. The data challenge has two tiers; beginner and advanced. You can learn more about the 2026 data challenge here.

Updated Roman Simulated Microlensing Light Curves (2024-2026)

An updated set of simulated microlensing light curves (which include lens orbital motion, parallax, and limb darkening) has been generated and is hosted by the RGES PIT. The dataset includes 10,000 microlensing light curves in total, of which approximately 2,000 have passed an initial ‘detectability’ cut. The dataset is split into several bins in planet mass (0.01, 0.1, 1.0 10.0, 100.0, and 1000.0 earth masses).

All of the raw light curve data is hosted by LSU and will be available to the public in the near future.

Catalog of Variable Star Light Curves

Working Group 6 has generated a self-consistent library of variable star lightcurves with a Roman noise model that can be used to test microlensing classifier software. Some primary objectives of this work are, (1) compile catalogs of real variable star light curves from prior surveys in the GBTDS footprint, (2) generate a catalog of simulated light curves, (3) resample the light curves to mimic GBTDS-like sensitivity, cadence, etc, and finally (4) transform the resampled light curves from the original passbands to the primary filters that the GBTDS will utilize. The stellar variability types that currently exist in the catalog are as follows:

  • RR Lyraes
  • Long Period Variables (i.e., Miras)
  • Ellipsoidal Binaries
  • Eclipsing Binaries

Several other variable star types are currently in progress to be added to the catalog.

Post-launch

Examples of post-launch data products include:

  • Best-fit light curve models and solution tables for planetary events in the GBTDS
  • Best-fit light curve models and solution tables for free-floating planet candidates in the GBTDS
  • Lens flux analysis best-fit model paramters and solution tables for most events.
  • Best-fit astrometric microlensing models, posteriors, and evidences.

And more.

For a programmatically-generated Table of all pre- and post-launch data products from the RGES PIT, please visit this global pipeline page.

We will continue to make our photometry, astrometry and event modeling data and software publicly available and user-friendly to enable broad participation in RGES science. The same is true for our exoplanet yield simulation tools, and this will enable those interested in doing science with the GBTDS to assess how the survey parameters affect their science yield, as well as that of the RGES.